5 of 5 players have a best-fit role from the data. 0 conflicts (the same best role, captain decides), 2 weak fits. Derived from the data, not set by anyone.
Full five record 3–1 in 4 matches
Squad rounds won
Full five56.2%±13.8 pp4 matches
At least 4 of 5 (incl. full five)55.9%±11.2 pp11 matches
A verdict on the squad needs 6 Full five matches.
How to read a card
How to read a card. The big number is the player's own FACEIT ELO, shown as FACEIT gives it. The role is the best fit to the data (or the captain's pick) and decides which stats appear; there is no overall score and no tier, because nothing in the data can weigh a support against a fragger. Blue = lifetime value, orange = the Full five matches, with its range. The dashed tick on a bar is the natural even point (50%, K/D 1.00). Stats marked Style, not quality say what a player does, not how well. Counters are summed, never the ratio fields FACEIT prints. A map is named best or worst only when it clearly differs from the same player's other maps; otherwise the card says so.
z and bar. On a map row, z says how far the map lies from the same player's other maps, in standard errors (0 = no difference). The bar is the z a map has to reach to count; it rises with the number of maps tested, so that testing many maps does not produce a mark by chance. A map under the bar gets no mark and no verdict.
Ranges. The orange Full five value comes with a 95% range from match-to-match spread. In Full five mode 0 of 5 shown stats lie outside their lifetime range; at 95% chance alone gives about 0.2. A ◆ mark is therefore a question to ask, not a finding.
Roles. Every player gets the role that fits him best compared with the other four, whatever roles the squad already has. When two or more fit the same role best, the app shows the conflict and the alternatives and leaves the decision to the captain (role_kapitana in config.json). Confidence (high, medium, low, weak) comes from the sample size and the gap to the next role, never from the size of the score. IGL cannot be measured from the data and appears only as a captain's pick. Nothing here is a verdict on a player.
Lifetime sample. 434–883 matches with extended stats each. Of 1 101–3 355 lifetime matches each, only 23–39% carry the extended stats. One averaged point with no time axis. FACEIT account history: lobbies differ by ELO, so numbers between players are not on the same difficulty.
Not graded. The IGL role is not determined by data (it would need about 150 matches and demos), so no card carries it. By convention an IGL on the AWP is often discouraged; the data neither confirm nor refute that. If the AWP player also calls the game, that load is the captain's concern, not something a card can show. A verdict on the squad needs 6 Full five matches.
Terms, sources and the minimum samples are on .
berkovskhy
AWPbest fit · weak
Lifetime n = 600 matcheswith extended stats, of 2 224 lifetime matches
2 404level 10FACEIT ELO
Best fit: AWP · Sniper share of kills 20.7% against 8.2% for the other four. Barely ahead of Fragger: nothing separates the two roles for him.
Also fits: Fragger, Anchor
Details
Confidence: weak. Nothing separates this role from the next one, or he does not stand out from the other four on any role. Still the best fit available, but only just. It comes from the sample size and the gap to the next role, never from the size of the score.
Based on 600 lifetime matches.
Fit = how far above the other four on the role's signals, in standard deviations of the five: Fragger +1.6, Anchor +1.6. A fit within this squad, not a grade.
Sniper kills / round600 matches
0.17
Five0.1995%: 0.02–0.364 matches
All0.17600 matches
Sniper share of killsStyle, not quality600 matches
21%
Five23%95%: 2%–44%4 matches
All21%600 matches
AWP share of killsStyle, not qualitydemos · 180 kills
24%
Details
First kill of the round with the AWPStyle, not qualitydemos · 24 AWP rounds
29%
AWP kills / deaths with the AWPdemos · 17 deaths w/ AWP
2.53
Maps (6)
Sniper share of kills · per map
DUST2 and MIRAGE are his best sniper maps (30% and 28% of kills, against 19% on his other maps). INFERNO and ANCIENT are the weakest (10% and 12% of kills, against 24% and 22% on his other maps).
Details
▲DUST2 29.5% vs 18.8% on the other maps1 925 kills · z +4.9
▲MIRAGE 28.3% vs 18.5% on the other maps2 360 kills · z +4.8
NUKE 24.6% vs 20.2% on the other maps1 288 kills · z +1.7
ANUBIS 17.4% vs 20.9% on the other maps720 kills · z -1.1
▼ANCIENT 12.1% vs 22.2% on the other maps1 629 kills · z -4.3
▼INFERNO 10.3% vs 23.6% on the other maps2 366 kills · z -6.5
6 maps tested; marked ones pass the corrected bar (z 2.6). Without a mark: no reliable verdict (strongest z 6.5).
Kills vs deaths · by map
No map stands out in kills against deaths: the gaps are small enough for chance to explain them.
Details
— no map with a reliable verdict6 maps with 50+ matches tested; strongest z 2.0, bar 2.6
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
_-Fresh
Second entrybest fit · weak
Lifetime n = 803 matcheswith extended stats, of 2 391 lifetime matches
2 342level 10FACEIT ELO
Best fit: Second entry · Trade kills per T round 0.16 against 0.13 for the other four. Barely ahead of Entry: nothing separates the two roles for him.
Also fits: Entry, Fragger
Details
Confidence: weak. Nothing separates this role from the next one, or he does not stand out from the other four on any role. Still the best fit available, but only just. It comes from the sample size and the gap to the next role, never from the size of the score.
Based on 11 demos and 18 counted events.
Fit = how far above the other four on the role's signals, in standard deviations of the five: Entry +1.5, Fragger +1.0. A fit within this squad, not a grade.
Trade kills per T roundStyle, not qualitydemos · 115 rounds
0.16
Own deaths traded (T)demos · 79 deaths
22%
From the demos · candidate test
Nothing separates him from the other four on Second entry. He is higher in 6 of 11 matches that differ, which chance could explain.
Details
Second entry: p = 1.00. trade kills per T round 0.16 vs 0.13 for the other four.
A role to observe, never an assignment. Sign test over matches (the unit is a match), each player against the other four in the same match; a candidate needs 6 matches that differ and p < 0.05. Several players and two roles are tested with no correction, so one false candidate is plausible.
Maps (6)
Kills vs deaths · by map
DUST2 is his best map in duels: 55.3% of his kills and deaths there are kills, against 53.1% on his other maps. No map is clearly weaker than the rest.
Details
▲DUST2 55.3% vs 53.1% on the other maps5 661 kills+deaths · z +2.7 · bar 2.6
worst map: none reliable (lowest z -1.6)
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
-michal
Lurkbest fit · low
Lifetime n = 883 matcheswith extended stats, of 2 381 lifetime matches
2 156level 10FACEIT ELO
Best fit: Lurk · T rounds spent in isolation 7.7% against 0.7% for the other four; T rounds survived 35% against 28% for the other four.
Also fits: Anchor, Second entry
Details
Confidence: low. Small sample or a narrow gap to the next role: a role to watch, not a settled one. It comes from the sample size and the gap to the next role, never from the size of the score.
Based on 7 demos and 6 counted events. Next best: Anchor.
Capped at low: only 6 counted events carry this fit (30 are needed for more), however many demos there are. The 30 is a convention, not a measurement. Capped at low: the sign test over matches has too few matches that differ to test yet (the candidate test is on this card).
Fit = how far above the other four on the role's signals, in standard deviations of the five: Anchor +1.2, Second entry +0.4. A fit within this squad, not a grade.
T rounds in isolationStyle, not qualitydemos · 78 rounds
8%
T rounds survivedStyle, not qualitydemos · 78 rounds
35%
Late-round kills (from 60 s)Style, not qualitydemos · 98 kills
28%
From the demos · candidate test
Too early to say if he fits Lurk. He is higher in 4 of 4 matches that differ, but 6 are needed before this can be tested.
Details
Lurk: not tested: fewer than 6 matches differ. rounds in isolation (T) 7.7% vs 0.4% for the other four.
A role to observe, never an assignment. Sign test over matches (the unit is a match), each player against the other four in the same match; a candidate needs 6 matches that differ and p < 0.05. Several players and two roles are tested with no correction, so one false candidate is plausible.
Maps (6)
Kills vs deaths · by map
No map stands out in kills against deaths: the gaps are small enough for chance to explain them.
Details
— no map with a reliable verdict6 maps with 50+ matches tested; strongest z 2.2, bar 2.6
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
MaT1337
Entrybest fit · low
Lifetime n = 784 matcheswith extended stats, of 3 355 lifetime matches
1 980level 9FACEIT ELO
Best fit: Entry · Openings per T round (without the AWP) 0.29 against 0.16 for the other four; Opening duels won on T (without the AWP) 48% against 56% for the other four.
Question for the captain: Who really takes the first contact on T?
Also fits: Anchor, Second entry
Details
Confidence: low. Small sample or a narrow gap to the next role: a role to watch, not a settled one. It comes from the sample size and the gap to the next role, never from the size of the score.
Based on 8 demos and 11 counted events. Next best: Anchor.
Capped at low: only 11 counted events carry this fit (30 are needed for more), however many demos there are. The 30 is a convention, not a measurement.
Lifetime FACEIT, context only (not scored): Entry rate (lifetime, both sides) 17.8% against 19.6% for the other four; Entry success (lifetime, both sides) 48.7% against 52.3% for the other four. It mixes both sides and AWP picks, so it does not decide the Entry fit.
Fit = how far above the other four on the role's signals, in standard deviations of the five: Anchor -0.5, Second entry -0.6. A fit within this squad, not a grade.
Entry success3 076 entries
49%
Five42%95%: 21%–62%4 matches
All49%3 076 entries
Entry rateStyle, not quality784 matches
18%
Five27%95%: 17%–37%4 matches
All18%784 matches
T openings / round (no AWP)Style, not qualitydemos · 79 T rounds
0.29
Details
T opening duels won (no AWP)demos · 23 openings
48%
Opening with the AWP (T)Style, not qualitydemos · 23 T openings
0%
Round won after an entry windemos · 11 won openings
82%
Time to first contact (T)Style, not qualitydemos · 23 contacts
13.7 s
Entries with a teammate's flashnot stored yetThe parser does not record this link between events, so no number is shown.
Maps (6)
Entry success · per map
No map clearly stands out from his other maps: the gaps are small enough for chance to explain them.
Details
AIM MAP FACEIT 58% vs 49% on the other maps45 entries · z +1.2
AIM PISTOLS FACEIT 53% vs 49% on the other maps30 entries · z +0.5
NUKE 53% vs 48% on the other maps284 entries · z +1.5
INFERNO 51% vs 48% on the other maps692 entries · z +1.1
ANCIENT 50% vs 49% on the other maps636 entries · z +0.4
DUST2 48% vs 49% on the other maps551 entries · z -0.3
MIRAGE 47% vs 49% on the other maps722 entries · z -1.3
ANUBIS 45% vs 49% on the other maps106 entries · z -0.7
TRAIN 37% vs 49% on the other maps38 entries · z -1.4
9 maps tested; marked ones pass the corrected bar (z 2.8). Without a mark: no reliable verdict (strongest z 1.5).
Kills vs deaths · by map
No map stands out in kills against deaths: the gaps are small enough for chance to explain them.
Details
— no map with a reliable verdict6 maps with 50+ matches tested; strongest z 2.4, bar 2.6
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
sqaar97
Support / flasherbest fit · high
Lifetime n = 434 matcheswith extended stats, of 1 101 lifetime matches
1 627level 8FACEIT ELO
Best fit: Support · Enemies flashed per round 0.70 against 0.36 for the other four.
Also fits: AWP, Entry
Details
Confidence: high. Large sample and a clear gap to the next role. It comes from the sample size and the gap to the next role, never from the size of the score.
Based on 434 lifetime matches. Next best: AWP.
Fit = how far above the other four on the role's signals, in standard deviations of the five: AWP +0.7, Entry -0.7. A fit within this squad, not a grade.
Enemies flashed / round434 matches
0.70
Five0.5295%: 0.25–0.784 matches
All0.70434 matches
Flash success8 772 flashes
51%
Five53%95%: 38%–68%4 matches
All51%8 772 flashes
Utility damage / round434 matches
6.6
Five6.795%: 1.7–11.74 matches
All6.6434 matches
Details
Enemy blind seconds / rounddemos · 222 rounds
2.32
Teamflash ratiodemos · 514 enemy blind s
0.44
Kills soon after his flash / rounddemos · 222 rounds
0.26
Maps (4)
Enemies flashed / round · per map
ANCIENT is his best flashing map (0.96 enemies flashed per round, against 0.67 on his other maps). NUKE and INFERNO are the weakest (0.41 and 0.53 enemies flashed per round, against 0.73 and 0.74 on his other maps).
Details
▲ANCIENT 0.96 vs 0.67 on the other maps1 061 rounds · z +6.0
ANUBIS 0.79 vs 0.68 on the other maps1 694 rounds · z +2.6
MIRAGE 0.75 vs 0.68 on the other maps2 754 rounds · z +1.9
DUST2 0.65 vs 0.71 on the other maps1 174 rounds · z -1.3
▼INFERNO 0.53 vs 0.74 on the other maps1 676 rounds · z -5.2
▼NUKE 0.41 vs 0.73 on the other maps800 rounds · z -5.8
6 maps tested; marked ones pass the corrected bar (z 2.6). Without a mark: no reliable verdict (strongest z 6.0).
Kills vs deaths · by map
INFERNO is his best map in duels: 53.3% of his kills and deaths there are kills, against 49.4% on his other maps. No map is clearly weaker than the rest.
Details
▲INFERNO 53.3% vs 49.4% on the other maps2 490 kills+deaths · z +3.1 · bar 2.5
worst map: none reliable (lowest z -1.4)
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
Team
3–1Full five · 4 matches
Record only: under 10 matches a percentage would read like a result and is not one. Only Full five matches.
89rounds played together
Latest 2026-09-23.
18matches with part of the squad
They stay out of the comparison: other opponents and maps would mix into it.
Map pooln = 4 matches · earlyINFERNO 1 / 2 won · DUST2 1 / 1 won · MIRAGE 1 / 1 won
Full five comparisonn = 4 matches · earlysame matches for everyone
Metric
berkovskhy
_-Fresh
-michal
MaT1337
sqaar97
Fragging
K/D
1.28
1.54
1.16
1.18
0.69
ADR
100
101
84
80
50▽
Headshot kills
33%▽
32%▽
59%△
61%
48%
Multi-kill rounds
19%
28%
17%
19%
9%▽
Opening
Opening duels taken
15%▽
24%
19%
27%
16%▽
Opening duels won
62%
81%△
65%
42%
36%
Sniper
Sniper share of kills
23%
14%
0%▽
1%▽
2%▽
Utility
Flashes per round
1.08△
0.74
0.44▽
0.37▽
0.67
Enemies flashed per round
1.08△
0.51
0.34▽
0.36▽
0.52
Flash success
68%
47%
51%
67%
53%
Utility damage per round
14.7
2.4▽
11.8
5.8
6.7
Late round
1vX situations
6.7%
3.4%
6.7%
4.5%
7.9%
1vX won
17%
0%
50%
25%
43%
Note
With 4 shared matches every arrow is only a hint (p = 0.125 at best): a clear arrow needs p < 0.05, and with this few matches no row can get there. A hint is closer to a coin toss than to a finding. Everyone is compared over the SAME matches: same rounds, same opponents, same maps — so the difference is between players, not between rosters. A full triangle (▲/▼) is a clear difference from the rest of the squad; an outlined one (△/▽) is only a hint. Green and red only where higher is better; ▲ marks a STYLE signal (opening duels, sniper, flashes, 1vX), where more means a different role, not a better one. The thin bar under a value only compares the five within that row. Source: FACEIT server stats, no demos needed. Hover a metric name for its definition, or a player for the number of matches stored for him.Details
The arrow is a sign test with a MATCH as the unit: in how many shared matches the player was above the average of the rest, and in how many below (rounds of one match are not independent). Full triangle: p < 0.05. Outlined and faded (a hint): p < 0.2.
Roles
Rolespick a role
Best role per playerfit to the role within this squad, not a grade
Player
AWP
Entry
Support
Lurk
Second entry
Anchor
Fragger
Confidence
Captain's pick
berkovskhy
best +1.7
-0.9
+0.8
-0.1
+0.3
alt +1.6
alt +1.6
weak
–
_-Fresh
+0.5
alt +1.5
+0.2
+0.3
best +1.7
-0.3
alt +1.0
weak
–
-michal
-1.4
-1.0
-1.1
best +1.8
alt +0.4
alt +1.2
-0.9
low
–
MaT1337
-0.9
best +0.4
-0.8
-0.6
alt -0.6
alt -0.5
-0.8
low
–
sqaar97
alt +0.7
alt -0.7
best +1.9
-1.1
-1.8
-1.5
-0.9
high
–
No conflicts: every player has a different best role.
Note
How to read this. Each number says how far the player is above the other four on that role's signals, in standard deviations of the five; it ranks roles for one player, it does not rank players. best is his best fit, alt the next two. A dash means no signal for that role (data missing or sample too small), never a zero. IGL is not measured: only the captain can set it.How confidence works
Confidence is high, medium, low or weak and comes from the sample size and the gap to the next role. The captain's pick (role_kapitana in config.json) comes before the data fit wherever the role matters, and the fit stays visible beside it.
Ranked from FACEIT
Described, not ranked
Nothing measured yet
Entry fragger (opener)T
Goes in first and takes the opening duel so the team gets space. Often dies first; a good entry wins more than half of those duels or gets traded. Not the number of kills, but who takes them and when.
#
Player
Entry rate
Entry success
Average place
1
berkovskhyAWP best fit
22%#195%: 21%–23%
58%#195%: 56%–59%
1.00
2
_-FreshSecond entry best fit
20%#295%: 19%–20%
55%#2≈95%: 53%–57%
2.00
3
-michalLurk best fit
19%#3≈95%: 18%–20%
48%#4≈95%: 46%–50%
3.50
4
MaT1337Entry best fit
18%#5≈95%: 17%–18%
49%#395%: 47%–51%
4.00
5
sqaar97Support / flasher best fit
18%#4≈95%: 17%–19%
48%#5≈95%: 45%–50%
4.50
Note
Best fit first. It compares the five with each other, not with a norm: there is no pro-play data set. There is no overall score.How the order is made
Each ranked stat gives the five a place (higher = fits the role better); the order is the mean of those places.
≈ means the 95% range overlaps the player ranked just above, so on that stat they cannot be told apart.
Lifetime counters, lobbies of different ELO. Stats marked (info) are shown but not ranked.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
Openings per T round (without the AWP) Style, not quality
0.13115 T rounds
0.23115 T rounds
0.1278 T rounds
0.2979 T rounds
0.15115 T rounds
Opening duels won on T (without the AWP)
47%15 openings
67%27 openings
4 / 9openings · too few events
48%23 openings
47%17 openings
Opening with the AWP Style, not quality
6%16 T openings
10%30 T openings
no AWP kills or deaths
0%23 T openings
0%17 T openings
Time to first contact Style, not quality
10.4 s7 contacts
10.1 s27 contacts
12.1 s8 contacts
13.7 s23 contacts
12.8 s26 contacts
Conversion after a won opening
6 / 8won openings · too few events
71%21 won openings
4 / 4won openings · too few events
82%11 won openings
6 / 8won openings · too few events
Median time of death Style, not quality
41.8 s151 deaths
34.8 s156 deaths
38.4 s104 deaths
29.9 s107 deaths
39.0 s157 deaths
Early deaths (before 30 s) Style, not quality
36%151 deaths
42%156 deaths
38%104 deaths
50%107 deaths
39%157 deaths
Median kill distance Style, not quality
18.9 m180 kills
17.5 m209 kills
14.1 m98 kills
15.5 m144 kills
15.1 m113 kills
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
Lifetime entry rate alone does not separate roles here: it sits near 0.18-0.20 for most of the five, and it mixes both sides and AWP picks. The Entry fit therefore comes from the demos (T openings without the AWP, per T round, and how often he wins them); the lifetime figures stay beside it as context. Success depends on the lobby and on teammates' flashes.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Entry rate
Total Entry Count / rounds with extended stats
FACEIT stats
style
measured · ranked
Entry success
Entry Wins / Entry Count
FACEIT stats
quality
measured · ranked
Openings per T round (without the AWP)
his T openings (the first death of the round, his kill or his death) without the AWP in his hand / T rounds; this is the volume that scores the Entry fit
demos
style
measured from the demos · info
Opening duels won on T (without the AWP)
his won T openings / his T openings, the same set; this is the success that scores the Entry fit
demos
quality
measured from the demos · info
Opening with the AWP
T openings in which he had the AWP in hand (his AWP kill or his death with the AWP) / all his T openings; counted apart, an AWP pick is not an entry
demos
style
measured from the demos · info
Time to first contact
median seconds from freeze end to first damage, T side, in the rounds he was part of the first contact
demos
style
measured from the demos · info
Conversion after a won opening
T rounds won / T rounds in which he won the opening duel
demos
quality
measured from the demos · info
Openings with a teammate's flash
won openings with an assisted flash / won openings
demos, not counted yet
quality
needs demos of all five (not computed yet)
Share of the team's openings
his openings / all team openings
demos, not counted yet
style
needs demos of all five (not computed yet)
Median time of death
median seconds from freeze end to his death by an enemy; only rounds he died in, so a survivor is not in it
demos
style
measured from the demos · info
Early deaths (before 30 s)
his deaths in the first 30 s of the round / his deaths (30 s is a convention)
demos
style
measured from the demos · info
Median kill distance
median distance of his kills, in the demo's own unit (meters per Valve's docs)
demos
style
measured from the demos · info
Second entry / tradeT
Enters seconds behind the entry and trades his death, takes the second angle or finishes the player the entry hurt. Looks for the second duel, not the first.
No ranking. None of this role's stats can be measured from FACEIT server stats, and the few that come from the demos describe what happened in those matches, so no one is put in order.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
Trade kills per round Style, not quality
0.14115 rounds
0.16115 rounds
0.1478 rounds
0.1379 rounds
0.11115 rounds
Own deaths traded
22%79 deaths
22%79 deaths
20%51 deaths
26%58 deaths
19%89 deaths
Kills soon after his flash per round
0.30222 rounds
0.21222 rounds
0.09146 rounds
0.12165 rounds
0.26222 rounds
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
Needs positions and events of all five (demos), which FACEIT does not hand out automatically.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Trade kills per round
trade kills / rounds, T side (the parser's definition: a kill shortly after a teammate's death, near him)
demos
style
measured from the demos · info
Share of teammates' deaths traded
his trades / teammates' deaths in his rounds
demos, not counted yet
quality
needs demos of all five (not computed yet)
Distance to nearest teammate at first contact
median units; 'close' = 500 or less
demos, not counted yet
style
needs demos of all five (not computed yet)
Delay behind the entry
median seconds between the two first contacts
demos, not counted yet
style
needs demos of all five (not computed yet)
Own deaths traded
T side: his deaths traded by a teammate / his deaths
demos
quality
measured from the demos · info
Kills soon after his flash per round
kills by teammates shortly after his flash landed / rounds (the parser's definition, not the game's assist credit)
demos
quality
measured from the demos · info
LurkT
Plays alone, away from the team: pulls rotations, catches backs, controls the rear of the map and often decides the end of the round. His role is time and place, not entries.
Player
1vX situations / round (info)
Clutch 1v1 + 1v2 won (info)
Entry rate (info)
berkovskhyAWP best fit
6.1%95%: 5.6%–6.5%
29%95%: 26%–32%
22%95%: 21%–23%
_-FreshSecond entry best fit
7.0%95%: 6.5%–7.4%
31%95%: 29%–34%
20%95%: 19%–20%
-michalLurk best fit
5.7%95%: 5.3%–6.1%
28%95%: 25%–30%
19%95%: 18%–20%
MaT1337Entry best fit
6.1%95%: 5.7%–6.5%
31%95%: 28%–34%
18%95%: 17%–18%
sqaar97Support / flasher best fit
6.8%95%: 6.3%–7.4%
33%95%: 30%–37%
18%95%: 17%–19%
Note
Not ranked. None of this role's FACEIT stats says who fits it better, so they are shown for information, in the same order as on Home, with no places and no average. There is no overall score.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
Rounds spent in isolation Style, not quality
0.0%115 rounds
1.7%115 rounds
7.7%78 rounds
0.0%79 rounds
0.9%115 rounds
Late-round kills (from 60 s) Style, not quality
23%180 kills
18%209 kills
28%98 kills
24%144 kills
29%113 kills
Survival Style, not quality
31%115 rounds
31%115 rounds
35%78 rounds
27%79 rounds
23%115 rounds
Median time of death Style, not quality
41.8 s151 deaths
34.8 s156 deaths
38.4 s104 deaths
29.9 s107 deaths
39.0 s157 deaths
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
Lurk does not come out of FACEIT server stats alone: a low entry rate looks the same on a passive player. The FACEIT stats here are shown for information and are not ranked: 1vX situations count being the last one alive, which is mostly a consequence of teammates dying, not of choosing to play away from the team. Playing in isolation is measured from the demos below, and the two can point different ways.
Method
Stat that describes the role
How it is counted
Source
Type
Status
1vX situations per round
(1v1 Count + 1v2 Count) / rounds with extended stats; last one alive against one or two, which mostly means teammates died first (info, not ranked)
FACEIT stats
style
measured · info
Clutches won (1v1 + 1v2)
Wins / Count
FACEIT stats
quality
measured · info
Low entry rate (a condition, not proof)
see Entry fragger; low only says he does not open
FACEIT stats
style
measured · info
Rounds spent in isolation
T rounds in which even the nearest teammate stayed beyond 1000 units (median distance over the round) / T rounds
demos
style
measured from the demos · info
Late-round kills (from 60 s)
kills 60 s or more after freeze end (a convention, not a measured break) / his kills
demos
style
measured from the demos · info
Kills behind the line
kills while isolated, on an enemy passing his spot
demos, not counted yet
quality
needs demos of all five (not computed yet)
Survival
T rounds without his own death / T rounds
demos
style
measured from the demos · info
Median time of death
median seconds from freeze end to his death by an enemy; only rounds he died in, so a survivor is not in it
demos
style
measured from the demos · info
AWPboth
Plays the sniper rifle: on T opens duels or holds an angle, on CT controls long lines. Carries the most expensive weapon, so losing it costs the team its economy.
#
Player
Sniper share of kills
Sniper kills / round (info)
Average place
1
berkovskhyAWP best fit
21%#195%: 19%–22%
0.1795%: 0.15–0.18
1.00
2
sqaar97Support / flasher best fit
15%#295%: 13%–17%
0.1195%: 0.09–0.12
2.00
3
_-FreshSecond entry best fit
14%#3≈95%: 12%–15%
0.1195%: 0.10–0.12
3.00
4
MaT1337Entry best fit
5%#495%: 4%–6%
0.0495%: 0.03–0.04
4.00
5
-michalLurk best fit
2%#595%: 2%–3%
0.0295%: 0.01–0.02
5.00
Note
Best fit first. It compares the five with each other, not with a norm: there is no pro-play data set. There is no overall score.How the order is made
Each ranked stat gives the five a place (higher = fits the role better); the order is the mean of those places.
≈ means the 95% range overlaps the player ranked just above, so on that stat they cannot be told apart.
Lifetime counters, lobbies of different ELO. Stats marked (info) are shown but not ranked.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
AWP share of kills in the parsed matches Style, not quality
24%180 kills
10%210 kills
0%98 kills
1%144 kills
2%114 kills
K/D with the AWP
2.5317 deaths w/ AWP
21 K / 4 Ddeaths w/ AWP · too few events
no AWP kills or deaths
2 K / 1 Ddeaths w/ AWP · too few events
2 K / 4 Ddeaths w/ AWP · too few events
First kill with the AWP Style, not quality
29%24 AWP rounds
31%13 AWP rounds
no AWP kills or deaths
0 / 2AWP rounds · too few events
0 / 2AWP rounds · too few events
AWP accuracy
46.1%102 AWP shots
46.9%49 AWP shots
no AWP kills or deaths
2 / 6AWP shots · too few events
2 / 15AWP shots · too few events
No-scope AWP kills Style, not quality
0 / 43AWP kills · too few events
0 / 21AWP kills · too few events
no AWP kills or deaths
0 / 2AWP kills · too few events
0 / 2AWP kills · too few events
Median AWP kill distance Style, not quality
29.4 m43 AWP kills
29.5 m21 AWP kills
no AWP kills or deaths
—2 AWP kills · too few
—2 AWP kills · too few
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
FACEIT's own field names for the sniper rate and share are swapped, so the counters are used. Lifetime is dominated by years of history, not current form, and a second AWP on the team lowers the share.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Sniper share of kills
Total Sniper Kills / Total Kills with extended stats
FACEIT stats
style
measured · ranked
Sniper kills per round
Total Sniper Kills / rounds (same counter as above, shown as info)
FACEIT stats
style
measured · info
AWP share of kills in the parsed matches
kills with the AWP / all his kills in the demos (weapon 'awp' only)
demos
style
measured from the demos · info
K/D with the AWP
AWP kills / deaths with the AWP in his hand (under 10 deaths shown as counts, e.g. 21 K / 4 D, with no ratio)
demos
quality
measured from the demos · info
First kill with the AWP
rounds whose first kill (of the whole round) is his AWP kill / rounds with an AWP kill by him
demos
style
measured from the demos · info
Saved AWPs
lost rounds alive with the AWP at round end
demos, not counted yet
quality
needs demos of all five (not computed yet)
AWP accuracy
damage events on enemies / AWP shots in the parsed demos (Steam's lifetime counters exist only for the owner)
demos
quality
measured from the demos · info
No-scope AWP kills
AWP kills without looking through the scope / his AWP kills
demos
style
measured from the demos · info
Median AWP kill distance
median distance of his AWP kills, in the demo's own unit (meters per Valve's docs)
demos
style
measured from the demos · info
Support / flasherboth
Throws for his teammates: flashes under entries, smokes and molotovs for executes, HE damage. His kill numbers are often average; his value shows in the teammates' entries.
#
Player
Enemies flashed / round
Flash success
Utility damage / round
Average place
1
berkovskhyAWP best fit
0.52#295%: 0.50–0.54
54%#195%: 53%–55%
7.6#1
1.33
2
sqaar97Support / flasher best fit
0.70#195%: 0.67–0.73
51%#395%: 50%–53%
6.6#2
2.00
3
-michalLurk best fit
0.25#595%: 0.24–0.26
54%#2≈95%: 53%–56%
6.1#3
3.33
4
MaT1337Entry best fit
0.28#495%: 0.27–0.30
49%#4≈95%: 48%–51%
5.9#4
4.00
5
_-FreshSecond entry best fit
0.43#395%: 0.41–0.44
44%#595%: 43%–45%
5.0#5
4.33
Note
Best fit first. It compares the five with each other, not with a norm: there is no pro-play data set. There is no overall score.How the order is made
Each ranked stat gives the five a place (higher = fits the role better); the order is the mean of those places.
≈ means the 95% range overlaps the player ranked just above, so on that stat they cannot be told apart.
Lifetime counters, lobbies of different ELO. Stats marked (info) are shown but not ranked.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
Kills soon after his flash per round
0.30222 rounds
0.21222 rounds
0.09146 rounds
0.12165 rounds
0.26222 rounds
Enemy blind seconds per round
3.38222 rounds
1.63222 rounds
0.97146 rounds
1.36165 rounds
2.32222 rounds
Teamflash
0.44751 enemy blind s
0.59362 enemy blind s
0.44142 enemy blind s
0.72225 enemy blind s
0.44514 enemy blind s
Flash share of assists Style, not quality
23%79 assists
16%63 assists
5%37 assists
6%36 assists
17%52 assists
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
The three ranked stats are related (all count blinding or grenades), so together they weigh as one support signal. A support is sometimes accidental: someone flashes because someone has to.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Enemies flashed per round
Total Enemies Flashed / rounds
FACEIT stats
style
measured · ranked
Flash success
Flash Successes / Flash Count
FACEIT stats
quality
measured · ranked
Utility damage per round
Total Utility Damage / rounds (HE + molotov)
FACEIT stats
style
measured · ranked
Assists per round
Assists / rounds
FACEIT stats
style
available from FACEIT, not collected yet
Kills soon after his flash per round
kills by teammates shortly after his flash landed / rounds (the parser's definition, not the game's assist credit)
demos
quality
measured from the demos · info
Enemy blind seconds per round
sum of blind duration on enemies / rounds
demos
quality
measured from the demos · info
Teamflash
teammate blind seconds / enemy blind seconds
demos
quality
measured from the demos · info
Flash share of assists
assists the game credits to his flash / all his assists
demos
style
measured from the demos · info
Utility specialistboth
Owns smokes, molotovs and executes, and the defence of a site. Overlaps with the support, but his field is smoke and fire, not flash.
#
Player
Utility used / round
Utility damage / round (info)
Average place
1
sqaar97Support / flasher best fit
0.61#195%: 0.59–0.63
6.6
1.00
2
berkovskhyAWP best fit
0.57#295%: 0.56–0.59
7.6
2.00
3
_-FreshSecond entry best fit
0.42#395%: 0.41–0.43
5.0
3.00
4
-michalLurk best fit
0.42#4≈95%: 0.41–0.43
6.1
4.00
5
MaT1337Entry best fit
0.37#595%: 0.36–0.38
5.9
5.00
Note
Best fit first. It compares the five with each other, not with a norm: there is no pro-play data set. There is no overall score.How the order is made
Each ranked stat gives the five a place (higher = fits the role better); the order is the mean of those places.
≈ means the 95% range overlaps the player ranked just above, so on that stat they cannot be told apart.
Lifetime counters, lobbies of different ELO. Stats marked (info) are shown but not ranked.
Note
Smoke quality is hard to measure: the demo shows where it landed, not whether it blocked the view.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Utility used per round (no flashes)
Total Utility Count / rounds
FACEIT stats
style
measured · ranked
Utility damage per round
Total Utility Damage / rounds (shown as info)
FACEIT stats
style
measured · info
Molotov damage per round
fire damage / rounds
owner's demos
style
needs own demos (owner only)
Smokes on a site before the plant
smokes landed in site callouts / smokes
demos, not counted yet
quality
needs demos of all five (not computed yet)
Molotov block time
seconds a thrown molotov blocks a path / rounds
demos, not counted yet
quality
needs demos of all five (not computed yet)
Utility left unused at death
grenade value in inventory at death / rounds
owner's demos
quality
needs own demos (owner only)
Fragger / riflerboth
No narrow function: shoots, trades and wins duels. His value is overall effectiveness in the fight. In this project a non-exclusive role, often a second one.
#
Player
K/D
ADR
Rounds with 3+ kills
Clutch 1v1 + 1v2 won (info)
Average place
1
berkovskhyAWP best fit
1.17#195%: 1.14–1.20
89#1
6.7%#195%: 6.3%–7.1%
29%95%: 26%–32%
1.00
2
_-FreshSecond entry best fit
1.15#2≈95%: 1.12–1.18
85#2
6.3%#2≈95%: 5.9%–6.6%
31%95%: 29%–34%
2.00
3
-michalLurk best fit
0.98#5≈95%: 0.96–1.01
79#3
4.5%#395%: 4.2%–4.8%
28%95%: 25%–30%
3.67
4
MaT1337Entry best fit
1.03#395%: 1.01–1.05
77#4
4.5%#5≈95%: 4.2%–4.7%
31%95%: 28%–34%
4.00
5
sqaar97Support / flasher best fit
1.01#4≈95%: 0.97–1.05
77#5
4.5%#4≈95%: 4.0%–4.9%
33%95%: 30%–37%
4.33
Note
Best fit first. It compares the five with each other, not with a norm: there is no pro-play data set. There is no overall score.How the order is made
Each ranked stat gives the five a place (higher = fits the role better); the order is the mean of those places.
≈ means the 95% range overlaps the player ranked just above, so on that stat they cannot be told apart.
Lifetime counters, lobbies of different ELO. Stats marked (info) are shown but not ranked.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
KAST
76%222 rounds
78%222 rounds
62%146 rounds
73%165 rounds
67%222 rounds
Accuracy
19.8%2 880 shots
22.3%2 764 shots
19.8%1 536 shots
18.9%2 175 shots
16.9%2 328 shots
Head share of hits Style, not quality
16%569 hits
16%615 hits
22%304 hits
25%411 hits
20%394 hits
Kills through smoke Style, not quality
7.2%180 kills
6.7%209 kills
5 / 98kills · too few events
11.1%144 kills
8.8%113 kills
Wallbang kills Style, not quality
6 / 180kills · too few events
6 / 209kills · too few events
3 / 98kills · too few events
6 / 144kills · too few events
5 / 113kills · too few events
Kills while blinded Style, not quality
1 / 180kills · too few events
1 / 209kills · too few events
1 / 98kills · too few events
1 / 144kills · too few events
1 / 113kills · too few events
Kills while airborne Style, not quality
1 / 180kills · too few events
5.7%209 kills
1 / 98kills · too few events
1 / 144kills · too few events
1 / 113kills · too few events
Deaths to a blinded enemy Style, not quality
0 / 151deaths · too few events
3 / 156deaths · too few events
1 / 104deaths · too few events
2 / 107deaths · too few events
0 / 157deaths · too few events
Deaths to an airborne enemy Style, not quality
1 / 151deaths · too few events
1 / 156deaths · too few events
1 / 104deaths · too few events
0 / 107deaths · too few events
0 / 157deaths · too few events
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
K/D and ADR move together. Lifetime mixes lobbies of different ELO, so two players' K/D is not on the same difficulty.
Method
Stat that describes the role
How it is counted
Source
Type
Status
K/D
sum of Kills / sum of Deaths over the per-map segments
FACEIT stats
quality
measured · ranked
ADR
Total Damage / rounds with extended stats
FACEIT stats
quality
measured · ranked
Rounds with 3+ kills
(3k + 4k + 5k) / rounds
FACEIT stats
quality
measured · ranked
Clutches won (1v1 + 1v2)
Wins / Count, a cross-cutting trait shown as info
FACEIT stats
quality
measured · info
Share of the team's damage
his ADR / team ADR in the same matches
FACEIT stats
quality
available from FACEIT, not collected yet
KAST
rounds with a kill, assist, survival or traded death / rounds
demos
quality
measured from the demos · info
Accuracy
damage events on enemies / shots, all firearms except shotguns (pellets give more hits than shots)
demos
quality
measured from the demos · info
Head share of hits
hits in the head / hits, the same weapons as accuracy (the neck is counted separately)
demos
style
measured from the demos · info
Kills through smoke
kills through a smoke / his kills
demos
style
measured from the demos · info
Wallbang kills
kills through a wall or object / his kills
demos
style
measured from the demos · info
Kills while blinded
his kills made while he was himself blinded by a flash / his kills
demos
style
measured from the demos · info
Kills while airborne
his kills made while he was in the air / his kills
demos
style
measured from the demos · info
Deaths to a blinded enemy
his deaths where the killer was blinded at that moment / his deaths
demos
style
measured from the demos · info
Deaths to an airborne enemy
his deaths where the killer was in the air / his deaths
demos
style
measured from the demos · info
AnchorCT
Holds one site alone or in a pair, usually with utility, to buy time for rotations. His success is not kills but that the site did not fall fast and the team made it in time.
No ranking. None of this role's stats can be measured from FACEIT server stats, and the few that come from the demos describe what happened in those matches, so no one is put in order.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
Time in his most-used CT callout Style, not quality
40%4 323 CT seconds
31%3 722 CT seconds
38%2 498 CT seconds
30%3 606 CT seconds
25%4 666 CT seconds
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
Context-dependent: every map has its own callouts. Does not come out of FACEIT counters.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Time in his most-used CT callout
share of his CT time (rounds survived 30 s or more) spent in the one callout he uses most, per map, weighted by time
demos
style
measured from the demos · info
Few rotations
site changes per round, rounds survived 30 s or more
demos, not counted yet
style
needs demos of all five (not computed yet)
Duels won on the site
kills / deaths in the site's callouts
demos, not counted yet
quality
needs demos of all five (not computed yet)
Time bought
median seconds from his first contact to a teammate arriving or the plant
demos, not counted yet
quality
needs demos of all five (not computed yet)
Site held
rounds the site did not fall / rounds it was attacked
demos, not counted yet
quality
needs demos of all five (not computed yet)
Defensive utility on the site
grenades thrown in site callouts / rounds
demos, not counted yet
quality
needs demos of all five (not computed yet)
Rotator / flexCT
Does not sit on one site, reacts to information and moves where the team needs help. A good rotator is in the right place at the right moment, not just a player who changes place often.
No ranking. None of this role's stats can be measured from FACEIT server stats, and the few that come from the demos describe what happened in those matches, so no one is put in order.
From the demos11 matches parsed
Stat
berkovskhy11 matches
_-Fresh11 matches
-michal7 matches
MaT13378 matches
sqaar9711 matches
Effective number of CT callouts Style, not quality
7.44 323 CT seconds
9.33 722 CT seconds
7.62 498 CT seconds
9.73 606 CT seconds
9.94 666 CT seconds
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Note
In our data a 'rotation' mostly measures survival, not map knowledge (CLAUDE.md, rule 36), so no verdict is announced for it.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Effective number of CT callouts
exp(entropy) of his CT time across callouts, rounds survived 30 s or more, per map, weighted by time
demos
style
measured from the demos · info
Site changes after information
rounds he changed site after a teammate's first contact / such rounds
demos, not counted yet
style
needs demos of all five (not computed yet)
Rotation delay
median seconds from a teammate's contact on another site to his arrival
demos, not counted yet
quality
needs demos of all five (not computed yet)
Retake help
his kills and defuses after an enemy plant / rounds with a plant
demos, not counted yet
quality
needs demos of all five (not computed yet)
Kills away from the start position
kills outside the starting callout / his kills
demos, not counted yet
style
needs demos of all five (not computed yet)
Mid / centre controllerboth
Controls the middle (Mirage, Dust2 and alike): opens or blocks passages, gathers information and only then decides where to send the team. On maps without a clear middle the role does not exist.
No ranking. None of this role's stats can be measured from FACEIT server stats or from the parsed demos yet, so no one is ordered.
Note
Needs a hand-written list of mid callouts per map and enough demos on each map.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Share of time in mid callouts
time in mid callouts / alive time
demos, not counted yet
style
needs demos of all five (not computed yet)
Share of openings in mid
his openings in mid / his openings
demos, not counted yet
style
needs demos of all five (not computed yet)
Duels won in mid
kills / deaths in mid callouts
demos, not counted yet
quality
needs demos of all five (not computed yet)
Utility into mid
grenades thrown in mid callouts / rounds
demos, not counted yet
style
needs demos of all five (not computed yet)
Time to first contact in mid
median seconds to first damage in mid
demos, not counted yet
style
needs demos of all five (not computed yet)
IGL (in-game leader)both
Calls strategy, economy and tempo and adapts after lost rounds. The one role whose behaviour is not in the demo: there is no voice in the files, only effects on the team that depend on all five.
No ranking. None of this role's stats can be measured from FACEIT server stats, and the few that come from the demos describe what happened in those matches, so no one is put in order.
after a lost round 37 / 93 = 40%after a won round 72 / 107 = 67%
Note
Descriptive, not ranked. 11 matches parsed: FACEIT matches in which at least 4 of the five played. A few matches and one or two stats do not justify putting five people in order, so there is no ranking and no overall score here. Style stats say what a player does, not how well. Under 10 events a value is shown as x / n; accuracy needs 30 shots first. Team results are the same for everyone: with a fixed five there is no "without him" to compare against, so they describe the team, not the person. Rounds of one match are not independent, so no interval is shown and no verdict is made.Details
Kill totals can differ by a kill or two between rows (for example 210 and 209): the AWP share counts from per-round counters, which also count a suicide as a kill, while the other kill rows count single events against enemies only.
Telling a 5-point difference from chance would take about 150 matches. The win rate after a lost round is lower for almost every team because the economy is weaker, so read it next to the figure after a won round, not as a grade for adaptation.
Note
All of these measure the team, not the person, so a role cannot be assigned from them: with a fixed five there is no 'without him' to compare against.
Method
Stat that describes the role
How it is counted
Source
Type
Status
Team rounds won
rounds won / rounds, in matches with this player
FACEIT stats
quality
available from FACEIT, not collected yet
Rounds won by side
as T and as CT separately (the same for the five: they played the same matches)
demos
quality
measured from the demos · info
Pistol rounds
pistol rounds won / pistol rounds
demos
quality
measured from the demos · info
Force and eco conversion
rounds won / rounds bought below the full-buy threshold (pistols excluded); full-buy rounds shown for scale
demos
quality
measured from the demos · info
Buy agreement
rounds the five bought at the same level / rounds
demos, not counted yet
quality
needs demos of all five (not computed yet)
Adaptation after a loss
win rate in the round right after a lost round, next to the win rate after a won round
demos
quality
measured from the demos · info
Maps
Team rounds by mapn = 11 demos · 222 roundsparsed FACEIT demos
Map
Demos
Rounds
As T
As CT
Pistol rounds
Force + eco
Full buy
MIRAGE
4
80
39%17 / 44
58%21 / 36
5 / 8too few rounds
35%9 / 26
52%24 / 46
DUST2
3
63
75%27 / 36
44%12 / 27
3 / 6too few rounds
29%5 / 17
78%31 / 40
INFERNO
3
59
61%14 / 23
56%20 / 36
5 / 6too few rounds
47%8 / 17
58%21 / 36
ANUBISOnly 1 match
1
20
58%7 / 12
6 / 8too few rounds
2 / 2too few rounds
3 / 8too few rounds
80%8 / 10
All parsed maps
11
222
57%65 / 115
55%59 / 107
68%15 / 22
37%25 / 68
64%84 / 132
Note
Descriptive, not a verdict. Rounds the team won in the parsed FACEIT demos (matches in which at least 4 of the five played). These are team results: they do not change when a single player is picked in the View bar. Rounds of one match are not independent, so there is no interval and no "strong side" mark. Under 10 rounds a cell shows x / n without a percentage, and a map with fewer than 3 demos is marked. The last three columns are subsets of the rounds (rounds with unknown equipment are left out), so they do not add up to the total.Details
Telling a 5-point gap from chance would take about 150 matches.
Pistol rounds are rounds 1 and 13; force + eco are the rounds bought below the full-buy threshold.
Maps by playerpick a player above for one column
Map
Full five record
berkovskhy
_-Fresh
-michal
MaT1337
sqaar97
INFERNO
1 / 2 won2 matches · too few for a bar
54.6%
54.2%
50.0%
52.0%
53.3%▲
DUST2
1 / 1 won1 match · too few for a bar
53.0%
55.3%▲
51.1%
48.8%
48.3%
MIRAGE
1 / 1 won1 match · too few for a bar
55.2%
53.2%
48.9%
50.8%
49.7%
ANCIENT
—
53.6%
52.7%
48.8%
51.1%
—
ANUBIS
—
53.9%
51.9%
49.9%
48.7%
50.1%
NUKE
—
52.9%
53.6%
50.2%
51.9%
—
Note
How to read this. The record counts Full five matches only. Each player's number is his share of kills among kills and deaths on the map (K/(K+D), from the FACEIT per-map segments; maps with fewer than 50 matches are left out). ▲/▼ mark a map clearly off the same player's other maps; grey numbers have no reliable verdict, so the gaps between them are within chance, not a ranking of maps.Details
A mark is given after a correction for the number of maps tested (the bar rises with the number of maps compared).
berkovskhy
Sniper share of kills · per map
DUST2 and MIRAGE are his best sniper maps (30% and 28% of kills, against 19% on his other maps). INFERNO and ANCIENT are the weakest (10% and 12% of kills, against 24% and 22% on his other maps).
Details
▲DUST2 29.5% vs 18.8% on the other maps1 925 kills · z +4.9
▲MIRAGE 28.3% vs 18.5% on the other maps2 360 kills · z +4.8
NUKE 24.6% vs 20.2% on the other maps1 288 kills · z +1.7
ANUBIS 17.4% vs 20.9% on the other maps720 kills · z -1.1
▼ANCIENT 12.1% vs 22.2% on the other maps1 629 kills · z -4.3
▼INFERNO 10.3% vs 23.6% on the other maps2 366 kills · z -6.5
6 maps tested; marked ones pass the corrected bar (z 2.6). Without a mark: no reliable verdict (strongest z 6.5).
Kills vs deaths · by map
No map stands out in kills against deaths: the gaps are small enough for chance to explain them.
Details
— no map with a reliable verdict6 maps with 50+ matches tested; strongest z 2.0, bar 2.6
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
_-Fresh
Kills vs deaths · by map
DUST2 is his best map in duels: 55.3% of his kills and deaths there are kills, against 53.1% on his other maps. No map is clearly weaker than the rest.
Details
▲DUST2 55.3% vs 53.1% on the other maps5 661 kills+deaths · z +2.7 · bar 2.6
worst map: none reliable (lowest z -1.6)
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
sqaar97
Enemies flashed / round · per map
ANCIENT is his best flashing map (0.96 enemies flashed per round, against 0.67 on his other maps). NUKE and INFERNO are the weakest (0.41 and 0.53 enemies flashed per round, against 0.73 and 0.74 on his other maps).
Details
▲ANCIENT 0.96 vs 0.67 on the other maps1 061 rounds · z +6.0
ANUBIS 0.79 vs 0.68 on the other maps1 694 rounds · z +2.6
MIRAGE 0.75 vs 0.68 on the other maps2 754 rounds · z +1.9
DUST2 0.65 vs 0.71 on the other maps1 174 rounds · z -1.3
▼INFERNO 0.53 vs 0.74 on the other maps1 676 rounds · z -5.2
▼NUKE 0.41 vs 0.73 on the other maps800 rounds · z -5.8
6 maps tested; marked ones pass the corrected bar (z 2.6). Without a mark: no reliable verdict (strongest z 6.0).
Kills vs deaths · by map
INFERNO is his best map in duels: 53.3% of his kills and deaths there are kills, against 49.4% on his other maps. No map is clearly weaker than the rest.
Details
▲INFERNO 53.3% vs 49.4% on the other maps2 490 kills+deaths · z +3.1 · bar 2.5
worst map: none reliable (lowest z -1.4)
Kill share = kills / (kills + deaths) on the map. z = how far the map lies from his other maps, in standard errors; bar = the z a map must reach to be named.
-michalNo reliable verdict yet
MaT1337No reliable verdict yet
Default positions
Default positionsconvention, not measured
convention, not measured
Player
Map
Player & role
MIRAGE
INFERNO
DUST2
ANUBIS
NUKE
berkovskhyAWP · best data fit · weakprovisional · weak fit
CTWINDOW
TTOP MID
CTARCH
TMID
CTMID DOORS
TLONG DOORS
CTHEAVEN
TMID
CTHEAVEN
TOUTSIDE
_-FreshSecond entry · best data fit · weakprovisional · weak fit
CTB SITE
TPALACE
CTB SITE
TB SITE
CTPIT
TLONG
CTCANAL
TWALKWAY
CTHUT
THUT
-michalLurk · best data fit · low
CTMID
TAPPS
CTSECOND MID
TAPPS
CTTOP MID
TUPPER TUNNELS
CTWALKWAY
TSTREET
CTSECRET
TSECRET
MaT1337Entry · best data fit · lowWho really takes the first contact on T?
CTAPPS
TPALACE ALLEY
CTBANANA
TBANANA
CTLONG
TOUTSIDE LONG
CTMID
TMAIN
CTOUTSIDE
TRAMP
sqaar97Support · best data fit · high
CTJUNGLE
TRAMP
CTAPPS
TBALCONY
CTB DOORS
TCATWALK
CTA SITE
TCANAL
CTA SITE
TLOBBY
no layout ANCIENT · OVERPASS · TRAIN · CACHE. The callout names have to be seen in a demo of the map first, otherwise a spot could be written down that the game calls something else.
Layout of the selected mapconvention
T defaultMIRAGE · convention
TOP MIDAWP · mid
berkovskhy
PALACE ALLEYEntry · A route
MaT1337
PALACESecond entry (trade) · A route
_-Fresh
RAMPSupport · A route
sqaar97
APPSLurk (flank) · B route
-michal
MIDControl spot · mid
open
CATWALKFlex · mid
open
CT setupMIRAGE · convention
WINDOWAWP · mid
berkovskhy
APPSForward · B route
MaT1337
B SITETrade behind the forward · B route
_-Fresh
JUNGLESupport · A route
sqaar97
MIDDelay (late rotation) · mid
-michal
A SITEAnchor · A route
open
CONNECTORRotator
open
T defaultINFERNO · convention
MIDAWP · mid
berkovskhy
BANANAEntry · B route
MaT1337
B SITESecond entry (trade) · B route
_-Fresh
BALCONYSupport · A route
sqaar97
APPSLurk (flank) · A route
-michal
LOWER MIDControl spot · mid
open
TOP MIDFlex · mid
open
CT setupINFERNO · convention
ARCHAWP · A route
berkovskhy
BANANAForward · B route
MaT1337
B SITETrade behind the forward · B route
_-Fresh
APPSSupport · A route
sqaar97
SECOND MIDDelay (late rotation) · mid
-michal
A SITEAnchor · A route
open
LIBRARYRotator · A route
open
T defaultDUST2 · convention
LONG DOORSAWP · A route
berkovskhy
OUTSIDE LONGEntry · A route
MaT1337
LONGSecond entry (trade) · A route
_-Fresh
CATWALKSupport · short
sqaar97
UPPER TUNNELSLurk (flank) · B route
-michal
TOP MIDControl spot · mid
open
OUTSIDE TUNNELSFlex · B route
open
CT setupDUST2 · convention
MID DOORSAWP · mid
berkovskhy
LONGForward · A route
MaT1337
PITTrade behind the forward · A route
_-Fresh
B DOORSSupport · B route
sqaar97
TOP MIDDelay (late rotation) · mid
-michal
B SITEAnchor · B route
open
SHORT STAIRSRotator · short
open
T defaultANUBIS · convention
MIDAWP
berkovskhy
MAINEntry
MaT1337
WALKWAYSecond entry (trade)
_-Fresh
CANALSupport
sqaar97
STREETLurk (flank)
-michal
BRIDGEControl spot
open
CONNECTORFlex
open
CT setupANUBIS · convention
HEAVENAWP
berkovskhy
MIDForward
MaT1337
CANALTrade behind the forward
_-Fresh
A SITESupport
sqaar97
WALKWAYDelay (late rotation)
-michal
B SITEAnchor
open
CONNECTORRotator
open
T defaultNUKE · convention
OUTSIDEAWP
berkovskhy
RAMPEntry
MaT1337
HUTSecond entry (trade)
_-Fresh
LOBBYSupport
sqaar97
SECRETLurk (flank)
-michal
GARAGEControl spot
open
SILOFlex
open
CT setupNUKE · convention
HEAVENAWP
berkovskhy
OUTSIDEForward
MaT1337
HUTTrade behind the forward
_-Fresh
A SITESupport
sqaar97
SECRETDelay (late rotation)
-michal
B SITEAnchor
open
RAMPRotator
open
ANCIENT: no layout. There is no convention to derive a spot from, and none is guessed.
OVERPASS: no layout. There is no convention to derive a spot from, and none is guessed.
TRAIN: no layout. There is no convention to derive a spot from, and none is guessed.
CACHE: no layout. There is no convention to derive a spot from, and none is guessed.
Measured from the demosmeasured, not a convention
What happened in the parsed FACEIT matches. It describes the past; it is not where anyone should stand.
Where each player actually stoodn = 4 demosFACEIT demos · top callouts by share of alive time
T sidemeasured
berkovskhy4 matches · 34 min alive habit differs
T SPAWN 16%PALACE ALLEY 10%A SITE 10%
_-Fresh4 matches · 36 min alive
PALACE 20%T SPAWN 10%CT SPAWN 9%
-michal2 matches · 24 min alive Only 2 demoshabit differs
PALACE 19%A SITE 16%T SPAWN 11%
MaT13373 matches · 21 min alive habit differs
MID 16%TOP MID 13%T SPAWN 11%
sqaar974 matches · 33 min alive habit differs
T SPAWN 11%PALACE ALLEY 9%MID 9%
CT sidemeasured
berkovskhy4 matches · 28 min alive habit differs
CT SPAWN 32%A SITE 20%CONNECTOR 9%
_-Fresh4 matches · 25 min alive habit differs
CT SPAWN 32%A SITE 27%PALACE 9%
-michal2 matches · 13 min alive Only 2 demoshabit differs
B SITE 34%CT SPAWN 16%CATWALK 15%
MaT13373 matches · 29 min alive habit differs
CT SPAWN 20%B SITE 15%CATWALK 13%
sqaar974 matches · 29 min alive habit differs
CT SPAWN 23%B SITE 22%A SITE 11%
Where each player actually stoodn = 3 demosFACEIT demos · top callouts by share of alive time
T sidemeasured
berkovskhy3 matches · 25 min alive habit differs
BANANA 38%B SITE 18%SECOND MID 8%
_-Fresh3 matches · 20 min alive
BANANA 31%B SITE 13%MID 11%
-michal2 matches · 14 min alive Only 2 demos
APPS 25%BANANA 22%T SPAWN 9%
MaT13373 matches · 20 min alive
BANANA 23%APPS 11%MID 9%
sqaar973 matches · 22 min alive habit differs
BANANA 23%B SITE 13%T SPAWN 11%
CT sidemeasured
berkovskhy3 matches · 37 min alive habit differs
B SITE 47%BANANA 15%RUINS 13%
_-Fresh3 matches · 32 min alive habit differs
A SITE 24%APPS 20%TOP MID 10%
-michal2 matches · 20 min alive Only 2 demoshabit differs
B SITE 48%BANANA 21%CT SPAWN 14%
MaT13373 matches · 34 min alive habit differs
A SITE 31%CT SPAWN 13%APPS 10%
sqaar973 matches · 39 min alive habit differs
B SITE 20%TOP MID 17%CT SPAWN 12%
Where each player actually stoodn = 3 demosFACEIT demos · top callouts by share of alive time
T sidemeasured
berkovskhy3 matches · 32 min alive habit differs
TOP MID 17%OUTSIDE LONG 16%UPPER TUNNELS 8%
_-Fresh3 matches · 35 min alive habit differs
UPPER TUNNELS 17%OUTSIDE LONG 10%B SITE 8%
-michal3 matches · 34 min alive
TOP MID 18%UPPER TUNNELS 14%OUTSIDE LONG 9%
MaT13371 match · 9 min alive Only 1 demohabit differs
LONG DOORS 14%TOP MID 10%B SITE 10%
sqaar973 matches · 29 min alive habit differs
OUTSIDE LONG 13%UPPER TUNNELS 10%TOP MID 9%
CT sidemeasured
berkovskhy3 matches · 24 min alive
MID DOORS 27%B SITE 14%SHORT STAIRS 12%
_-Fresh3 matches · 23 min alive habit differs
UNDER A 26%B SITE 15%MID DOORS 13%
-michal3 matches · 21 min alive habit differs
A SITE 13%UNDER A 12%B SITE 12%
MaT13371 match · 6 min alive Only 1 demo
B SITE 31%LONG 13%LOWER TUNNELS 12%
sqaar973 matches · 21 min alive habit differs
B SITE 17%MID DOORS 12%LONG 10%
Where each player actually stoodn = 1 demo · earlyFACEIT demos · top callouts by share of alive time
T sidemeasured
berkovskhy1 match · 8 min alive Only 1 demohabit differs
OUTSIDE LONG 22%RUINS 15%MAIN 12%
_-Fresh1 match · 9 min alive Only 1 demohabit differs
RUINS 24%BRIDGE 13%OUTSIDE LONG 11%
MaT13371 match · 9 min alive Only 1 demohabit differs
STREET 16%OUTSIDE LONG 15%T SPAWN 12%
sqaar971 match · 8 min alive Only 1 demo
CANAL 19%OUTSIDE LONG 12%RUINS 10%
CT sidemeasured
berkovskhy1 match · 5 min alive Only 1 demohabit differs
MID 35%WALKWAY 11%MAIN 10%
_-Fresh1 match · 4 min alive Only 1 demo
MID 39%CANAL 11%BRIDGE 8%
MaT13371 match · 8 min alive Only 1 demohabit differs
A SITE 32%MAIN 24%WALKWAY 10%
sqaar971 match · 8 min alive Only 1 demohabit differs
ALLEY 29%PALACE 18%B SITE 17%
No parsed FACEIT demo of this map yet, so nothing is measured.
No parsed FACEIT demo of this map yet, so nothing is measured.
No parsed FACEIT demo of this map yet, so nothing is measured.
No parsed FACEIT demo of this map yet, so nothing is measured.
No parsed FACEIT demo of this map yet, so nothing is measured.
Note
How to read this.Convention, not measured. A player's spot comes only from his role: the captain's pick when there is one, otherwise his best data fit (see Home and Roles). Where each role stands on each map is written down by hand, and no pro-play data set is available; nothing is guessed from where he usually plays. The block at the bottom is measured: where the five actually stood in the parsed FACEIT demos.Details
With two players on one role, the better fit keeps the slot and the other takes the slot of his next best fit if it is free, otherwise an open one. When several players fit one role best, the captain decides. Anchor, lurk and flex mean something different on each side (see each map).
Training plan
Training plantargets are conventions, not measurements
targets are conventions, not measurements
Player
berkovskhyPlan role: AWP · best data fit · weak provisional: nothing separates this role from the next one for him, so the captain decides600 lifetime FACEIT matches with extended stats · 11 parsed demos
Week budget: about 30 min if every dose below is followed ([P] a proposal of this app, not a measurement; doses the source gives are counted as one session a week, a daily dose as five days).
1First kill of the round with the AWPbaselinemain stat of the roleparsed FACEIT demos n = 11 demos · 24 AWP rounds
The main stat of his role. There is nothing to compare it with yet (fewer than three players have enough data for it and no published reference was found), so the value is the baseline to watch.
Now29%
The other four togethernot comparable yetFewer than three players have enough data for this stat.
Targetno published reference found
ExerciseAWP peeks from spawn on Inferno (Mid, Boiler, Banana) with a planned way backDose: 15 min x 2 a week on one map ([P] proposal: the source gives no dose)Week budget: about 30 min ([P] a proposal, not a measurement)Source: BLAST (AWP guide)B · Dignitas (AWP tips)B · critfeed (CS2 roles)C
Check progressSlow measure (demos): First kill of the round with the AWP. Baseline 29% from 24 AWP rounds. Re-measure on at least 30 new AWP rounds that come after this plan, never on the baseline period. About ~14 demos to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
No data-driven gap. The other stats of this role are not below the other four players or the published convention, so none is listed as something to fix. Of the other stats of this role: 1 stat is at or above the other four and the published convention; 2 could not be compared yet (fewer than three players with enough data for them, and no published reference). Team focus below has the points of the whole team.
_-FreshPlan role: Second entry · best data fit · weak provisional: nothing separates this role from the next one for him, so the captain decides803 lifetime FACEIT matches with extended stats · 11 parsed demos
Week budget: about 30 min if every dose below is followed ([P] a proposal of this app, not a measurement; doses the source gives are counted as one session a week, a daily dose as five days).
1Trade kills per T roundmaintainmain stat of the roleparsed FACEIT demos n = 11 demos · 115 rounds
The main stat of his role, and he is above the other four players here: keep it where it is.
Now0.16
The other four together0.13he is 0.03 above them
Targetno published reference foundNo published figure for trade percentages was found; only the trade window differs between sources (HLTV 5 s, ours 3 s).
Exercise2v2 post-plant on Mirage A (the swing for the entry)Dose: 10 min x 3 sets (proposal)Week budget: about 30 min ([P] a proposal, not a measurement)Source: critfeed (CS2 roles)CThe source gives no dose.
Check progressSlow measure (demos): Trade kills per T round. Baseline 0.16 from 115 rounds. Re-measure on at least 100 new rounds that come after this plan, never on the baseline period. About ~10 demos to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
No data-driven gap. The other stats of this role are not below the other four players or the published convention, so none is listed as something to fix. Of the other stats of this role: 2 stats are at or above the other four and the published convention. Team focus below has the points of the whole team.
-michalPlan role: Lurk · best data fit · low883 lifetime FACEIT matches with extended stats · 7 parsed demos
Survival also moves with how fast the team wins a round, and a lurk is meant to be away from the others, so read it as a description of how he plays, not as a score.
Week budget: about 60 min if every dose below is followed ([P] a proposal of this app, not a measurement; doses the source gives are counted as one session a week, a daily dose as five days).
1T rounds survivedmaintainmain stat of the roleparsed FACEIT demos n = 7 demos · 78 rounds
The main stat of his role, and he is above the other four players here: keep it where it is.
Now35%
The other four together28%he is 7 pp above them
Targetno published reference foundHLTV's 20.6% counts survivors in LOST rounds only (CS:GO, 2023), a different measure from rounds survived overall.
ExerciseAnubis B Water: walk and pinch at 0:55 after a utility pack on A, recording the sessionsDose: 15 min x 2 a week, recorded ([P] proposal: the source gives no dose)Week budget: about 30 min ([P] a proposal, not a measurement)Source: critfeed (CS2 roles)C · BLAST (lurker guide)BThe source's example is Anubis B Water, and Anubis has few parsed demos. Practise the same timing on the maps with the most parsed demos: DUST2 (3 demos), INFERNO (2 demos).
Check progressSlow measure (demos): T rounds survived. Baseline 35% from 78 rounds. Re-measure on at least 100 new rounds that come after this plan, never on the baseline period. About ~9 demos to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
2KAST (rounds with a kill, assist, survival or traded death)gapparsed FACEIT demos n = 7 demos · 146 rounds
A real gap: he is below the other four and below the convention.
Now62%
The other four together74%he is 12 pp below them
Targetconvention, not a measurementpro median about 72% (the middle 80% of pro player-maps: about 58-83%)esportsodds.gg (KAST)B · 2026-07-10Pro players; the method is not disclosed. Lurks and supports often have the highest KAST.he is 10 pp below the target
ExerciseWatch demos of pro lurkersDose: 30 min a week: one demo of a pro lurker ([P] proposal: the source gives no dose)Week budget: about 30 min ([P] a proposal, not a measurement)Source: critfeed (CS2 roles)C
Check progressSlow measure (demos): KAST (rounds with a kill, assist, survival or traded death). Baseline 62% from 146 rounds. Re-measure on at least 100 new rounds that come after this plan, never on the baseline period. About ~5 demos to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
No data-driven gap. The other stats of this role are not below the other four players or the published convention, so none is listed as something to fix. Of the other stats of this role: 1 stat is at or above the other four and the published convention. Team focus below has the points of the whole team.
MaT1337Plan role: Entry · best data fit · low784 lifetime FACEIT matches with extended stats · 8 parsed demos
Week budget: about 150 min if every dose below is followed ([P] a proposal of this app, not a measurement; doses the source gives are counted as one session a week, a daily dose as five days).
1Entry success (opening duels won)gapmain stat of the roleFACEIT lifetime n = 784 matches · 3 076 entries
The main stat of his role, and he is below the other four and below the convention.
Now48.7%
The other four together52.3%he is 3.6 pp below them
Targetconvention, not a measurementpro entries typically win 52-58% of opening duelsrecoilanalytics.com (opening duels)C · 2026-03-17A blog without citations. The average over all players is about 50% by definition: every opening duel has exactly one winner.he is 3.3 pp below the target
ExercisePrefire practice on one map (Yprac)Dose: 15 min a day (from the source)Week budget: about 75 min ([P] a proposal, not a measurement)Source: VStreet (aim training routine)D
Check progressQuick measure (server stats; they arrive by themselves after every FACEIT match): Entry success from FACEIT match stats (Entry Wins / Entry Count). Baseline 47.2% from 36 entries in 8 FACEIT matches. Re-measure after 10 new FACEIT matches that come after this plan.
2Accuracy (hits per shot, shotguns excluded)gapparsed FACEIT demos n = 8 demos · 2 175 shots
A real gap: he is below the other four.
Now19%
The other four together20%he is 1 pp below them
Targetno published reference found
ExerciseAim Botz, moving targetsDose: 15 min (from the source)Week budget: about 75 min ([P] a proposal, not a measurement)Source: VStreet (aim training routine)D
Check progressSlow measure (demos): Accuracy (hits per shot, shotguns excluded). Baseline 19% from 2 175 shots. Re-measure on at least 90 new shots that come after this plan, never on the baseline period. About ~1 demo to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
No data-driven gap. The other stats of this role are not below the other four players or the published convention, so none is listed as something to fix. Of the other stats of this role: 2 stats are at or above the other four and the published convention; 1 could not be compared yet (fewer than three players with enough data for it, and no published reference). Team focus below has the points of the whole team.
sqaar97Plan role: Support · best data fit · high434 lifetime FACEIT matches with extended stats · 11 parsed demos
Week budget: about 50 min if every dose below is followed ([P] a proposal of this app, not a measurement; doses the source gives are counted as one session a week, a daily dose as five days).
1Enemy blind seconds per roundmaintainmain stat of the roleparsed FACEIT demos n = 11 demos · 222 rounds
The main stat of his role, and he is above the other four players here: keep it where it is.
Now2.32
The other four together1.96he is 0.36 above them
Targetconvention, not a measurementpros typically 1.5-2.0+ seconds of enemy blindness per roundHLTV (flashbang statistics)A · 2022-10-25Pro players; HLTV's unit may differ from the demo seconds counted here (rule 25).he is at or above the target
ExerciseFlash lineups per mapDose: at least 3 flashes per map (5-6 grenades per map as a beginner) (from the source)Week budget: about 30 min ([P] a proposal, not a measurement)Source: rankforge.gg (flashbang guide)D · csdb.gg (player roles)C
Check progressQuick measure (server stats; they arrive by themselves after every FACEIT match): Enemies flashed per round from FACEIT match stats (a count of enemies; the demo measure is seconds of blindness, so this is a stand-in, not the same number). Baseline 0.71 from 4 827 rounds in 216 FACEIT matches. Re-measure after 10 new FACEIT matches that come after this plan. It counts enemies, the demo counts seconds, so it is a stand-in for the direction, not the same number.Slow measure (demos): Enemy blind seconds per round. Baseline 2.32 from 222 rounds. Re-measure on at least 100 new rounds that come after this plan, never on the baseline period. About ~5 demos to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
2KAST (rounds with a kill, assist, survival or traded death)gapparsed FACEIT demos n = 11 demos · 222 rounds
A real gap: he is below the other four and below the convention.
Now67%
The other four together73%he is 6 pp below them
Targetconvention, not a measurementpro median about 72% (the middle 80% of pro player-maps: about 58-83%)esportsodds.gg (KAST)B · 2026-07-10Pro players; the method is not disclosed. Lurks and supports often have the highest KAST.he is 5 pp below the target
ExerciseReview your own deaths: was a teammate close enough to trade?Dose: 20 min a week, 2-3 repeating patterns (from the source)Week budget: about 20 min ([P] a proposal, not a measurement)Source: cs2.eu (demo review)C
Check progressSlow measure (demos): KAST (rounds with a kill, assist, survival or traded death). Baseline 67% from 222 rounds. Re-measure on at least 100 new rounds that come after this plan, never on the baseline period. About ~5 demos to download by hand from the match rooms (rule 28: FACEIT demos do not download themselves). Fewer than that proves nothing; a week is rarely enough.
No data-driven gap. The other stats of this role are not below the other four players or the published convention, so none is listed as something to fix. Of the other stats of this role: 3 stats are at or above the other four and the published convention. Team focus below has the points of the whole team.
Team focusn = 11 demos · 222 rounds
Ordered by the squad's OWN comparisons, not by distance from a published figure: first the points where its own results differ (the lowest p first), then the points that only describe (utility, retake and post-plant, buys, pistols), then comparisons that found no difference. Published base rates are only context next to a number: for one team a market average is not a goal. It describes; it does not judge.
Trades: not dying for free
The first duel is won or traded; a death without a trade is a 4v5 from the start of the round.
From the parsed demos
Team
Context: published base rate, not a goal
Own deaths traded by a teammate, T side
21% 76 / 356
no published reference found
Own deaths traded by a teammate, CT side
7% 23 / 322
no published reference found
Own deaths traded by a teammate, all five
15% 99 / 678
no published reference found
Own comparison: 14 pp below the other side · p = 0.000 (a difference that chance rarely produces)The weaker side is CT.
No published figure for trade percentages was found; only the trade window differs between sources (HLTV 5 s, ours 3 s).
ExerciseEntry plus trader pair: two players, different angles, the trader right behind the entryDose: 10 min x 3 sets a week ([P] proposal: the source gives no dose)Week budget: about 30 min ([P] a proposal, not a measurement)Source: critfeed (CS2 roles)C
Check progressSlow measure (demos): share of own deaths traded (3 s window). Re-measure on at least 30 new deaths after this plan, never on the period above. About ~1 demo to download by hand (rule 28). Fewer than that proves nothing.
The round after a lost round
A fixed protocol for the next round (plan A, an emergency exit) and a regular debrief.
From the parsed demos
Team
Context: published base rate, not a goal
Round won after a lost round
40% 37 / 93
no published reference found
Round won after a won round
67% 72 / 107
no published reference found
Own comparison: 28 pp below the round after a won round · p = 0.000 (a difference that chance rarely produces)
A lost round also means a weaker buy in the next one, so part of the gap is economy, not the reaction to the loss.
ExerciseDebrief after a scrim dayDose: about 30 min (from the source)Week budget: about 30 min ([P] a proposal, not a measurement)Source: skin.club (scrim culture)C
Check progressSlow measure (demos): round won after a lost round. Re-measure on at least 100 new rounds after a lost round after this plan, never on the period above. About ~12 demos to download by hand (rule 28). Fewer than that proves nothing.
The weakest map and side
Find out why one map-side loses more than the rest of that side: go through the matches that carry it before changing the setup.
From the parsed demos
Team
Context: published base rate, not a goal
Mirage T rounds won
39% 17 / 44
no published reference found
All other maps, T rounds won
68% 48 / 71
context T wins about 48.7% of rounds overall (CT about 51.3%) esportsodds.gg (map win rates)B About 2 million pro rounds from the CS and CS2 eras mixed together.
Own comparison: 29 pp below the rest of that side · p = 0.004 (a difference that chance rarely produces) Per match, the T rounds won there: 5/5, 1/12, 9/15, 2/12. Below the rest of the side in 3 of 4 matches, so look at the matches before the map.
7 map-sides were checked; with that many comparisons one result under 0.05 can be chance.
ExerciseScrims with a goal (test ideas, not the result), then a debriefDose: debrief 15-30 min (convention from the source)Week budget: about 30 min ([P] a proposal, not a measurement)Source: skin.club (scrim culture)C
Check progressSlow measure (demos): T rounds won on Mirage from parsed FACEIT demos. Re-measure on at least 100 new T rounds on that map after this plan, never on the period above. About ~10 demos to download by hand (rule 28). Fewer than that proves nothing.
Utility: flashes that blind your own team
Grenades thrown under a specific entry, not on spec; few teamflashes; no grenade dies in the inventory.
From the parsed demos
Team
Context: published base rate, not a goal
Teamflash, all five: teammate blind s per enemy blind s
0.50 999 / 1 995 enemy blind s
no published reference found
MaT1337: teammate blind s per enemy blind s
0.72 162 / 225 enemy blind s
no published reference found
_-Fresh: teammate blind s per enemy blind s
0.59 216 / 362 enemy blind s
no published reference found
berkovskhy: teammate blind s per enemy blind s
0.44 334 / 751 enemy blind s
no published reference found
-michal: teammate blind s per enemy blind s
0.44 63 / 142 enemy blind s
no published reference found
sqaar97: teammate blind s per enemy blind s
0.44 225 / 514 enemy blind s
no published reference found
Enemy blind seconds per player-round, all five
2.04 1 995 / 977 player-rounds
context pros typically 1.5-2.0+ seconds of enemy blindness per round HLTV (flashbang statistics)A Pro players; HLTV's unit may differ from the demo seconds counted here (rule 25).
Seconds, not events: there is no test, only the ratio. Lower is better here. Comparing players with each other says who throws more flashes, not who throws them worse.
ExercisePractice commands: sv_grenade_trajectory_prac_pipreview 1, sv_rethrow_last_grenade, bot_placeDose: 15 min a week in a practice server ([P] proposal: the source gives no dose)Week budget: about 15 min ([P] a proposal, not a measurement)Source: Leetify (practice commands)A
Check progressSlow measure (demos): teammate blind seconds per enemy blind second. Re-measure on at least 100 new rounds after this plan, never on the period above. About ~6 demos to download by hand (rule 28). Fewer than that proves nothing.
Retake and post-plant
After a plant the T side closes the round; the CT side has a retake set with utility and a crossfire.
context pro retakes win about 16-22% depending on the site HLTV (which bomb sites are hardest to hold)A Pro teams with over 3250 in equipment; the hardest are Mirage B and Dust2 A.
A round with the bomb down is a round of its own kind: post-plant is mostly won and retake mostly lost, so they are not set against each other or against the overall win rate.
Exercise2v2 post-plant on Mirage A (the swing for the entry)Dose: 10 min x 3 sets (proposal)Week budget: about 30 min ([P] a proposal, not a measurement)Source: critfeed (CS2 roles)CThe source gives no dose.
Check progressSlow measure (demos): post-plant and retake win rate from parsed FACEIT demos. Re-measure on at least 100 new rounds with the bomb down after this plan, never on the period above. About ~9 demos to download by hand (rule 28). Fewer than that proves nothing.
How to read this. A player's plan follows his plan role: the captain's pick when there is one, otherwise his best data fit (see Home and Roles). The first area is always the main stat of that role: maintain when he is above the other four players, gap when below, baseline when there is nothing to compare it with. The next areas are the real gaps: stats of the role where he is below the other four and below the published convention. When there is no such gap the plan says no data-driven gap instead of listing a strength as a task; a stat without enough data shows what to collect. A target is a convention, not a measurement: it comes from a public text listed in CELE-DRUZYNY.md, carries its source and a source quality from A (an analysis with a described sample) to D (a seller's page), and is not a norm for FACEIT. Where no figure was found the plan says so and gives no number. An exercise's dose is either from the source or a [P] proposal where the text gives none, and the week budget is a proposal too. The app cannot know whether an exercise was done, so progress is a new measurement of the stat on data that comes after the plan, in a period that does not overlap the baseline: a quick measure from FACEIT match stats where one exists, and a slow measure from demos that needs about five demos downloaded by hand. No verdict is given from less than the stated sample, and one week is rarely enough. There is no overall rating of a player.
How to read
How to readthe numbers on this page come from the same constants the app uses
Terms
Best fit
The role whose signals stand out most for a player when he is compared with the other four. A comparison inside this squad, not a grade and not an assignment.
Captain's pick
A role chosen by the captain (role_kapitana in config.json). It wins wherever a role matters, and the data fit stays visible next to it.
Confidence: high, medium, low, weak
How far to lean on a best fit. It comes from the sample and from the gap to the next role, never from the size of the score: a gap under 0.30 (in standard deviations of the squad) is weak, under 0.60 low, under 1.00 medium, above that high. A small sample lowers it further (see the minimums).
Conflict
Two or more players have the same best role. The app shows it with each player's alternative and leaves the decision to the captain.
Full five
FACEIT matches in which all 5 of the squad played on one side. At least 4 of 5 also counts the matches in which one of them was missing.
Early
A sample under the minimum for its kind (see the table). The number is still shown, with its n; a verdict is not.
x / n
x out of n: a count shown instead of a percentage when n is under its minimum (wins out of matches, hits out of shots). Never a fraction of something else.
A dash (—)
No data. It is not zero: nothing was measured, or the source does not carry it.
Style, not quality
Stats that say what a player does, not how well. More of it is a different way of playing, not a better one.
Convention, not a measurement
A number or a position taken from a public text or written down by hand, not measured from this squad's games. Targets on Plan and layouts on Defaults are conventions.
z and bar
On a map row, z says how far the map lies from the same player's other maps, in standard errors (0 = no difference). The bar is the z a map has to reach to count; it rises with the number of maps tested, so that testing many maps does not give a mark by chance.
Note and Details
A Note is a caveat that stays visible. Details folds the method behind a sentence; the caveat itself is never inside it.
Where the data comes from
FACEIT Open API. Lifetime counters of each player's account and the per-match stats of the matches the five played together. Read by python tracker.py with a FACEIT key; the app only reads, it never writes to FACEIT.
FACEIT demos. Match files downloaded by hand (FACEIT blocks automatic download) and parsed on this computer. Every parsed value is stored in the local database, so the file itself can be deleted. Events, shots, positions and round results come from here, and only from the matches that have a demo.
Conventions. Role signals, position layouts, training targets and the minimums below are chosen in advance or taken from public texts (CELE-DRUZYNY.md lists them with a source quality from A to D). Where they appear they are marked convention, not a measurement.
Not used. No Leetify, Steam or Valve-match data appears on any screen, and nothing about one owner: every screen is about the squad or one of its players. The page is one local file and loads nothing from the network.
Minimum samples
What is counted
Minimum
Below it
Full five matches (a verdict on the squad)
6 matches
the sample is marked early and the squad gets no verdict
Full five matches (record without a percentage)
10 matches
only the W-L record is shown, no percentage
Full five matches on one map (bar on Maps)
3 matches
the map shows x / n won and no bar
Lifetime matches of an account (extended stats)
20 matches
the sample is marked early
Matches of a player on one map (K/(K+D) on a card)
50 matches
the map is left out and gets no verdict
Parsed demos on one map
3 demos
the numbers describe, they are not a pattern, and the sample is marked early
Events behind a percentage
10 events
x / n (x of n) is shown instead of a percentage
Shots behind an accuracy
30 shots
hits / shots is shown instead of a percentage
Rounds in one situation
12 rounds
the sample is marked early
Rounds behind a demo signal of a role (Anchor: 600 seconds on CT)
20 rounds
the signal is left out of the role fit
Events behind a demo signal of a role
30 events
role confidence is at most low
Demos of a player (role confidence)
6 demos
role confidence is at most low; below 10 demos it is at most medium
Note
Every minimum is a convention chosen in advance, not a measured gap in the data. A sample under its minimum is marked early and shows counts with their n, never a verdict.Details
Where a small sample changes a word and not only a number: a role fit drops to low or medium confidence, and a squad with fewer Full five matches than the minimum gets no verdict.
What this app does not say
No overall rating or ranking of a player. A bigger number usually means more rounds played, not better play.
No verdict from a small sample. Under the minimums above it shows counts with their n.
No professional norms. There is no source of pro-level data for this app, so a target is a convention from a public text and is marked as one.
No role assignment. A best fit is a comparison inside the squad. IGL cannot be measured and appears only as the captain's pick.
No habits behind a position. Positions on Defaults follow the role (a convention); where players actually stood in the demos is shown separately, with its sample.
No causes. A number that goes together with winning (traded deaths, for example) is not shown as the reason for winning.
No time axis for lifetime stats. They are one averaged point over years of FACEIT matches; lobbies differ by ELO, so numbers between players are not on the same difficulty.
No analysis for a match without a demo file. Such a match carries only what FACEIT's server stats give.
Data status
Data statuscounts only · the key itself is never shown
NextDownload more FACEIT demos for -michal: 7 of 10 parsed demos. python tracker.py --faceit-linki lists the match rooms without a demo.FACEIT blocks automatic download, so demos are fetched by hand. Under 10 demos a player's role confidence is held back by the sample.
Squad5 of 5 playerscomplete
FACEIT keypresentokOnly whether it exists is shown, never the key.
Database viewread-only · values and samples, no ratings
Values
127 fields
Raw. Sums straight from the tables of the parsed FACEIT demos, one number per player, with its sample (n) under it. A dash is no data, not zero. Nothing here is rated or ranked: a bigger number usually means more rounds played, not better play.Derived. The indicators the other screens use, computed by the app from these tables. Below the threshold of events a percentage is replaced by x / n; some of them describe style, not quality. Still no rating and no ranking. From 11 parsed FACEIT demos in the database; the five are in the order of Home.
Field
Source
berkovskhy11 demos
_-Fresh11 demos
-michal7 demos
MaT13378 demos
sqaar9711 demos
Demosn = rounds in them
Roundsgracz_demko.plik
11n 201
11n 208
7n 136
8n 153
11n 192
Rounds on CTn = demos
Roundsgracz_demko.rundy_ct
101n 11
98n 11
62n 7
77n 8
88n 11
Rounds on Tn = demos
Roundsgracz_demko.rundy_t
100n 11
110n 11
74n 7
76n 8
104n 11
Rounds recordedn = demos
Roundsgracz_runda.runda
222n 11
222n 11
146n 7
165n 8
222n 11
Rounds with round countersn = rounds recordedA round known only from the equipment has no counters; it is left out of every rate.
Roundsgracz_runda.zgony
222n 222
222n 222
146n 146
165n 165
222n 222
Rounds wonn = rounds with a result
Roundsgracz_runda.wynik
124n 222
124n 222
79n 146
95n 165
124n 222
Rounds won as Tn = T rounds with a result
Roundsgracz_runda.wynik, gracz_runda.strona
65n 115
65n 115
45n 78
46n 79
65n 115
Rounds won as CTn = CT rounds with a result
Roundsgracz_runda.wynik, gracz_runda.strona
59n 107
59n 107
34n 68
49n 86
59n 107
Pistol roundsn = rounds recorded
Roundsgracz_runda.runda
22n 222
22n 222
14n 146
16n 165
22n 222
Pistol rounds wonn = pistol rounds with a result
Roundsgracz_runda.runda, gracz_runda.wynik
15n 22
15n 22
9n 14
14n 16
15n 22
First contact was hisn = rounds with a first contact
Roundsgracz_runda.kontakt_moj
32n 222
49n 222
17n 146
31n 165
49n 222
First kill went to his teamn = rounds with a first kill
Roundsgracz_runda.otwarcie_moja
119n 222
119n 222
75n 146
95n 165
119n 222
Rounds with the bomb plantedn = rounds with the bomb status knownEmpty when the demo was read by an older parser.
Roundsgracz_runda.plant_ktos
129n 222
129n 222
91n 146
96n 165
129n 222
Deathsn = rounds with counters
Roundsgracz_runda.zgony
151n 222
157n 222
104n 146
108n 165
158n 222
Deaths tradedn = rounds with counters
Roundsgracz_runda.zgony_odtradowane
24n 222
21n 222
15n 146
19n 165
20n 222
Trade killsn = rounds with counters
Roundsgracz_runda.trade_kille
23n 222
25n 222
13n 146
18n 165
16n 222
Clutch situations (1vX)n = rounds with countersEmpty clutch field means there was no such situation, not a lost one.
Roundsgracz_runda.clutch_przeciw
25n 222
25n 222
16n 146
14n 165
26n 222
Clutches wonn = clutches with a result
Roundsgracz_runda.clutch_wygrany
5n 25
6n 25
4n 16
4n 14
4n 26
Damage dealtn = rounds with counters
Roundsgracz_runda.obrazenia
20 250n 222
22 751n 222
11 219n 146
14 463n 165
12 294n 222
Bombs plantedn = rounds with counters
Roundsgracz_runda.plant
22n 222
15n 222
7n 146
6n 165
12n 222
Bombs defusedn = rounds with counters
Roundsgracz_runda.defuse
4n 222
4n 222
4n 146
3n 165
5n 222
Enemy blind secondsn = rounds with counters
Roundsgracz_runda.flash_wrog
751.3n 222
362.4n 222
142.1n 146
224.6n 165
514.2n 222
Teammate blind secondsn = rounds with counters
Roundsgracz_runda.flash_kolega
333.7n 222
215.6n 222
62.6n 146
162.3n 165
225.2n 222
Kills soon after his flashn = rounds with countersKills by teammates shortly after his flash; the parser's definition, not the game's assist credit.
Roundsgracz_runda.flash_zabojstwa
66n 222
46n 222
13n 146
19n 165
57n 222
Team equipment value (sum of rounds)n = rounds with a value
Roundsgracz_runda.sprzet_druzyny
4 290 200n 222
4 290 200n 222
2 839 300n 146
3 219 950n 165
4 290 200n 222
Enemy equipment value (sum of rounds)n = rounds with a value
Roundsgracz_runda.sprzet_wrog
4 260 500n 222
4 260 500n 222
2 874 200n 146
3 126 550n 165
4 260 500n 222
Own equipment value (sum of rounds)n = rounds with a value
Roundsgracz_runda.sprzet_gracza
932 500n 222
866 800n 222
549 750n 146
658 700n 165
798 600n 222
Own bank at round start (sum of rounds)n = rounds with a value
Roundsgracz_runda.konto
383 600n 222
393 200n 222
272 750n 146
222 000n 165
399 050n 222
Killsn = demos
Eventsgracz_zdarzenie.rodzaj
180n 11
209n 11
98n 7
144n 8
113n 11
Deaths (to an enemy)n = demos
Eventsgracz_zdarzenie.rodzaj
151n 11
156n 11
104n 7
107n 8
157n 11
Assistsn = demos
Eventsgracz_zdarzenie.rodzaj
79n 11
63n 11
37n 7
36n 8
52n 11
Kills with a time in the roundn = kills
Eventsgracz_zdarzenie.t_s
180n 180
209n 209
98n 98
144n 144
113n 113
Kills from 60 s into the roundn = kills with a time
Eventsgracz_zdarzenie.t_s
41n 180
37n 209
27n 98
34n 144
33n 113
Deaths with a time in the roundn = deaths
Eventsgracz_zdarzenie.t_s
151n 151
156n 156
104n 104
107n 107
157n 157
Deaths before 30 sn = deaths with a time
Eventsgracz_zdarzenie.t_s
54n 151
65n 156
39n 104
54n 107
61n 157
Opening killsn = kills
Eventsgracz_zdarzenie.otwarcie
21n 180
37n 209
12n 98
17n 144
17n 113
Opening deathsn = deaths
Eventsgracz_zdarzenie.otwarcie
17n 151
20n 156
11n 104
19n 107
18n 157
Kills with the AWPn = kills
Eventsgracz_zdarzenie.bron
43n 180
21n 209
0n 98
2n 144
2n 113
Deaths with the AWP in handn = deathsWhat he held when he died, not what killed him.
Eventsgracz_zdarzenie.bron_ofiary
17n 151
4n 156
0n 104
1n 107
4n 157
Headshot killsn = kills with the flag recorded
Eventsgracz_zdarzenie.hs
72n 180
84n 209
54n 98
84n 144
55n 113
Kills through smoken = kills with the flag recorded
Eventsgracz_zdarzenie.przez_dym
13n 180
14n 209
5n 98
16n 144
10n 113
Wallbang killsn = kills with the flag recorded
Eventsgracz_zdarzenie.przebicie
6n 180
6n 209
3n 98
6n 144
5n 113
No-scope AWP killsn = AWP kills with the flag recorded
Eventsgracz_zdarzenie.bez_celownika
0n 43
0n 21
—
0n 2
0n 2
Kills while blindedn = kills with the flag recorded
Eventsgracz_zdarzenie.zabojca_slepy
1n 180
1n 209
1n 98
1n 144
1n 113
Deaths to a blinded enemyn = deaths with the flag recorded
Eventsgracz_zdarzenie.zabojca_slepy
0n 151
3n 156
1n 104
2n 107
0n 157
Kills while airbornen = kills with the flag recorded
Eventsgracz_zdarzenie.zabojca_w_powietrzu
1n 180
12n 209
1n 98
1n 144
1n 113
Deaths to an airborne enemyn = deaths with the flag recorded
Eventsgracz_zdarzenie.zabojca_w_powietrzu
1n 151
1n 156
1n 104
0n 107
0n 157
Assists by a flashn = assists with the flag recorded
Eventsgracz_zdarzenie.flash_asysta
18n 79
10n 63
2n 37
2n 36
9n 52
Kills not against an enemyn = demosTeamkills and the like; left out of every other number here.
Eventsgracz_zdarzenie.wrog
0n 11
0n 11
0n 7
0n 8
1n 11
Deaths not to an enemyn = demosThe world, a suicide or a teamkill; left out of every other number here.
Eventsgracz_zdarzenie.wrog
0n 11
1n 11
0n 7
1n 8
1n 11
Shots firedn = rounds with shotsNo knife, grenades, bomb or taser.
Shotsgracz_strzaly.strzaly
2 882n 208
2 806n 202
1 544n 129
2 181n 146
2 329n 177
Hitsn = shotsA hit is a damage event on an enemy; a shotgun can give more hits than shots.
Shotsgracz_strzaly.trafienia
574n 2 882
736n 2 806
309n 1 544
426n 2 181
400n 2 329
Hits to the headn = hits
Shotsgracz_strzaly.tr_glowa
93n 574
118n 736
71n 309
105n 426
78n 400
Hits to the neckn = hits
Shotsgracz_strzaly.tr_szyja
3n 574
4n 736
3n 309
5n 426
1n 400
Hits to the chestn = hits
Shotsgracz_strzaly.tr_klatka
274n 574
339n 736
122n 309
183n 426
175n 400
Hits to the stomachn = hits
Shotsgracz_strzaly.tr_brzuch
110n 574
146n 736
48n 309
59n 426
68n 400
Hits to the armsn = hits
Shotsgracz_strzaly.tr_rece
65n 574
81n 736
51n 309
51n 426
48n 400
Hits to the legsn = hits
Shotsgracz_strzaly.tr_nogi
29n 574
48n 736
14n 309
23n 426
30n 400
Hits elsewheren = hits
Shotsgracz_strzaly.tr_inne
0n 574
0n 736
0n 309
0n 426
0n 400
AWP shotsn = shots
Shotsgracz_strzaly.strzaly, gracz_strzaly.bron
102n 2 882
49n 2 806
0n 1 544
6n 2 181
15n 2 329
AWP hitsn = AWP shots
Shotsgracz_strzaly.trafienia, gracz_strzaly.bron
47n 102
23n 49
—
2n 6
2n 15
Kills by weapon tablen = rounds with a weapon rowOlder counter: it also counts a kill of the wrong team, so it can differ from Events by a few.
Weaponsgracz_bron.zabojstwa
180n 213
210n 214
98n 140
144n 155
114n 199
Deaths by weapon tablen = rounds with a weapon row
Weaponsgracz_bron.smierci
151n 213
157n 214
104n 140
108n 155
158n 199
Damage by weapon tablen = rounds with a weapon row
Weaponsgracz_bron.obrazenia
20 250n 213
22 751n 214
11 219n 140
14 463n 155
12 294n 199
Weapons usedn = demos
Weaponsgracz_bron.bron
23n 11
21n 11
19n 7
22n 8
21n 11
AWP killsn = kills
Weaponsgracz_bron.zabojstwa, gracz_bron.bron
43n 180
21n 210
0n 98
2n 144
2n 114
Deaths to the AWPn = deaths
Weaponsgracz_bron.smierci, gracz_bron.bron
13n 151
8n 157
10n 104
7n 108
18n 158
AWP damagen = damage
Weaponsgracz_bron.obrazenia, gracz_bron.bron
3 792n 20 250
1 789n 22 751
0n 11 219
200n 14 463
195n 12 294
Seconds alive at a calloutn = rounds with a position
Positionsgracz_miejsce.sekundy
11 590n 222
11 055n 222
7 531n 146
8 142n 165
11 408n 222
Kills at calloutsn = rounds with a position
Positionsgracz_miejsce.zabojstwa
180n 222
210n 222
98n 146
144n 165
114n 222
Deaths at calloutsn = rounds with a position
Positionsgracz_miejsce.smierci
151n 222
157n 222
104n 146
108n 165
158n 222
AWP kills at calloutsn = rounds with a position
Positionsgracz_miejsce.awp_zab
43n 222
21n 222
0n 146
2n 165
2n 222
AWP deaths at calloutsn = rounds with a position
Positionsgracz_miejsce.awp_sm
13n 222
8n 222
10n 146
7n 165
18n 222
Opening duels won at calloutsn = rounds with a position
Positionsgracz_miejsce.otwarcia_w
21n 222
37n 222
12n 146
17n 165
17n 222
Opening duels lost at calloutsn = rounds with a position
Positionsgracz_miejsce.otwarcia_p
17n 222
20n 222
11n 146
19n 165
18n 222
Different calloutsn = rounds with a position
Positionsgracz_miejsce.miejsce
53n 222
58n 222
49n 146
61n 165
59n 222
Heatmap cells visitedn = demosOne cell is 64 game units; the same cell on two maps counts twice.
Positionsgracz_cieplna.gx, gracz_cieplna.gy
10 920n 11
11 698n 11
8 115n 7
8 719n 8
12 518n 11
Heatmap ticksn = demos
Positionsgracz_cieplna.ticki
741 776n 11
707 534n 11
481 976n 7
521 068n 8
730 092n 11
Partner-roundsn = rounds with a partner
Duetsgracz_duet.partner
888n 222
888n 222
584n 146
660n 165
888n 222
Rounds with a partner aliven = demos
Duetsgracz_duet.runda
222n 11
222n 11
146n 7
165n 8
222n 11
Different partnersn = demos
Duetsgracz_duet.partner
11n 11
11n 11
7n 7
8n 8
11n 11
Ticks togethern = partner-rounds
Duetsgracz_duet.ticki_razem
3 719 514n 888
3 638 406n 888
2 376 068n 584
2 716 029n 660
3 634 021n 888
Own deaths in those roundsn = rounds with a partnerCounted once per round, not once per partner.
Duetsgracz_duet.moje_zgony
151n 222
157n 222
104n 146
108n 165
158n 222
Own deaths traded by that partnern = partner-rounds
Duetsgracz_duet.on_nadrobil
24n 888
21n 888
14n 584
19n 660
20n 888
Partner deathsn = partner-rounds
Duetsgracz_duet.jego_zgony
626n 888
620n 888
405n 584
453n 660
619n 888
Partner deaths traded by himn = partner-rounds
Duetsgracz_duet.ja_nadrobilem
26n 888
27n 888
13n 584
19n 660
19n 888
1v1 Win Raten = matches with extended stats
FACEIT lifetimefaceit_konto
0.36n 600
0.43n 803
0.36n 883
0.41n 784
0.38n 434
1v2 Win Raten = matches with extended stats
FACEIT lifetimefaceit_konto
0.22n 600
0.22n 803
0.2n 883
0.22n 784
0.29n 434
ADRn = matches with extended stats
FACEIT lifetimefaceit_konto
89.34n 600
85.11n 803
79.01n 883
77.36n 784
76.52n 434
Average Headshots %n = matchesAn average of per-match averages, not headshots over kills.
FACEIT lifetimefaceit_konto
39n 2 224
44n 2 391
48n 2 381
50n 3 355
41n 1 101
Average K/D Ration = matchesAn average of per-match averages, not kills over deaths.
FACEIT lifetimefaceit_konto
1.17n 2 224
1.24n 2 391
1.05n 2 381
1.07n 3 355
1.07n 1 101
Current Win Streakn = matches
FACEIT lifetimefaceit_konto
0n 2 224
2n 2 391
0n 2 381
1n 3 355
0n 1 101
Enemies Flashed per Roundn = matches with extended stats
FACEIT lifetimefaceit_konto
0.52n 600
0.43n 803
0.25n 883
0.28n 784
0.7n 434
Entry Raten = matches with extended stats
FACEIT lifetimefaceit_konto
0.22n 600
0.2n 803
0.19n 883
0.18n 784
0.18n 434
Entry Success Raten = matches with extended stats
FACEIT lifetimefaceit_konto
0.58n 600
0.55n 803
0.48n 883
0.49n 784
0.48n 434
Flash Success Raten = matches with extended stats
FACEIT lifetimefaceit_konto
0.54n 600
0.44n 803
0.54n 883
0.49n 784
0.51n 434
Flashes per Roundn = matches with extended stats
FACEIT lifetimefaceit_konto
0.63n 600
0.64n 803
0.3n 883
0.39n 784
0.91n 434
K/D Ration = matchesA sum of per-match ratios, not kills over deaths; divide by Matches for the mean.
FACEIT lifetimefaceit_konto
2 591.52n 2 224
2 955.28n 2 391
2 498.59n 2 381
3 602.32n 3 355
1 180.3n 1 101
Longest Win Streakn = matches
FACEIT lifetimefaceit_konto
11n 2 224
10n 2 391
10n 2 381
12n 3 355
8n 1 101
Matchesn = matchesAll matches of the account, with or without extended stats.
FACEIT lifetimefaceit_konto
2 224n 2 224
2 391n 2 391
2 381n 2 381
3 355n 3 355
1 101n 1 101
Recent Resultsn = matches
FACEIT lifetimefaceit_konto
0 1 1 0 0n 2 224
1 1 0 1 1n 2 391
0 0 1 1 0n 2 381
0 0 0 0 1n 3 355
1 1 0 1 0n 1 101
Sniper Kill Raten = matches with extended statsFACEIT's name is swapped: this is sniper kills per round.
FACEIT lifetimefaceit_konto
0.17n 600
0.11n 803
0.02n 883
0.04n 784
0.11n 434
Sniper Kill Rate per Roundn = matches with extended statsFACEIT's name is swapped: this is the share of kills.
FACEIT lifetimefaceit_konto
0.21n 600
0.14n 803
0.02n 883
0.05n 784
0.15n 434
Total 1v1 Countn = matches with extended stats
FACEIT lifetimefaceit_konto
385n 600
540n 803
523n 883
485n 784
304n 434
Total 1v1 Winsn = matches with extended stats
FACEIT lifetimefaceit_konto
140n 600
233n 803
187n 883
199n 784
116n 434
Total 1v2 Countn = matches with extended stats
FACEIT lifetimefaceit_konto
416n 600
685n 803
596n 883
562n 784
352n 434
Total 1v2 Winsn = matches with extended stats
FACEIT lifetimefaceit_konto
91n 600
150n 803
121n 883
124n 784
103n 434
Total Damagen = matches with extended stats
FACEIT lifetimefaceit_konto
1 181 673n 600
1 496 665n 803
1 546 527n 883
1 336 823n 784
733 938n 434
Total Enemies Flashedn = matches with extended stats
FACEIT lifetimefaceit_konto
6 906n 600
7 485n 803
4 935n 883
4 921n 784
6 734n 434
Total Entry Countn = matches with extended stats
FACEIT lifetimefaceit_konto
2 893n 600
3 449n 803
3 722n 883
3 076n 784
1 720n 434
Total Entry Winsn = matches with extended stats
FACEIT lifetimefaceit_konto
1 664n 600
1 894n 803
1 784n 883
1 499n 784
819n 434
Total Flash Countn = matches with extended stats
FACEIT lifetimefaceit_konto
8 305n 600
11 340n 803
5 857n 883
6 792n 784
8 772n 434
Total Flash Successesn = matches with extended stats
FACEIT lifetimefaceit_konto
4 498n 600
5 010n 803
3 172n 883
3 343n 784
4 501n 434
Total Headshots %n = matchesA sum of per-match percentages, not a count of headshots.
FACEIT lifetimefaceit_konto
86 710n 2 224
104 443n 2 391
114 620n 2 381
167 960n 3 355
45 016n 1 101
Total Kills with extended statsn = matches with extended stats
FACEIT lifetimefaceit_konto
10 680n 600
13 833n 803
13 861n 883
12 385n 784
6 621n 434
Total Matchesn = all matchesMatches that carry the extended stats: the sample of most fields below.
FACEIT lifetimefaceit_konto
600n 2 224
803n 2 391
883n 2 381
784n 3 355
434n 1 101
Total Rounds with extended statsn = matches with extended stats
FACEIT lifetimefaceit_konto
13 226n 600
17 585n 803
19 575n 883
17 280n 784
9 592n 434
Total Sniper Killsn = matches with extended stats
FACEIT lifetimefaceit_konto
2 210n 600
1 868n 803
326n 883
643n 784
1 011n 434
Total Utility Countn = matches with extended stats
FACEIT lifetimefaceit_konto
7 564n 600
7 418n 803
8 126n 883
6 317n 784
5 846n 434
Total Utility Damagen = matches with extended stats
FACEIT lifetimefaceit_konto
100 848n 600
88 478n 803
119 423n 883
101 819n 784
63 285n 434
Total Utility Successesn = matches with extended stats
FACEIT lifetimefaceit_konto
2 887n 600
2 614n 803
3 172n 883
2 804n 784
1 739n 434
Utility Damage Success Raten = matches with extended stats
FACEIT lifetimefaceit_konto
13.33n 600
11.93n 803
14.7n 883
16.12n 784
10.83n 434
Utility Damage per Roundn = matches with extended stats
FACEIT lifetimefaceit_konto
7.62n 600
5.03n 803
6.1n 883
5.89n 784
6.6n 434
Utility Success Raten = matches with extended stats
FACEIT lifetimefaceit_konto
0.38n 600
0.35n 803
0.39n 883
0.44n 784
0.3n 434
Utility Usage per Roundn = matches with extended stats