Team highlights
Compared with the 16 Champions teamsPlayers change agents
mean share of a player's maps spent on their main agent
39%
29 maps
field 56%
Larger agent pool
distinct agents fielded, resampled to a common exposure
18.4
29 maps
field 15.2
Changes its role mix
share of maps using their most common role mix: the number of duelists, initiators, controllers and sentinels
33%
29 maps
field 51%
Collision card
NS
H2H · 11 maps
T1
H2H · 4 maps
100T
none in Stage 2
EDG
none in Stage 2
FUT
none in Stage 2
G2
none in Stage 2
JDG
none in Stage 2
KC
none in Stage 2
LOUD
none in Stage 2
NRG
none in Stage 2
PRX
none in Stage 2
TYL
none in Stage 2
TL
none in Stage 2
VIT
none in Stage 2
XLG
none in Stage 2
clear
3 of the 13 fingerprint dimensions separate
GE from
NRG on this sample
full matchup →
Players change agents
structure
-14 pts
← Players change agentsPlayers keep one agent →
gap -22 to -5 · bar length is each team's position on that axis against the sixteen-team field; the hairline is the field mean
mean share of a player's maps spent on their main agent
Fewer assists per 100 kills
texture
-11.8
← Fewer assists per 100 killsMore assists per 100 kills →
gap -18.2 to -5.3 · bar length is each team's position on that axis against the sixteen-team field; the hairline is the field mean
assists per 100 kills (utility-dense play)
Larger agent pool
structure
+4.4
← Smaller agent poolLarger agent pool →
gap +2.9 to +6.0 · bar length is each team's position on that axis against the sixteen-team field; the hairline is the field mean
distinct agents fielded, resampled to a common exposure
All 13 dimensions, separated or not
| Dimension | GE | Sample | NRG | Sample | Gap | Estimate range |
|---|---|---|---|---|---|---|
| Opening kills concentrated | 32% | 29 maps | 32% | 22 maps | -0 | -8 to +7 pts no separation |
| More wins by defuse | 43% | 173 defence round wins | 40% | 125 defence round wins | +3 | -10 to +15 pts no separation |
| Higher headshot % | 28% | 29 maps | 26% | 22 maps | +2 | -0 to +4 pts no separation |
| More assists per 100 kills | 35.9 | 29 maps | 47.7 | 22 maps | -11.8 | -18.2 to -5.3 |
| Kills concentrated | 20.2 | 29 maps | 20.1 | 22 maps | +0.1 | -0.6 to +0.7 no separation |
| Top fragger takes more | 22% | 29 maps | 21% | 22 maps | +1 | -2 to +3 pts no separation |
| Players keep one role | 71% | 29 maps | 68% | 22 maps | +3 | -5 to +11 no separation |
| Players keep one agent | 39% | 29 maps | 53% | 22 maps | -14 | -22 to -5 |
| Repeats its role mix | 33% | 29 maps | 33% | 22 maps | +0 | -15 to +16 no separation |
| Uses unusual role mixes | 46% | 29 maps | 55% | 22 maps | -9 | -33 to +15 no separation |
| More distinct comps | 11.1 | 29 maps | 9.2 | 22 maps | +1.8 | -1.5 to +5.1 no separation |
| More role mixes | 4.9 | 29 maps | 5.0 | 22 maps | -0.1 | -0.8 to +0.7 no separation |
| Larger agent pool | 18.4 | 29 maps | 14.0 | 22 maps | +4.4 | +2.9 to +6.0 |
Results by game state
13 conditions · compared with the 16 Champions teams
won the first pistol (R1)
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
69% across 233 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Does the first pistol decide their map?
whole field does this
When they do
88%
16 maps
When they don't
38%
13 maps
True of the field, not a read on them. Their attack-round-1 setup is the lever. Deny the R1 pistol and the whole half re-prices: force the duel you are best at on round 1, and keep a save plan so their R2 bonus is contested.
Difference +50 pts
estimate range +14 to +72 pts
field 19 pts
What that means
Does it change their result?
Yes, +50 pts (+14 to +72 pts).
Is it a read on them?
No. The field lift is 19 pts and theirs sits +30 against it, so this is the game speaking about every team at once.
won the second pistol (R13)
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
74% across 210 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Does the second pistol decide their map?
When they do
69%
13 maps
When they don't
63%
16 maps
Difference +7 pts
estimate range -26 to +37 pts
field 28 pts
What that means
Does it change their result?
The +7 pts gap sits inside an estimate range of -26 to +37 pts: real or noise, this sample cannot say.
Is it a read on them?
No. Their rate with the condition is within noise of their rate without it. For scale, the field averages 74% with this condition.
won BOTH pistols
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
78% across 129 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
When they take both pistols, do they close it out?
When they do
7 of 8
maps
When they don't
57%
21 maps
Difference +31 pts
estimate range -9 to +53 pts
field 26 pts
What that means
Does it change their result?
The +31 pts gap sits inside an estimate range of -9 to +53 pts: real or noise, this sample cannot say.
Is it a read on them?
Not answerable yet: one of the two arms is 7 of 8 maps, too thin to carry a verdict.
won NEITHER pistol
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
36% across 101 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Can they win a map after losing both pistols?
whole field does this
When they do
3 of 8
maps
When they don't
76%
21 maps
True of the field, not a read on them. This is their resilience floor — how often they still win after losing both openers. A high number here means pistol prep alone will not beat them.
Difference -38 pts
estimate range -66 to -0 pts
field -33 pts
What that means
Does it change their result?
The -38 pts gap sits inside an estimate range of -66 to -0 pts: real or noise, this sample cannot say.
Is it a read on them?
No. The field lift is -33 pts and theirs sits -6 against it, so this is the game speaking about every team at once.
led at half-time
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
84% across 209 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Do they hold a half-time lead?
When they do
85%
13 maps
When they don't
50%
16 maps
Difference +35 pts
estimate range -0 to +59 pts
field 48 pts
What that means
Does it change their result?
The +35 pts gap sits inside an estimate range of -0 to +59 pts: real or noise, this sample cannot say.
Is it a read on them?
No. Their rate with the condition is within noise of their rate without it. For scale, the field averages 84% with this condition.
reached half-time tied 6-6
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
58% across 59 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Who wins the coin-flip maps?
When they do
6 of 7
maps
When they don't
59%
22 maps
Difference +27 pts
estimate range -14 to +50 pts
field -3 pts
What that means
Does it change their result?
The +27 pts gap sits inside an estimate range of -14 to +50 pts: real or noise, this sample cannot say.
Is it a read on them?
Not answerable yet: one of the two arms is 6 of 7 maps, too thin to carry a verdict.
trailed at half-time
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
28% across 147 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Can they come back from a half-time deficit?
whole field does this
When they do
2 of 9
maps
When they don't
85%
20 maps
True of the field, not a read on them. Their comeback rate. If it is low, a first-half lead is close to match point; if high, do not relax a 8-4.
Difference -63 pts
estimate range -81 to -24 pts
field -50 pts
What that means
Does it change their result?
The -63 pts gap sits inside an estimate range of -81 to -24 pts: real or noise, this sample cannot say.
Is it a read on them?
No. The field lift is -50 pts and theirs sits -12 against it, so this is the game speaking about every team at once.
playing their OWN map pick
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
63% across 175 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Is their own pick actually their advantage?
When they do
83%
12 maps
When they don't
53%
17 maps
Difference +30 pts
estimate range -5 to +55 pts
field 5 pts
What that means
Does it change their result?
The +30 pts gap sits inside an estimate range of -5 to +55 pts: real or noise, this sample cannot say.
Is it a read on them?
No. Their rate with the condition is within noise of their rate without it. For scale, the field averages 63% with this condition.
playing the OPPONENT'S pick
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
59% across 174 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
How do they travel on someone else's pick?
When they do
58%
12 maps
When they don't
71%
17 maps
Difference -13 pts
estimate range -43 to +20 pts
field -2 pts
What that means
Does it change their result?
The -13 pts gap sits inside an estimate range of -43 to +20 pts: real or noise, this sample cannot say.
Is it a read on them?
No. Their rate with the condition is within noise of their rate without it. For scale, the field averages 59% with this condition.
playing a decider (neither team picked it)
Denominator: maps. Both bars are observed frequencies over this team's
Stage 2 maps, with no model behind them.
The white hairline on each track is the field's rate for this condition,
56% across 66 Champions-team maps.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
How do they handle the map neither side chose?
When they do
2 of 5
maps
When they don't
71%
24 maps
Difference -31 pts
estimate range -62 to +11 pts
field -5 pts
What that means
Does it change their result?
The -31 pts gap sits inside an estimate range of -62 to +11 pts: real or noise, this sample cannot say.
Is it a read on them?
Not answerable yet: one of the two arms is 2 of 5 maps, too thin to carry a verdict.
won the previous round
Denominator: non-pistol regulation rounds. Both bars are observed frequencies over this team's
Stage 2 rounds, with no model behind them.
The white hairline on each track is the field's rate for this condition,
64% across 4149 Champions-team rounds.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Do they string rounds together?
whole field does this
When they do
61%
293 rounds
When they don't
45%
261 rounds
True of the field, not a read on them. Break the chain: their strongest round is the one after a win, so the highest-value round to spend utility on is the one you are about to lose anyway.
Difference +16 pts
estimate range +7 to +24 pts
field 22 pts
What that means
Does it change their result?
Yes, +16 pts (+7 to +24 pts).
Is it a read on them?
No. The field lift is 22 pts and theirs sits -6 against it, so this is the game speaking about every team at once.
lost the previous round
Denominator: non-pistol regulation rounds. Both bars are observed frequencies over this team's
Stage 2 rounds, with no model behind them.
The white hairline on each track is the field's rate for this condition,
42% across 3684 Champions-team rounds.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Do they answer a lost round?
whole field does this
When they do
45%
261 rounds
When they don't
61%
293 rounds
True of the field, not a read on them. Their reset speed. A low number means a won round is worth more than one round — press immediately.
Difference -16 pts
estimate range -24 to -7 pts
field -22 pts
What that means
Does it change their result?
Yes, -16 pts (-24 to -7 pts).
Is it a read on them?
No. The field lift is -22 pts and theirs sits +6 against it, so this is the game speaking about every team at once.
have lost 2+ rounds in a row
Denominator: non-pistol regulation rounds. Both bars are observed frequencies over this team's
Stage 2 rounds, with no model behind them.
The white hairline on each track is the field's rate for this condition,
42% across 2039 Champions-team rounds.
The pale band inside each bar is that rate's estimate range: where the true rate most likely sits, given this sample.
Can they stop a slide?
When they do
49%
138 rounds
When they don't
55%
416 rounds
Difference -6 pts
estimate range -15 to +4 pts
field -14 pts
What that means
Does it change their result?
The -6 pts gap sits inside an estimate range of -15 to +4 pts: real or noise, this sample cannot say.
Is it a read on them?
No. Their rate with the condition is within noise of their rate without it. For scale, the field averages 42% with this condition.
Fingerprint
GE compared with the 16-team averageSides
Loses attack rounds Wins attack rounds 52% 54% 307 attack rounds Loses defence rounds Wins defence rounds 55% 53% 317 defence rounds Loses rounds Wins rounds 53% 53% 624 rounds Defence-sided Attack-sided -2.5 pp 1.0 pp 624 roundsOpeners
Loses opening duels Wins opening duels 50% 51% 624 opening duels Opening kills spread Opening kills concentrated 32% 32% 29 maps Loses pistols Wins pistols 50% 53% 58 pistolsHow rounds end
Fewer wins by elimination More wins by elimination 68% 68% 333 round wins Fewer wins by detonation More wins by detonation 15% 19% 160 attack round wins Fewer wins by defuse More wins by defuse 43% 40% 173 defence round wins Fewer wins on the clock More wins on the clock 5% 6% 173 defence round wins Fewer losses to detonation More losses to detonation 19% 20% 144 defence round losses Converts fewer plants Converts more plants 31% 35% 77 planted rounds Wins fewer retakes Wins more retakes 73% 69% 102 retakesMomentum
Lower after a win Higher after a win 61% 63% 293 rounds after a win Lower after 2+ losses Higher after 2+ losses 49% 43% 138 rounds after 2+ lossesTexture
Lower KAST Higher KAST 72% 72% 29 maps Lower damage per round Higher damage per round 135 133 29 maps Lower headshot % Higher headshot % 28% 28% 29 maps Fewer assists per 100 kills More assists per 100 kills 35.9 39.6 29 maps Kills spread evenly Kills concentrated 20.2 20.1 29 maps Top fragger takes less Top fragger takes more 22% 23% 29 mapsStructure
Players change roles Players keep one role 71% 81% 29 maps Players change agents Players keep one agent 39% 56% 29 maps Changes its role mix Repeats its role mix 33% 51% 29 maps Uses the usual role mix Uses unusual role mixes 46% 43% 29 maps Fewer distinct comps More distinct comps 11.1 9.3 29 maps Fewer role mixes More role mixes 4.9 3.7 29 maps Smaller agent pool Larger agent pool 18.4 15.2 29 mapsCounts of distinct comps and agents grow with maps played, so each team's figure is resampled to a common 18-map exposure before comparison. The sample shown is the team's real number of maps.
Kills spread evenly ←→ Kills concentrated
kill-load concentration (100x Herfindahl over 5; even=20)
20.2
29 maps
field 20.1
No per-map decomposition for this one.
Its season value is not a sum over maps: splitting it would produce rows that
do not add back up to the number above, so none are shown. What the number is
built from is below instead.
Where this number comes from separates teams style_dimension
SourceMatch records, every map in the archive
Sample29
Separates teams?
Yes
separates teams; a stable read · measured reliability 0.88
Defined asA team-level style or structure dimension. value_text carries the reliability TIER that governs where it may be rendered: TEAM-SCOPED (comparable team to team), MAP-SCOPED (comparable only within a map; team-level renders greyed), REJECTED (not reproducible at this sample and never rendered as a comparison).
GE
one of the other 15 teams
estimate range
"fld" = 16-team average