CS2 Map Specialists
Which tracked professional teams actually perform on each Active Duty map — and, read the other way, the map every one of them is strongest and weakest on. Every win rate is shown next to the number of maps behind it.
How these rankings work
A team's raw win rate on a map is unusable at a small sample: play two Nuke games and win both and you read as the best Nuke team alive. So two things happen before anything is ranked. First, a team that has played a map fewer than 5times is left out of that map's table entirely rather than ranked with a caveat. Second, the teams that remain are ordered by an adjusted win rate: their record with 6 imaginary coin-flip maps added to it, which drags a thin record towards 50% and leaves a large one closer to where it started — a 5-0 lands on 72.7% rather than 100%, while 60% over 20 maps only eases to 57.7%. That is the whole method. The raw win rate and the maps played sit in the same row, so you can always see what the adjustment did and disagree with it.
The CT and T columns work differently, on purpose. Rounds are a separate sample from maps played — a map is about twelve rounds per side, so a team that has only just cleared the 5-map threshold has roughly sixty rounds of each side, and the maps gate says nothing about whether that is enough. Those columns have their own minimum of 100 rounds, the point where the margin of error on a near-even rate drops to about five points; under it the cell is a dash. Past it the rate is shown raw, with its round count beside it. It is not put through the adjustment above, because six phantom maps against several hundred rounds would move the number by a fraction of a point while implying a correction had been made — a gate is the honest tool at that sample size, not a prior.
A bare side rate is close to meaningless on its own, because Nuke is CT-sided for everyone. So each side rate is also shown as points above or below that map's pooled ratefor the same side, listed in each map's header. Both the team figure and the baseline are printed, so the subtraction is yours to check. One caveat worth knowing: a team is itself part of the pool its baseline is drawn from — roughly one thirtieth of it — so this is a comparison against the field including itself, not against its opponents.
Pistol rounds are pooled, never shown per map, and that is a deliberate refusal rather than an omission. A map has about two pistol rounds, so even a team with twenty games on one map has only forty pistol rounds there — a margin of error of roughly eight points, wide enough to swallow any difference worth naming. Reaching the same confidence the side columns hold would take fifty games on a single map, which nobody in this sample has. Added up across the whole pool the same team has several hundred pistol rounds, which is a usable sample for the weaker claim actually being made: this team wins pistol rounds, not this team wins pistol rounds here.
Picks and bans are counts of what happened, taken straight from each team's tracked veto record — never inferred from how well a team plays a map. One gap is worth knowing: a map is only counted when the provider carries a line for it, so a team that bans a map in every single veto and therefore never plays it may be missing from that column rather than showing a large number. Picks are not affected, because a map has to be played to have been picked.
Professional statistics are temporarily unavailable, so no team records can be shown. Nothing is estimated in their place — the tables return when the provider data does. The map comparison and individual map pages carry our own callout and utility data, which is unaffected.
What this page does not tell you
The sample is the tracked teams playing recent tier-S and tier-A matches — around thirty of them, not the whole professional scene. A team missing from a table has not been judged and found wanting; it either is not in the sample or has not played that map enough to rank.
Nor is opponent strength accounted for. A 70% record built against weaker opposition and one built at a Major both read as 70% here, because the per-map records the provider exposes do not say who the opponent was. Read these tables as "how has this team fared on this map", not as a strength rating.
Only the 7 Active Duty maps are ranked. Reserve maps (Overpass, Vertigo, Train) stay playable in Competitive matchmaking but cannot appear in a professional veto, so a record on one is not a specialism worth acting on.
Frequently asked
- Which pro team is best on each CS2 map?
- Rankings are generated from tracked professional results across the 7 Active Duty maps (Mirage, Inferno, Dust 2, Anubis, Ancient, Nuke, and Cache). Provider statistics are temporarily unavailable, so no team is listed right now.
- Why is a team with a 100% win rate not top of the table?
- Because a perfect record over two or three maps is not evidence of a specialism. Teams under 5 maps on a given map are not ranked at all, and the teams that are ranked are ordered by an adjusted win rate that pulls thin records towards 50%. The raw win rate and the number of maps behind it are both shown, so you can see exactly what was adjusted.
- What do the CT and T columns mean?
- They are the team's own round win rate on each side of that map, shown against the map's pooled rate for the same side. Nuke is CT-sided for everybody, so a 55% CT record there is unremarkable — what the "vs map" column shows is how far above or below the map's own baseline a team sits. A side figure appears only once a team has at least 100 rounds on that side, and the round count is printed next to every percentage.
- Why do some teams show a win rate but no CT or T figure?
- Because rounds are a different sample from maps played. A CS2 map is about twelve rounds per side, so a team that has just cleared the 5-map threshold has only around sixty rounds of each side — below the 100-round minimum the side columns require. Rather than show a rate the sample does not support, those cells are left as a dash.
- How is the adjusted win rate calculated?
- Each team's record on a map has 6 imaginary coin-flip maps added to it before the percentage is taken. A 5-0 record becomes 72.7% rather than 100%, while a 60% record over 20 maps only eases to 57.7%. It is one rule applied to one pair of numbers — wins and maps played — not a weighted score of several ingredients. It is applied to map records only, never to the CT and T round rates.
- What counts as a team's worst map?
- The Active Duty map with its lowest adjusted win rate, among maps it has played at least 5 times. Maps a team has barely played are excluded rather than assumed bad, and a team that clears the threshold on only one map gets no best-versus-worst comparison at all — with one map, both would be the same row.
- Why is there no per-map pistol round win rate?
- Because the sample does not support one. A CS2 map has about two pistol rounds, so a team with twenty games on a map has roughly forty pistol rounds there, giving a margin of error near eight percentage points — bigger than any gap that would be interesting. Matching the confidence of the CT and T columns would need about fifty games on a single map, which no team in this sample has. Rather than print a noisy number we pool each team's pistol rounds across all 7 Active Duty maps, where the sample reaches several hundred rounds, and label it as a pool-wide figure.
- Are the pick and ban numbers real, or inferred?
- They are counts of tracked vetoes — how many times a team actually picked or banned each map. Nothing is inferred from win rates or from how well a team plays a map. Shares are only shown once a team has at least 50 of that action, so a team with three tracked picks gets counts but no percentages. One known gap: a map only appears if the provider carries a line for that team and map, so a map a team always bans and consequently never plays can be under-counted.
- Where does this data come from?
- PandaScore competitive match data, aggregated per team per map — the same sample behind our map meta and individual map pages. It covers the tracked teams playing recent tier-S and tier-A matches, not every professional team in the world.