Bottom line: Counter-Strike is mildly CT-sided overall: across roughly two million professional team-rounds (from more than 20,000 Counter-Strike matches, spanning the CS and CS2 eras), the Counter-Terrorist side wins 51.3% and the Terrorist side 48.7%. The map-by-map spread is wide: Nuke (54.1% CT) is strongly CT-sided, while Anubis (45.0% CT) is firmly T-sided. Because the active-duty pool keeps changing, we lead with the data on the maps that actually have a sample, not with pool membership. This is a data study built on the EsportsOdds CS2 data API, with figures as of 16 July 2026.
The overall balance
Every round of Counter-Strike is won by exactly one side, so across a large enough sample the CT/T split tells you which side the game favours. The answer is the CT side, but only just.
That 51.3% / 48.7% edge is small, about one round in forty tipping CT's way, but it is consistent enough over two million team-rounds to be real rather than noise. It reflects the structural advantage of defending: CTs hold known angles, force the Ts to commit, and only have to stop the bomb.
Side balance varies enormously by map
The overall number hides the real story, which is how differently each map plays. Here is the side lean of the maps in heaviest rotation:
| Map | CT win | T win |
|---|---|---|
| Overpass | 55.3% | 44.7% |
| Nuke | 54.1% | 45.9% |
| Mirage | 52.8% | 47.2% |
| Ancient | 51.2% | 48.8% |
| Dust2 | 50.2% | 49.8% |
| Inferno | 49.6% | 50.4% |
| Anubis | 45.0% | 55.0% |
Nuke is the classic CT fortress: tight, vertical, easy to hold and hard to retake. Anubis is the outlier the other way, an open, connector-heavy map where the T side can take space and trade. Dust2 and Inferno sit on the coin-flip line. One caveat before reading this as "the current pool": it changes, and Overpass in particular has since rotated out.
The pool changes, so read the data, not the pool
The active-duty pool turns over regularly, so any list framed as "the current pool" goes stale fast. The clearest recent example: on 6 July 2026 Valve replaced Overpass with Cache. Cache had been gone for years and only returned to CS2 in April 2026, so it has essentially no round-level history, and nobody has a meaningful CS2-era read on how it sides. That is the case for leading with the data. Here is the CT round-win rate for every map with a real sample, including two (Overpass and Train) that have since rotated out but carry years of rounds:
Read as long-run tendencies, Train (55.6% CT) and the now-rotated Overpass (55.3%) are the most CT-sided maps on record. Anubis (45.0% CT) is the T-sided outlier, and Vertigo, Inferno and Dust2 sit within a point of even. One caveat: these are all-time aggregates spanning map reworks, so treat them as each map's long-run lean, not this exact patch.
The extremes
The gap between the most CT-sided and most T-sided map on record is about ten percentage points, a huge difference in a game decided round by round:
Train edges even Overpass at the top (55.6%), as the fuller chart above shows. This spread is why the map veto matters so much and why teams drill side-specific setups: a side that wins its "good half" comfortably on Nuke banks a lead a balanced map like Dust2 never offers.
The pistol round is the fairest round
There is one round that resists all of this map-by-map skew: the pistol round that opens each half. With both teams on the same minimal buy, it is nearly a coin flip regardless of map:
At 50.3% CT / 49.7% T, the pistol round is the closest thing Counter-Strike has to a level start, which is why it is so pivotal. Win the pistol and the follow-up rounds it funds (thanks to the economy), and a team can snowball a 3-0 before the rifle rounds begin.
Why maps lean the way they do
Two forces mostly explain the spread:
- Map geometry. Tight, layered maps with strong defensive holds (Nuke) favour CTs; open maps with space to trade (Anubis) favour Ts.
- The AWP. As our weapon study shows, about 64% of AWP kills come on the CT side. Maps with long, AWP-friendly sightlines hand the defence a low-risk pick every round, a big part of why those maps skew CT.
Getting map and side data from the API
Every figure here is computed from per-round, per-team data (side, outcome, economy) available for every map of every match through the EsportsOdds CS2 data API. That is what you would build a map-meta tracker, a side-win-rate model, or a map-specific study on, as clean JSON. You compute these numbers yourself rather than trusting someone else's aggregate, and our methodology page documents how side and round outcomes are recorded.
Counter-Strike is mildly CT-sided overall (51.3% / 48.7%), but it ranges from Nuke and Train near 55% CT to Anubis at 45% CT, a ten-point spread. The pool keeps changing (Cache is the newest map, with no data yet), so read the per-map data, not the pool. The pistol round, on equal buys, is the one near-even round on every map.
Frequently asked questions
Is CS2 CT-sided or T-sided?
Mildly CT-sided overall: across roughly two million team-rounds in our data, the CT side wins 51.3% and the T side 48.7%. It varies a lot by map: Nuke is strongly CT-sided, Anubis is T-sided.
What is the most CT-sided map in CS2?
On the historical record, Train (55.6% CT) and Overpass (55.3%) are the most CT-sided, though both have rotated in and out of the active-duty pool. Among maps in the current pool, Nuke (54.1%) leads.
What is the most T-sided map?
Anubis, at 45.0% CT win (55.0% T), is the most Terrorist-sided map by a clear margin, and it is in the current pool.
Is the pistol round balanced?
Nearly. Across roughly 87,000 pistol team-rounds, the CT side wins 50.3% and the T side 49.7%, the closest to even any round type gets, because both teams start on the same buy.
Where can I get CS2 map and round data via an API?
Through the EsportsOdds CS2 data API: per-round, per-team side and outcome data for every map, as JSON. Recompute any map's CT/T split, or track it as the pool changes, straight from the rows.