Pro map meta
Across every Active Duty map, the CT side has gained ground since February 2024 — the earliest month whose sample is thick enough to trust (the 4 months before it, October 2023 through January 2024, are slivers of 29 to 74 maps): from 50.0% CT to 55.2% CT in June 2026 — a swing of 5.2pts across the whole map pool, measured month by month rather than by Premier season.
Scope, sample, and method
Measured across Tier 1 events — HLTV Majors and international LAN events with a prize pool between $600,000 and $2,000,000: 1,948 matches (4,369 maps played) between 2023-10-16 and 2026-08-02. Side counts are regulation rounds only — overtime alternates sides in short halves and can't be attributed to CT or T from the totals this data provides, so it is excluded from every share on this page.
Where a Valve season holds more than one Active Duty pool state, seasons are told apart with a letter — 1a, 1b, and so on. Valve names the whole span “Season One”; it never split it into 1a/1b itself — that split is this site's own, to distinguish two different pools within one Valve season.
the trendHow the sides moved, month by month
Every Active Duty map's regulation rounds, summed by calendar month — 35 buckets, not the three Premier-season buckets this page used to cut the same rounds into. Three points can show that the meta differs season to season; they can't show where inside a season it actually moved. Season boundaries are marked on the axis below rather than used to bucket it, and each still links to its own season.
CT share first reaches 50% in February 2024 (50.0% CT), though 5 later months still fall back below it — inside Season 1a, 516 days before Season 3 began (2025-07-15). That is also the first month this page reads at all, so there is no earlier reading here for the share to have climbed from.
35 months of figures, 129 updates linked, and method notes
2023-10 50.1%† · 2023-11 49.2%† · 2023-12 50.3%† · 2024-01 51.9%† · 2024-02 50.0% · 2024-03 51.7% · 2024-04 50.4% · 2024-05 50.5% · 2024-06 46.7%† · 2024-07 49.4%† · 2024-08 46.3% · 2024-09 52.2% · 2024-10 52.1% · 2024-11 49.8%† · 2024-12 49.3% · 2025-01 48.6%† · 2025-02 49.2% · 2025-03 46.3% · 2025-04 46.9% · 2025-05 50.8% · 2025-06 51.1% · 2025-07 52.7%† · 2025-08 52.7%† · 2025-09 52.8% · 2025-10 53.5% · 2025-11 53.0% · 2025-12 53.5% · 2026-01 51.1% · 2026-02 55.1% · 2026-03 53.5% · 2026-04 52.3% · 2026-05 54.4% · 2026-06 55.2% · 2026-07 52.2%† · 2026-08 49.8%†
† marks a month whose sample is under a quarter of this window's richest (318 maps, October 2025) — thin enough to read as noise, the hollow, dashed point above. Median 123 maps played per month; tournament schedules cluster, so the thinnest month (August 2026, 10 maps) and the richest differ by more than 31×. A thin point still counts toward every figure on this page — it's marked, not dropped.
129 updates in this window touched a weapon, the economy, a map, its utility, or player movement — the categories that could plausibly move a side's balance — ticked along the axis above and linked by day below. Marking an update next to a shift in this line is showing where the two sit side by side, not asserting one caused the other: this site does not claim any Valve update moved the CT share, here or anywhere else on this page.
October 2023: 17, 19, 25November 2023: 02, 08, 13, 16, 17, 20, 30December 2023: 04, 06, 13, 18January 2024: 04February 2024: 06, 07, 08, 12, 15, 16, 28March 2024: 13, 19April 2024: 16, 25, 26, 30May 2024: 07, 09, 23, 24, 29June 2024: 04, 10, 13, 25, 26July 2024: 02, 11, 22August 2024: 08, 14, 19, 21September 2024: 03October 2024: 02, 03, 04, 09, 23, 24, 28, 29November 2024: 06, 13, 14, 15, 19, 27December 2024: 03, 18, 19January 2025: 07, 09, 14, 21, 28, 29February 2025: 03, 27March 2025: 31May 2025: 07, 08, 12, 13July 2025: 15, 28, 29August 2025: 01, 14, 18, 26September 2025: 03, 08, 17, 23, 25October 2025: 01, 02, 03, 13, 14, 15, 22, 23, 29November 2025: 04, 12, 13, 19January 2026: 21, 22, 26, 29February 2026: 04, 23, 25March 2026: 04, 11, 18April 2026: 01, 14, 20, 21, 24, 28, 29, 30May 2026: 07, 14, 18, 20, 28July 2026: 08, 09, 14, 20, 28
by mapThe same rounds, one map at a time
The same rounds, one map at a time and still cut by Premier season — the shape of each line is the finding; the all-time figure beside each name is the same aggregate this page led with above, kept here as context rather than the headline.
Mirageall-time 53.7% CTS1a 53.3% · S1b 52.3% · S2 49.2% · S3 53.5% · S4 57.7% · S5 49.8%†
Trainall-time 53.2% CTS1a no data · S1b 54.3%† · S2 48.8% · S3 56.3% · S4 no data · S5 no data
Infernoall-time 50.1% CTS1a 49.5% · S1b 47.6% · S2 49.6% · S3 50.5% · S4 52.6% · S5 48.0%†
Dust IIall-time 50.0% CTS1a no data · S1b 49.7% · S2 46.7% · S3 51.6% · S4 51.3% · S5 48.0%†
Nukeall-time 54.9% CTS1a 53.5% · S1b 53.4% · S2 53.0% · S3 55.8% · S4 57.8% · S5 62.0%†
Ancientall-time 50.0% CTS1a 49.8% · S1b 48.5% · S2 48.6% · S3 50.5% · S4 52.3% · S5 51.8%†
Anubisall-time 44.6% CTS1a 43.5% · S1b 45.2% · S2 43.5% · S3 45.7%† · S4 46.4% · S5 44.5%†
Overpassall-time 55.6% CTS1a 56.1% · S1b 63.5%† · S2 no data · S3 55.2% · S4 54.8% · S5 no data
Vertigoall-time 48.8% CTS1a 49.2% · S1b 48.6% · S2 no data · S3 no data · S4 no data · S5 no data
Cacheall-time 56.6% CTS1a no data · S1b no data · S2 no data · S3 no data · S4 no data · S5 56.6%
Season-by-season figures and notes
A hollow, dashed point (and the † beside its number in the line below the chart) marks a season whose sample is under a quarter of that map's own largest season here — thin enough to read as noise, not signal. A blank season is not zero: the map was outside the Active Duty pool then.
veto orderWhat teams ban, pick, and leave over
veto_actions.seq orders every ban and pick inside a match, which pulls apart three signals a plain picked/removed count blurs together: the first ban (the map a team most wants gone), the first pick (the map it most wants), and the decider (the map neither side chose, left over once banning finished). Every rate below divides by only the matches where that map actually sat in the Active Duty pool that month — a map outside the pool cannot be banned, picked, or left over, so counting it against every match played that month would understate it. A map with zero such matches shows no rate at all, not a rate of zero.
Normalising this way matters, and this window has a map that shows it: Train spent part of this window outside the Active Duty pool, so a share computed against every match played that month (rather than only the ones it could be banned in) makes its first-ban rate look like it climbed after 2025-03-01 (2.3% → 11.2%) when, normalised, it fell (25.7% → 19.2%). That's why every figure on this page is opportunity-normalised rather than a share of all matches.
Split at 2025-03-01 — roughly the midpoint of this window's own coverage — first-ban rate moves the most for Train (25.7% → 19.2%, −6.5pts), Dust II (9.1% → 14.0%, +4.9pts), Anubis (12.9% → 17.2%, +4.3pts), and Mirage (17.2% → 13.2%, −4.0pts), in that order. Train is the most first-banned map in the second half of this window. The full set, with sample sizes, is below.
Per-map figures, the line legend, and notes
Anubis and Vertigo are the only maps in “the trend” above with a T-sided regulation split (44.6% and 48.8% CT). How often a map is banned and which side wins on it are separate measurements; this page draws no line between them.
Lines below: solid — first ban · dashed — first pick · dotted — decider. A gap is a month the map was outside the Active Duty pool, not a rate of zero.
Trainfirst ban 25.7% → 19.2%first ban 25.7% (18/70) before March 2025 → 19.2% (132/686) after · first pick 7.5% (n=756) · decider 8.9% (n=756) overall
Dust IIfirst ban 9.1% → 14.0%first ban 9.1% (47/517) before March 2025 → 14.0% (164/1174) after · first pick 15.8% (n=1691) · decider 15.9% (n=1691) overall
Anubisfirst ban 12.9% → 17.2%first ban 12.9% (100/774) before March 2025 → 17.2% (132/767) after · first pick 10.7% (n=1541) · decider 10.7% (n=1541) overall
Miragefirst ban 17.2% → 13.2%first ban 17.2% (133/774) before March 2025 → 13.2% (155/1174) after · first pick 13.7% (n=1948) · decider 16.2% (n=1948) overall
Nukefirst ban 11.6% → 15.0%first ban 11.6% (90/774) before March 2025 → 15.0% (176/1174) after · first pick 12.4% (n=1948) · decider 18.4% (n=1948) overall
Infernofirst ban 15.9% → 12.7%first ban 15.9% (123/774) before March 2025 → 12.7% (149/1174) after · first pick 12.4% (n=1948) · decider 14.9% (n=1948) overall
Overpassfirst ban 10.5% → 12.7%first ban 10.5% (27/257) before March 2025 → 12.7% (110/864) after · first pick 10.8% (n=1121) · decider 9.7% (n=1121) overall
Ancientfirst ban 12.3% → 13.0%first ban 12.3% (95/774) before March 2025 → 13.0% (153/1174) after · first pick 12.1% (n=1948) · decider 16.9% (n=1948) overall
Cachetoo thin either side of the splitfirst ban no data before March 2025 → 9.7% (3/31) after · first pick 12.9% (n=31) · decider 16.1% (n=31) overall
Vertigotoo thin either side of the splitfirst ban 20.0% (141/704) before March 2025 → no data after · first pick 8.8% (n=704) · decider 5.8% (n=704) overall
by tournamentWhere each event’s own numbers live
Everything above is this window averaged across every Tier 1 event it holds. Each event on its own — which maps it played, its own veto order, its own CT/T split against the season it sat in, and a champion where its own format supports naming one — is inside the season it sat in, at /seasons/.
Match data via HLTV, through a committed point-in-time snapshot — not a live feed. Every figure on this page is regulation rounds only; the full measurement conditions are at the top.