2,370 matches. 48,500 rounds. 54% headshots, against a measured median of 51% at Level 10 — the best mechanical number in this set of examples. ADR 81.3, a shade under the 82.0 median for the band above theirs. Opening duel win rate of 50.8% against a 50.8% median for their ELO band — exactly at it — and their entry balance across 11,085 duels is positive: 5,635 first kills to 5,450 first deaths.

There is nothing wrong with how this person plays Counter-Strike.

Their lifetime win rate is 50%.

What the report found

The top card, at severity 72, is about sessions. Across the seven in their recent history:

• Two ran long and ended negative.
• The two-loss rule was broken three times.
• 29% included matches after 23:00 — the lowest late-night share in this set.

The detail is in the session history itself. Two evenings stand out:

• A 7-match session finishing 1W–6L, estimated at −125 ELO.
• An 8-match session finishing 2W–6L, estimated at −100 ELO.

That is 225 ELO across two sittings, from a player whose per-match performance is at or above the median for their level in almost every dimension.

The asymmetry that defines a drawdown

A 50% win rate over 2,370 matches means this player has won and lost roughly 1,185 times each. ELO gained and lost should cancel. It does — that is the problem. Two thousand three hundred and seventy matches of median-or-better Level 10 play has produced a net of nothing, because the good stretches are slow and spread across dozens of matches while the bad ones are concentrated into single evenings.

Climbing is a win-rate effect. A drawdown is a handful of sessions. The two do not happen at the same speed, and that is why a player can be mechanically better than their rank and stay at it indefinitely. There is more on the pattern in the Level 9 drawdown guide, which describes the same shape a level below.

Why the late-night number is a red herring here

29% is the lowest late-night share of any example in this set, and it would be easy to read that as exoneration. It is not the mechanism.

Compare the Level 6 session profile, where 88% of sessions ran past 23:00 and the sessions were mostly short. That player’s problem is when. This player’s problem is length — the 7-match and 8-match evenings — and the two need different rules. The card gives both, but the one that matters here is:

Cap the session at five matches even when you are winning.

Neither of their two disaster sessions would have been possible under that rule. The 1W–6L would have been cut at 5 matches; the 2W–6L likewise. Roughly half of the 225 ELO does not happen.

The utility card, and what it is actually saying

Second finding at severity 50, and it is a mix problem rather than an output problem: 18,822 non-flash grenades against 13,931 flashes, a ratio of 1.35 where the report wants 0.6 or below.

Their utility damage is 5.0 per round, above the 4.7 median for their band, so the report does not ask them to raise it. It asks them to convert nades into flashes. Their flash success rate is 56% — the highest in this set — which is the argument: the flashes they do throw work, and they throw 0.29 per round.

The map card

Nuke at 46% over 142 matches, against Ancient at 54% over 323. One ban, on a large enough sample to be certain. Ranked third at severity 39, correctly — it is worth 1–3 points of win rate against the 2–4 the session card offers, and the session card is free.