How to analyze your CS2 matches: what the demo shows and what your hand shows

What happened vs why it happened

After a match you usually have two sources. One is the demo, the replay of the game. It shows your positions, your team's movement, utility and every death from your point of view. The other is stat sites such as Leetify or Scope.gg. They calculate ADR, K/D, headshot %, utility usage and often aim metrics derived from the demo. Both mostly tell you what happened.

The why is harder. A demo shows you lost the duel. It won't show whether your hand stopped 5 mm short, whether you fired before the crosshair settled, or whether you lifted the mouse halfway through a turn. For that you need data from the mouse itself, recorded during the match. There's a longer comparison in FPS Performance Lab vs Leetify.

None of these replaces the others. If you lose rounds on rotations, hand data won't fix that. If you lose duels from good positions, a demo won't tell you what to do with your hand.

How to analyze your CS2 matches step by step

Here's a routine that takes about 15 minutes after a match. I'll use FPS Performance Lab's CS2 match analysis as the example, but you can do steps 4–6 with just a demo.

1. Get your settings right. Make sure the app has your mouse's real DPI and your CS2 sensitivity. If you're not sure the DPI on the box is accurate, check it with a payment card or a ruler (real DPI guide). Otherwise the in-game degrees will be off.
2. Record the match. Start recording in the app and play normally. The mouse is read through Windows Raw Input in the background, but only while recording. Game events come from Valve's official Game State Integration. Video is optional. If you turn it on, you keep short clips after the match and the full recording is deleted unless you choose to keep it.
3. Read the match summary. At the top: one strength, one main problem and a "For the next match" tip. Then the numbers: fights, flicks, headshot %, ADR and rounds won/lost, deaths without a shot. Don't start with the table. Start with the sentence.
4. Watch 2–3 clips. Pick clips that match the problem from the summary. Play them slowed down and watch the hand-speed chart under the video: where the first move ends, whether there are extra corrections, when the shot lands.
5. Compare the match with your test. "Match vs your test" shows whether the problem from your aim test also shows up in the game. If you stop short in the test but not in matches, check it over a few more matches before you change your training plan. The comparison only makes sense with the same settings in both.
6. Pick one thing for the next match. One, not five. Ideally something you can check with a number, like the share of fights with the crosshair placed before the fight, or deaths without a shot.
7. Repeat. After 3–5 matches in the same mode you'll see whether the number moves. The Player card puts this match against your average from previous matches.

What to look at in hand data

• First movement: short of the target / no correction / past the shot point. Tells you whether you tend to undershoot or overshoot. Across the developer's 180 analysed flicks from 13 matches, 39% stopped short of the target, 54% were on point and 8% went past.
• Shot on the move. Clicking before the crosshair and hand stopped. That was 28% of the developer's flicks. It's not a counter-strafing score; the app doesn't measure character movement.
• Big turns (over 50°). How often enemies show up to your side and what the correction costs you. More in losing duels at wide angles.
• Crosshair placed before the fight. How many fights you start without a flick. See crosshair placement in CS2.
• Deaths without a shot. Deaths where your hand never got to act. Covered in dying without shooting.

Common mistakes when reviewing matches

• Conclusions from one match. One game doesn't have enough fights to tell a habit from chance. Look for problems that keep coming back.
• Mixing modes. Deathmatch and Premier are different situations. Compare matches in the same mode, ideally on the same map.
• Changing sensitivity after a bad game. Once you change it, comparisons with earlier matches stop meaning much. Change it rarely and on purpose, then take a new test.
• Five goals at once. In a match you're thinking about the game, not a checklist. One thing per match is enough.

Start with one match

The free FPS Performance Lab demo on Steam includes a sample CS2 match from the developer and lets you record one match of your own. Match analysis in 2.0 works in Counter-Strike 2 only. How it works explains exactly how the data is measured, and you can look at your hand outside a match with the aim test. Keep watching demos and checking your stats. Hand data adds the answer to "why".

Sources

• Leetify
• Scope.gg

Questions and answers

Read next