Do not watch the whole match at normal speed. Start with one question, shortlist three to five decisive rounds, watch each moment first from player POV and then from free camera, inspect teammate spacing and utility, and write one change you can test. The viewer gives you the round timeline, POV, follow, top-down and free cameras, an event list, a utility view and a walk mode to check sightlines yourself.
Prepare one review question
A vague goal produces a vague review. "Play better" is not a question; it is a mood. Pick one specific question before you open the file: Why did our B holds collapse? Why were my opening deaths untradeable? Did our flashes create safe fights, or just noise? Were our post-plant positions supporting each other or duplicating each other?
Upload the .dem to the demo viewer, let it parse, then press R to open the review panel. Check the Summary tab first and confirm the map, score and player names. If you opened the wrong match — and everyone does eventually — you want to find out now rather than after twenty minutes of confident analysis.
Choose a small set of decisive rounds
Use the Rounds and Events tabs to shortlist rather than scrubbing blindly. Reliable starting points: the first death in a close round, a lost man advantage, a failed site entry, a failed retake, and a round where expensive utility bought no space. Three to five rounds is a full session.
Also pick one round that went well on the same side and site. Comparing a failed execute against a successful one controls for the map and the opponent, and the difference that remains — timing, spacing, utility order — is the thing you can actually name.
Start watching several seconds before contact. A death is usually the last symptom of a spacing or timing decision made earlier, and if you start at the kill you will diagnose the symptom. Shift+arrow moves by round and Shift+Q/E moves between deaths; the rest are in the controls reference.
Compare POV against the full situation
- Watch in Player POV and say out loud what information was available. Not what happened — what was knowable.
- Pause before the decision, not after the kill. The frame-step keys , and . put you exactly on the moment of commitment.
- Switch to free or top-down camera and inspect teammate spacing, enemy routes and utility.
- Return to POV and ask whether the alternative was realistic with the information available. If it required knowing where the enemy was, it is not a lesson.
This ordering is the single most important discipline in demo review, and it exists to defeat hindsight bias. Once free camera has shown you the enemy behind the box, you cannot un-see it, and every subsequent judgement is contaminated. POV first, always.
When a conclusion depends on "they should have seen that", verify it before writing it down. Press B to enter walk mode and stand where the player stood, at player height. Free camera floats; players do not. A sightline that exists for a hovering camera may not exist for anyone in the round. If the map is multi-level, check the floor too — the per-map notes in the CS2 map review guide cover Nuke, Train and Vertigo specifically.
Check the utility, not just the fight
Most rounds are decided before anyone shoots. Open the Utility tab, jump to each relevant grenade in the shortlisted round and draw its complete trajectory in the world — the release point, the bounces and the destination, not the icon in someone's inventory. Then compare the effect against movement: was the entry player ready when the flash went off, or already dead?
Two questions are worth asking every single time. Did two teammates spend utility on the same job while another lane stayed open? And was useful utility still in someone's inventory when they died? The full method is in how to review CS2 smokes, flashes and grenades.
Ask questions the demo can answer
- Positioning: Could a teammate trade the first fight? Were two players exposed to the same angle?
- Timing: Did contact happen before the team was ready? Was a rotation based on information visible in the replay?
- Utility: Did the smoke block the intended sightline? Did the flash precede the peek? Was useful utility still carried at death?
- Economy and equipment: Did the plan fit the weapons and money visible in that round?
- Decision: What lower-risk option existed at the exact moment of commitment?
Avoid diagnosing personality or intent from coordinates. "The second player was too far away to trade within the fight" is reviewable and checkable. "The player did not care about trading" is a story about someone's character that the demo contains no evidence for. The boundary between those two sentences is set out in what a demo analyzer can and cannot tell you.
Leave with one action
End with a rule that can be performed, not a label that can only be felt. Replace "bad positioning" with "after the first A-ramp contact, hold until the connector player can trade." Replace "use utility better" with "throw the entry flash before the first player crosses the choke."
Then test it. If the situation is available as a Replay Moment, play it: the recorded positions, utility, health and clock are rebuilt and you take control, so the alternative you just claimed would work either works or does not. That is the difference between a review that changes a match and a review that produces a well-phrased regret.
Four ways demo review goes wrong
| Mistake | What to do instead |
|---|---|
| Watching the full match at 1× | Skim at 4× with a 30-second seek interval to find rounds, then drop to 0.25× on the three that matter. |
| Starting in free camera | POV first. Free camera is for the second pass, once you have recorded what was knowable. |
| Reviewing only your own deaths | Review one round you won on the same side and site. The contrast is where the transferable lesson lives. |
| Ending with a list of six fixes | End with one. A list of six rules is a list of zero rules by the third round of the next match. |
Frequently asked questions
How long should a CS2 demo review take?
Twenty to thirty minutes on three to five rounds beats two hours on a whole match. A full-match rewatch at normal speed produces recognition rather than change, because attention is spread evenly across rounds that were never in doubt.
Should you review your own POV or the free camera first?
POV first, always. Free camera shows you information the player did not have, so opening with it contaminates every judgement you make afterwards. Establish what was knowable, then widen the view to explain what actually happened.
What should you look for when reviewing a CS2 demo?
Look for the decision before the death rather than the death itself. Deaths after an advantage, late utility, site entries, post-plant spacing and retakes carry most of the improvable signal, while one-off aim misses carry almost none.
Can you review a CS2 demo without opening the game?
Yes. Uploading the demo file to a browser demo viewer gives you the same round timeline, camera modes, scoreboard and utility inspection without a Counter-Strike 2 installation, which also means you can review on a machine that cannot run the game.