Where do CS2 demo statistics come from?
A demo is a record of events, not a report card. Each kill carries an attacker, a victim, a weapon and flags such as headshot or wallbang. Each damage event carries an amount and a hit group. Rounds carry a start, a freeze-time end and an outcome. Everything you see on a scoreboard is arithmetic performed on those records after the fact.
That is why two tools can show different numbers for the same match without either being wrong: they are making different choices about round boundaries, self-damage, or what counts as an assist. It is also why a statistic is only as trustworthy as the definition behind it. The rest of this page gives the definitions.
In Memorin these live in the match review panel, which has four tabs — Summary, Rounds, Events and Utility. Summary is the per-player table; Rounds is a card per round; Events is a searchable, filterable log of every kill, damage, bomb, utility and reload event; Utility is the grenade review. The Summary tab can also export the full match as JSON or the player table as CSV, if you would rather do your own arithmetic.
What does ADR mean in CS2?
ADR is average damage per round: the total damage a player dealt to opponents, divided by the number of rounds played. It rewards damage that did not convert into a kill, which is why it separates players who create openings from players who only finish them.
The practical use of ADR is not comparing yourself to a benchmark. It is comparing your ADR to your kills. A player with high kills and low ADR is finishing damaged opponents — valuable, but dependent on teammates. A player with low kills and high ADR is doing the setup work and losing the finish, which is usually an aim or positioning problem you can attack directly. The two profiles need completely different practice.
What does KAST measure?
KAST is the percentage of rounds in which a player got a kill, an assist, survived, or was traded. It measures how often someone contributed something rather than how much, so it exposes players who disappear in the rounds they do not frag.
Note the fourth term. Being traded counts, which means KAST credits dying usefully. A player who dies first every round but always in a spot where a teammate immediately punishes the killer has a respectable KAST; the same player dying alone off-angle does not. That single asymmetry makes KAST the best available proxy for whether someone is playing with their team.
What is the rating number?
Ratings combine several per-round quantities into one figure so that players can be ordered. Memorin's Summary tab shows a deterministic Rating 2.0 estimate computed from KAST, kills per round, deaths per round, an impact term and ADR, and labels it with an asterisk for exactly that reason: it is a reproduction of a public formula from demo events, not an official number issued by anyone.
Treat any rating as a sorting key, not a verdict. It is useful for "which of my ten teammates this month was actually carrying" and useless for "am I a 1.15 player", because a single match is far too small a sample for a composite statistic to be stable.
What are opening duels and traded deaths?
The opening duel is the round's first kill. Every round has exactly one, and it is the most informative single event in Counter-Strike because it is the only fight both teams enter with full information and full equipment. Memorin's Summary tab shows each player's opening kills and opening deaths under their name, and the Rounds tab names the opener and the victim for each round with buttons to jump straight into either player's POV.
| Pattern | What it usually means | What to change |
|---|---|---|
| Many opening kills, many opening deaths | You are the entry, with roughly coin-flip duels | Improve the setup: utility before the peek, not aim |
| Few opening kills, many opening deaths | You are taking duels you are not set up to win | Stop peeking first, or peek with support |
| Few of either | You arrive after the round is decided | Move earlier; late information is worth little |
| Opening deaths rarely traded | You die where nobody can punish it | Positioning, not aim — this is the cheapest fix in the game |
The last row is the one to take seriously. Trade rate is almost entirely a function of where you stand relative to a teammate, and it is the change with the shortest distance between "I understand it" and "it works".
Does the demo show utility usage?
Yes, in two forms. The Summary tab counts utility damage per player, and the Utility tab plus the event log record every grenade thrown, with trajectories you can inspect in 3D. Flash assists and utility damage are recorded events; whether a smoke helped is not.
This is the sharpest example of the evidence-versus-interpretation line. A molotov that does zero damage can be the best utility of the round if it denied a rotation, and a grenade that does 40 damage across three players can be worthless if it broke your own timing. Only the round shows you which. The utility review guide works through the method for judging that fairly.
What about the economy?
The Rounds tab classifies each team's buy for each round — full buy, force buy, light buy or eco — from team money and equipment value, alongside the round outcome, the objective site, multikills, clutches and the number of utility throws. Reading buy type next to outcome is the fastest way to find the rounds that actually cost you the half: an eco you lost is noise, an anti-eco you lost is a problem, and a full buy lost without a plant is a problem with a cause you can watch.
Which CS2 stat should I actually try to improve?
Opening duels and traded deaths, because both are decisions rather than outcomes. You control when you take a first contact and whether you die somewhere a teammate can punish it.
Compare that with the alternatives. Headshot percentage is mostly a function of the weapons you buy and the ranges you fight at. Kills are a function of how much of the round you survive. Rating is a function of all of the above. Opening duel selection and trade positioning are choices you make before any aim is involved, which makes them the only entries on the list that respond to a deliberate decision.
The workflow that follows from this is short:
- Open the Summary tab and find your worst gap — usually opening deaths, or an untraded-death pattern.
- Use the Rounds tab to jump to those specific rounds rather than watching the match end to end.
- Watch each one from your POV, then from free camera, and note what was actually available.
- Take the recurring one and open it as a Replay Moment, where the situation is rebuilt and handed to you so you can try the other option until it becomes the default.
Can a demo tell me why I lost the round?
No. A demo records what happened, not why anyone chose it. Statistics point you at the rounds worth examining, and watching those rounds is what supplies the reason.
Some concrete things no statistic in any tool can recover from a demo file: what was said on comms, what anyone intended, what a player could actually see on their monitor, whether the crosshair was placed deliberately or by luck, and whether a good decision simply lost. Playback also runs at integer-tick precision, so sub-tick command ordering is not reconstructed. There is no automatic explanation of a loss here, and you should be suspicious of any tool that claims one — the demo does not contain the information such a claim would require. The capabilities and limits guide is the full version of this boundary.
FAQ
What does ADR mean in CS2?
ADR is average damage per round: the total damage a player dealt to opponents, divided by the number of rounds played. It rewards damage that did not convert into a kill, which is why it separates players who create openings from players who only finish them.
What does KAST measure?
KAST is the percentage of rounds in which a player got a kill, an assist, survived, or was traded. It measures how often someone contributed something rather than how much, so it exposes players who disappear in the rounds they do not frag.
Can a demo tell me why I lost the round?
No. A demo records what happened, not why anyone chose it. Statistics point you at the rounds worth examining, and watching those rounds is what supplies the reason.
Which CS2 stat should I actually try to improve?
Opening duels and traded deaths, because both are decisions rather than outcomes. You control when you take a first contact and whether you die somewhere a teammate can punish it.