mLkoach / Guides
FACEIT level: what really separates the levels
Climbing is often explained by “game sense” or “communication”. Those may be good explanations; they're not the ones the demos show. Here's what really sets a level 3 apart from a level 7, measured on 1,015 FACEIT players.
The ranking, by separating power
For each measure, we computed its rank correlation with FACEIT level: how much it rises as the level rises. Zero means “no link”; the higher the number, the better the measure separates levels.
| Measure | Link with level |
|---|---|
| ADR | +0.34 |
| Win impact | +0.29 |
| Duel (aim, openings, trades) | +0.27 |
| Post-plant (clutches, kills and saves once the bomb is down) | +0.17 |
| Utility (utility return) | +0.15 |
| KAST | +0.15 |
| Game sense | −0.01 |
Rank correlation between the measure and the most recent known FACEIT level, over 1,015 players measured from their demos.
Three things this table says
What rises is firepower
The first three rows — ADR, impact, duel — all speak about the same thing: what you're worth when you shoot. That's where most of the gap between levels lies. Nothing romantic about it, but it's what the data shows.
KAST barely separates anything
From the bottom to the top of the ladder, median KAST goes from 70 % to 75 % — five points, while ADR gains thirty-two. A level 2 player is useful to their team almost as often as a level 7: they get a kill, an assist, they survive or they get traded. What changes is the weight of what they do, not how often. KAST measures presence, not contribution.
“Game sense” isn't where you think
Our game sense measure — traded deaths, KAST, avoidable deaths, utility taken
to the grave, survival — does not correlate with level (−0.01). In
other words: on these points, a level 8 does no better than a level
3.
Two possible readings, and we don't pick one: either these habits don't improve as you rank up — you climb by aiming better, not by dying smarter; or they do improve but are offset by tougher opponents, who punish the same mistakes harder. Either way, the practical consequence is the same: it's the area where a player can win rounds without improving their aim, and almost nobody works on it.
The level 5 dip. Median ADR drops slightly below level 4 (78 against 80). It isn't a measurement error: at level 5, opponents get clearly better while the player has only just arrived. Every level reached comes with a dip — and that's exactly when you think you're getting worse.
What to do with it
- To climb: ADR is the number to follow, because it's the one that moves with the level. The benchmarks are on the what a good ADR is page.
- To win rounds right now: the measures that don't separate levels are the ones almost nobody is good at. Utility taken to the grave, avoidable deaths, team damage — a player who fixes them gets ahead of their whole level without touching their mouse.
- To avoid getting discouraged: a dip after climbing is normal, and it shows up in everyone's numbers.
Where you stand, measure by measure. mLkoach reads your FACEIT demos and gives each of your axes the level it matches — and says so plainly when a measure isn't enough to conclude.
Read next: how to analyse your CS2 demo.