CSStatLab
Players

How to Compare Two Players Fairly

Comparing players is easy to get wrong. How to do it in a way that actually means something.

1 min readBy CSStatLab EditorialUpdated
This Academy article is maintained under the site's editorial standards for sourcing, updates and corrections. Read our Editorial Policy for more on how CSStatLab reviews and updates content.

Comparing two players by a single stat-usually K/D-is how most arguments start and how most of them go wrong. A fair comparison accounts for role, sample size and context.

Compare Like Roles

An entry fragger and a support player do different jobs, so comparing their raw frags is meaningless. An entry should have more deaths; a support should have more assists and utility. Compare players in the same role, or compare each against what their role is meant to do-see The Roles in a Counter-Strike Team.


Use More Than One Number

No single stat captures a player. Look at a spread:

  • ADR for consistent damage-see What is ADR?
  • KAST for round-to-round usefulness-see Understanding KAST
  • Opening duels, trades and impact for role and moment-to-moment value.

This is exactly why a profile splits into four pillars instead of one score-see Understanding Statistical DNA.


Mind the Sample and the Level

  • A comparison over a handful of games is noise-see Why One Match Doesn't Matter.
  • Stats are relative to the level of play; the same number means different things at different ranks.

Takeaway

Compare players by role, across several stats, over a large sample. Anything less-especially a single K/D-tells you almost nothing.

Was this article useful?

See this concept on real profiles

Open a live player page and compare the article's ideas with real trends, ranks and statistical signals.

Learn how to read a profile