Entrying, Trading, AWPing, Flashes: What affects HLTV rating the most?

In competitive Counter-Strike, a player’s HLTV 2.1 rating is often used as the gold standard for individual performance. But which teamplay statistics affect this number and by how much? To answer this, I gathered a dataset of top 50 matches since 2024 (minimum 2,000 rounds, n=287), separated them by side and whether the player was a primary Rifler or AWPer and ran linear regression on all permutations. 

I gathered data on every statistic on a players HLTV page that could reasonably be considered a teamplay metric. I removed any stats that had no effect across every model. I also found that most tests generally agreed which statistics were most important. Because of this, I didn’t have to make Riflers/AWPers only models. The only 2 models that are necessary were a T-side only model and CT-side only model. 

The CT-side model:
Adjusted R^2: 0.405
RMSE: 0.061
Variables (In order of importance): Trading, Sniper Kill%, Save % on round losses, Entrying, Assisted Kill%, Last Alive%, Flash assists.

The T-side model:
Adjusted R^2: 0.303
RMSE: 0.069
Variables (In order of importance): Trading, Flash assists, Sniper Kill%, Entrying, Save % on round losses, Assisted Kill%.

The #1 takeaway was trading. Trading is a “main category” on HLTV that uses subcategories Saved teammate per round, Trade kills per round, and trade kills percentage to provide a number from 0-100. Every test agreed Trading is a top 2 variable. AWPers have appeared less important since CS2 came out, but every test also agreed that AWPers have higher HLTV ratings still. My last notable takeaway was Flash assists barely matter on CT side, but are near the top of T side. 

Using these models, I created “adjustments”. Adjustments are estimates of how much a player’s HLTV rating would change if their teamplay stats (roles) were average. For example, a CT-side adjustment of –0.04 means a player’s HLTV rating would be expected to be 0.04 lower in a vacuum where their CT role was neutral.

I calculated the adjustments for some of the Tier 1 teams in 2025.
Here is the top 3 for the CT side model:

#1 Techno4k: +0.18
#2 kyxsan: +0.16

#3 Aleksib: +0.16

CT side Bottom 3:

#1 FL4MUS: -0.17

#2 donk: -0.14

#3 XANTARES: -0.14

Top 3 T side model:

#1 ultimate: +0.17

#2 apEX: +0.15

#3 karrigan: +0.13

Bottom 3 T side model:

#1 ICY: -0.15

#2 FL1T: -0.14

#3 Sh1ro: -0.13

Here are the top 10 in overall adjustment (CT and T side models results averaged):

#1 apEX: +0.13

#2 Techno4k: +0.13

#3 Aleksib: +0.12

#4 kyxsan: +0.11

#5 Brollan: +0.11

#6 MAJ3R: +0.11

#7 karrigan: +0.10

#8 FalleN: +0.09

#9 ultimate: +0.08

#10 chelo: +0.08

Bottom 10:

#1 sh1ro: -0.13
#2 FL1T: -0.13

#3 donk: -0.12

#4 FL4MUS: -0.11

#5 XANTARES: -0.11

#6 NiKo: -0.09

#7 woxic: -0.08

#8 ICY: -0.08

#9 m0NESY: -0.07

#10 ropz: -0.07