They decide it by breaking every duel, trade, and clutch into weighted sub-scores that get averaged against a fixed baseline, not by comparing raw kill counts. HLTV's Rating 3.0, introduced for Counter-Strike on August 20, 2025 per the site's own release notes, blends six separate components — kills, damage, survival, KAST, multi-kills, and a new metric called Round Swing — into one figure built to hover around 1.00 for an average performance. VLR.gg's Valorant rating, built independently by a community statistician known as FerahgoTheGreat and explained on the site in December 2022, runs a parallel but distinct formula. Both exist to answer the same fan argument — who actually carried that match — with math instead of vibes.
That argument is older than either site. Box scores have always undersold role players and oversold star fraggers, and competitive gaming inherited the problem the moment tournaments started publishing stat sheets. Rating systems were built to correct for it — to reward the player who traded a teammate's death or won a 3-on-3 clutch as much as the one who padded a scoreboard against a losing team's economy round. Whether they succeed is the more interesting question, and it's one both HLTV and VLR.gg keep revising their own answers to.
How does HLTV's Rating actually work in Counter-Strike?
Rating 3.0 starts from the same instinct as its predecessors — reward impact, not just kills — but per HLTV's own August 2025 announcement, it adds context that earlier versions never had. The headline addition is Round Swing, which measures how much a single kill shifts a team's win probability given the map, the side, the economy, the players remaining, and whether the bomb is planted. A kill that flips a 3-on-3 into a 3-on-2 counts for more than the same kill in a lopsided 5-on-2, because it demonstrably mattered more to the outcome.
The second major addition is economy adjustment. Per the same HLTV release notes, every kill in Rating 3.0 is now weighted against the value of the gun and gear the victim was carrying — a rifle duel and a pistol-round pickoff no longer count the same, which was one of the loudest complaints about Rating 2.1. HLTV's own numbers illustrated the shift on release: aggressive riflers who padded stats against low-buy rounds saw their ratings dip, while support players who traded teammates and survived clutch rounds saw theirs climb.
The system didn't stay static. Per HLTV's October 29, 2025 update notes, the site walked back some of Round Swing's initial weight after finding it let clutches and highlight-reel plays overshadow the kind of consistent fragging that actually wins maps. HLTV said the fix restored roughly a 60-40 split between what it calls output metrics — kills and damage — and cost metrics like KAST and survival, specifically so that, in the site's words, a player has to be strong in both output and impact rather than coasting on one.
How is VLR.gg's rating built for Valorant, and how does it differ?
VLR.gg's rating answers a different design brief. Per the site's own explainer, published in December 2022, the formula combines five weighted components: kill contribution, death contribution, assists per round, adjusted damage per round (damage with kill-related damage stripped out, so it doesn't double-count with kills), and a survival rating tied to how much gear was in play when a round ended.
The distinguishing idea, per that same explainer, is that kill and death contributions are not flat values. A kill is weighted by how many players remained on each side when it happened, whether it came from a trade within roughly three seconds of a teammate's death, and the economic gap between the two teams at that moment — the same “not all kills are equal” logic HLTV eventually built into Round Swing, arrived at through an independent formula years earlier. VLR.gg's system was designed by a single community statistician rather than an organization's staff, and the site has continued to host open forum threads where readers ask how specific figures were produced — a level of public math-showing that HLTV's proprietary formula doesn't offer in the same way.
Why do these rating systems keep changing?
Because the games underneath them keep changing, and because both sites have openly treated their formulas as works in progress rather than fixed instruments. HLTV rebuilt its rating twice in under a decade — Rating 2.0 arrived in 2017, Rating 2.1 recalibrated the formula's average back to 1.00 in October 2024 after CS2's round format and damage-assist rules had quietly drifted it to 1.06, and Rating 3.0 followed less than a year later with the economy and Round Swing overhaul. Then came the October 2025 correction, all per HLTV's own release history.
That pattern isn't a flaw fans should read as instability. A rating system is a model of what winning rounds actually requires, and when a publisher patches core mechanics — new weapon costs, new round structures, new agent abilities — the old model stops describing the new game accurately. VLR.gg's formula has stayed comparatively stable since 2022, but Valorant's own competitive meta and Riot's patch cadence mean fans should expect the same kind of recalibration eventually, not treat either number as a permanent scientific constant.
What should fans keep in mind when reading a rating number?
- A rating is a summary of what happened in the rounds played, not a prediction of what will happen next — neither HLTV nor VLR.gg presents it as a forecasting tool.
- The number is calibrated to a specific era of a specific game. Comparing a Rating 3.0 figure to an old Rating 2.0 number, or a 2023 VLR rating to a 2026 one, means comparing two different rulers.
- Both formulas reward role-appropriate play over raw stat-padding by design, so a support player's lower kill count and solid rating aren't a contradiction — that's the systems working as intended.
- Neither site claims its formula is beyond debate. HLTV has revised its own model twice inside fourteen months, and VLR.gg's forums are full of the site's own users asking it to explain itself — a healthier sign than a number nobody questions.
None of this makes the debate over who actually carried a series go away — it just gives fans a sourced, transparent starting point instead of a scoreboard alone. Academia Hagi is an online publication covering the competitive scenes these numbers describe, not a broadcaster, a tournament organizer, or the operator of either rating system.
For a related reviews perspective, read What is the talk show Ari Global on Gossip Stone all about?.
For more context, read What is the talk show Ari Global on Gossip Stone all about?.