A new tool has emerged in Counter-Strike statistics, allowing users to search for data that was previously difficult or even impossible for ordinary users to access. The HLTV Brain AI system is already uncovering dozens of unusual records — from the highest number of knife kills to matches won with an extremely low individual rating.
AI gains access to hidden statistics
The main feature of the new tool is the ability to formulate queries in natural language. Users can ask who has killed the most teammates, how many matches a particular player has won with a rating below a certain value, or who has the most frags with a specific weapon.
The system is currently in beta and supports only English. In addition, there is a limit of 30 queries per hour, so for now, it is more of an experimental tool for exploring unusual statistical categories than a full replacement for a traditional database.
JW dominates with the knife, while Spiidi leads with the Zeus
Some of the most interesting results involve less commonly used weapons. According to the AI, JW has recorded 267 knife kills throughout his career, putting him far ahead of SHiPZ, who ranks second with 161 such frags. JW has also knifed twist four times — the same number of knife kills that Attacker has recorded against aumaN.
In the ranking of Zeus kills, first place reportedly belongs to Spiidi with 115 frags. Another unusual statistic belongs to brutmonster: the system counted 60 Glock kills against fully equipped opponents carrying rifles. Meanwhile, flusha leads the list for HE grenade frags with 870 grenade kills.
Unusual records of top players
Another category of statistics focuses on matches where individual performance strongly contrasted with the team’s result. For example, donk has reportedly won only 12 matches with a rating below 1.00, while losing 44 such matches. ZywOo, on the other hand, has 18 losses in matches where he achieved a rating of 1.50 or higher — according to the system, this is the highest number recorded.
karrigan’s statistics look even stranger. The AI claims that the Dane has won 563 matches while being the statistically worst-performing player on his team. Meanwhile, HEN1 is credited with 384 lost 1v1 clutches throughout his career, while Maka reportedly has 95 teammate kills.
There are even records for disastrous performances
The system also allows users to search for extremely specific statistics. According to its calculations, no player has won more than two matches with a rating of 0.50 or lower. B1ad3, bodyy, Kyivstoner, and byek each have two such victories.
American player Danejoris has reportedly lost a map 108 times while remaining the best player on his team. frgd is credited with a streak of 34 consecutive matches with a rating below 1.00, while nirvana_sf reportedly played three maps without recording a single frag. Among professional players, pyth, Kristou, and poiii each have two such maps.
The results should still be treated with caution
Despite its interesting capabilities, not all answers from the new AI should be trusted blindly. The tool is still in beta, and some extremely specific queries may return incorrect or misinterpreted results.
The most striking example is the system’s claim that SHOCK has killed 168,445 chickens, which looks clearly unrealistic. Therefore, for now, such numbers should be treated as a reason for further verification rather than as automatically confirmed historical records.
A new level of Counter-Strike statistics
The very concept behind HLTV Brain opens up an enormous number of possibilities for statistics enthusiasts. Users can now search not only for standard ratings, K-D, or ADR, but also for highly specific historical statistics that previously required manually processing massive amounts of match data.
If the system’s accuracy improves after beta testing is completed, such queries could reveal an entirely new layer of Counter-Strike statistics. For now, the most interesting part of the experiment is discovering what other strange records are hidden throughout the long history of the professional scene.
