Reports are the one input Valve named in 2017 and never took back. They are also the input players understand least, because nothing happens when you press the button, and something happens weeks later to somebody. This post is about the stretch in between, using the only part of it that is public: the bans that eventually land, in our own data, day by day.
Reports Are on the List, Since 2017
The original Trust Factor announcement described the inputs by example: “the overall amount of time they had spent playing CS:GO, how frequently they were reported for cheating, time spent playing other games on their Steam account”. Frequency of cheating reports is right there, second on the list. The current FAQ no longer lists inputs at all, on the stated grounds that “any list of factors would become outdated very quickly”, but nothing in it retracts the 2017 example.
Notice the wording. “How frequently they were reported” is a count of reports received, not a count of bans. A report is an input the moment it is filed. Whether it later turns into anything is a separate pipeline.
A Report Is Not a Ban: What Sits in Between
We are going to stay inside what is public here, because the internals are not. A report goes to Valve. Some reports, in some categories, feed automated systems; some feed review; VAC itself detects cheat software independently of reports. Which report triggers which review, with what thresholds, is not published, and the popular explanations of it are reverse engineering from forum anecdotes. What is public is the output: a VAC ban or a game ban appears on the profile, with a “days since last ban” counter that lets anyone estimate the date.
That counter is how we see the pipeline from the outside. Read enough profiles, and the dates the bans landed form a series. Most days the series is flat. Some days it is not.
Ban Waves in Our Data
Below is the last 180 days of estimated ban dates across our study pool. The median day in that window has 72 bans in it. The days that stand out have ten times that or more.
Show the numbers
| Day | Bans | × median |
|---|---|---|
| 2026-03-26 | 6,078 | 84 |
| 2026-06-19 | 7,315 | 102 |
| 2026-07-25 | 1,905 | 26 |
| 2026-08-06 | 2,722 | 38 |
| 2026-08-25 | 2,033 | 28 |
The shape is the point. Bans do not arrive as a steady trickle proportional to reports; they arrive in waves, with long quiet stretches between them. In the 180 days on the chart there are 5 days above ten times the median, and the biggest of them alone accounts for more bans than a typical month. If you were reported in March and nothing happened, that tells you nothing about whether something will happen in June.
Two limits of the method travel with that chart. The date is estimated from a counter that measures distance to the most recent ban, so an account banned twice only shows its latest one, and older waves look smaller than they were. And the pool grew from Steam groups and from banned accounts' friend lists, so it over-represents accounts near bans by design. It is good for shape and timing. It is not a count of how many players Valve bans.
Bans Get Reversed: 105 Accounts in Our Pool
One thing we did not expect to find: bans that disappear. Our collector re-reads ban records over time, and as of September 10, 2026 the pool held 105 accounts that had a game ban on one read and no ban on a later one. Steam lifted them. We noticed because our own audit flagged one of them as a data error for three days running, until we looked and found the record had changed on Valve's side, not ours.
105 out of over 1.6 million is a small number, but it is not zero, and it says something about the pipeline: there is a review step somewhere after the ban, and it sometimes decides the other way. If you believe you have been banned by mistake, the route is Steam Support, not a forum thread. Our data says the route occasionally works.
False Reports: What the Community Claims and What Valve Says
A recurring worry is that reports filed against you by angry opponents, or a lobby that mass-reports the enemy after losing, count the same as legitimate ones. The community answer is that Valve weights reports by the reporter's own standing, so spam from low-trust accounts is discounted. That is plausible. It is also unsourced: Valve has never described how reports are weighted, and every statement about it in the guides traces back to other guides. We will call it what it is, a community claim, and leave it at that.
What Valve does say is narrow and worth repeating. The FAQ's advice to a player who thinks their matches are being ruined by Trust Factor is to email the CS2 team with the subject “Trust Factor Feedback”, a description, and a Steam ID. There is no appeal button for reports, no way to see them, and no acknowledgement. If a report campaign is genuinely hurting your matchmaking, that email is the only documented door.
What This Means for Your Own Account
Reports are the one lever that runs both ways, and the one you control most directly: finish your matches, keep the mic civil, do not damage your team. That is not our advice, it is the practical reading of “how frequently they were reported”. Everything else about your account that is public, its age, its company, its ban records and the records around it, you can read yourself. Our Steam Profile Checker reads 74 of those fields and scores them against the study behind the chart above. It does not read reports, because nobody outside Valve can, and it does not read Trust Factor, because that does not exist as a thing to read. It reads what is visible, and it says so.
If your record already has a problem, the practical list is in how to fix a red Trust Factor. For the whole system in one place, start with Steam Trust Factor explained.
