Dota 2 Player Count Hits 1.1 Million: What Changed

The Numbers: What Steam Recorded

Start with the headline figures, because they come from two trackers that disagree by a rounding error. SteamCharts shows a 30-day peak of 1,096,364, reached on Sunday, October 25 at about 14:00 UTC. SteamDB, which GosuGamers quoted in its report, put the same day at 1,097,854. Both trackers sample Steam’s public player counter, so a difference of about 1,500 is expected.

The weekend before the record, Saturday, October 24, peaked at 1,080,756. Then the part that matters: Monday, October 26 peaked at 1,039,289 and Tuesday, October 27 at 1,069,502. In the daily data SteamCharts gives us for the window starting September 14, those are the first weekdays above one million.

For scale, the same counter showed 473,708 players at roughly 05:00 UTC this morning, October 28. That is the quiet part of the cycle, and we will come back to why a single “player count” number depends so much on the hour you check it.

Weekend Peaks, Week by Week

The cleanest way to see the trend is to compare weekends, because weekends are when the game peaks and the comparison is like for like. The treasure launched on a Wednesday, so the weekend of October 3-4 is the last “before” reading.

Look at the daily average column, not only the peak. On Sunday, October 4, the average across the day was 641,381 players. On Sunday, October 25 it was 836,747. That is a 30.5% rise in the average against a 23.2% rise in the peak. The curve did not just get taller at the top, it got fatter all day.

The first jump came the weekend of October 10-11, three days after the treasure. The second weekend dipped slightly on Saturday (969,198) and still cleared a million on Sunday. The third weekend, October 24-25, is where it broke away.

The Weekday Proof: Tuesday to Tuesday

Weekend spikes can be explained away by a good weather forecast or a school holiday. Weekdays are harder to fake, so we lined up five consecutive Tuesdays, one of them before the treasure launched.

Between September 29 and October 27 the Tuesday peak rose 33.5%, the Tuesday average rose 50.8% and the lowest hour of the day rose 62.5% (from 281,828 to 457,853). The floor lifting faster than the ceiling is the single most interesting detail in the whole data set, and it is central to the bot argument later in this post.

Peak vs Average: Why 1.1M Is One Hour

It is easy to read “1.1 million players” as “1.1 million people queue for Dota every day”. They do not. The peak is the single busiest sampled hour. In the last 30 days SteamCharts lists an average of 653,654 concurrent players against that 1,096,364 peak.

That average is still significant. It is up 14.4% on September’s average of 571,294 and above August’s 617,078. Here is the longer monthly record, using the same source.

Two things stand out. First, September 2025 had a 953,869 peak, which tells you the game was already healthy around that time of year before this surge. Second, the “Last 30 Days” row is not a calendar month, so treat it as a rolling window: it covers roughly September 28 to October 27 and includes both the pre-treasure and post-treasure weeks.

Also remember what Steam’s counter measures. It tracks accounts that currently have the game running. That includes people sitting in the menu, in a party lobby, in a custom game or in the Arcade, not only players inside a ranked match. A million “players” is therefore not a million ranked games in progress.

Is It Really a Record?

No, and the sources disagree on the wording, so here is the careful version. The all-time peak is from March 2016: 1,291,328 in the SteamCharts monthly table, and 1,295,114 according to GosuGamers citing SteamDB. Neither number is close to being threatened.

The more common headline is “highest since June 2016”. GosuGamers uses that framing. Our check against the SteamCharts monthly table finds a wrinkle: the monthly peaks recorded after June 2016 include 1,117,519 in August 2016 and 1,141,191 in October 2016, both higher than this week’s 1,096,364. The two trackers sample at different moments, so the exact “since when” depends on which one you trust.

The safe, accurate version is this: Dota 2 has not posted a peak this high in roughly a decade. The 30-day average of 653,654 is also the highest rolling average in the SteamCharts monthly table since about August 2016, when the monthly average was 666,013.

For the other times the game crossed a million in the modern era, the same table shows 1,033,925 in March 2019 (the month the hero Mars launched, per GosuGamers) and 1,038,848 in October 2022 (the “Swag Bag” release with a free Arcana). Both of those were short event spikes in months whose averages stayed far lower, 586,506 and 456,303. The current run is different because three weekends in a row crossed the line and the weekday floor moved too.

Why It Happened: Trove of Terror

Valve has not said anything about the player count, so the cause is an inference. But the timing is hard to ignore. On October 7, 2026 Valve released the Trove of Terror treasure. According to GosuGamers’ coverage, it is the first brand-new Arcana for a hero in more than two years: Tidehunter’s “Maw of the Megalodon”.

The contents, as reported by GosuGamers:

• Two Arcanas: the new Tidehunter Arcana, and a rerun of Drow Ranger’s 2021 Arcana.
• 15 Immortal-rarity sets, including a free Persona (Anti-Mage “Daemon Hunter”).
• Chest milestones: a Mythical Tormentor skin after opening 12 chests and a Legendary Roshan set (“R’shan the Eternal”, with its own courier) after 40.
• Tidehunter Arcana styles: Maw of the Megalodon (blue-toned shark), The Hammer of Maelrawn (brown hammerhead) and Unfathomable Hunger (green-striped).

Our full breakdown of what is inside, and whether it is worth buying, is in our Trove of Terror guide. For this post the point is the sequence: the treasure arrives October 7, the first million-player weekend follows on October 10-11, and each weekend since has stayed at or near that level.

Correlation is not proof. School calendars, other events and plain word of mouth could all contribute, and we cannot separate those from the treasure with public data. What we can say is that nothing else Valve shipped in this window looks like a candidate. The official patch list ends at 7.41f, released September 15, and the latest item in Valve’s Steam news feed is the October 7 Trove of Terror post.

The likeliest read: players are returning for cosmetics, a famous hero and a seasonal event, not for a balance patch. That matters, because the gameplay itself has been standing still for a long time.

The Patch Drought Behind It

GosuGamers points out that the game has gone more than 210 days without a new gameplay patch. We checked that against Valve’s own patch datafeed, counting only numbered releases (7.32, 7.33 and so on) and treating the lettered patches as hotfixes inside a cycle. The 7.41 cycle began on March 24, 2026. As of this morning it is 217 days old, with lettered patches 7.41a through 7.41f in between (7.41f was the last, on September 15).

Here is how that compares with the longest numbered-patch cycles in the datafeed, which runs back to 7.08 in February 2018.

The 7.41 cycle is already the second longest on record in that data, behind only the 7.32 cycle at 240 days. If 7.42 has not shipped by about November 19, 2026, it passes that record. We found no announced date. Our own read of Valve’s official sources this morning matches OOPS DOTA’s check from October 25: the patch notes list ends at 7.41f and the Steam news feed has no 7.42 announcement.

That is an unusual backdrop for a record player count. It also sets up the most interesting question of the winter, which we cover at the end: what happens to these numbers when a real patch finally lands?

The Bot Debate

Whenever a number jumps this fast, someone asks whether it is real. GosuGamers reports that the question is already live in the community: some players on Reddit complain about “an abundance of bots”, particularly in Unranked matches, and others claim bot accounts are used in the Dota 2 Arcade to push custom games up the “Most Played” list. Sentiment is mixed, with other players defending the surge as genuine and noticing shorter queues.

Here is what we can responsibly say.

What the data can show

• The daily rhythm is intact. Averaged over October 19-27, the count falls to about 540,000 at 02:00 UTC and climbs to about 985,000 at 15:00 UTC. That swing is what you would expect from players in many time zones going to bed and waking up.
• The rise is broad. Weekends, Tuesdays and the daily average all moved, not only a single hour.

What the data cannot show

• Whether the extra accounts are human. The lowest hour of the day rose faster (+62.5%) than the peak (+33.5%) between the two Tuesdays. A wave of new and returning players across all time zones would do that. So would a population of accounts that stay online around the clock. The Steam counter cannot tell those apart.
• Which game mode the extra accounts are in. Steam counts “in game”, not “in a ranked match”.
• Anything about Valve’s detection. We found no official Valve statement on bots or on the player count as of this morning.

Our position as a Dota 2-only service: treat the bot claims as unproven in both directions. If bots are a real share of the growth, they matter most in Unranked and the Arcade, which is where the complaints are concentrated. Ranked is gated: GosuGamers notes new players must meet playtime requirements before they can queue it. That is a reason to think the ranked pool is harder to inflate than the headline number.

When Dota 2 Is Busiest

Because we have hourly data, we can show the shape of an ordinary day. This table averages every hourly reading from October 19 through October 27 (nine days, zero-readings excluded). Dubai time is UTC+4.

The busiest window is 14:00 to 16:00 UTC, which is 18:00 to 20:00 in Dubai, when about 980,000 accounts are online on an ordinary day in this period. The quietest is the early hours UTC, with a trough near 540,000. The ratio between the best and worst hour is about 1.8 to 1.

This is global data. It tells you when the most accounts are online worldwide, not when your regional queue is shortest. A European player and an Asian player will not have the same best hour. Use it as a rough guide, then check your own queue times at different hours for a week.

What It Means for Your Ranked Games

Now the part players actually care about. Be skeptical of anyone who gives you a confident answer here, because we have no public queue-time or match-quality data from Valve. What follows is reasoning from the facts above, not measured results.

Queue times may be better

GosuGamers notes that some players report shorter queue times. A bigger pool generally helps matchmaking find games faster, especially at off-peak hours and in thinner regions. If you queue at awkward hours, you are the player most likely to feel it.

Match quality is the open question

The same GosuGamers report records the opposite complaint: players saying match quality is “getting worse” and blaming an influx of new or returning players. That is plausible. Returning players often start with rusty mechanics and a view of the game that predates years of patches, and a cosmetic-driven surge could bring in players who are less invested in climbing. Whether that shows up in your bracket depends on your MMR and your region.

If you are in the lower and middle brackets, expect more variance, not more skill gap. The pool has more people who are rusty, new or experimenting. If you want to understand where your rank sits in the whole population, our rank percentile breakdown by region is the place to start.

Three practical adjustments

1. Narrow your hero pool. When teammates are less predictable, heroes you know cold carry more of the game. We cover how big your pool should be in our hero pool guide.
2. Pick for the current patch. The meta has been frozen on 7.41f for six weeks, so the data is stable. Our 7.41f hero tier list shows which heroes are over and under-performing.
3. Know the account-side tools. If smurfs or a penalty queue are part of your problem, read our explainer on smurfing in Dota 2 and, if you are stuck in a penalty queue, our low priority removal service.

If you are tired of fighting the variance yourself, that is what our MMR boost and coaching teams exist for. For the long version of what a boost involves and how safe it is, read the MMR boosting guide and our honest is Dota 2 boosting safe breakdown, or compare providers in our best Dota 2 boosting services review.

What to Watch Next

Three things will tell us whether this is a lasting recovery or a seasonal bump.

• The next two Sundays. If the Sunday peaks hold near or above one million without a new content drop, the rise is not just novelty. A fall back toward 900,000 would suggest the treasure effect is fading.
• The weekday floor. The 457,853 low on Tuesday is the number to track. A floor that stays above 450,000 means the all-day population really grew.
• Patch 7.42. The longest cycle on record ends only when Valve ships a numbered patch. A major balance update could pull lapsed players back and push the curve higher, or it could do little if the audience arrived for cosmetics. Nobody knows yet. We will refresh this post when the data is in.

In the meantime, the PARI Universe playoffs, which started on October 28, are the main competitive event on the calendar. Our PARI Universe playoffs preview has the full draft data from the group stage.

If you want to check the numbers yourself, the primary sources are SteamCharts for Dota 2, SteamDB’s Dota 2 chart and Valve’s official patch notes list. The original news reports we cited are GosuGamers’ pieces on the 1.1 million peak and the Trove of Terror treasure.

Frequently Asked Questions