GGPoker traffic: how many players are online

The snapshot

We opened PokerScout's traffic report on 14 October 2026; the page stamps its data "14 October 2026 at 2:55 pm GMT", a Wednesday afternoon. A copy of the page is kept in our files. The columns below are PokerScout's own, and the last one is our division.

GGPoker's 166,761 is 64 percent of the 258,985 players the report counts across all 28 sites. It is also 15.6 percent above its own seven-day average. Among the seven rows above, only CoinPoker (22.8 percent above) sits further from its average; PokerStars is 0.5 percent below, PokerKing 3.0 below, Ignition 2.1 above, and ACR and WPT Global are level. So the reading caught GGPoker on an unusually busy hour, and a number like that should not be quoted as "the" traffic. A weekday afternoon in GMT can sit well above a quiet Sunday night, and this table is a single reading. The second one, below, was taken in the early morning GMT and came in 4.9 percent lower.

The same page carries a 13 October headline about a $5 knockout on GGPoker passing 420,000 entries. Entries accumulate over a registration window, while the report counts people seated at one moment. The entry total and seated-player count cannot be compared directly.

The second reading, 26 October

We opened the same report on Monday 26 October 2026 and kept a copy. The page stamps its data "October 26, 2026 at 5:55 am GMT". This table adds the 24H Peak column, which the 14 October report carried too (GGPoker 4,108 then) and we left out of the first table.

GGPoker's own count is 4.9 percent lower than on 14 October. Two readings at different hours cannot explain the gap. Its share of the report fell from 64 percent to 43.7 because of the HighStakes row. HighStakes showed 2,378 online on 14 October and shows 122,682 now, more than fifty times higher, with seven-day and thirty-day averages of 76,935 and 21,585 that rise with it. We do not know what drives it. Take that one row out and GGPoker holds 65.9 percent of the report, against 64.4 on 14 October.

Some rows do not move at all. ACR shows 6,677 online on both dates, WPT Global 4,613, JackPoker 523 and 1Win 293, twelve days apart and once in the afternoon, once in the early morning. A count of people seated at one moment does not repeat to the last digit four times, so we read those rows as fixed or estimated values. That is our inference; the report does not say. GGPoker's row does move, but its seven-day average of 158,617 sits one player below its Online figure, where on 14 October the average was 15.6 percent under it. PokerStars shows the same one-player gap (37,915 and 37,914). We cannot tell how the averages are produced, so read them with the doubt we raised on 14 October.

The 24H Peak column does not look like a peak either. For GGPoker it reads 5,390 against 158,618 online. In no row is it larger than Online, and it is smaller than the Cash figure in 21 of the 28 rows. The tile above the table gives a separate site-wide peak of 379,653, above the 363,358 online, as a peak should be. We do not know what the column measures.

The third reading, 2 November

We opened the same report on Monday 2 November 2026 and kept a copy. The page stamps its data "November 2, 2026 at 4:55 am GMT", one hour earlier in the day than the 26 October reading and on the same weekday. Both land in the quiet small hours GMT, so this pair sits close enough in time of day to compare week to week, which the Wednesday afternoon reading never could.

GGPoker shows 154,860 online, 2.4 percent below the 158,618 it showed on 26 October. The two Monday readings are an hour apart, so the hour alone could account for a gap that small. We read it as a weaker week. Its share of the whole report is 43.2 percent, close to the 43.7 of 26 October, and for the same reason. HighStakes again fills a second large row, 122,063 online against GGPoker's 154,860. Take that one row out and GGPoker holds 65.4 percent of the field, where it held 65.9 on 26 October and 65.0 on 14 October. That non-HighStakes share has stayed inside one and a half points across three readings, so it is the steadier figure to quote.

The rows we flagged as fixed on 26 October have not moved. ACR shows 6,677 online for a third reading running, WPT Global 4,613 for a third, and 1Win 293 on both Monday readings. A live count of seated players does not land on the same digits three readings apart, so we still read those rows as held or estimated values. GGPoker's own 7-day average reads 154,859, one player below its Online figure again, the same gap as on 26 October. A trailing average that tracks the current count to the last player is not behaving like one, so we lean on the 30-day column instead.

The weekly series so far

Three readings are now on record. This table keeps GGPoker's own columns across the dates, with our share division on the end. The hours are not comparable: the first reading was a Wednesday afternoon and the other two fell early on a Monday.

How to read it

We use the 30-day average to track changes across readings. It has risen at every reading, from 140,576 to 146,841 to 151,383, about 7.7 percent over the nineteen days, while the instantaneous Online figure fell across the same span. A smoothed average can rise while snapshots fall when readings are taken at different and often quieter hours. We cannot see how PokerScout builds the 30-day column, so this holds only if it is a real trailing average and not another derived value.

The Online figure and the 7-day average move together to within one player, so treat them as one number and read it as a snapshot of the hour. Leave the Cash column out of any series, since it is Online times a constant for 27 of 28 sites and carries nothing the Online column does not. The "Share, all 28" column swings from 64 to 43 percent, but that swing is one row, HighStakes, going from a few thousand players in mid-October to more than 120,000. The share with HighStakes removed is the one that has held, inside the band 65.0 to 65.9 percent, so that is the figure to track week to week.

To extend the series, open the same report on the same weekday near the same hour, write down the timestamp the page prints rather than your own clock, and keep Online, the 30-day average and the all-sites total. Each later reading goes under this table with its own stamp, and the earlier rows stay as they were recorded. Website visits to ggpoker.com are a different measure from seated players. Panel tools such as Similarweb estimate visits; this report does not cover them.

The cash column is a formula

The report defines Cash Players as the number "currently playing poker online for real money at ring game tables". For GGPoker that is 10,839, or 57.5 percent of the 18,846 cash players in the report's headline tile. The figure reads like a count of cash-table seats, but it does not behave like one.

We multiplied each site's Online figure by 0.065 and rounded down. The result equals the Cash figure in 27 of the 28 rows, the zero rows included. GGPoker: 166,761 × 0.065 = 10,839.4, shown as 10,839. PokerStars: 37,583 × 0.065 = 2,442.9, shown as 2,442. PartyPoker: 1,059 × 0.065 = 68.8, shown as 68. The one row that breaks the pattern is CoinPoker, at 24 percent. The 26 October reading repeats the result: 27 of 28 rows again, with CoinPoker the exception at 1,422 cash against 2,912 online, 49 percent.

We cannot see how PokerScout builds the column, and the page does not describe it. A count of ring-game seats would not land on one ratio at a site with 166,000 players and at one with 400. Cash appears to be derived from Online for 27 sites. If you have seen PokerScout explain it, tell us and we will correct this section.

Anyone can repeat the check from the table on PokerScout's front page. Take Betfair Poker: 132 online, 132 × 0.065 = 8.58, shown as 8. Upoker: 765 online, 49.7, shown as 49. Tigergaming: 161 online, 10.5, shown as 10. Rounding is always down. That is also why the ratio looks like 6.2 or 5.6 percent on the smallest sites.

The only public cash figure we found for GGPoker is an assumption of roughly 6.5 percent, so nobody can say from it how many people are at ring-game tables right now.

Bot counts and traffic

People who ask about GGPoker traffic are often asking how many of those players could be bots. The snapshot cannot answer that.

A traffic counter counts seated accounts. A script holding a seat adds one to Online exactly like a person does, so bots sit inside the number and cannot be subtracted from it. The choice of denominator changes the headcount by a factor of fifteen. One percent of 166,761 is about 1,668 accounts; one percent of the modelled 10,839 is about 108. A claim such as "5 percent of GGPoker is bots" says nothing until it names the column, the hour and the source of the percentage, and we have not found a primary source for any such percentage. The dated statements the operator did publish are collected in what GGPoker has said publicly about bots.

Where automation sits depends on the format rather than total traffic. Our view of that, with the reasons, is in the format table on the home page, and the detection architecture note describes what the operator can see once an account is seated. Why that same volume feeds the rake and Ocean Rewards economics a bot farm runs on is in GGPoker rake and bots.

What stays open

We still have no breakdown of GGPoker's players by format. Nothing outside the operator separates cash, Spin and tournament seats, and the PokerScout column that appears to do it does not. We also do not know how PokerScout gathers the Online figure itself, whether the 7-day and 30-day averages are built from real readings, or whether every row is a live count at all, since several rows repeated to the digit across our three readings and GGPoker's 7-day average keeps landing one player under its Online figure. The developer FAQ covers the engineering side of why shallow formats are the ones watched first.

Where the figures were read

• PokerScout, Online Poker Traffic Report, pokerscout.com, read 14 October 2026, data stamped 2:55 pm GMT
• PokerScout, news item on the $5 Ocean KO field, 13 October 2026, same site
• PokerScout, Online Poker Traffic Report, pokerscout.com, read 26 October 2026, data stamped 5:55 am GMT
• PokerScout, Online Poker Traffic Report, pokerscout.com, read 2 November 2026, data stamped 4:55 am GMT