Poker Cheating: Bots Usage by Poker Site!

DoingPoker using bots to fill up the tables? The evidence seems overwhelming.

A routine observation of the traffic at DoingPoker revealed some unusual activity. After further investigation, it seems likely that the site uses robotic players, or bots, to fill up the games and create the illusion of an active player base.

It should be noted that this conclusion is an inference based solely on outside observation. It cannot be proven to a certainty. DoingPoker’s management claims that the players are human testers. Some of the evidence that led to the conclusion that they are bots is presented here, and readers are invited to investigate for themselves and make their own judgments.

Not Your Typical Poker Site

The first remarkable thing about DoingPoker is the number of players. Every online poker site has a daily cycle, with traffic building throughout the day, peaking in the evening, and bottoming out in the early morning hours. Every poker site, that is, except DoingPoker.

The site averages about 90 real money ring game players, but what is unusual is that the site has the same number of players at all times of the day and night. Over the span of 48 hours, observations at many different times showed that DoingPoker always had between 90 and 95 real money ring game players. (At the time of this writing the table arrangement had been changed slightly and player counts were consistently in the mid-80s around the clock). This fact by itself is enough to indicate that something highly unusual is gong on, but it is certainly not the end of the story.

The site offers five different poker variations: hold’em, omaha, omaha high/low, stud and stud high/low. At most sites of a similar size, players naturally gravitate toward hold’em, the most popular game, and the other games will run only sporadically, if at all. DoingPoker, however, has at least one active game of every type around the clock. Such a high level of game diversity is unheard of for such a small site.

The Players

Taking omaha high/low as an example, at any time of day, an eight-seat table with two seats open can be found. At one point in time, six particular players were observed playing at the table. Two hours later, the same six players were still playing at the table, but every single one of them had changed seats. Twelve hours later, the same six players were still at the table, once again in different seats. In the following days, the same six players were always observed at the table.

The player names also show patterns suggestive of a common origin. Among the relatively small group of players are “sasha-us,” “sasha-00” and “SUMO-00.” Also present are “jangsangv” and “jangsang-.” Many of the names end with a period and/or contain a hyphen, as if management were trying to create unusual character strings that wouldn’t interfere with the name choices available to their actual customers.

The Play

The players at DoingPoker often seem to make decisions that no rational human player would make. This applies not only to the betting strategies, but even to the way the players enter and leave the games.

One player in a limit hold’em game was observed to leave the table after playing a hand in early position, only to return within a minute and post a blind immediately in front of the big blind. After playing through both blinds, the player left and immediately returned again, posting a blind again when the button had passed. The player played every hand, but instead of staying seated and posting just two blinds, small and large, the player posted three big blinds and a small blind.

Many other occasions were witnessed in which a player left and then returned within a minute, for no apparent reason related to bankroll management or otherwise. The maneuver, repeated at random intervals, did serve the purpose of ensuring that an observing human player would have regular opportunities to grab an open seat at an otherwise full table.

One might expect robotic players to have a tight, value-oriented style of play in which percentages are carefully calculated, but that does not seem to be the case. Based on limited observation, the players play loose and aggressive, happily check-raising and bluffing each other while waiting for human competition.

In one 10-handed limit hold’em hand, at least five players saw the turn. A third club hit the board on the turn, making a flush possible. The pot was then bet, raised and reraised by the first three players. In all, five players ended up calling four bets to see the river. A blank hit, and the turn bettor, raiser and reraiser all checked. The pot was bet and raised behind them, and the turn bettor, raiser and reraiser all folded to two bets on the river. The river bettor called the raise and then mucked, because the raiser showed a monster hand -- top two pair.

In another limit hold’em hand, four players called two bets on the turn. The board was 10-9-9-5 with three clubs, making a flush and full house possible. On the river, an offsuit king also made a straight possible, but otherwise made no difference. The pot was then bet, raised and reraised on the river by three different players. The bettor mucked and the raiser called the third bet. The reraiser showed an unimproved pocket pair of sevens, and the raiser won, not with a full house, flush, straight or even three of a kind, but with J-10, for two pair.

In no limit hold'em, the players can be observed making seemingly endless strings of minimum raises. Players are seen to raise and reraise, putting in more than five bets, only to fold to one or two more bets. No bets or raises of any size other than the minimum were observed.

The Denial

DoingPoker does not warn its players that they are playing against bots. The website makes no mention of them. The website contains the following quotes:

Responsible Gaming: DoingPoker, acts solely as host, insuring Fair and Honest games.

Action: DoingPoker has non-stop poker action 24-hours a day against real people from all around the world.

The site’s terms and conditions specifically forbid players from using any kind of artificial assistance, including bots.

The site’s management seems to make some effort to disguise the existence of the bots. Even at a table with seats open, the robot players will regularly get up from the table and change seats. Unsuspecting human players who do not pay close attention to the names might not realize that they are facing the same cast of characters they faced two hours or two days ago.

In an apparent effort to maintain the illusion of normalcy, some tables are kept full while others have seats kept open. At the full tables, departing players will be replaced from time to time by new players. And yet despite all the turnover, the total number of seated players hardly changes at all.

The site’s support staff, when contacted, denied only that bots are used in real money games.

DoingPoker’s management was asked to comment on this story. In response, Carlos Miranda, General Manager of Dagoweb S.A., sent the following statement:

DoingPoker.com has not been officially launched yet. We are still testing our software and for that purpose we have both modes available, play money and real cash. We have a testing facility that gives jobs to a lot of people and those we use, for playing money mode and also real cash games as well. We also use our own programmers in India, programmers in China and friends to help us test our software. Most of the people we have playing are not that much of poker savvy, but we get great feed back. DoingPoker.com does not have any kind of Bots as mentioned in your email to fill in spaces.

It should be noted that the site is not only open for testing. According to the website, real money players and deposits are being accepted. There is no indication on the website that the site is in a testing phase. The only indication of testing is in the site's downloadable client software, where the word "beta" appears below the name of the site.

A Word Of Caution

Given the reckless playing style of the bots (or human testers), DoingPoker might seem like a poker player’s dream. However, it is unknown whether the bots play the same style against human opponents. No evidence for collusion among the bots has been seen, but In a game where all but one player are house robots, the possibility for collusion is obvious. If the bots are sharing their hole card information, a whole array of cheating options is available. If the bots also know the human player’s hole cards, the hapless human doesn’t stand a chance. This is what allegedly happened at the now-defunct ProPoker.com, which went out of business when players began to suspect they were being cheated.

Pokertropolis also allegedly used bots to prop up the games, though without necessarily cheating the players. The only fraud alleged in that case was management’s repeated denials that bots were used in the games. Pokertropolis is now also defunct.

Some players may consider the use of bots to be a harmless extension of the concept of proposition players, or props. Most players have no problem with playing against human props, but not everyone is comfortable playing against robots, especially when those robots are programmed and managed by the house.

Even if the bots are not cheating and are easy to beat, there is another thing to consider. A site which has almost no real players also has very little cash flow, and may have difficulty paying the players who do win. This is especially true when any money won on the site is likely to have been lost by the site’s robot players. In other words, unlike every other poker site, at DoingPoker the players are playing against the house, giving the house even more incentive not to pay out. There has been at least one unsubstantiated claim that the site refused to pay out winnings to a player.

As stated above, the existence or non-existence of bots cannot be conclusively proven by observation. Players are urged to do their own investigation and to use caution when deciding to play at a new site.