The online poker industry, and the poker-playing public, is at war with bots – that is, computer software that plays poker without human intervention. At first glance, this is a somewhat illogical position. After all, playing without human intervention is not necessarily cheating, because a poker bot plays by the same rules as everybody else. A typical bot doesn’t collude, abuse other players in the chat, or multi-account, but it does keep games running and generates revenue. What’s more, a typical bot is extremely predictable and should therefore be simple for a skilled player to beat. If you didn’t know much about poker, you might think that a bot would be exactly the kind of player that the public would like to play against and that poker businesses would want as a customer.
However, the vast majority of the poker playing public does not want to play against bots. This is partly due to an irrational fear of bots brought on by the success of computer programs at playing chess (IBM’s Deep Blue computer) and backgammon (Snowie), which has led players to believe that sophisticated poker bots are going to win their money by playing perfect, unexploitable poker. As a result every top poker site prohibits bots, not because they are against the rules of the game, but to protect their reputation.
But there is another, more philosophical reason not to allow bots. Some bots are designed by computer programmers as a sort of intellectual challenge – the same sort of motivation that drives programmers into hacking or virus creation. Others are used by people who either can’t resist the urge to cheat, or who are looking for the next ‘get rich quick’ scheme. But neither of these types of people is a poker lover, and no bot is a true participant in the game of poker. A bot does not contribute to the growth or popularisation of the game like a human player might. It doesn’t socialise with other players and create a fun environment like a human player might. A bot doesn’t engage in the game beyond simply taking action – it doesn’t enrich the lives of others or create joy. Every bot which breaks even or wins simply drains money and resources from the online poker community without giving anything back. That’s the real reason not to allow them.
In 2004, I was approached by a businessman who wanted to create an online poker ‘sweatshop’. He proposed that I would develop a simple strategy for beating the game, and he would then hire workers in India at very low cost to play the game using that strategy, with all of the profits going into the business. He didn’t see this as unethical because the players wouldn’t be colluding with each other, and no bots would be used which were against the Terms and Conditions of most sites. Indeed, what he was proposing was totally within the rules. I declined, and no doubt the businessman went on to find someone else to develop his strategy. This was my first encounter with a ‘gold farm’.
The term ‘gold farming’ derives from the popular online role playing game ‘World of Warcraft’, developed by Blizzard Entertainment. The game’s virtual currency, known as ‘gold’, is bought, sold, and traded for real money outside the game. As a result, a similar problem as in online poker arose, with bots designed to play automatically and earn gold which would then be exchanged for real money, and teams of workers in countries like China and India being paid to earn gold from the game. These gold farms could make money because of the extremely low cost of labour and the relatively high price that the in-game currency attracted in the real world.
In online poker, a gold farm consists of a group of workers paid to play online poker. Their accounts are funded by their leader, and the workers are paid a salary for the number of hours they put in, with profits going to the leader. Workers don’t need to be poker players, as they can simply follow a strategy from a chart, or from one of the many advisory programs available to assist players as they play. Workers might assemble in an office, or work from home.
In recent years, gold farming has attracted some mainstream attention. In 2008 and 2009, a notorious gold farm from China made a lot of money playing Seven Card Stud online. Although no cheating was ever proven, the fallout from the case resulted in a lot of negative publicity, a great deal of suspicion, and much resentment from players.
It’s abundantly clear then, that players don’t want to play against gold farms. What’s more, gold farms detract from the game in the exact same way as bots do, by draining money and resources from the community without contributing anything in return. For those reasons, I believe that gold farms should be prohibited.
I can understand the industry’s reluctance to do so, however. After all, it’s generally not a good idea to create a rule which you cannot effectively police or enforce, and detection of gold farms is potentially very difficult. Unlike bots, gold farms interact with the games in exactly the same way as a normal player, because there is a human at the controls. The difference between a human gold farmer and a human poker player is small and difficult to detect.
Automatically analysing the playing style of players and comparing them against known strategies, which is a technique used by online backgammon providers, could potentially identify those players who follow strategy charts, or use an identical strategy to other players – a potential indicator of a gold farm or a bot. However, this method’s reliability is questionable if used in isolation and expensive to implement.
Another alternative is to not spread games that can be played well by following a simple strategy chart. Two obvious examples of such games are short-stacked No Limit Hold’em cash games, and tournaments with very fast betting structures such as Hyper-Turbo Sit & Gos. By not offering such games, you make it harder for gold farms to develop a simple winning strategy.
Also worth considering is prohibiting the use of third-party tools which provide advice about what action to take in every situation. It’s one thing for a tool to tell you what the pot odds are, but it shouldn’t be acceptable for a tool to tell a player that they should ‘raise all-in’. Yet such tools exist and are allowed by some sites.
Lastly, we can make gold-farming more difficult by placing stricter rules on player-to-player transfers, which are often used to fund the accounts in gold farms, and on the sharing of accounts by multiple players.
These are just ideas. It is worth investigating how Blizzard and other developers of online role-playing games deal with the same problem, and of course the best solution will be one that the whole industry adopts together. It’s clear to me that gold farms are just as much of a threat to online poker as bots are, and are just as unwanted by the poker community. It will be a brave but virtuous business who takes a stand against them first.
This article was published in InsidePokerBusiness, Mar-Apr 2011.