Battle the botsGambling against computer players, known as bots, has proved an unexpected hit with online gamers. Robert Blincoe meets a man who is the brains behind the bots
Gambling against bots has a pretty bad image problem. The feeling and fear is that these software grinders are just working the online poker games and winning a rich haul back for their code supremo masters.
Think again my friend. In a legit environment, bots are available to teach you the ropes of a game, and can help you practice. They can be very good players and raise your skill level. And best of all you can even take money off them. They're beatable.
One of the world's top artificial intelligence (AI) experts builds the bots for the major online sports books and casinos – Gala Coral, Rank, Stan James, BoyleSports, Sportingbet Europe, William Hill, BetFred, PaddyPower, BlueSquare, SkyBet, CasinoRip. The money games on offer are PokerDice, Dominoes, Backgammon, BlackJack, and heads up Texas Hold'em and Pineapple Poker.
Dr Martin Smith is head of technology for GameAccount Global, which provides the online games for these companies. He's been a top British chess player and a poker pro who worked the card rooms of Los Angeles before he was legally old enough to. He got himself a PhD in AI, and even developed the AI element of the classic Elixir Studios video game Republic.
For Smith, it comes down to a personal heads-up with everyone who takes on the bots, and they play at levels up to £1,000. Each morning, straight out of bed, he's up and checking no one has carved them up overnight.
You might think “with this guy's background, what chance do I have?'” Plenty. To keep players interested there have to be new games, new variations, new deals. Players are waiting, and looking for the edge. They have an advantage. Smith's bots have to be out there all the time and have to play. Customers can walk away. “My strategies have to work against every player with every conceivable strategy. All they have to do is fine tune their strategy to just beat me,” says Smith.
Entire university departments spend years working on AI projects to play a single game. Smith works in the real world, has to keep developing bots to play several games, and will lose his clients serious money if he gets it wrong. There are players who do make money against his bots, and he has suffered the odd serious-spanking.
At the time of writing his Gin Rummy bots have been getting a pounding. It only takes one £1,000 beating to wipe out the profits of 999 £1 winning games. The dominoes game is also vulnerable to a seriously good player.
In this world all the bots are clearly labelled. Probot is the highly skilled player that will play for the top stakes. He also has little brothers. Smith has created a family, in which Probot sits alongside Rookiebot and Amateurbot, who are programmed to play strongly but to make a statistically modelled number of stupid errors, which mimics the way humans play. You can't play these bots for the top stakes, and the system keeps track of how good a player you are, so you can't just take money off the easy bots.
There’s also Beginnerbot. His job is to play you for free to help you increase your confidence and sharpen your skills. They're good teachers because they play well both strategically and tactically, unlike a good human player who in backgammon, for example, will tend to deviate from best play in order to punish a lesser opponent’s mistakes.
Various bots are beatable at various levels. All the games involve enough skill for people to make money at them. Smith's aware of it, he just makes sure they just don't win large amounts.
Smith's battles with the good human players take place on constantly shifting terrain. Smith releases his latest version of a program. It wins for a couple of weeks. Then the humans figure out how to beat it and they win for a couple of weeks while Smith goes away and works on his software – algorithms, probabilistic calculations, search techniques – and scratches his head. Then he comes back with a new version of the program that wins for a couple of weeks, while the humans go away and think about it. “We have this ratcheting up,” Smith says. “It’s a very intellectually rewarding thing.”
A short while back, a bug in Probot meant it had been doing badly at Hypergammon, a cut-down version of backgammon. The smart humans had been beating it for £500 stakes. Smith figured out his fix and waited to spring it on them at a busy time to earn some money back. He caught them by surprise and won his money back.
“On with the fun and games,” he grins, loving the constantly evolving challenge of the artificial intelligence arms race.
The bots were initially used to help build up enough customers, and get the liquidity flowing on the gaming sites. It turned out that people liked playing them.
Although Smith’s software has turned out to be part of the reason punters like to come and play, this wasn’t the original reason that GameAccount hired him. Instead, he was originally brought in to work out whether rogue players were using backgammon software to beat other customers. They were – it was easy.
It's easy to find backgammon software (Snowie 4, or the free, open source GNU Backgammon) that can match or beat the best human players. It was obvious that people would try a little online hustling, sitting at their PCs with champion software running alongside to make their moves. One gaming site discovered that its top 14 backgammon players were software programs.
If you want to cheat against GameAccount’s bots, Smith says, “Bring it on. I rate my bots as being pretty decent at these games.” Cheating against other punters, however, is another matter. Smith’s team have developed software that detects rogue players who are using software. “We have an automated system which plays over the games and looks for similarities in the play to the way computers play. It’s almost like fingerprinting. Over enough games the only way they could play the moves they have done is with certain bits of software.”
It is not illegal to use this software, but it is against the gaming sites’ terms and conditions. Smith will confiscate winnings before showing an offender the door.
Sometimes a human is so good that Smith thinks they're either cheating or are a bot. A guy called ‘Smellysocks’ has beaten Probot at Naval Wars, the site’s version of battleships. The problem was the game wasn't popular enough to warrant the time Smith would need to spend improving his code. The game is back on the shelf for the time being.
Smellysocks was upset. No only because of the money, but for game players, bots can be the best friends, partners or parents. Smith understands this really well.
Like in Terminator 2, Sarah Connor thought the terminator was the best dad for John Connor because it “wouldn't stop, it would never leave him. It would never hurt him or shout at him or get drunk and hit him or say it was too busy to spend time with him.”
The reason people like to take on his bots is precisely because they do not think and behave as if they were human.
“When you’re beating them they’re not bitching or moaning,” says Smith. “When they’re beating you they’re not boasting. They’ll never stop playing when they’re in front or stop paying when they’re behind. They’ll play fast. They won’t say ‘I've got to go to bed now’. They won’t be saying, ‘I'm just going to make a cup of tea’ and not come back. There’s none of those annoying human frailties.”