The World Computer Bridge Championships

The World Computer Bridge Championships

By Alex J. Coyne © 2018 Great Bridge Links

The World Computer Bridge Championships is the only competition in the world of its type, where bridge playing AI bots are put against each other to find the best bridge software of the bunch. Bridge software has made huge leaps in recent decades, and we’re excited to see what the entrants for 2018 have to offer.

Bridge and AI

Artificial Intelligence (AI) is behind what makes bridge robots be able to play real players in the first place. It is, in essence, the same technology that drove Deep Blue to beat Kasparov in 1996/1997 and the same technology that drives Facebook’s facial recognition technology, or Google’s searches. For more on the AI behind games like chess, Go and bridge, read Bridging the AI Gap from Great Bridge Links.

The 22nd Annual World Computer Bridge Championships

The 22nd Annual Robot Bridge Championships will be taking place from September 29th to October 4th, 2018. Most of the contestants players will know from past competitions, though there are two new participants to the championship that’ll make you look.

The first is Meadowlark Bridge, created by Rodney Lugwig from Meadowlark Software; the second is the TCS (Tata Consultancy Services) BridgeBot, developed by Prasad Ramesh Bokil. There is little available information about TCS BridgeBot out there, though we can say Tata Consulting Services is an international information technology branch of the larger Tata group. It’s going to be interesting to see how it competes with other well-known bridge bots like Bridge Baron and WBridge 5.

The competitors for the 22nd World Computer Bridge Championships (2018) are listed below. You can find all their links and information on our Bridge Software page.

  • WBridge 5
  • Shark Bridge
  • MicroBridge
  • Q-Plus Bridge
  • Bridge Baron
  • RoboBridge
  • Synrey Bridge
  • Meadowlark Bridge
  • TCS BridgeBot

Previous champions of the competition for the last five years have been Jack (2013); Shark Bridge (2014); Jack (2015); WBridge5 (2016) and WBridge5 (2017).

How does it work?

We asked tournament founder and organizer Al Levy to update us on how the tournament works from a technical point of view. Here is his update:

A bridge “table” consists of a central server, or Table Manager (TM), that distributes the cards to four connected identical computers, each of which contains a bridge-robot. Before a match begins the operators enter into their robot’s memory pertinent information about their opponent’s system. During the bidding, the meaning of certain alertable bids are manually input. Play then proceeds automatically with the TM receiving and passing information from and to the robots, as well as recording the play. This year Intel Core 3.4GHz/8GB Ram desktop computers will be used. The elapsed computer time averages approximately 45 seconds per pair per deal, well under that of human play.

The First Champions

The very first World Computer Bridge Championship was held at the ACBL North American Bridge Championship (1997) in Albuquerque, New Mexico. You can still find the championship’s information results online at their website. They were sponsored by Baron Barclay Bridge Supplies, the world’s biggest online bridge supply store. It’s come a really long way since then. Oh, and who won? Bridge Baron was the very first bridge bot champion.

Bridge Robots

One of the greatest things about Bridge AI is that we can now pick up our phone or tablet at any time and play hands against robots. It’s great practice and has removed barriers to learning for many people. And not just that – the robots are getting better and better! As evidence, you can watch matches between expert player Milan Macura and the FunBridge robots on their blog. I’ve been playing the Series Tournaments again the Funbridge robots for a few months now – and I’d say almost 80% of the time, I’m ok with my partner’s play. My only complaint is, the defenders never make a mistake – it’s hard to fool them and they never lose focus.

More About Bridge AI

AI has made great leaps in game such as chess; it can even, these days, organise battlefield tactics in combat-themed games. AI technology powers a lot of what’s on the internet – and isn’t – including face recognition technology, games and search engines.

But bridge still offers a massive challenge for programming artificial intelligence. Want to find out more about the AI behind bridge bots? Here are some articles, studies and resources to keep you busy.

AI Magazine, 1998

NYU Computer Science, 2009

Matthew L. Ginsberg

X Mei, 2012

CY Ho, 2015

CK Yeh, 2016

V Ventos, 2017

Al Levy, 2015, 2016, 2017