Join Bridge Winners
Bots, Funbridge, GIB etc

The real Bridge player sneers at the Bots on BBO. Bots don't play real Bridge. Their bidding is haphazard, their play is weak, they don't signal, their defence is non-existent. And yet, it appears that the Bot games on BBO are wildly popular and profitable.

Funbridge has been mentioned in this forum before. I had never heard of it and paid no attention since there are dozens of Bridge playing sites but a person only needs one and I don't really need any. 

 

The other day somebody from Funbridge announced  a merger with BBO. There were 5 responses to the post, each saying essentially: "so what?"  Without any inside information from either company, I will try to answer that question.

In view of the fact that the people at BBO generally know what they are doing, I googled around a bit in order to figure out what Funbridge is. Once you get past the profuse  self-praise on their web pages you can figure out that this is a site where you pay to play bridge against bots. The bot is a program written by Alexis Maugat and Jérôme Rombaut. I have no idea whether it is any good, whether it is better than GIB or whether it has a future. That's not the point.

GIB is 20 yrs old and was by far the strongest program when Ginsberg stopped working on it. The ACBL not only missed the opportunity to support this development but, in the shape of Al Levy, excluded Ginsberg from the Computer Bridge World Championship. Ginsberg went on to do more interesting and more important things and nobody maintained GIB. 

BBO acquired GIB at some point and it has been available as a seat filler for years. That it could become one of the main attractions is obvious in retrospect but it certainly surprised me. This may be a form of Gresham's law or the McDonald Principle. (Who would have predicted that making the most disgusting, gruesome, cheap hamburgers would make the perpetrators wildly rich and bankrupt every real hamburger joint in the country?) In any case, playing solitaire Bridge is much less stressful than exposing your every error to the 3 critics at the table.   

But now consider this: Suppose we had a bot that played excellent bridge; that bid, played and defended better than the LOL and the FOM (=fat old man) at the club, better than the local tyros at the sectional, better than the class B experts at the regional, and ultimately met Zia's challenge.  Playing against such a bot would be real bridge and could be made into a teaching tool equivalent to modern chess software. To develop such software is now definitely within range and will inevitably happen. What you-all need to do is make sure that the effect of such a development is positive and doesn't exhaust itself in a mad scramble to make money. The current boom in Scholastic Chess should be an indication of what can happen. The computer makes it possible  to become good fast, since accurate information becomes instantly accessible. And it makes it possible to degenerate as rapidly when fiction becomes indistinguishable from fact.

An optimistic view of the current situation is that conceivably the profit made on the games against the imperfect bots will finance the development of superior bots, which will lead to a general revival of the game.                              

24 Comments
Getting Comments... loading...
.

Bottom Home Top