I think the question was raised because I guess someone put a “No Computers Allowed” in a TLA and people reading it were confused about what they meant (understandable if no computer has actually played in a human tournament in almost 30 years).
Computer members were assigned the last name “Compx” in the ratings database. The last active computer membership (National Master B P Compx) expired in 2001.
Going by expiration dates, it looks like the last computer to play a rated tournament game against a human opponent might have been Spector, which played in my 4th Concord Open in June 1995. B P played its last rated game in March 1995. B P was much higher rated, though.
http://www.uschess.org/msa/MbrDtlTnmtHst.php?12516851
http://www.uschess.org/msa/MbrDtlTnmtHst.php?12490909
Non-existent like button pressed. I was going to look at that tonight, and now I don’t have to.
It might be time to retire the ‘Computers Allowed’ or ‘Computers Not Allowed’ designations, since it appears no computer has appeared in a rated event in around two decades.
Yes – and also because it is no longer clear (if it ever was) just what a “computer” is. An experimental project, provided that its creator is the one entering the “computer”? Or just any hand-held device that can play a respectable game of chess? Etc.
Plus, it is pretty clear, by now, that computers are not permitted to enter tournaments, period.
Bill Smythe
Computers were also allowed as house players to avoid giving players a bye, but it had to be a program that had a published rating, and these days I don’t know how many do, at least at the time controls in common usage.
Maybe it’s time to strike all the language about computers and just replace it with a simple statement that computers aren’t allowed, period?
Computers have evolved to the point where we don’t really need (or want) them in our tournaments. Some people might find that an ominous development in AI.
+1
Bill Smythe
If the owner of a computer program that met the existing membership and other requirements showed up at one if my events, I would let it play.
But I am more likely to beat Hikaru Nakamura in bullet than an owner of a computer program that meets requirements is likely to show up at anyone’s event.
Just because there is an unlikelihood that a certified computer program will enter one of your tournaments is not enough reason to abandon the “no computers” designation for organizers to use. If AlphaZero or some other AI program were to appear at an event with all of the i’s dotted and the t’s crossed, many of the other participants would likely withdraw, costing the organizer thousands of dollars. Remember that these devices involve more than an operator and a keyboard but also connection to a off site supercomputer and handlers who can tweak a game in progress. As a TD, you have no way of knowing what is going on off site. Let’s just keep that little “NC” available for use rather than give the AI rats any ideas.
If we’re going to do something like that, I’d rather establish “NC” as the default, and have a different designation, such as “CA”, for computers allowed.
A long time ago, “NS” was used for no smoking, but now that’s the default. I’m not even sure if there is an abbreviation, such as “S” or “SP”, for smoking permitted.
Bill Smythe
I would like to think that if some AI company got there machine certified for play and they obtained a rating, that fact would become public knowledge before they showed up at any tournaments.
NC is the default now. I used to say “NS NC W” at the end of my TLAs, and at some point the US Chess office started stripping off the “NS NC”.
Not that I know much about it, but it seems to me that back in the day, computer programs entering tournaments may have been the only way for the developers to get an idea how strong their creations actually were. That’s no longer the case. I see no reason why computers should be allowed in US Chess rated events these days. I suppose you could make an exception to allow them as house players to avoid a bye, but that would only be against the lowest rated human player who would have received the bye, and you’d still need an operator to translate the computer moves to a real board. I just don’t see the point.
I think a much more important reason has always been so that chess players who are thinking of buying a computer chess program have an objective way of knowing how good the program is. The two cheapest chess computers that are for sale in the store attached to this site have USCF ratings of over 1600. Another more expensive one has a rating of over 2200. I’d be willing to pay the extra money needed to buy the 2200 computer, because I wouldn’t consider the cheaper models to be challenging enough to allow me to improve as a player.
Are the still more expensive sets better than these? It makes sense that they should be, but is a more expensive product always better than a cheaper one? One of the reasons Consumer Reports magazine has stayed in business is because the answer to this question isn’t always yes. I can’t imagine why someone who creates a chess computer wouldn’t want it rated (so they could advertise the rating), or why anyone who is considering buying a chess computer (particularly an expensive one) wouldn’t want to know its rating before deciding whether to buy it. So allowing computers to play in USCF-rated tournaments is a benefit both to manufacturers and to players.
Bob
With many devices, including computers, manufacturers often produce a number of product lines, usually with overlapping price ranges. So the high end of line A might be a little higher priced than the lower end of line B, etc. And since they’re often sold through different channels, you may not see those in the same place.
Without taking something apart, something stores frown upon, it’s difficult to determine what the inner components of something are, much less their properties and quality.
Manufacturers have computer tournaments to test their devices in. They can get in more official games for a computer rating. With the number of programs available, one can test his program or device against Stockfish, Houdini, Komodo, Shredder, or a number of other programs to find out how it stacks up. However, if he wants to drop down a six figure entry fee, we will gladly allow the device to play. Might even buy it a silicon sandwich.
Have often wondered why computer programs aren’t given a five digit rating, not only to distinguish them from human players but also to allow an opportunity to more finely compare their performances with each other.
I don’t think five digit ratings would be more accurate or precise, even for computers.