Luke skywalker has done it again.

[Terry McCracken]

http://chessbase.com/newsdetail.asp?newsid=8047

King's gambit "weakly solved" by Vas. Admit it. This guy definately knows what he does

Solved is the wrong word. Maybe the best move the computer could come up with is more like it.

This whole thing comes down to the validity of the assumption that if Rybka scores 5.12 or more, it is a win with 99.99999999% certainty and that it follows that this happens to be the same certainty for the entire results.

I really have a difficult time with both those assumptions and I want to know how he came up with that value. This is about 1 in 10 billion positions! That means if you sacrifice a queen, or a rook plus a pawn or two that you automatically lose (except once every 10,000,000,000 times.)

I have to say that I think this is utter nonsense, my years of experience in computer chess and other games tells me that no matter what the "score" reported by the program (other than Mate) there are holes in the knowledge and search that can make this go wrong.

The number seems to be a poetic license picked in a conversation with a journalist, rather than a number you write in a paper. Like a teen ager saying I am super-hyper-duper-sure, so I am 99.9999.... etc.

The bottom line is, there is a 100% probabilities to reach a position +5.12 in rybka scale for black after 1.e4 e5 2. f4 exf4 3. Nf5 d6.

Or, this system returns

Code: Select all

3... d6 :-) >+5.12 

(I hope some people pick my joke)

Miguel

Yes, I see it...and why are people giving this article any credence??? I mean really...

I am not so sure you got my joke (the smiley output), only a Gaviota fan will get it, and there only very few.

Anyway, +5.12 I am pretty sure was an old Rybka bug when the output got stuck in that number. I do not have rybka but I remember the complaints.

Miguel
EDIT: A further article should refute this one because Be2 is loses based on an underpromotion

I got it. But it would mean more for a fan.

[Terry McCracken]

I hope this is just an April's Fool's joke.

It can't be, as it was not published on April the 1st... So it's either true or a regular hoax.

The game is dated April 1st.

[Daniel Shawul]

Terry postmortem doesn't get you credits.
You were almost 4 hours late to reply

[Rein Halbersma]

In the book One Jump Ahead, Jonathan Schaeffer at some point thought that it might be good enough to say that if you were N checkers ahead you could write the position off as a win - and much to his surprise this was not a valid assumption even for a fairly large number of checkers, and in checkers a single pawn (or checker) ahead is a huge advantage. It's been my experience that no simplistic rule can be reliably used to stop a search without introducing scalability issues - because you will ALWAYS be able to find a position where it is badly wrong! In this study Vas it treating 5.12 as a forward pruning rule to represent a complete search to the end of the game.

Apart from the lame April fools date obfuscation in the piece, the actual numbers already made the story incredible without reading further!

First, the solution space of checkers was 10^22 (the search space was 10^40), which was reduced to 10^14 by a bidirectional search. The back-end search built 10^14 database positions, and the front-end search built 10^14 opening positions. Schaeffer (http://ilk.uvt.nl/icga/journal/pdf/toc30-4.pdf) estimates it would take 200 core years to re-create this solution. Second, Schaeffer also estimates that the solution space for chess is about the square of that of checkers.

How does Vas's claim stack up against this? Hm, about 10 times the computing power but about the square of the search space (10^80 vs 10^40). Being liberal, let's suppose the actual solution space is the square root of that (10^40). However, the efficient bidirectional search (giving almost another square root reduction), was dependent on 10-piece databases which were already reachable from shallow root searches. Without the equivalent chess databases, the 10^40 solution space will not be reduced by another square root. And covering 26 orders of magnitude with 10 times more computing power...

BTW, the way checkers was solved by Schaeffer et al. was by iterating over the threshold value. So a real proof would take the 5.12 as the first step in such an iteration, and stepwise increase it all the way to a mate score.

[Terry McCracken]

Terry postmortem doesn't get you credits.
You were almost 4 hours late to reply

I didn't see it till you posted it. I was the first in the thread to reply. I rarely look at Chessbase anymore. Probably that article to attack the ICGA was the final straw.

I believe you could solve an opening in chess only a little more probable than FTL neutrinos.

I'll win the lottery first by finding the winning numbers in a dumpster and I don't look through dumpsters.

[IanO]

Whether this is a joke or not, I have long expected computer science research like this to have come out of computer chess. Are any other academic groups working on automatic discovery of opening theory?

[Dan Honeycutt]

Terry postmortem doesn't get you credits.
You were almost 4 hours late to reply

Where do you get 4 hours? My post headers say 12 minutes.

Good call, Terry.

Best
Dan H.

[playjunior]

There are 2 things about this article that make it lame. First, it was posted on April 2nd. Second, Chessbase uses Fischer's name in almost every April fools article.

Still, will be fun to read the readers' feedback

[Jouni]

This was good joke, when some "experts" here fell for it!

[Daniel Shawul]

Where do you get 4 hours? My post headers say 12 minutes.

Look the headers since our replies to him that it is real NOT his first post. 12:34 - 3:30
Time starts from there and ends on the first evidence Referee. The PGN tag was his input ofcourse. Even he said he didnt see the chessbase article your post is nonsense. Many things have been talked about by that time.