I like the way this comes up in every second thread in this forum and is totally and utterly wrong.

In fact if you were to say "there are about ten legal chess positions" you would be closer to the truth.

A) I don't live in the chess forum.
B) This is a LC thread
c) suck my nuts


http://en.wikipedia.org/wiki/Shannon_number. I'm not sure what estimations are right, but it's certainly a statement a lot of very smart people have made. I have seen 10^120 for game-tree complexity whatever that means, but also some as low as 10^43 which obviously isn't even close.

My friend who is an IM really likes the 5. a3 line and thinks that its very comfortable for White. I saw him recently pick apart a 2350-2400 player in that line over the board.

In fact it is a statement a lot of very stupid people have made. More charitably, it is a statement made by people who are ignorant or who don't really understand numbers above about 1000.

"Game tree complexity" is something totally different from the number of legal positions. If you actually read the source given on your wikipedia page you would find the number 10^43 given by Claude Shannon himself.

"A lot of very smart people" would say just that. In fact any "very smart person" would either say that 10^43 is very close, or be smart enough to say that he doesn't know.

Never did I imply that numbers such as 10^43 or 10^44 were close. And yes, I did read the wikipedia page also, I didn't need you to summarize it for me.

This chess forum is mucho beefy

Chess is insanely complex

/argument

brag: Beat a 2100 rated computer on fics(standard) as black when he played a novelty opening.(1. c4 d5 2. cxd5 e6 3. Qa4+)
beat: he force-blundered away his rook for 2 pawns at move 5.

was the most entertaining game i have played. never had so much fun going over lines after the game as i did after this one. missed some extremely cool attacking lines, including a mate in 11.

Pretty sure he's just getting confused between chess positions and chess games. I've heard the comment about chess games many times but I'm not going to pretend to know whether it's actually true or not.

You could try 1. d4 e6 and invite a transposition into the french defence. If white plays c4 instead of e4, then you can transpose into another opening like the Bogo Indian or w/e.

I know, you implied that they weren't, which is clearly incorrect.

You know what, there are 64 squares on a chessboard. Each square can contain either a pawn, knight, bishop, rook, queen, king, of either colour, for 12 possibilities. Add in an empty square, which means that each square can have 13 possible units.

The total number of possible configurations would be 13^64. This would mean around 10^71 possible configurations. The vast majority of these would be nonsense illegal positions with 17 white kings and 33 black knights on the board, so this is an absolute upper bound to the number of positions.

There are around 10^80 atoms in the universe.

I think that just using elementary logic, and not knowing any hard math, it's possible to figure this out.

That's a good start, although multiply it by two since it's a different position depending on whether white or black is to move. A few technicalities with whether or not en passant captures are possible, castling options where applicable (have the king and rooks moved yet), and I guess if you want to get super technical you'd have to look at the count for the 50 move rule and what the move history is for determining threefold repetition.

Also you can tighten the upper bound a little, no pawns are possible on the 1st or 8th rank. So disregarding the technicalities I mentioned, the new upper bound on possible positions is 2 * 11^16 * 13^48. Easiest way to calculate the actual value would be random sampling. Randomly generate a large sample of those positions, see what percentage of them are actually legal (1 white king, 1 black king, only the side to move can be in check, you don't have an impossible number of pieces for either side, etc). Multiply that percentage by the upper bound and you've got your estimate.

got sick after drinking way too much on friday.
started playing chess saturday morning while hungover and on 4 hours sleep.
finished playing chess sunday afternoon.

rating at start of session: 2154
rating end of session: 1951

not sure what parts of that are beats and what parts are brags.

sure do wish poker was as fun as chess.

Needs more graphs

was trying to figure out a good way to get a graph. there doesn't seem to be any way to graph results in chessbase 10, and ICC's rating graph thing blows since it's 3 months minimum.

but I did find out I played a minimum of 110 games of 5-minute. I missed alot of games from having my database open and it not writing them to file. so probably 130ish games.

hmmm.. chess endurance prop betting suddenly seems like a really fun idea.

lol Dire that's like at least 12 hours of chess straight. Yikes.

If you don't mind doing a bit of data entry you could just sort your database chronologically, and then type your pre-game rating for each game into Excel and graph that.

I once played over 21 hours of bullet in an ICC 24h bullet marathon.
Beat: Someone played the whole 24h and I lost to him

im sure he meant games and not positions!!

Well since he made three posts about it and claimed he read the wikipedia page thoroughly, I'm sure he meant positions. But he would have been correct if he said games.


In suicide chess, the object is to lose all your pieces, including your king. If on any move you can make a capture, then you must make a capture.

there are almost as many "legal"* game positions in suicide chess as in normal chess, but the game tree is much much smaller because there is often only one legal move. Although the game isn't solved, it is solved after certain first moves. For example 1. e4? has proven to lead to a loss by a brute force analysis. So the sheer number of legal positions isn't enough to say that chess is unsolvable.

*by "legal" I mean "can't be trivially proved to be impossible".

how can you have a blitz rating of 1381 and do this

http://www.ficsgames.com/cgi-bin/sho...44;action=show

this just blew my mind basically

LMAO http://www.ficsgames.com/cgi-bin/sho...26;action=show

haha

i can kinda believe a 1082 has never seen it before, but a 1381? bizarre. even if he's somehow never ever been faced with it it's pretty intuitive where the king should go.

LOL copycat.

Here's a game I've had a couple of times over the board, and many times online as well:

http://www.ficsgames.com/cgi-bin/sho...38;action=show

Oh and those games remind me of this:

http://www.ficsgames.com/cgi-bin/sho...52;action=show