Search found 116 matches

by Pio
Fri Dec 15, 2017 7:51 pm
Forum: Computer Chess Club: General Topics
Topic: recent article on alphazero ... 12/11/2017 ...
Replies: 106
Views: 18967

Re: recent article on alphazero ... 12/11/2017 ...

You do not understand how smart he is. I am a physicist and understand how smart he is :P
by Pio
Fri Dec 15, 2017 7:42 pm
Forum: Computer Chess Club: General Topics
Topic: recent article on alphazero ... 12/11/2017 ...
Replies: 106
Views: 18967

Re: recent article on alphazero ... 12/11/2017 ...

I find Muller very funny.

By the way Physicists usually do not make assumptions, they make educated guesses :idea:
by Pio
Thu Dec 14, 2017 10:18 am
Forum: Computer Chess Club: General Topics
Topic: Tornado blog
Replies: 4
Views: 1864

Re: Tornado blog

Hi Engin!

I sent you a PM long time ago but it seems you have missed it. Look at your message box :)

Good luck with Tornado

/Pio
by Pio
Thu Dec 14, 2017 9:47 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Google's AlphaGo team has been working on chess
Replies: 203
Views: 66542

Re: Google's AlphaGo team has been working on chess

I would love to see that engine. I think you would make a great one :)
by Pio
Sun Nov 26, 2017 9:56 pm
Forum: Computer Chess Club: General Topics
Topic: AlphaGo's evaluation function
Replies: 14
Views: 3748

Re: AlphaGo's evaluation function

How do you know that the network does not do a shallow search ;) Actually you do need to search at all if you have a perfect evaluation. And even better you only have to know what move is the best at every position. I think you could do really cool things with a neural network in chess and I have so...
by Pio
Fri Aug 04, 2017 8:50 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Ways to avoid "Draw Death" in Computer Chess
Replies: 121
Views: 13924

Re: Ways to avoid "Draw Death" in Computer Chess

Hi Kai! I have watched every post in this thread even though I cannot remember everything. Of course you can adjust the time odds so that it will give you the most information (depending on your rating model). Another gain by using time odds is that you will be able to inspect the games where the pl...
by Pio
Fri Aug 04, 2017 7:57 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Ways to avoid "Draw Death" in Computer Chess
Replies: 121
Views: 13924

Re: Ways to avoid "Draw Death" in Computer Chess

I just love your posts HGM. You are of course correct about that the engines will shift their distribution if they are tested very close to the edge. I do not understand why people do not use your time odds idea that is brilliant.

/Pio
by Pio
Sat Jul 15, 2017 5:21 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Equidistributed draft
Replies: 4
Views: 1051

Re: Equidistributed draft

I sent you a pm

/Pio
by Pio
Thu Jul 06, 2017 8:40 pm
Forum: Computer Chess Club: General Topics
Topic: Answer to Ultimate Question of Life, Universe & Everythi
Replies: 5
Views: 1131

Re: Answer to Ultimate Question of Life, Universe & Ever

Hi Dann!

I am not a chess expert but would never play d5 since you trap your white squared bishop for a very long time and the opponent can take his knight through to the king side and use it in an attack. I would consider d5 almost as bad as giving up a pawn

/Pio
by Pio
Wed Jun 28, 2017 9:35 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Move size & ordering ?
Replies: 6
Views: 1236

Re: Move size & ordering ?

Hi Mahmoud! There is no problem storing the move in 16 bits and if you need to compress the scores to 16 bits, you can always transform the normal centipawn or millipawn scores to probability based scores. The advantage of storing the score as a probability is that the resolution is very high where ...