19/08/2019 New Poll

Post Reply
User avatar
Kevin
Site Admin
Posts: 26
Joined: Thu Aug 01, 2019 11:47 am
Location: Cumnor,Oxford, UK
Contact:

19/08/2019 New Poll

Post by Kevin » Mon Aug 19, 2019 6:15 pm

What is your prediction for the No. 1 chess engine in 2030? More info and vote at http://www.infinitychess.com/Default.aspx
Last edited by Kevin on Mon Aug 19, 2019 6:17 pm, edited 1 time in total.

User avatar
KhalidOmar
Posts: 99
Joined: Thu Aug 01, 2019 12:21 pm
Contact:

Re: 19/08/2019 New Poll

Post by KhalidOmar » Mon Aug 19, 2019 7:54 pm

Nice poll, really got me thinking!

I have few comments about this poll :idea:
  1. I think the next few years (maybe 5) will have very close competition between AB and NN with both engines winning it for a few years.
  2. After those 5 years, NN is going to rule AB after completing enough training to overcome any AB engine
  3. Considering computer chess history since its "true" beginning, which wasn't that many years ago; I think a new type of chess engines is going to emerge which is going to be better and faster than both AB and NN. My own expectation for this new type is that it may start at 5000 ELO points if not even more!
  4. Who knows, by 2030 computer chess engine may reach its end as someone manages to generate 64-MAN TB :D or an engine manages to use parallel computing to calculate definite mate/draw from start position :D :D !
For information about Eman engine, check http://eman.zohosites.com

User avatar
Kevin
Site Admin
Posts: 26
Joined: Thu Aug 01, 2019 11:47 am
Location: Cumnor,Oxford, UK
Contact:

Re: 19/08/2019 New Poll

Post by Kevin » Thu Aug 22, 2019 1:34 pm

I think for the next ten years it will be close between the AB's and NN's but also Hybrids will also be in the mix which could end up being able to prolong the tide of when the NN's start dominating.

Hopefully we will see more competition like 5 (maybe more?) different engines trying to dominate the field instead of the two we have now, Stockfish/derivatives and LC0/derivatives. :D

DavidG
Posts: 11
Joined: Wed Aug 28, 2019 12:44 am

Re: 19/08/2019 New Poll

Post by DavidG » Fri Aug 30, 2019 5:26 pm

Gentlemen,
I believe that in 10 years, computer chess will have advanced to the stage where every game will be a draw. Lines will be so perfect, and blunders so rare, such as to render the game of chess "solved". This will increase the popularity of ancillary chess games such as Fischer Random chess.
David

Ozymandias
Posts: 6
Joined: Thu Aug 01, 2019 7:35 pm

Re: 19/08/2019 New Poll

Post by Ozymandias » Sat Aug 31, 2019 12:33 pm

DavidG wrote:
Fri Aug 30, 2019 5:26 pm
Gentlemen,
I believe that in 10 years, computer chess will have advanced to the stage where every game will be a draw. Lines will be so perfect, and blunders so rare, such as to render the game of chess "solved". This will increase the popularity of ancillary chess games such as Fischer Random chess.
David
As long as players like Olivier Evans keep being involved, tournaments with a 100% draw rate will never happen. Flukes aside, when it comes to Freestyle, that's already been the situation for years.

User avatar
KhalidOmar
Posts: 99
Joined: Thu Aug 01, 2019 12:21 pm
Contact:

Re: 19/08/2019 New Poll

Post by KhalidOmar » Sat Aug 31, 2019 12:44 pm

DavidG wrote:
Fri Aug 30, 2019 5:26 pm
Gentlemen,
I believe that in 10 years, computer chess will have advanced to the stage where every game will be a draw. Lines will be so perfect, and blunders so rare, such as to render the game of chess "solved". This will increase the popularity of ancillary chess games such as Fischer Random chess.
David
I highly doubt that actually! I agree that in 10 years we will have supercomputers that run hundreds if not thousand times faster than now, but the complexity of chess and the number of possible moves is just insane and too much to handle even for much faster computers.
For information about Eman engine, check http://eman.zohosites.com

Olivierevan
Posts: 9
Joined: Thu Aug 01, 2019 6:23 pm

Re: 19/08/2019 New Poll

Post by Olivierevan » Fri Sep 13, 2019 4:19 pm

Hello everyone.

The only way is NN before quantum computers.
AB = End of era !
NN = New Era !

AB brute force do not know how to create a game strategy.
NN an man can create a strategy in the game.

A strong player (GM) who prepares with a NN engine can beat a powerful AB machine, I'm sure !
All analyzes of AB engines will be refuted by NN engines.

Take the example of the king's gambit accepted C33 1.e4 e5 2.f4 exf4... The statistics humans for winnings percentage :
- White 44 %/ Draw 19.4 %/ Black 36.5 %
After 10 minutes of infinite analysis...
AB Engine Eval (Eman 4.10) / NN Engine Eval (Lc0 22)
=+ -0.60 depth 35 / 3.Bc4 // =+ -0.60 depth 40 / 3.Nf3
The 2 chess engines give the same evaluation with different moves ! How to play this opening pattern with e4 f4 and avoid the accepted gambit of the king ?

Nothing's easier, the Vienna Game is an opening in chess that begins with the moves : 1.e4 e5 2.Nc3! Nf6 3.f4...At grandmaster level, the gambit move 3.f4 is considered too risky an opening // Keene, Raymond. The Times 1 June 2012 - wikipedia ??

Now 3...exf4?! += / better 3... d5=.

Return on opening line 1.e4 e5 2.Nc3 Nf6 3.f4 d5 4.Nf3! ...
Winnings percentage (humans games) after white move 4.Nf3... White 33.3% / Draw 13.4 % / Black 53.3 %.

After 10 minutes of infinite analysis...
AB Engine Eval (Eman 4.10) / NN Engine Eval (Lc0 22)
=+ -0.74 depth 37 / 4... exf4 // =+ -0.42 depth 28 / 4... exf4

AB = End of era !
NN = New Era !

Enjoy

Tuxpower
Posts: 7
Joined: Sat Aug 03, 2019 12:45 pm

Re: 19/08/2019 New Poll

Post by Tuxpower » Sun Oct 13, 2019 11:30 pm

NN is great for the serious chess player like me :D :P

I have big database with NN games, I did learn alot from Alphazero chess and Leela games.

They did play many times also a novelty in the opening.

8-)

User avatar
Kater Carlo
Posts: 8
Joined: Mon Nov 04, 2019 6:40 pm

Re: 19/08/2019 New Poll

Post by Kater Carlo » Mon Nov 04, 2019 7:02 pm

Make me think, how to achieve ELO 5000 in engine tournaments?? Corr chess offers even today enough evidence that chess is a draw by perfect play on both sides. This leads to a constant deflation of ELO values in ICCF practise. Regularly 9 of ten games end in a draw in engine based games with long thinking time. Even today chess is nearly solved. It can still be a sport in human play with human tension, time trouble and so on.. :idea:

Post Reply