Surprise: Lc0 is beating SF in TCEC

[Jouni]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

[Milos]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Year right, it's really stunning that after 8 coin tosses one gets 5 heads and 3 tails instead of 4 heads and 4 tails or even 3 heads and 5 tails (with tail-biased coin) .
Gee, how difficult is for some ppl to understand how meaningless it to draw any conclusions based on few coin tosses.

[lkaufman]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Why is it that in this FRC event a majority of games at the top are decisive, but when chess.com had an FRC event the final between SF and Dragon had about 99% draws?? Something must be very different, but I don't know what.

[dkappe]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Why is it that in this FRC event a majority of games at the top are decisive, but when chess.com had an FRC event the final between SF and Dragon had about 99% draws?? Something must be very different, but I don't know what.

Very unbalanced exits, much like uho in standard.

[lkaufman]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Why is it that in this FRC event a majority of games at the top are decisive, but when chess.com had an FRC event the final between SF and Dragon had about 99% draws?? Something must be very different, but I don't know what.

Very unbalanced exits, much like uho in standard.

I thought that might be it, but with 99% draws in chess.com I didn't think there were enough sufficiently unbalanced FRC positions to get a majority of decisive results!

[connor_mcmonigle]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Why is it that in this FRC event a majority of games at the top are decisive, but when chess.com had an FRC event the final between SF and Dragon had about 99% draws?? Something must be very different, but I don't know what.

Very unbalanced exits, much like uho in standard.

I thought that might be it, but with 99% draws in chess.com I didn't think there were enough sufficiently unbalanced FRC positions to get a majority of decisive results!

TCEC is using a "FRC book" to insure imbalanced positions and lower draw rates.

[lkaufman]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Why is it that in this FRC event a majority of games at the top are decisive, but when chess.com had an FRC event the final between SF and Dragon had about 99% draws?? Something must be very different, but I don't know what.

Very unbalanced exits, much like uho in standard.

I thought that might be it, but with 99% draws in chess.com I didn't think there were enough sufficiently unbalanced FRC positions to get a majority of decisive results!

TCEC is using a "FRC book" to insure imbalanced positions and lower draw rates.

I see. Well, I guess it was necessary, but it seems to defeat the point of FRC to randomize the start position and then stipulate book moves to be played. I don't have a better suggestion, it's just a bit sad that this should be needed. I suppose just selecting the most unbalanced FRC positions wasn't sufficient to avoid draws. I think we need a more unbalanced version of FRC rules, but it's hard to think of sensible rules for this. Bobby Fischer's rules are really great for human chess, even at 2800 level, but no one foresaw that engines would reach the level where they were virtually unbeatable from most if not all of the 960 positions if free to choose their own moves from the start.

[alex67a]

In chess tactics and strategies are well established and the days when engines used only brute force have passed for decades ...
The classic engines could be little inclined to the inevitable strategic / tactical upheavals of the FRC ...
The best demonstration of this are engines such as crystal and crystallite which resolve positions that classic Stockfish does not resolve, but in a match with the latter they would lose.
This match will tell who is strongest in FRC, but little or nothing about who is strongest in classical chess

[RubiChess]

https://tcec-chess.com/

Yes FRC and 16 games so far. But still stunning! Dragon was third again. But You need two A100 10000€ each.

Year right, it's really stunning that after 8 coin tosses one gets 5 heads and 3 tails instead of 4 heads and 4 tails or even 3 heads and 5 tails (with tail-biased coin) .
Gee, how difficult is for some ppl to understand how meaningless it to draw any conclusions based on few coin tosses.

Meanwhile we have 9 tails and 8 heads.

[Uri Blass]

In chess tactics and strategies are well established and the days when engines used only brute force have passed for decades ...
The classic engines could be little inclined to the inevitable strategic / tactical upheavals of the FRC ...
The best demonstration of this are engines such as crystal and crystallite which resolve positions that classic Stockfish does not resolve, but in a match with the latter they would lose.
This match will tell who is strongest in FRC, but little or nothing about who is strongest in classical chess

The match is not going to tell who is the strongest in FRC

FRC is with no unbalanced opening books.