Predateur 0.1.1 : 1498

[Le Fou numerique]

Hi,

Predateur 0.1.1 first rating in the UEL: 1498.

PGN on the Accession & Réserve page.

Regards,
Patrick

[hgm]

Kind of arbitrary, isn't it?

As it did lose each and every single game it played, its rating is just as likely to be 1000, or 500, or 0... You really have no idea about its strength at all, except that it must be very, very weak.

This is as good as a rating based on zero games.

[Tony Thomas]

Kind of arbitrary, isn't it?

As it did lose each and every single game it played, its rating is just as likely to be 1000, or 500, or 0... You really have no idea about its strength at all, except that it must be very, very weak.

This is as good as a rating based on zero games.

How about the one below?? I dont feel like using Bayesian, so its just a quick and dirty elostat list..The problem with the engine is mainly its time allocation, the author is aware of it. It uses way too much time in the beginning of the game, about 50% of the time within the first ten moves and then at the end it just plays too quick. You might enjoy seeing some obscure engines that you hate on the list of opponents.

Code: Select all

71 Predateur 1.1             : 1426   19 (+  6,=  0,- 13), 31.6 %

Neg03d                        :   1 (+  1,=  0,-  0), 100.0 %
DelphiMax 2.9                 :   1 (+  0,=  0,-  1),  0.0 %
Dreamer 0.10??                :   1 (+  1,=  0,-  0), 100.0 %
Ghost_v0_14                   :   1 (+  0,=  0,-  1),  0.0 %
Embracer_112                  :   1 (+  0,=  0,-  1),  0.0 %
Genesis 2.01                  :   1 (+  0,=  0,-  1),  0.0 %
Hagrid0.7.56                  :   1 (+  0,=  0,-  1),  0.0 %
Geko 0.4.3                    :   1 (+  1,=  0,-  0), 100.0 %
Cpp1 0.1038                   :   1 (+  1,=  0,-  0), 100.0 %
Pos 1.18                      :   1 (+  1,=  0,-  0), 100.0 %
Golem 0.1                     :   1 (+  1,=  0,-  0), 100.0 %
Heracles 0.5.14               :   1 (+  0,=  0,-  1),  0.0 %
Joana_DC                      :   1 (+  0,=  0,-  1),  0.0 %
Jester_083DC                  :   1 (+  0,=  0,-  1),  0.0 %
Hoplite 2.1.1                 :   1 (+  0,=  0,-  1),  0.0 %
Holmes 0.8.29e                :   1 (+  0,=  0,-  1),  0.0 %
IQ23 .003                     :   1 (+  0,=  0,-  1),  0.0 %
MSCP 1.41                     :   1 (+  0,=  0,-  1),  0.0 %
LaDameBlanche 2.0c            :   1 (+  0,=  0,-  1),  0.0 %

[hgm]

Well, as long as it does not actually lose on time, changing the time management won't work miracles. Even if it would play with half the time, it should lose only about 70 Elo. And the other half is not totally wasted, it did 10 moves in it. And these moves were actually better than what they otherwise would have been.

If you take 5% of the remaining time for each move in a sudden-death game, in stead of 2.5%, it takes until move 28 before you actually use less time than you would otherwise have done for a move. And by that time the game might already be decided. It takes until move 54 before you use less than half the time per move (causing a theoretical weakening of 70 Elo points).

Winning against Pos and NEG also doesn't mean an awful lot: they are asymptotically weak, and due to lack of enough intermediate weak engines their rating is usually highly overestimated. Based on analyzing the results of the Chess-War Promo with a less-distorting prior than the BayesElo default, I came to the conclusion that the true rating of Pos must be approximately zero. (Actually even a little below it.) Of course at faster TC they gain compared to the others, as the others get weaker, and they always play the same, independent of TC.

[Dr.Wael Deeb]

Kind of arbitrary, isn't it?

As it did lose each and every single game it played, its rating is just as likely to be 1000, or 500, or 0... You really have no idea about its strength at all, except that it must be very, very weak.

This is as good as a rating based on zero games.

Know what H.G.Muller,one thing I like about you is that you are always kind and your words are music for the ears

[hgm]

Yes, I know, I missed my vocation: become a politician or a diplomat.

[Dr.Wael Deeb]

Yes, I know, I missed my vocation: become a politician or a diplomat.

Definitely the case

[Tony Thomas]

I think I might have underestimated the time usage a little bit. By the 20th move or so, Predator only has 1 second left on the clock and it plays rest of the games with the increment.

[Tony Thomas]

I came to the conclusion that the true rating of Pos must be approximately zero. (Actually even a little below it.) Of course at faster TC they gain compared to the others, as the others get weaker, and they always play the same, independent of TC.

I suspect that CPP1 (his previous engine) might be stronger. I played few games of cpp1 vs pos, and in mostly all the matches, CPP1 was ahead on material but let POS get a stalemate, three fold rep, or 50 move draw...

[hgm]

I can see that would be a problem when increment = 0... Early Smirf versions used to behave the same.

The funny thing is that between engines of approximately equal basic strength, this behavior is not so easy to punish. If the opponent starts thinking only after a book opening, say move 12 or so, he will greatly outsearch you in an important part of the game if you just stick to your own schedule. And being a Rook ahead, he will not have too much trouble finishing you off even with only 1 sec for the remaining game.

I think the important lesson is: use about as much time as your opponent. If you think he is way too slow, just be a little bit faster than him, (say 20%). But don't allow him to outsearch you by a factor 2 or more. If you think he is way to fast, just be a little bit slower, but prevent that he builds up a time reserve that will allow him to outsearch you by a factor 2 later in the game.