Would someone please explain to me the meaning of "Ptnml(0-2):" and what each of the five numbers following it mean ?
It is used in describing an engine's performance. For example:
Passed LTC:
https://tests.stockfishchess.org/tests/ ... 625bce0c0e
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 184664 W: 49933 L: 49344 D: 85387 Elo +1.11
Ptnml(0-2): 111, 17977, 55561, 18578, 105
thanks,
Norm
Ptnml(0-2):
Moderator: Ras
-
Norm Pollock
- Posts: 1087
- Joined: Thu Mar 09, 2006 4:15 pm
- Location: Long Island, NY, USA
-
Uri
- Posts: 529
- Joined: Thu Dec 27, 2007 9:34 pm
Re: Ptnml(0-2):
All I know is that Ptnml is an abbreviation of Pentanomial so maybe these numbers have something to do with match statistics.
Check out this page for instance: https://www.chessprogramming.org/Match_Statistics
Check out this page for instance: https://www.chessprogramming.org/Match_Statistics
-
RubiChess
- Posts: 669
- Joined: Fri Mar 30, 2018 7:20 am
- Full name: Andreas Matthies
Re: Ptnml(0-2):
They look at game pairs (two games starting with the same opening positions and both engines play once white and once black).
Five (penta) different results for a game pair are possible 0-2, 0.5-1.5, 1-1, 1.5-0.5 and 2.0.
In the example we had
111 game pairs with result 0-2
17977 game pairs with result 0.5-1.5
55561 game pairs with result 1-1
18578 game pairs with result 1.5-0.5
105 game pairs with result 2-0
See also https://www.chessprogramming.org/Match_ ... l_Analysis
Regards, Andreas
Five (penta) different results for a game pair are possible 0-2, 0.5-1.5, 1-1, 1.5-0.5 and 2.0.
In the example we had
111 game pairs with result 0-2
17977 game pairs with result 0.5-1.5
55561 game pairs with result 1-1
18578 game pairs with result 1.5-0.5
105 game pairs with result 2-0
See also https://www.chessprogramming.org/Match_ ... l_Analysis
Regards, Andreas
-
Norm Pollock
- Posts: 1087
- Joined: Thu Mar 09, 2006 4:15 pm
- Location: Long Island, NY, USA
Re: Ptnml(0-2):
Thanks Uri and Rubi.
The data for the match between side_1 and side_2 is from the perspective
of side_1 for both sets of data.
The trinomial data:
side_1 has 49933 wins, 85387 draws and 49344 losses.
The net wins for side_1 = 589 (49933 - 49344).
The pentnomial data:
111:
side_1 lost 0-2, 111 times
This implies 222 net wins for side_2 (-222 net wins for side_1).
17977:
side_1 lost 0.5-1.5, 17977 times
This implies 17977 net wins for side_2 (-17977 net wins for side_1).
55561:
There were 55561 results of 1-1.
This is not broken down into 2 draws or each side winning once.
This implies that the number of wins for side_1 and side_2 are equal.
0 net wins for side_1.
18578:
side_1 won 1.5-0.5, 18578 times.
This implies 18578 net wins for side_1.
105:
side_1 won 2-0, 105 times.
This implies 210 net wins for side_1.
Cumulative net wins for side_1:
(-222) + (-17977) + 0 + (18578) + (210) = 589
589 equals the net wins from the trinomial data.
The data for the match between side_1 and side_2 is from the perspective
of side_1 for both sets of data.
The trinomial data:
side_1 has 49933 wins, 85387 draws and 49344 losses.
The net wins for side_1 = 589 (49933 - 49344).
The pentnomial data:
111:
side_1 lost 0-2, 111 times
This implies 222 net wins for side_2 (-222 net wins for side_1).
17977:
side_1 lost 0.5-1.5, 17977 times
This implies 17977 net wins for side_2 (-17977 net wins for side_1).
55561:
There were 55561 results of 1-1.
This is not broken down into 2 draws or each side winning once.
This implies that the number of wins for side_1 and side_2 are equal.
0 net wins for side_1.
18578:
side_1 won 1.5-0.5, 18578 times.
This implies 18578 net wins for side_1.
105:
side_1 won 2-0, 105 times.
This implies 210 net wins for side_1.
Cumulative net wins for side_1:
(-222) + (-17977) + 0 + (18578) + (210) = 589
589 equals the net wins from the trinomial data.