Komodo Dragon 2 released.

[JohnW]

No Lc0 in your list?

[pohl4711]

No Lc0 in your list?

Not in that list. You can not play 7000 games with lc0, because only one lc0 game on a machine can be played.

Look here:
https://www.sp-cc.de/nn-vs-sf-testing.htm

Or Andreas Strangmuellers excellent website:
http://www.fastgm.de/16-60-0.6.html

[lkaufman]

7000 games testrun of KomodoDragon 2.0 MCTS finished.

https://www.sp-cc.de

(Perhaps you have to clear your browsercache or reload the website)

Nice to see that you are showing even larger elo gains from Komodo 14 to Dragon 2 in both standard and MCTS mode than we claimed; you show +223 elo for MCTS and +209 for standard mode, both somewhat above our estimates. In general you show larger rating differences between engines than other lists, perhaps because the book you use for that list is less apt to lead to draws? Anyway I see that as a good thing.

[pohl4711]

7000 games testrun of KomodoDragon 2.0 MCTS finished.

https://www.sp-cc.de

(Perhaps you have to clear your browsercache or reload the website)

Nice to see that you are showing even larger elo gains from Komodo 14 to Dragon 2 in both standard and MCTS mode than we claimed; you show +223 elo for MCTS and +209 for standard mode, both somewhat above our estimates. In general you show larger rating differences between engines than other lists, perhaps because the book you use for that list is less apt to lead to draws? Anyway I see that as a good thing.

No, the HERT openings by Thomas Zipproth, I use for my main list are not "low-draws" openings, but classical openings.
The rating differences in my list are higher, because Komodo plays vs. 7 strong opponents, only. In other lists, there are more (and weaker!) opponents, which always makes it harder to gain Elo. IMHO.

[lkaufman]

7000 games testrun of KomodoDragon 2.0 MCTS finished.

https://www.sp-cc.de

(Perhaps you have to clear your browsercache or reload the website)

Nice to see that you are showing even larger elo gains from Komodo 14 to Dragon 2 in both standard and MCTS mode than we claimed; you show +223 elo for MCTS and +209 for standard mode, both somewhat above our estimates. In general you show larger rating differences between engines than other lists, perhaps because the book you use for that list is less apt to lead to draws? Anyway I see that as a good thing.

No, the HERT openings by Thomas Zipproth, I use for my main list are not "low-draws" openings, but classical openings.
The rating differences in my list are higher, because Komodo plays vs. 7 strong opponents, only. In other lists, there are more (and weaker!) opponents, which always makes it harder to gain Elo. IMHO.

Yes, that makes sense. It's interesting that you get larger rating gains from Komodo 14 to Dragon 2 than we get in a direct match, perhaps because there are more draws between engines that are at least somewhat related, or else because the gap is already too large for a direct match to be fully accurate.

[Krzysztof Grzelak]

When the new version of the program, because Komodo 2 is so weak, it is a pity to use it. Gentlemen of Komodo, accelerate the pace of your work.

[cc2150dx]

When the new version of the program,

"What is the 'one year subscription'?

This option is for chess players who need the strongest possible engine. We release updates to Dragon three to five times per year. The subscription is basically Dragon plus these 3-5 updates for a discount."

So in other words, the next update can be within 2.5-4 months roughly.

[Cornfed]

When the new version of the program, because Komodo 2 is so weak, it is a pity to use it. Gentlemen of Komodo, accelerate the pace of your work.

LOL!

1. Komodo 2 is ancient, goes back many years. Try Dragon 2.
2. It (Dragon 2) is clearly NOT weak.
3. Don't use it then...what do you use engines for anyway? Engine vs Engine play? If so...sorry to hear that - there is no sport in it...it's like 'watching a basketball game' the only sport is on the court. If you use them to hide behind while playing, thinking YOU have accomplished something at the end of the day....well, most across the world find better things to do with their time. Try actual human vs human play...it's the bees knees!

[Chessqueen]

When the new version of the program, because Komodo 2 is so weak, it is a pity to use it. Gentlemen of Komodo, accelerate the pace of your work.

From Komodo 2 to Komodo 14 the difference is like 900 to 1000 rating points, as a matter of fact you can get Komodo 12 for Free at Komodo website and later came Dragon 1 and then Dragon 2, if the name remained the same it would have become Komodo 15 instead of Dragon 1 and Komodo 16 instead of Dragon 2, but after Komodo 14 it changed to Dragon
https://komodochess.com/dragon-2.htm

[cc2150dx]

When the new version of the program, because Komodo 2 is so weak, it is a pity to use it.

I also think you mean Dragon 2, not Komodo 2. http://talkchess.com/forum3/viewtopic.p ... on#p887873