Anyone have any first hand experience with this? There are rumors circulating that it's bug ridden and weak and that's why TCEC 10 hasn't begun. Again: Rumors. I have absolutely no personal knowledge of any of this.
When I look at the CCRL lists and things, though...it just looks like SF Development is running all over K11.2 in both playing and analysis. Is this generally where things are at?
$100 for 11.2 + 12 months of updates is quite a bit for a chess engine. I asked them a question about buying the engine now and then the subscription later (via email) and haven't heard anything back in days. Apparently (I got an automated response) there's one person who handles customer support for Komodo Chess and he's on vacation. He said it might take "a couple" of days to get back to me but it's been 3+ now.
I currently use the best and fastest Stockfish Development versions along with Fritz 15.
Actually there aren't rumors there's just one troll who keep spamming stupid stuff in the TCEC chat.
Komodo is a strong engine however if you're not near the top level or playing correspondence chess you'll be fine without it. (although it can still be nice at times to see a second opinion on positions, SF and K synergize very well and will often have different ideas)
TCEC is probably a dead project by now and it has nothing to do with Komodos strength. Some bad decisions were made:
groups.google.com/d/msg/fishcooking/QOYhAnmEJCI/APBLwLLUBgAJI don't know anything about Komodo 11.2 because I don't own it. All I can say is, that - in my experience - Larry Kaufman and especially Mark Lefler are nice and competent people and very close to their (potential) customers.
Interesting stuff, guys. Thanks for filling me in. I looked into it and sure enough, it does look as if one troll has spent the last 3 weeks doing just about nothing but spamming anti Komodo 11.2 propaganda. Unreal.
I'm thinking about grabbing it today. The thing is that there are already so many different development versions of Stockfish available and I came across BrainFish earlier which has an unbelievably high Elo. With the open source nature of Stockfish and the fact that it's fully integrated into this site, I just don't see how Komodo can keep up in the long run. This isn't really a knock against Komodo (as I still want it) but more a compliment to Stockfish.
I'll be playing in the next season of Team 45+45 on ICC and thought that Komodo would help with the longer time controls but now there's CorrChess which is basically Stockfish tailored for classical and correspondence chess. Tough decision.
Heh...as an engine Brainfish is just Stockfish, nothing more. If you just use it to analyze your games it won't be any better than just using Stockfish.
What sets it apart on the lists where it's tested is the fact that it has a massive opening book created from minimaxing SF evaluations.
That's the only reason it's higher rated than SF. It has a book basically tuned by SF for SF and is allowed to use it.
A very nice and interesting idea, for sure, and one that obviously succeeds quite nicely in gameplay, but it's not going to make its analysis any stronger :)
Ah, I see. So Brainfish's strength is in playing vs other engines, not in analyzing games/positions.
Anyone here have K11.x and use SF Dev, as well?
http://chess.ultimaiq.net/stockfish.htmlhttp://abrok.eu/stockfish/These two sites full of links are what's keeping me from dropping the money on K11.2. That and the fact that you can download all sorts of endgame tablebases for free. Even if I bought a tablebase, it would cost less or around the price of K11.2.
I mean, They just released 11.2. SF 9 will be here this fall (almost for sure). Komodo will just be playing catchup again at that point, no? Meanwhile, we're all $100 poorer and SF continues to truck along for $0.
You should get your stockfish here
github.com/lantonov/asmFishThis version is written in assembler thus faster but otherwise the same.
Very, very cool. Thanks.
What am I primarily looking to download - The ArmFish executable files?
I use Komodo 11.01 yet to download 11.2 but I will probably get it soon. My subscription ended a week or so before 11.2 came out.
Also I use stockfish as well. Both handle positions slightly differently. And Komodo is easier to set up to get wider evaluations. For example you can increase search depth or you can increase search accuracy with Komodo.
In the settings you can also turn Null moves on/off, change selectivity and turn on/off LMR unlike Stockfish. Allowing for stronger width or depth searches.
When searching wider move fields Komodo takes longer to get to n ply than stockfish but is generally more accurate.
I do fairly in depth opening work for people (not just myself) and find that both can give small inaccurate moves once in a while. Occasionally I find fairly strong novelties with one engine over another where one thinks a move is strong but the other doesn't. So I make them play out the position and find what one is best among other ways to tell what one is best. Sometimes I just have to feed the moves to it and the evaluation will change back to normal.
As for these rumors I have not herd anything about them. I wouldn't really worry about it. Sounds fishy to me.
For analysis depending on the depth and time you care to consume using two engines is a good idea.
To summarize.
I use:
SF for depth searches
K for width searches
