My personal preference is Stockfish 6 for open tactical middle game and endgame positions, and Komodo 9 for closed/semi-closed positions. I also use Komodo 9 for endgame, I actually use a mix of the two engines even though they 9/10 are evaluating the endgame just about the same way. As for openings, you might as well use a database because engines follow their opening books. My personal belief is that while in opening lines, an engines evaluation is generally useless until around the opening sequence is over. Sometimes I'll also use Deep Junior just to see if it can find any tactical errors and maybe a sacrifice somewhere and use the Stockfish to make sure it'll work. Deep Junior (atleast on my part) is known to make some pretty foolish sacs, still most of the time at a 2500-2600 computer (Not human) level. Honestly, the 'best' is all a matter of opinion. Some people consider the highest rated engine the best, some people call an engine the best if it relates to their personal style of play, some people may say a certain engine because it finds the best defenses, it's the most aggresive etc. But my personal suggestion is to test the engines out yourself. Stockfish is one of the strongest engines in the world and it's free to download at https://stockfishchess.org/ Fritz 14 is good. It's pretty much good enough that it can still improve your game a lot. It's stronger than the strongest human grandmasters (2700-2800+) which means that you can still get a lot out of the engines analysis.
Best Chess Engine for analysis lines
oh ok thanks i will take a look and also i have heard people talk a lot about Houdini , how is that?
The latest version of Houdini, which is Houdini 4, has an elo rating of about 3200. It's the 3rd strongest engine behind Stockfish 6 and Komodo 9.1. It's stronger than Fritz 14 and definetly stronger than any human. Here's the rating list: http://www.computerchess.org.uk/ccrl/4040/rating_list_pure.html
Houdini 4 pro is available for purchase here: http://shop.chessbase.com/en/products/houdini_4_pro_multiprocessor_version
To my knowledge, Houdini has been known to have really strong endgame analysis and is excellent at defending a position.
I'd take Stockfish as main engine. Then if you are ready to pay, Komodo as secondary, and if you aren't, then Gull
Thanks for all the info. I am planning to use Stockfish 6 and considering to buy one of Komodo 9.1 or Houdini 4? which one do you suggest ?
FYI, I am a positional player so looking for an engine that provides solid positional mid and end game analysis. Also, planning to use the engine to test some of my new opening lines (I know its not such a good idea to use engine for opening analysis).
Komodo is stronger than Houdini. Houdini is not updated lately
oh cool thanks.
Sorry for not answering sooner, I've been pretty busy.
So, to answer your question:
I personally would suggest Komodo 9.1 over Houdini 4 because it is a better engine overall. They both have their highlights, but there is a reason Houdini 4 is under Komodo 9.1
But on the other hand, If I were to have you play Houdini 4, and Komodo 9.1 without you knowing which engine is which, you wouldn't be able to tell the difference. It's like a fly getting hit by truck and then a plane. It can't tell which one hit it harder.
So think of it this way, do your research on both engines, and choose the one that better compliments your playing style. I would definitely say get Stockfish 6 regardless simply because it's free. And if you need a walkthrough on how to install it simply send me a message and I'll help you. I only know how to put install it into ChessBase though.
Sorry again for taking so long to reply.
I go over my games with stockfish, and have a question on how to configure the strength ==
1. Is it better to set it up for the number of moves it considers, or for length of time.
2. Assuming that I set it up for time, is ten seconds too much?
3. Also I notice that when I run an analysis on the same move multiple times, I get different results for each analysis. What does that mean?
Thanks
I found an entire forum for discussing chess engines
http://www.chesspub.com/cgi-bin/chess/YaBB.pl?board=chesscomp
1) Number of moves. When you put it for length of time, the engine is on a time limit. So it won't have as much time to analyze.
2) I don't think 10 seconds is too much... It's more of too little. Engines analyze so many positions per second (I forgot the number, but it's huge). But you should still give your engine at LEAST 10 minutes to look at a position. That's what I do.
3) That means that there is more than one good move in the position, any of those moves is accurate. It's pretty much up to you to decide which move you like the best.
Thanks for your suggestions and i have already started using Stockfish 6 with chessbase, found it really helpful.
I'm glad you found it helpful! If you need any extra help just send me a message. I'll help you out!
Sure Thanks.
I usually give an engine to go over entire game, like half a minute per move, with good CPU it's enough for me, it finds all tactical and positional mistakes, than I manually look deeper, up to hour sometimes. All this is done after I went over game by myself, without looking at engine at all.
interesting.
Komodo 9, it is also openware. https://komodochess.com/
Sorry I forgot version 7 is free; 9 you have to pay for.
Hello
Can anyone suggest a chess engine that is best for game analysis (openining and positional) ? Also, Is Fritz 14 good? i heard it had some bugs.