Engine popularity
Engine popularity
My yearly look at German chess magazines (December issues of Schach 64, Schach Magazin, Rochade, KARL) shows that Houdini is still the most mentioned engine in game analysis (by far), with Komodo easily second, and Stockfish tied for 3rd with 0 citations. Part of the possible reasons: inertia, not everyone buys/downloads the latest and greatest more than once a year or two; Houdini/Komodo are [or at least were] easier to use with analysis, and/or their evals are more believable (particularly pawn structures is a Komodo edge I think); psychological effect of thinking that something you pay for must be better.
-
- Posts: 616
- Joined: Thu May 19, 2011 1:35 am
Re: Engine popularity
I use all three.
Stockfish is the best for quiet moves.
Houdini is the best for tactical moves.
Komodo will sometimes come up with a good idea that neither Stockfish nor Houdini found.
I am probably a special case because I have piles of machines at my disposal.
My current favorite is Stockfish (especially since it comes with source and I can modify it to suit my purposes).
I use a tweaked version that creates EPD records as it analyzes.
Stockfish is the best for quiet moves.
Houdini is the best for tactical moves.
Komodo will sometimes come up with a good idea that neither Stockfish nor Houdini found.
I am probably a special case because I have piles of machines at my disposal.
My current favorite is Stockfish (especially since it comes with source and I can modify it to suit my purposes).
I use a tweaked version that creates EPD records as it analyzes.
Re: Engine popularity
Obviously, you're looking at the wrong measure. Commercial websites or ches magazines, will only advertise products that are commercial and that they can sell. They all have a vested interest in keeping their audience oblivious to the fact that the best engine in the world is currently an open source and free program... Anyway, how many people actually read these magazines ? Now compare that to the number of people using Google...BB+ wrote:My yearly look at German chess magazines (December issues of Schach 64, Schach Magazin, Rochade, KARL) shows that Houdini is still the most mentioned engine in game analysis (by far), with Komodo easily second, and Stockfish tied for 3rd with 0 citations. Part of the possible reasons: inertia, not everyone buys/downloads the latest and greatest more than once a year or two; Houdini/Komodo are [or at least were] easier to use with analysis, and/or their evals are more believable (particularly pawn structures is a Komodo edge I think); psychological effect of thinking that something you pay for must be better.
So let's look at Google. Simple search count for:
* Houdini chess: 415,000 results
* Stockfish chess: 344,000 results
* Komodo chess: 247,000 results
So, you can see that SF and K are not that far behind H, in fact.
"Talk is cheap. Show me the code." -- Linus Torvalds.
-
- Posts: 616
- Joined: Thu May 19, 2011 1:35 am
Re: Engine popularity
Houdini sat at #1 for a long time.
I guess that if you check inquiries over the past month things will be more in favor of both Komodo and Stockfish.
Eventually, the strongest engine will always float to the top.
I guess that if you check inquiries over the past month things will be more in favor of both Komodo and Stockfish.
Eventually, the strongest engine will always float to the top.
-
- Posts: 1242
- Joined: Thu Jun 10, 2010 2:13 am
- Real Name: Bob Hyatt (Robert M. Hyatt)
- Location: University of Alabama at Birmingham
- Contact:
Re: Engine popularity
User923005 wrote:Houdini sat at #1 for a long time.
I guess that if you check inquiries over the past month things will be more in favor of both Komodo and Stockfish.
Eventually, the strongest engine will always float to the top.
This points out "the problem" with LFU (least frequently used) page replacement in virtual memory systems. Once a page has been referenced a bunch of times, it will always be more frequently used than a brand new page. One has to somehow age the results so that the frequency counters are halved (or something) periodically so that pages with high counters, but no recent activity, eventually get dumped. Takes a good while for something new to replace something old when counting past history.