I think there's some validity to your Observations 1 and 2.
However, even beyond that, the reduced strength settings for many engines are notoriously inaccurate. Some are too high, some are too low. It's just too much work for the engine programmer to calibrate the LimitStrength feature, so most either don't do it or don't do it well enough.
I've indeed noticed that the Rybka LimitStrength feature results in a too-strong engine, so it's not just your imagination.
Thanks for your input.
I notice with Rybka at 1200 ELO it's very strong at opening theory, but in the endgame it performs badly. So maybe it has an opening of 1600 ELO and an endgame of 800 ELO and they average out to 1200 ELO. Trade off pieces and get to the endgame has been my strategy for a lot of games against it.
I was wondering also, do you know of any engines where you can set the ELO and play from a pre-set position? With Rybka you can only play from a set position with it at maximum strength, and if you play a training game where you can set the ELO you can't, and what's more, it makes you play black and white alternately. To consistently beat Rybka at a lower ELO someone need to familiarize themselves with hundreds of openings and variations. I'm not trying to learn all the openings right now, I just want to focus on 2 or 3.
I'm not sure. That problem might be because of your graphical interface rather than Rybka. What GUI are you using?
EscherehcsE nailed it.
Computer ELO values are just rough, rough estimates. Setting it to 1200 doesn't necessarily mean its strength is truly 1200, or even near 1200. It's possibly much more.
Rybka 4 was the last commercial version of Rybka. If you wish to purchase it, that's the version you should buy.
Rybka 2.3.2a was the last free version.
Please note, of course, the latest version of Stockfish is stronger than that of Rybka 4, and Stockfish is free. Critter, Fire, and even Gull are also stronger engines than Rybka 4, and these engines are also free. (Rybka 4 is already a few years old.)
And yes, all of these engines, and many, many more, will work fine in Chessbase 9 and Fritz 13.
choose "Handicap and fun" and it'll be very stupid engine.
If you have Fritz 14 (and possibly other versions) go to ENGINE and then CREATE UCI ENGINE. Click the button and navigate through your folder structure and point to the Stockfish exe file that you downloaded. Optionally, click the PARAMETER button to change some of the default options.
I never play full games against any of my engines, but I do often "spar" with them for a half dozen moves or so. And yes, I also use engines to help examine my own games.
EDIT - Just realised this thread is 13 months old and the OP's account is closed...
don't play computers dude. If you program them to an ELO level of 1200, it plays flawless openings and does complete blunders towards the endgame....and then when you program them to an ELO rating of 2000+, their opening repetoire might seem bizarre, but their endgame is close to flawless and the game will be over before you know it, and you will be frustrated lol.
don't play computers dude. If you program them to an ELO level of 1200, it plays flawless openings and does complete blunders towards the endgame....and then when you program them to an ELO rating of 2000+, their opening repetoire might seem bizarre, but their endgame is close to flawless and the game will be over before you know it, and you will be frustrated lol.
I think you're somewhat misinformed about how chess engines work. It all depends on which engine you're using and how that engine was designed. Most chess engines can have their opening book disabled, either from within the GUI, or at the actual engine level. And when the engine IS using an opening book, SOME engines are programmed to only use the first several moves from an opening book when it's playing at lower elo levels.
Regarding the endgame quality at reduced elo levels, that also is a mixed bag. Some engines play the endgame poorly for the chosen elo level (Shredder is notorious for not being able to execute simple mates like KR vs K at levels around 1400 or lower), but some engines play the endgame believably at lower levels. Very few engines play the engame nearly flawlessly, unless the programmer mistakenly allows the full use of endgame tablebases at lower elo levels.
I play Rybka set to low ELO settings, and I've noticed for some reason I play a lot worse against Rybka than human opponents at the same ELO or even a hundred points higher.
I've been trying to figure out the reason why. Here are some ideas I've considered to explain it, please tell me if you think these ideas are correct.
First, even if I set Rybka at 1200 ELO which is the lowest setting, it still uses a very wide range of openings and follows opening theory correctly for at least the first 7 or 8 moves. I don't think there are many 1200 ELO human players who would posess that amount of opening knowledge.
Secondly, I set Rybka to play instantaneously, both because I don't like to wait, and when it calculates for long periods of time it heats up my computer's processor. And I find when it plays instantanously, it seems to trick me into playing faster too.
Finally, since Rybka isn't a human, I don't feel the fear of messing up as badly so I have less motivation to beat it.
Does anyone else play worse against Rybka or Houdini or whatever engine you use? It seems like Rybka acts as my krptonite and when I play against it I usually do poorly.