radmagichat wrote:
Also I play a random bot and set it so the name or rating is not revealed to me until after the match
How do you do that? Is the rad magic hat somehow involved in the process?
How strong is Chessmaster 10th Edition?
Sort:
zBorris wrote:
Here's a list of chess rating strength ranked in order of ELO from CCRL and IPON:
Houdini 4 --- 3121 Stockfish DD --- 3071 Komodo TCECr --- 3059 Gull 2.2 --- 2987 Critter 1.4a --- 2983 Equinox 2.02 --- 2978 Deep Rybka 4.1 --- 2969 Deep Fritz 14 --- 2900 Chiron 2 --- 2897 Magnus Carlsen --- 2881 Hannibal 1.4b --- 2879 Naum 4.2 --- 2840 (rank 59th) Chessmaster 11 2CPU --- 2722CCRL is worthless. They use 40/40 time controls. TCEC is the real deal; Stockfish and Komodo have been better than Houdini for a long time now.
Here's Stockfish crushing Houdini in 29 moves: http://tcec.chessdom.com/archive.php?se=6&st=4&ga=54
Jun 4, 2014
0
#64
btickler wrote:
zBorris wrote:
Here's a list of chess rating strength ranked in order of ELO from CCRL and IPON:
Houdini 4 --- 3121 Stockfish DD --- 3071 Komodo TCECr --- 3059 Gull 2.2 --- 2987 Critter 1.4a --- 2983 Equinox 2.02 --- 2978 Deep Rybka 4.1 --- 2969 Deep Fritz 14 --- 2900 Chiron 2 --- 2897 Magnus Carlsen --- 2881 Hannibal 1.4b --- 2879 Naum 4.2 --- 2840 (rank 59th) Chessmaster 11 2CPU --- 2722CCRL is worthless. They use 40/40 time controls. TCEC is the real deal; Stockfish and Komodo have been better than Houdini for a long time now.
Here's Stockfish crushing Houdini in 29 moves: http://tcec.chessdom.com/archive.php?se=6&st=4&ga=54
The CCRL testers will be disheartened to hear that their work is all for naught. 
But seriously, both CCRL and TCEC provide useful but different information. CCRL provides information on more engines and more games at the cost of faster time controls, while TCEC provides information on fewer engines and fewer games, but at a longer time control. The critical difference is that CCRL also provides statistical data to help the reader decide if the information is statistically significant; TCEC doesn't provide that information.
EscherehcsE wrote:
btickler wrote:
zBorris wrote:
Here's a list of chess rating strength ranked in order of ELO from CCRL and IPON:
Houdini 4 --- 3121 Stockfish DD --- 3071 Komodo TCECr --- 3059 Gull 2.2 --- 2987 Critter 1.4a --- 2983 Equinox 2.02 --- 2978 Deep Rybka 4.1 --- 2969 Deep Fritz 14 --- 2900 Chiron 2 --- 2897 Magnus Carlsen --- 2881 Hannibal 1.4b --- 2879 Naum 4.2 --- 2840 (rank 59th) Chessmaster 11 2CPU --- 2722CCRL is worthless. They use 40/40 time controls. TCEC is the real deal; Stockfish and Komodo have been better than Houdini for a long time now.
Here's Stockfish crushing Houdini in 29 moves: http://tcec.chessdom.com/archive.php?se=6&st=4&ga=54
The CCRL testers will be disheartened to hear that their work is all for naught.
But seriously, both CCRL and TCEC provide useful but different information. CCRL provides information on more engines and more games at the cost of faster time controls, while TCEC provides information on fewer engines and fewer games, but at a longer time control. The critical difference is that CCRL also provides statistical data to help the reader decide if the information is statistically significant; TCEC doesn't provide that information.
I'd be fine with CCRL if they didn't keep perpetuating the Houdini is #1 myth...maybe put a link to TCEC on their front page ;).
Jun 5, 2014
0
#66
btickler wrote:
I'd be fine with CCRL if they didn't keep perpetuating the Houdini is #1 myth...maybe put a link to TCEC on their front page ;).
Check out the June 4th list for 40/4:
http://www.computerchess.org.uk/ccrl/404/index.html
(The 40/40 list will probably be slower to reflect the changes due to taking longer to play the longer time control games.)
EscherehcsE wrote:
radmagichat wrote:
Also I play a random bot and set it so the name or rating is not revealed to me until after the match
How do you do that? Is the rad magic hat somehow involved in the process?
Hmm an example would be. I click play rated game. When the menu comes up asking you which personality you would like to play. Check the two boxes that say "random" and "hidden". The point of this is so you don't get distracted by knowing who you are playing and what they are rated. You Just play chess to the best of your ability. Its not really that impressive being able to beat the computers up to that level.. When I was offline for a while and had a rating increase in the game.. I thought I was gonna be able to dominate real people. Not the case. If you blitz at about 1900 lvl in chessmaster. I would say you should be able to blitz anywhere between 1100 and 1400 here at any given point.
Jun 12, 2014
0
#68
radmagichat wrote:
EscherehcsE wrote:
radmagichat wrote:
Also I play a random bot and set it so the name or rating is not revealed to me until after the match
How do you do that? Is the rad magic hat somehow involved in the process?
Hmm an example would be. I click play rated game. When the menu comes up asking you which personality you would like to play. Check the two boxes that say "random" and "hidden". The point of this is so you don't get distracted by knowing who you are playing and what they are rated. You Just play chess to the best of your ability. Its not really that impressive being able to beat the computers up to that level.. When I was offline for a while and had a rating increase in the game.. I thought I was gonna be able to dominate real people. Not the case. If you blitz at about 1900 lvl in chessmaster. I would say you should be able to blitz anywhere between 1100 and 1400 here at any given point.
Well, I guess that'll teach me to make fun of someone's magic hat! Yes, you're correct about the random and hidden boxes; I'd never noticed those before. (I don't play a lot of rated games in Chessmaster.) I'll have to try out those features sometime.
Jun 12, 2014
0
#69
radmagichat wrote:
In all fairness.. I usually get crushed by any bot over 2200. However, I change the difficulty settings to maximum
and give myself and the bot 10 minutes a piece. Also I play a random bot and set it so the name or rating is not revealed to me until after the match
I don't play for tactics at all because it does not work. Eventually, if
my position is solid they will make a blunder trying for a tactical shot or start trading down pieces
and then I get into an end game situation with the bot...
I don't know if you ever played an end game against a chess master bot rated 2100ish or less but there is an obvious rating drop by atleast 1000 points.
They start to give up material with out much effort or thought.
Your observation about the personalities being weak in the endgame intrigues me. The 1900-2100 range is above my pay scale, so I haven't really played against those personalities in that range. I just wonder what could be the reason for this weakness? Does the King engine just not have much endgame knowledge programmed into it, or did the programmers disable some of that knowledge at the lower elo levels?
If I can find some time, I might try to duplicate your observations.
One question though about the personalities' weak endgame - Do you recall if all of the personalities in that 1900-2100 range had the weak endgame, or was it just some of the personalities? Any particular personality names come to mind? Also, have you noticed this effect in the lower ratings below say 1800 or 1900?
EscherehcsE wrote:
radmagichat wrote:
In all fairness.. I usually get crushed by any bot over 2200. However, I change the difficulty settings to maximum
and give myself and the bot 10 minutes a piece. Also I play a random bot and set it so the name or rating is not revealed to me until after the match
I don't play for tactics at all because it does not work. Eventually, if
my position is solid they will make a blunder trying for a tactical shot or start trading down pieces
and then I get into an end game situation with the bot...
I don't know if you ever played an end game against a chess master bot rated 2100ish or less but there is an obvious rating drop by atleast 1000 points.
They start to give up material with out much effort or thought.
Your observation about the personalities being weak in the endgame intrigues me. The 1900-2100 range is above my pay scale, so I haven't really played against those personalities in that range. I just wonder what could be the reason for this weakness? Does the King engine just not have much endgame knowledge programmed into it, or did the programmers disable some of that knowledge at the lower elo levels?
If I can find some time, I might try to duplicate your observations.
One question though about the personalities' weak endgame - Do you recall if all of the personalities in that 1900-2100 range had the weak endgame, or was it just some of the personalities? Any particular personality names come to mind? Also, have you noticed this effect in the lower ratings below say 1800 or 1900?
Almost every single bot 2100 below has terrible end game except for josh. If you read his profile it says he is good at endgame.. which oddly enough.. he is really not that great rated 2100 or lower. However if you set up an endgame with 1600 josh and 2097 rick.. josh will win 9 times out of ten.
I haven't really studied the end games in depth per bot. What I have studied is the easiest way to beat most of the bots. Play positional and when you are ready just simplify.. Key feature.. make sure the queens are off.
Here is something to try out if you want a win. 9 times out of ten the computer does not want a double pawn.. set up a tactic in a pre end game doubling one of its pawns.. guess what the computer does? It lets you win another pawn in order to not have a double pawn.
Next thing to consider.. if you have a pawn chain that looks strong.. (example e2 to b5) and are able to hold at least two pawns that far into there position for a few moves.. almost every single bot will give up material.
An additional thing that I would like to point out is bots that are way above their rating level. (not speaking about endgame just play) Vlad, Max, and J.T are pretty good. I remember one time I actually questioned, is vlad really an 1800. I paired vlad against several bots way higher then his lvl and found he was wiping the floor with the gms. Those three bots will beat others out of their rating often.. its pretty interesting
I owned Chessmaster 10th Edition and CM9000 on my old computers, but recently I just installed CM9000 back again into my PC over CM10th Edition. The reason? Because the layout and functions are just more stable and better looking.
CM10th Edition is basically CM9000 but with a stronger Chessmaster engine (a bit more only). That is all. I consider CM9000 to be the best Chessmaster by Ubisoft (haven't tried CM11).
The Vlad personality is fairly strong, because it is actually the main King engine at strength of play 100, just with some tweaks and programmed to play 1 move per second.
As for the usual criticisms that the personalities hang pieces and then play strong, well, actually they are just playing as how computers ALL play. All are strong tactically, but the nice part about the weaker engines is that you too, get to unleash your tactical skill on the engine, because at many points they will allow you to.
Besides, it is good to get used to beating a tactical opponent once you have a material advantage. I don't see that as a negative at all.
Jun 13, 2014
0
#72
In my experience, Vlad makes positional errors reasonably early in every game. It is a tactical monster if it gets an advatage, but in quiet, balanced positions, it can blunder. See 12...Qe7?? below.
Jun 13, 2014
0
#73
Debistro wrote:
I owned Chessmaster 10th Edition and CM9000 on my old computers, but recently I just installed CM9000 back again into my PC over CM10th Edition. The reason? Because the layout and functions are just more stable and better looking.
CM10th Edition is basically CM9000 but with a stronger Chessmaster engine (a bit more only). That is all. I consider CM9000 to be the best Chessmaster by Ubisoft (haven't tried CM11).
I mostly agree with you about CM9000 being preferable to CM10th. CM10th does have the Christiansen tutorials that CM9000 lacks, and CM10th has about 20% more personalities than CM9000. But other than that, CM9000 is less buggy, it's quicker to load, it has a bigger and better manual, and its tournament system isn't limited to only running one game at a time like CM10th.
The only reason I hesitate to recommend CM9000 to people is that I had a devil of a time getting all of the patches installed in my Win7 machine so that I didn't have to run the program with a disk in the drive. (I was successful, but I think I was lucky on this count.)
Hy guys!
I am 23 years old, intermediete chessplayer. I have Chessmaster 11 about 4-5 years... so, a very long time. I agreed Bobby DK "please stop talking about chessmaster 10. why not have the latest and greatest 11.
on my computer Chessmaster® Grandmaster Edition is powered by version 3.50 of The King chess engine. On this system, Chess master is rated at 2999 U.S.
I think the reason you can easily beat the lower rated players in cm is that it will give you some sort of edge to simulate a lower rating and if you are suddenly a piece up it is basically trading down all the time till you won."
So, I have Chessmaster 11 about 5 years, and I noticed that some personality is in CM 11 is much more stronger than the others.... and their ratings isn't important...
For example: I had very hard times againt ORIN, who has USCF rating between 1200-1240 ( their rating always changes), but I can easily beat RISA, who had rating between 1250-1280 ( so, she has much more higher rating than Orin).
Futhermore, I have never beat Will ( rating around: 1390- 1440), my best result againts him is a draw, but I can easily beat almost every personalities between Orin and Will.
In conlusion, In my opinion, the strength of each CM 11 personality can play depends on WHICH personality you choose to play, and their ratings are not important.... Futhermore, I often saw Will (CM personality, USCF rating about: 1390-1440) plays game againts much higher rated CM XI players ( I see a computer vs. computer match on CM XI), and WILL easily beat another personalities, who has much more higher rating ( ex.: 1500, 1535, 1600) etc.
Is chessmaster better then Fritz?
Dec 3, 2015
0
#76
RuneTheKid wrote:
Is chessmaster better then Fritz?
Quite the reverse.
See http://chessskill.blogspot.com/2009/12/chessmaster-versus-fritz-analysis.html
most of the people saying here how in accurate CM rating is but they didnt tell how much time they have given CM to think i conducted the experiment with CM 3 years ago and the results were very interesting i chose 2400+ player and game time was 25min/ 10 sec increment i played 10 games and won by me with score 9.5-0.5 in that time format 2400+ only managed one draw against me then i have played with same opponent but now time control was different now it was 90min/ 30 sec increment which gave CM 2 hours approximately to think now again i played 10 games with 2400+ CM player now the result was entirely different now the result was 8-2 in favour of 2400+ so i concluded it is not the CM ratings r wrong it is actually the fact that we dont give enough time to CM personalities we gave them time like 10 mins and we expect that software will play according to that rating well it is really impossible
Sep 22, 2017
0
#78
Let's put it this way GM Benjamin FInegold lost a play off match with an earlier version of Chessmaster ( I forgot which one) and the program has been upgraded many times since then! It may not be as strong as Komodo, Stockfish, or Houdini but on the grandmaster levels it can definitely handle GMs! However the programmers have admitted to inflating the lower level ratings for a more user friendly experience. LOL..... So a 1600 elo rated player could beat the program at say 1800. Just keep learning from the fee advice given online on sites like chess.com, chesstempo.com, and chessbounty.com, and you should be able to handle the software on the lower levels. 
I put face to face chessmaster 10 vs carlsen app(26 years old) and chesmaster won ,even when the match was fischer character of chessmaste vs carlen app age 26 ,fischer won. I am sure that Chessmaster must have 2900elo
Chess master gm edition is mochary. I have beatin every single bot up to 2100 lvl but am only ranked 1250 against human players. My rating in that game is around 1900 against those bots... big difference. I also played the intermediate study serious, which evaluates all of your skills from, tactics, checkmates, strategy ect... In every single skill set I am ranked above 1600. Then I go and play a game against a human way lower then what my strength indicates.. and get my butt kicked...chess master gm is good for raising your self esteem maybe but has nothing to do with rating accuracy.
just to be curious how long do you let chessmaster think with the below 1900. if you just force it to move it will not be nearly as strong as when you let it think.
I give meself 2 hours and chessmaster 10 minutes in order to avoid the game being too long. if I set chessmaster to use 2 hours it will really use mentor lines to come up with a good move.
In all fairness.. I usually get crushed by any bot over 2200. However, I change the difficulty settings to maximum
and give myself and the bot 10 minutes a piece. Also I play a random bot and set it so the name or rating is not revealed to me until after the match
I don't play for tactics at all because it does not work. Eventually, if
my position is solid they will make a blunder trying for a tactical shot or start trading down pieces
and then I get into an end game situation with the bot...
I don't know if you ever played an end game against a chess master bot rated 2100ish or less but there is an obvious rating drop by atleast 1000 points.
They start to give up material with out much effort or thought.