Chessmaster 10 ratings

Hello,

  In the old game Chessmaster 10 you can find plenty of character with 'ratings' assigned to them. Do you think this ratings can be in any meaningful way associated with the FIDE ones, or is it just rubbish?

I think there is some association.  The number of mistakes that the characters make statistically resembles that of a play with that FIDE rating.  However, the moves are so egregiously bad, that I don't think there can be any true correlation to FIDE and USCF ratings.  I started gaming the system against one of the 2000 characters when I was only 1300, by making the game go enough moves to induce the major error.  The bad moves often have no rationale.  Maybe they are programmed to randomly select the nth best choice on move m, but the nth best choice doesn't fit into any plan and is sometimes reduces the computer's position to rubble in a single ply.  This has been my experience. 

I do enjoy the styles that the characters mimic, though.

No relation. In general, there's no good elo estimate through games against engines, because none of them can emulate human play well enough.

Thank you very much. I think you both are right (as for Milliern remark I had a similar impression when a 1600 opponent after the series of good, as far as I can assess, moves just blundered a knight...)

In chessmaster you can edit the personalities, they have one control called random or randomness, if you put that to zero it should eliminate the weird blunders they sometimes make.

But doesn't that also change their overall strength?  I was chatting to a friend, a professor of software design at Pitt, and he said that the chief difficulty is matching the kinds of errors made by, say, a player rated USCF 1600; so, without those random mistakes with a given frequency, the personalities become much stronger.

Also, that's not a feature on CM9000, right?  Is it on the GM edition?  I recently sold my CM9000 and got the GM edition.

They correlate in this respect, an 1800 "personality" is stronger than a 1400 "personality". But these personalities are so non-humanlike that you should not otherwise infer any similarity. I've played several dozen games with Vlad. "He" always makes a positional error on move four or five. Very few USCF players under 1300 make similar positional errors. Once Vlad gets an advantage, he will make no tactical errors. Most FIDE players up to and including 2300 will make occasional tactical mistakes in winning positions.

Hello, Milliern, would you mind unblocking me please?  I apologise for any childish stuff I might have done that lead to this block. Thank you. 

To what end?

Excuse me?

Editing the personalities does change the strength. Thirty years ago, programmers were grappling with the challenge to create software that could play at master level. They accomplished that task. There is far less motivation for programmers to realistically dumb down their software. Chessmaster's programmers came up with tweaks that have satisfied many casual users, but that fails to meet the needs of serious players looking for training. Fritz and family has a much simpler method for reducing engine playing strength, and it results in much more realistic play.

For what purpose?

So I'll be able to post in your threads. Sometimes I come across an interesting one and I'm not able to post in it.

True, however there are other configurable options like skill and how many moves ahead it can look, you could lower those to reduce it's difficulty without letting it make random blunders that do not make sense. Trial and error there may be worth considering.

This is definitely a feature in the GM edition, however I am unsure if it is present in any other edition.

Those features have been present, as I recall, since the first Windows version of Chessmaster. The problem then and now is that Chessmaster''s dumbing down features have never been as good as those in Fritz, but that the "personalities" appeal to casual chess players who lack the guidance of a competent coach.

I'm an empiricist, and the empirical data I've collected thus far indicates that your comments will not be missed.  That's why you are on the list.  Thanks for your interest.

You're the most passive agressive person I ever met.  

On an unrelated note, I knew immediately you were a philosopher just by the content of that comment.

I felt I was straightforwardly aggressive enough.

That statement did not reflect my philosophical position.

It reflected what you are as a person.

Well, have a nice day. Try to go for a walk.

And yet I was the one with the sense to block and ignore you, until your request.  That the lower minded should block and ignore the higher minded, what a wonderful world the the higher minded would live in!  You may return to the masses of nameless, the ignored, the forgotten  --in short, the blocked.