Spacious Mind rating test reloaded

[Tibono2]

... and the King Performance levels as well re-uploaded after fix (fun 3 &4 levels did play this move in game 4).
The timer for getting back updating a previous post is so short (or maybe I am a slow man ), I needed to post a new short one instead of completing the previous one, sorry for that.
Cheers,
Eric

[spacious_mind]

Hi Nick,

I am stepping forward with my analysis tool, now duplicating it for test game number 4.
It enabled me discovering an issue in this game 4 with move #25 White: Rxa5 is granted 0 points whilst the G1-DATA is set to 27,90.
The issue is with the score search area:$'G1-DATA'.$CM$2:$CN$39 (should expand to $40)
BR,
Eric

Yes lookup was one short on it. i checked the rest in game 4 and they all seem to be ok. It means we are going to have to re-paste the games played on game 4.

[DaMaBu]

Regarding the King Performance, I frequently experience an interesting phenomenon when playing the lower ELO levels, particularly let's say 1000. Several moves into the game it will make a serious blunder, at which point I know that as long as I play carefully, I'll win. But then it puts up a very solid defense and it's a real struggle to win. It's almost like the old days of chess computers lowering their playing strength by making big blunders a certain proportion of the time but otherwise playing at full strength. But supposedly the KP's Easy levels are accomplished by simply limiting the number of nodes examined. I'm just curious as to how that results in the phenomenon I've described.

[spacious_mind]

Regarding the King Performance, I frequently experience an interesting phenomenon when playing the lower ELO levels, particularly let's say 1000. Several moves into the game it will make a serious blunder, at which point I know that as long as I play carefully, I'll win. But then it puts up a very solid defense and it's a real struggle to win. It's almost like the old days of chess computers lowering their playing strength by making big blunders a certain proportion of the time but otherwise playing at full strength. But supposedly the KP's Easy levels are accomplished by simply limiting the number of nodes examined. I'm just curious as to how that results in the phenomenon I've described.

The King programs have always done this since Chessmaster 3000 started in 1993. If you have the Chessmaster software, you can look up the personalities which I think you might be able to also do with King Performance? I know you can do it with King Element. There are several options which control the strength or weaknesses of a personality. For example in Chessmaster you have sliding scales for:

Style
Attacker/Defender
Strength of play (1-100)
Randomness (1-100)
Max search depth
Selective search
Contempt for a draw

Positional
Material/Position (1-100)
Control of Center
Mobility
King Safety
Passed Pawns
Pawns Weakness

Material
Here you set up yours and your opponent's values for:

Queen
Rook
Bishop
Knight
Pawn

Therefore, depending how you set these you can manipulate the decision making of the program and as a result the strength or weakness of the program. I imagine you can do the same with King Performance. I have two King Element boards therefore for fun I created personalities in the past to see if I can create them stronger then the native King program. Not just with King Element but pretty much with all of Johan De Koning's Chessmaster programs etc.

Therefore, in summary weak opponents have existed since Chessmaster 3000 and all that has happened is that the King Program in Millennium which is probably closely related to Chessmaster 11 adopted some of these for your enjoyment and probably given you the ability to create your own personalities.

I hope this explanation helps you.

Best regards
Nick