Author: KarinsDad
Date: 22:14:05 01/17/99
Go up one level in this thread
I agree with John. The commercial programs that I have used have done a "lousy job" of coming up with different playing levels to challenge the consumer. It has mostly been: 1) Get a high powered engine 2) Throw in some graphics, short sets of annotated games, etc. 3) Add some neat interfaces to databases and ways to play against other computers. 4) And oh yeah, make a couple of ways to make the program weaker so that weaker players can play against it. For example, the annotated games by Josh Waitzkin on CM6000 are a joke. Not in their ability to instruct, but rather in the fact that you can run through them all in a few hours. Why give instruction if your average player doesn't gain more than a limited set of ideas and then runs out of games to look at. If you are going to give instruction, give THOUSANDS of annotated games. Who needs dozens of different chess pieces? Nobody. This functionality is silly and a waste of hard disk space. If you are going to make some form of levels (like CM6000s multiple personalities), at least have the 1200 one play 1200 chess in the opening, the middlegame, and the endgame. Not 1000 opening and middlegame and then 1500 endgame (which is different than at the 2400 level where some of them play 2500 until the endgame when they suddenly get dumb unless they have tablebases). How can this be done you may ask? Research. Just like any other part of the program. Have the program examine 2000 1200 level (and 1300, 1400, etc.) games and do an analysis to find out what types of mistakes are made. Then, code the chess program to make similar mistakes when playing at that level. Is this a lot of work? You bet. But it is work aimed at improving the usability of the program for 99% of the people who buy it? Yup. Do you really think that 99% of the people who use chess program really gain an advantage of the analysis engine of a 2550 program versus a 2575 program? The engine strength wars are kind of like the cold war. Not needed by the majority of the people on the planet, but the part focused on by the majority of interested people. KarinsDad On January 16, 1999 at 15:13:55, John Coffey wrote: >On January 16, 1999 at 14:36:33, Laurence Chen wrote: > >>On January 16, 1999 at 14:23:10, John Coffey wrote: >> >>>I think that programs would benefit from playing different levels on the >>>internet and looking at the ratings obtained there. >>> >>>With recent programs there has been too much emphasis on using ratings >>>as a method of determining the level of difficulty. This is a good >>>marketing feature, but programs have shown an inability to predict >>>ratings with any accuracy. >>> >>>I think that some programs should consider going back to having >>>"levels of difficulty." These should start at such a easy setting >>>that any player, no matter how bad, could find an equal opponent. For >>>example, level 1 might only look at 30 possibilities. Level 2 might >>>look at 60. Level 3 would look at 120 and it would continue to increase >>>like this with each increase in level. This idea would make for a very >>>interesting chess program. >>> >>>John Coffey > > >>I think not. If you want a weak chess engine then go buy a chess engine which is >>weaker. Nowadays all the top 10 chess engines in the SSDF plays at GM strength >>and I can not see why would anyone want to buy a chess engine which plays at >>amateur strength. So the best solution for you either to give up chess or start >>studying the books and improve your game. Come now, people buy strong chess >>engine rated at GM to help them in chess analysis, and if you happen to get one >>of the top 10 chess engines, and cannot beat them, so the problem is you not the >>engine. > >This is incorrect for lots of different reasons. I am glad that chess >programs can play at top levels because I too use them to analyze games and >I do learn from them. I *have* been studying books for the last 24 years, but >to say that I should give up chess because I cannot yet beat a GM equivalent >program is just absurd. To say the problem is with me because I can't beat them >is hard to understand. > >*Most* people buy chess programs to play games with them. Most commercially >successful programs make some pretense of offering levels that most people >can enjoy playing, but they have not done a very good job of providing a wide >range of levels. Instead they have been more concerned with beating other >chess programs than they have with providing a good experience for their >customer... which might be a low ranked tournament player. (Even most weaker >chess engines, say Chess Master 3000, has too few levels that would be enjoyable >to the average tournament player.) > >A chess program should be a fun experience for all who purchase it. It has >been my observation that 90% of the people who bought programs do not play >those programs despite the fact that it was their intention to do so. Why? >Because the programs play too tough. They face nothing but frustration. > >The previous poster seems to think that i am saying that chess programs should >not be top level. I have said nothng of the sort. It is my contention that >it should not be too difficult to provide levels that would allow a program >to play all the way from 0 to whatever the top strength is. The fact that >most programs do not do this is to me missing a marketing opportunity and a >chance to bring more players into the chess world. > >The previous poster misses the point that one of the best ways to improve >one's chess is to be able to select a computer opponent equal to one's >strength and then try to learn how to beat it. Once that goal is obtained >then the player can learn how to beat the next level. I used to be able >to do this with dedicated units that i purchased, and I saw my rating rise >considerably. > >The previous poster is most rude in his response, and seems to imply that I >should be a master or senior master before I consider buying a chess program. >How many programs are you going to sell that way? > >Best wishes, > >John Coffey
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.