Author: Stephen A. Boak
Date: 05:26:56 02/14/01
Go up one level in this thread
===>> I wonder what other users think about the Game Analysis change? <<=== Embedded below are my detailed thoughts. >On February 13, 2001 at 21:37:04, John Merlino wrote: [snip] My post game analysis mind and routine works like this (maybe it is just my conditioning based on how my programs have operated in the past, including CM6000, several Rebel versions, and Fritz6, or maybe it is a simple and natural way of reviewing a game after Game Analysis has been prepared by the computer): 1. Look at a real position from a played game. If the position was challenging during the actual game or is very interesting at time of review, I rehash mentally what are the candidate moves/lines of play (per memory of my thoughts during the actual game or by 'fresh' look during post game analysis). 2. Note (without moving) what the actually played move was after that position was reached--preparing to compare it to the computer suggested best move in that position). I say to myself something like 'In this position, I played xxx, trying to ...', or 'In this position my opponent played yyy, I guess because he was trying to...' At this moment I don't normally make the actual game move on a real board or on a board on the screen. Why? Partly because I remember it from the actual game and/or I see it in the move score as part of the Game Analysis. Partly because I am more interested in what another stronger player (the chess program!) would have done in the situation, if facing the same real game position that either I or my opponent had to handle over the board. 3. I then look at the computer suggested best move and see if it agrees with the noted move that was actually played in the real game. I treat this move as though a much stronger player has just told me 'Here's what I would play in this position, and what I think it would lead to'. 4. To carry out the Step 3 comparison, I want to readily look on the *same line* as the played move shown in 2, above--to make an instant comparison of the move actually played with the best move suggested by the program. This is the purpose of the program (for me)--to show what the best move is, for direct, immediate, and easy comparison with the actual move played. It is much easier and totally different to simply slide my eyes to the right of the noted move played in the actual game, to compare it with the best move (suggested by the computer) [as in CM6000 and my other chess programs], than to move my eyes up a line, to the prior game move, and then slide my eyes right to the computer analysis shown there [as I have to repeatedly do with CM8000]. As you see, I wish to directly (easily) compare the actual move that was made in the game with the best move/line/score suggested by the computer, over and over and over, as I go through a previously played game using the assistance of Game Analysis by a program. 5. After comparing the suggested best move with the noted actual move played in the real game: A. If the suggested best move is the same as the actual move, I assume the player indeed made the best available move (trusting the computer in most cases). I make that move and go to Step 1, above, examining the next position from the real game. B. If the suggested best move is *not* the same as the actual move, I play out mentally, or on a real board or screen board, the computer suggested move & best line. I try to figure out the 'idea(s)' behind the suggested computer move and best line of play (whether tactical, positional, or a combination of both) and I compare and contrast what I reason out (based on a strong program showing me its own suggestions) versus my own thinking during Step 1. After cogitating on understanding the computer line of play and making the desired move/line/idea comparison as much as desired, I back up the computer moves (if made on a real board or screen board) and then play the actual move from the game and go to Step 1, above, examining the next position from the real game. > >2) It was decided that, of all possible ways of displaying the data, this was >the least confusing. The line can simply be described as "best play" from the >current position. Maybe I missed some mention of the pending Game Analysis change for CM8000 (to be included in the patch), but I think it may be an example of a simple change by a development team, not without reason (it was genuinely believed to be an improvement, and more straightforward to program), of a basic aspect that may not be in accord with general customer reliance and satisfaction. It would have been great to raise the proposed change to allow discussion in advance. If that was done, sorry, I simply overlooked that facet. A couple of additional points-- That customer reliance may not exist the way I perceive it may exist (because I realize my own views may not be the same as most users), but I note: 1) All other programs that I have owned until now (including CM6000) do it the way I have been used to, which is a method very pleasing and easy to follow during review of Game Analysis. I guess other programmers thought the same way was best (or it became a de facto standard), and I have never heard users complain about that method. 2) Even the Continuous Analysis modes in those programs continually show the suggested best move/line/score for the shown board position, without *first* making the actual move played in the game being reviewed. In other words, the suggestion is made before the move is made (not afterwards, as it were). I realize this is a bit of 'which came first, the chicken or the egg', but I think it supports my thoughts a bit at least. > >3) No standard. It was just decided this was the best way to do it. It also >corresponds exactly to engine analysis output, which made it easier to program. Maybe there is a de facto standard, if not a formal one, or a more natural way for the users, even if display philosophies are debatable (with pros and cons for more each plausible method). Simply showing the suggested best move/line/score on the *subsequent* move line, after the actually played next move, would do the trick with virtually no programming fuss at all! Then the all-important comparison can be made between the move played and the move suggested as best--all on the same line! > >4) The score that is shown IS for the actual move played, but it also assumes >the "best play" line as well. So, to be specific, it is the score for 8...Bxf3. > Thanks for your thorough and honest explanation--as usual! I hope I have shown my point of view as clearly as your answer. I can adapt to the changes--they are not the end of the world, and we have always had to adapt to eccentricity of programs, oftentimes with enthusiasm as some programs had improved features over others!--but these changes now make CM8000 different than the other major programs I use to do overnight analysis and subsequent review of my games. In my opinion the program has become more difficult and cumbersome to use to review Game Analysis output, because of the chosen new method of display; but, I'd also like to hear what some other users think. My joy in analyzing (overnight analysis is my favorite use of a chess program) is greatly increased when the software tools make it easy to learn better moves and better ways of handling positions I have faced over the board or when helping friends analyze their most difficult or interesting games. >You're correct. It can be confusing either way, depending on HOW you want to >look at it. But, combined with the move list in the Game Status window, it was >felt that the CM8000 way was better overall. > >I hope that answers your questions. If not, please feel free to ask more! Take care, and thanks for such a great job in pre-patch discussions with users. You and your development team have made such a supercalifragilisticexpialidocious :) (sp?) effort to handle so many bug reports, criticisms and suggestions (including my own request for thinking times to show in Game Analysis output--which you have incorporated!) that I can't complain. However, if you are willing to listen, I am willing to contribute suggestions! --Steve > >jm
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.