Game analysis using statistics

Hello, Kostja approached me with an exciting idea: almost half a million chess games and almost the same number of Russian checkers games have been played in Vint to date. This is a large enough number of games to draw statistical conclusions based on them. How to use the available amount of information sensibly? One idea is to add additional functionality to the game replay: in each game state, it would be possible to see what the most popular move in that state has been (either the most popular numerically or the most games won by continuing with that move). This could be applied to all board games (Gomoku, Rendzu, Fischer's chess, Othello, etc.). Such functionality would provide a good learning opportunity: When analyzing your own games, you could see after which move your chances of winning became close to 0, which move would have been the best instead, etc. Such an analysis would require a lot of server resources (you can imagine that you need to analyze 500,000 games with all the obstacles in these games), which is why our server would not be able to handle it if all users started using this functionality. Therefore, this would be an additional functionality for VIP users. What do you think of the idea?

Then it would be a good idea to organize tournaments only for VIPs in the future - what's the point of playing if you open the index and only play the best moves. I think that's definitely not justified on a portal like this.

In addition, many players have developed their own variations and quirks over the past 10 years of playing. Those who have analyzed and worked hard have naturally found a cure for them, but letting the computer waste all the work done in previous years is unacceptable.

It should be relatively easy to implement so that it cannot be used during the game. As for the "tricks" invented by one user over the past 10 years, I can only say the following: Due to the spread of computers, professional chess / checkers has become more of a memory game. If Keres went home after the game and analyzed the position with friends for weeks, today a computer can analyze the position in seconds. So there is less creativity - the one who can remember the opening tree the longest wins. Play Fischer chess! Maybe we should also make Russian checkers / rv checkers with draw openings? Let's take the first row empty of buttons, for example :)? This is going too far from the original topic, let's still focus on the original proposal.

I don't dispute that computers have taken over, but let's respect players who have never used the program, and don't plan to!

In this regard, I'm in the same position as you - I've never used a checkers/chess program. However, the question "What went wrong / what could have been done better" has sometimes arisen. This new functionality would help answer these questions. This new functionality would still be for learning, not cheating!

It would be really exciting to see the results!

I very much agree with Kostja's suggestion. It would certainly add a lot to the analysis. Vint is also a place where players go to practice and later, when analyzing their games, it is interesting for beginners (and also advanced players) to see where they could have played differently and what moves were good. I also agree with the idea of making opening moves in checkers with a draw. At least in my case, it would make playing more interesting and test the players' skills more. But I don't really understand the train's opinion. It doesn't matter whether you can analyze games on the Vint page or use some other computer program where you put in the position and let the computer choose the best move. There are certainly users who use programs when playing and the fact that Vint adds the option to analyze games doesn't affect anything. This is my opinion and it is certainly not the same as everyone's, but I believe that if Vint also allows games to be analyzed, it won't change anything. It simply makes analysis more convenient for users because they no longer have to use other programs.

I think the ability to analyze lots in Vindis is good. Those who want to use the opportunity immediately, those who want to later after their own analysis.

Lies were once classified in ascending order as follows: Small lies, big lies, and statistics. That's it, but there are several misleading reasons in the statistical analysis of games: 1. Games that ended in a loss of time, which does not determine the quality of the game 2. Player level. Beginners can play many low-level games. 3. How to statistically determine at which stage of the game success was achieved? Is it in the opening, middle game, or end game? Most games are won in the end game. 4. There are even more of these reasons. What is the use of misleading statistics? Rather, it is a loss for those who are learning to play.