Advice on an electronic board or chess software
Sort:
itrenix
Dec 10, 2020
0
#21
Yes that’s one thing I am not liking about the DGT Centaur at the moment. Hoping it gets better as it “adapts” to my strength.
Dec 10, 2020
0
#22
The Centaur in my case does not do it every game. The problem the Centaur faces in regards my ability is that I am inconsistent. I can find a couple of best moves and can play good sound moves or three but then I play a series of poor moves.
This means that the combined positional and piece score can vary wildly and the Centaur’s programming is not sophisticated enough to deal with this other than dropping a piece to even up the score.
In other words, when the Centaur makes a kamikaze move, it is because my play is poor that particular game - I am playing inconsistently.
Proof of concept is that if I play the Centaur against one of my other chess computers set to a normal play mode, the Centaur keeps the position and piece score very even throughout and does not make any Kamikaze moves.
So, yes, the Centaur will not play those strange moves if you are able to play more consistently.
This means that the combined positional and piece score can vary wildly and the Centaur’s programming is not sophisticated enough to deal with this other than dropping a piece to even up the score.
In other words, when the Centaur makes a kamikaze move, it is because my play is poor that particular game - I am playing inconsistently.
Proof of concept is that if I play the Centaur against one of my other chess computers set to a normal play mode, the Centaur keeps the position and piece score very even throughout and does not make any Kamikaze moves.
So, yes, the Centaur will not play those strange moves if you are able to play more consistently.
Dec 10, 2020
0
#23
Yes I agree. What I have been strict to find out is if the Centaur adapts during each game, or get time with many games, as that would make a difference here.
I am currently enjoying playing against other engines at the moment, and although I returned my DGT e-board I am now awaiting delivery of the plastic version of the DGT smart board as I miss using it for game entry etc. Aw well, I am sure you will improve and have many enjoyable games to come :)
I am currently enjoying playing against other engines at the moment, and although I returned my DGT e-board I am now awaiting delivery of the plastic version of the DGT smart board as I miss using it for game entry etc. Aw well, I am sure you will improve and have many enjoyable games to come :)
Dec 10, 2020
0
#24
I would progress faster if I played more often and made efforts to study the game. However, I am happy to amble along and enjoy the game as a game rather than an educational pursuit. I have improved lately since going back through my games with the Analysis mode on the King Performance and Chess.com.
Whether the DGT Centaur alters its’ difficulty based on a single game or over a series of games has been hotly discussed on the forums. People have contacted DGT to ask, including myself. They say that it adjusts each game, starting afresh each time. However, there is circumstantial evidence that suggests the Centaur retains its strength setting from the previous game - particularly if people of widely varying abilities try to share the same Centaur.
I noted that when I played my Centaur against one of my chess computers that the next game it played against me it played a more capable game. Then I lost a series of games so used what seems like a level reset cheat by taking back a move during a game. Then the Centaur lost a game and after that played a more even game.
The occurrence could have been circumstantial of course but my conclusion is the level setting system of the Centaur is a bit of both - within a game and carried forward to the next game.
Whether the DGT Centaur alters its’ difficulty based on a single game or over a series of games has been hotly discussed on the forums. People have contacted DGT to ask, including myself. They say that it adjusts each game, starting afresh each time. However, there is circumstantial evidence that suggests the Centaur retains its strength setting from the previous game - particularly if people of widely varying abilities try to share the same Centaur.
I noted that when I played my Centaur against one of my chess computers that the next game it played against me it played a more capable game. Then I lost a series of games so used what seems like a level reset cheat by taking back a move during a game. Then the Centaur lost a game and after that played a more even game.
The occurrence could have been circumstantial of course but my conclusion is the level setting system of the Centaur is a bit of both - within a game and carried forward to the next game.
Dec 10, 2020
0
#25
Really? That’s very interesting. I wanted to run some experiments myself but it looks like you have already proved what I wanted to discover. Either way I do like the concept of the Centaur for what it is as I feel we are lacking somewhat in the chess computer market considering how far we have progressed with technology in general, in my opinion.
I may even consider the KP as my next purchase :)
I may even consider the KP as my next purchase :)
Dec 11, 2020
0
#26
What we have so far is confirmation bias - no actual experiments to prove or disprove either way.
If I were to design an experiment to test the theory then I would pitch the centaur against a capable though adjustable chess engine under these three scenarios:
Single average strength setting of the engine across the entire set of games (The control)
Same strength setting at the start of the game changed to Easy strength midway through the game for the set of games - noting any differences in the style of play, specifically how Centaur opens and plays the games compared to the first set.
Easy strength at the start of each game changing to original setting midway through the game for the set of games, noting any differences in style of play of the Centaur from the first set.
The sample size (number of games) does not have to be massive though less than three each section would perhaps be too small a sample.
If DGT have it right that the centaur starts a fresh each game then the Centaur should play the same standard of opening in the second series to the first. If it plays a weak opening or plays blunders early on in the games of the second series then this would strongly suggest the Centaur has retained the level setting of the previous game.
If in the third series the Centaur surges ahead into the mid game each game then that to me would again suggest it set out with a stronger level setting from the previous game in that series.
Perhaps the experiment would be incomplete without a second control of a series of games with the chess engine set to easy to test of random variability in Centaur’s play at weaker levels from game to game.
At the moment I am hard pressed for time to carry out the experiment but if I wasn’t, that’s how I personally would do it.
If I were to design an experiment to test the theory then I would pitch the centaur against a capable though adjustable chess engine under these three scenarios:
Single average strength setting of the engine across the entire set of games (The control)
Same strength setting at the start of the game changed to Easy strength midway through the game for the set of games - noting any differences in the style of play, specifically how Centaur opens and plays the games compared to the first set.
Easy strength at the start of each game changing to original setting midway through the game for the set of games, noting any differences in style of play of the Centaur from the first set.
The sample size (number of games) does not have to be massive though less than three each section would perhaps be too small a sample.
If DGT have it right that the centaur starts a fresh each game then the Centaur should play the same standard of opening in the second series to the first. If it plays a weak opening or plays blunders early on in the games of the second series then this would strongly suggest the Centaur has retained the level setting of the previous game.
If in the third series the Centaur surges ahead into the mid game each game then that to me would again suggest it set out with a stronger level setting from the previous game in that series.
Perhaps the experiment would be incomplete without a second control of a series of games with the chess engine set to easy to test of random variability in Centaur’s play at weaker levels from game to game.
At the moment I am hard pressed for time to carry out the experiment but if I wasn’t, that’s how I personally would do it.
Dec 11, 2020
0
#27
Sounds interesting. Reminding me of my university days setting up experiments :)
It would certainly be interesting to see how it goes, unfortunately I don’t have the time at the moment as well but maybe after Christmas. Could be a fun project.
It would be interesting to see what technology is being used inside? Perhaps there will be a tear down video on YouTube at some point although it’s likely to be a mass produced RISC or similar very specific chip. Who knows.
I still find it odd why the Centaur neither displays info on current ELO or provides such extra options regarding levels etc. I get that they want an easy to use model, however, at that price point surely they could have at least provided some functionality buried in an advanced options menu?
It would certainly be interesting to see how it goes, unfortunately I don’t have the time at the moment as well but maybe after Christmas. Could be a fun project.
It would be interesting to see what technology is being used inside? Perhaps there will be a tear down video on YouTube at some point although it’s likely to be a mass produced RISC or similar very specific chip. Who knows.
I still find it odd why the Centaur neither displays info on current ELO or provides such extra options regarding levels etc. I get that they want an easy to use model, however, at that price point surely they could have at least provided some functionality buried in an advanced options menu?
Dec 11, 2020
0
#28
There is a thread on here about Centaur where someone opened up a Centaur and answered the question about replacing the battery while about it. They post details about what’s inside.
Dec 11, 2020
0
#29
If the updates that Millennium have announced pan out for them I am sure the guys at DGT will give themselves a couple of face palms over not making the Centaur updatable.
Dec 11, 2020
0
#31
Well I am off to take a look at my new arrival. Just this sec took delivery of my Zegrab set :)
Scwashere wrote:
I haven't seriously played chess since my high school years. I'm not a beginner, but in those days, I won the high school contest, travelled to other schools, did well, and once played one of those series where many of us played a ranked player, and I won. However, I haven't played much at all since those days. I watched "The Queen's Gambit," and it's reignited my interest.
I want something that will help me relearn and improve my skill level. I'm finding with the apps on my phone, I either win every time, or on higher levels, it wins(same with computer versions). Either way, it does nothing to improve my game, and that's what I'd like to do versus just passing the time.
I prefer an electronic board versus staring at a computer screen, and I looked at the DGT Centaur(read a post here about it), but I'm not familiar with other offerings or companies for that matter.
Any recommendations?
I recommend an e-board that you can hook up to a PC. This way, you have more flexibility with its chess engine selection. Something like the regular DGT e-boards that can play with Fritz and interface with many online chess sites. This will allow you upgrade the chess software easily.