Same here (app user), really annoying. I want puzzles where I actually need to think. Back to Lichess for puzzles I guess. Their app is improving...
Puzzles too easy
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.
The same algorithm is used. The algorithm is all server side
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.
The same algorithm is used. The algorithm is all server side
Are you sure about that?
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.
The same algorithm is used. The algorithm is all server side
Are you sure about that?
Yes. Puzzles are pulled from the back end. What version is your app?
I will say the app will cache some puzzles, to make it possible to solve some if there are connection issues, so if the app had some cached puzzles then you could see something weird like that.
I was included in the pilot and just switched from "standard" to "hard" and went from seeing puzzles around 1200 to seeing puzzles around 1800. I think I can see why they tried the pilot. I did like the repetition. No idea the affect on my playing game, and that's not why I like to do puzzles either way. But it let me get up to 2700 puzzle rating before I went looking through the app puzzle settings, saw nothing was listed as a difficulty setting in the app settings, and then contacted support and being directed to a chess.com browser page where I could change the puzzle difficulty. I am still getting these new puzzles right. But my thoughts really are these: it should be obvious to the chess.com people by now that, when introducing a new feature, they should not override and eliminate the previous feature. Why can't they co-exist? They butchered game review that way, and I dropped from Diamond to Gold because of that. I also think that the puzzles should automatically increase in difficulty setting once hitting whatever metric. They were getting boring. And lastly, I question if the puzzle ratings are tied to any combination of rapid, blitz, and/or bullet playing ratings. Because the puzzles I was getting are in the range of my playing ratings. And if that's the case, wouldn't it make more sense for the puzzles to be slightly above one's playing rating?
... But my thoughts really are these: it should be obvious to the chess.com people by now that, when introducing a new feature, they should not override and eliminate the previous feature. Why can't they co-exist? ... I also think that the puzzles should automatically increase in difficulty setting once hitting whatever metric. They were getting boring. And lastly, I question if the puzzle ratings are tied to any combination of rapid, blitz, and/or bullet playing ratings. Because the puzzles I was getting are in the range of my playing ratings. And if that's the case, wouldn't it make more sense for the puzzles to be slightly above one's playing rating?
The update has multiple goals. One was a change in the selection algorithm using new ratings. The site recalculated all the puzzle ratings in the back-end and are using the new data for puzzle selection.
The site has specific goals and the old system isn't compatible with those goals
- Provide a more meaningful rating which correlates more closely to player’s Blitz and Rapid ratings.
- Serve puzzles better suited for each player’s level.
- Offer options to member who are looking for different puzzle-solving experiences.
The update has multiple goals. One was a change in the selection algorithm using new ratings. The site recalculated all the puzzle ratings in the back-end and are using the new data for puzzle selection.
The site has specific goals and the old system isn't compatible with those goals
- Provide a more meaningful rating which correlates more closely to player’s Blitz and Rapid ratings.
- Serve puzzles better suited for each player’s level.
- Offer options to member who are looking for different puzzle-solving experiences.
Why not add an "adaptive" option to the puzzle setting options? Since the new puzzle ratings exist on the server, it should be easy to serve up puzzles within a few hundred points of the player's current puzzle rating. This would be almost trivial to implement, and would bring back the functionality that almost everyone who has weighed in on the issue wants.
Why not add an "adaptive" option to the puzzle setting options? Since the new puzzle ratings exist on the server, it should be easy to serve up puzzles within a few hundred points of the player's current puzzle rating. This would be almost trivial to implement, and would bring back the functionality that almost everyone who has weighed in on the issue wants.
It still is adaptive as it is now.
The update has multiple goals. One was a change in the selection algorithm using new ratings. The site recalculated all the puzzle ratings in the back-end and are using the new data for puzzle selection.
The site has specific goals and the old system isn't compatible with those goals
- Provide a more meaningful rating which correlates more closely to player’s Blitz and Rapid ratings.
- Serve puzzles better suited for each player’s level.
- Offer options to member who are looking for different puzzle-solving experiences.
I apologize if I missed it, but how do you know what the new system's goals are? Regarding goal #1, am I correct in understanding that this means the old rating system is completely being scrapped, and the end goal will be to have new puzzle ratings closer to playing ratings? (e.g., no more players having 2000 puzzle ratings with 1200 playing ratings?) Regarding goal #2, how is a puzzle determined to be better suited for each players level? Does players level refer to playing ratings? Or is this generally just a comment toward the beneficial effect of having more accurately rated puzzles upon rerating? Regarding goal #3, I don't understand how the update can be considered to be offering more options with different experiences if it is the one option for players which is replacing the previous option for players. In other words, isn't this update just replacing the puzzle-solving experience, and not increasing the amount of puzzle options available?
I apologize if I missed it, but how do you know what the new system's goals are? Regarding goal #1, am I correct in understanding that this means the old rating system is completely being scrapped, and the end goal will be to have new puzzle ratings closer to playing ratings? (e.g., no more players having 2000 puzzle ratings with 1200 playing ratings?) Regarding goal #2, how is a puzzle determined to be better suited for each players level? Does players level refer to playing ratings? Or is this generally just a comment toward the beneficial effect of having more accurately rated puzzles upon rerating? Regarding goal #3, I don't understand how the update can be considered to be offering more options with different experiences if it is the one option for players which is replacing the previous option for players. In other words, isn't this update just replacing the puzzle-solving experience, and not increasing the amount of puzzle options available?
Tha information came directly from staff. The same information was posted to social media as well.
Yes, the goal is to end up with ratings closer to player game ratings. All the puzzles have been re-rated on the backend (the value showing in the UI is the older rating).
The next updates will be to expose the new ratings and the selection algorithm will continue to pull puzzles from around the player puzzle rating which is planned to be more aligned with game ratings.
I'm not 100% what that looks like but my understanding is it won't be too much longer before those changes are implemented
I don't know how to directly reply, but thank you for your responses Martin. I also would like to add that I just got a puzzle and when I made the first (correct) move, it flashed a brilliancy (!!) next to the move and then faded. So I'm guessing that this might also be a part of the new puzzle experience.
interesting - I guess this throws a wrench in things and explains a lot .. I had been using chess.com predominantly for puzzles, training, library etc and really don't play on the platform very much - I guess I should work on developing a game elo now if that's the basis for determining the level.
what's interesting is it seems (I'm guessing here) that the website uses an overall game elo, but the mobile app (can only speak to the ios app) uses the puzzle elo - it made for an interesting balance in puzzles since I would bump up the puzzle rating at my desktop and then crash it when I was mobile - eventually it kind of balances out, and it is nice getting a mix of both systems to see what patterns and vision I'm weak on .. actually made it more exciting for me since i was kind of stuck in a puzzle rut there and really need to work on my overall game across many different levels .. so thanks (i think) - I like change and the new puzzle experience seems to be a good one that can work out well in the long run
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.
The same algorithm is used. The algorithm is all server side
Are you sure about that?
I have the same experience. Which is good, because the new system is awful, so now I just use my phone
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.
The same algorithm is used. The algorithm is all server side
Are you sure about that?
I have the same experience. Which is good, because the new system is awful, so now I just use my phone
There was a change yesterday. Now puzzles both on my phone and computer are equally low ELO
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.
The same algorithm is used. The algorithm is all server side
Are you sure about that?
I have the same experience. Which is good, because the new system is awful, so now I just use my phone
There was a change yesterday. Now puzzles both on my phone and computer are equally low ELO
Same
-- I think chess-dot-com has become wealthy and is trying new program gadgets (to appeal to younger players?). Often they make the changes "live" without beta-testing.
-- They do respond to feedback, if you persist. When they added an image of a bearded white boy over Game Review, after some message exchanges they made it optional.
-- They are too eager to fix what isn't broken, in my view. Still, chess-dot-com is the best game in town. Be well!
Same problem, but a weird thing is that when I do puzzles on the chess.com app on my phone, the puzzle ratings are way higher.