Something is off in puzzle ratings

Sort:
goodspellr

I've gotten a number of puzzles lately that have absurdly low pass rates for their rating, and it can not be chalked up to low sample size.  The number of attempts is large enough that any functioning elo-like system for rating puzzles should have stabilized to its rating and a near-50% pass rate.  Here is an example puzzle (996888) with a screenshot of the rating as of Nov 29, 2023

In the comments of the puzzle, another user says that the rating was 2978 at 1533 attempts on October 6, 2022.  That means that the 1669 attempts between 10/6/22 and 11/29/23 somehow brought the rating down (and presumably, but not necessarily, the pass rate up, but still only to 6%). 

What is the reason for this?  Are some more-recent, low-rated users solving this puzzle easily and driving the rating down, or is there some flaw in the algorithm that calculates the rating?

justbefair
goodspellr wrote:

I've gotten a number of puzzles lately that have absurdly low pass rates for their rating, and it can not be chalked up to low sample size. The number of attempts is large enough that any functioning elo-like system for rating puzzles should have stabilized to its rating and a near-50% pass rate. Here is an example puzzle (996888) with a screenshot of the rating as of Nov 29, 2023

In the comments of the puzzle, another user says that the rating was 2978 at 1533 attempts on October 6, 2022. That means that the 1669 attempts between 10/6/22 and 11/29/23 somehow brought the rating down (and presumably, but not necessarily, the pass rate up, but still only to 6%).

What is the reason for this? Are some more-recent, low-rated users solving this puzzle easily and driving the rating down, or is there some flaw in the algorithm that calculates the rating?

It is interesting to follow the comments made in older puzzles to try and understand the ratings of puzzles.

I think this one has an easy first move and a hard second move. I would guess that 90+% of people get the first move and only a very low percentage get the second move.

It's entirely possible that the rating reflects a manual readjustment by Chess.com and not one brought about by algorithm adjustments.

https://support.chess.com/article/2016-how-do-puzzle-ratings-work

How is the rating of each puzzle decided? 

You may notice that when you get a puzzle wrong or right, your Puzzle Rating changes, but the rating of the puzzle does not. This is because the rating of each puzzle is locked after a certain point. 

When a puzzle is newly added, its rating is determined by who is able to solve it, but after a set amount of time, the puzzle rating becomes locked in place and does not change anymore. 

In some cases, if a puzzle rating is found to be too high or too low, Chess.com may adjust the rating to be more accurate. 

Martin_Stahl

This puzzle has been referred to staff and should be out of rotation soon.

goodspellr

Thanks, @Martin_Stahl. Here is another massively underrated puzzle: 1222553. As of 1/16/2024, it is rated 2200 with 3% pass rate and 6211 attempts. I realize it's only 2 data points, but both of the ridiculously underrated puzzles mentioned in this thread have ratings of exactly 2200. Maybe there is a pattern there.

I think that chess.com should allow an option to report the puzzle for being improperly rated. At present, they give the instruction that "You should only report problems where the answer given is incorrect or there is more than 1 correct answer." However, an improper rating is a valid reason to report a puzzle as improper ratings harm the user experience. Maybe provide a dialog box that allows the reporter to select the reason they are reporting the puzzle.

Also, it seems like this would be an easy enough problem for chess.com to root out. If the pass rate is significantly different from 50% with a large number of attempts, then the rating is probably inaccurate and should be adjusted.

justbefair

The one you posted originally is still available and the rating hasn't changed.

It does look like the number of attempts has only increased by 2. Maybe it was taken out of the active puzzles list. (I can't tell.)

It would be nice to know if they have affirmed the 2200 rating.

Martin_Stahl

Looks like that one may have multiple solutions. I'll report it.

Martin_Stahl
justbefair wrote:

The one you posted originally is still available and the rating hasn't changed.

It would be nice to know if they have affirmed the 2200 rating.

Puzzles aren't removed from the database, just made so they won't be chosen during puzzle training

Martin_Stahl
Martin_Stahl wrote:

Looks like that one may have multiple solutions. I'll report it.

https://www.chess.com/puzzles/problem/1222553

Has been removed from rotation

RyanZ_MD

wait a minute, I'm 90 percent sure that I've done that puzzle before...