Suggestion for acceptable problem settings.

[tomohawk]

Hi Richard,

I have noticed a rather large percentage of the last batch of problems I did (latest ones: 38889, 42330, 35282) had easily winning moves that either got marked wrong or were given "computer found a better move" notifications. In my experience, wasting time looking for the perfect move when you have some move that wins easily is an error. My goal for working on this site is not to become a chess problem solver, but rather to train my eyes and brain to find reasonable - not perfect! ;-) - solutions within an acceptable time limit in order to help my tournament chess.

My suggestion is to allow users the option of just marking all moves that score say 3.5+ as right. You don't even have to rate the problems where people have activated this option, just let us go to the next one. I often feel when I have gotten this notification that the problem is over and I want to click on any old move just to move on to the next problem. I somehow doubt this is what you intended. ;-)

[richard]

Hi Tomohawk,

If I understand your idea correctly you are talking about making alternative moves a "pass". I agree that looking for a better move when you know your candidate wins is not a good model of what you would do in a tournament.  However I think one of the drawbacks to this is similar to one of the issues with making alternatives the same as correct moves.  That is if user X gets an alternative move at move 1 of a very hard 4 move problem where perhaps move one was easy and moves 2,3,4 all very hard and user Y gets the first 3 moves correct and then move 4 wrong (non-alt) then I think user Y might feel hardly done by compared to user X under both the "pass" and "win" treatment of alternatives (where user X would either come out even or ahead compared to user Y).

Having said that, raising the level for the "pass" above the current alternative level from +1.75 to +3.5 does slightly mitigate the above effect and I'd like to find a way of allowing users to solve problems in a manner that doesn't annoy them. However it is very tricky to add options to the rated problem solving without providing an advantage to some users over others.  Some users would prefer to have problems with alternatives over a problem set which was cleaned of all problems with alternatives as it gives them a chance to see other aspects of the problem.  Providing the option you mention would give users a choice between  the current treatment of alternatives (which some may prefer) and an optional "pass" feature which would give them a rating advantage. Users who care about their ratings may therefore be forced to choose an option they don't want in order to stay competitive.

I have some code not yet uploaded which would change the current "mates on length X will be rejected if the shortest mate is of length Y" so that ALL mates less than a certain length (say 9) would be allowed as alts no matter how short the shortest mate.  Unfortunately this doesn't help you much as it sounds like the notion of a winning move being given "a try again message" is what is annoying you as much as anything.

One way of dealing with some of my concerns with adding optional features would be just to make a completely separate rating system which treated alts as you'd like.  I've avoided adding another rating system yet as there is also a problem with not enough hard problems at the top end of the problem set and introducing another rating system means it takes longer for new problems to settle into the set which means I can't introduce them as fast as I'd like. There are a few user suggestions as well as some of my own ideas to speed up the integration of new problems into the set.  Once this issue is resolved I'll be more inclined to introduce a new rating system (or two). In the mean time I'm open to suggestions of working around the above issues.

Regards,
Richard.