Recent Posts

Since this thread was first made we added two non-spaced repetition drilling modes, 'review in order' and 'least  recent/unseen first' which would seem to address your issue. These are the first setting shown when you open the opening trainer settings.

It is on the todo list, but no estimate on when it will happen. It should be relatively easy to add, so it has a better chance than some other more difficult items of seeing development time in the short to medium term.

Yup, MINUS-SIGN definitely makes more logical sense, so it was a good suggestion, and I'm glad you're happy with the outcome!

@6081ch
Nalimov 6 piece Tablebase agrees that 1.Qb4 is a draw with perfect play.

The problem has been disabled.

prairiedoc

Houdini considers 1.Qb4 an alternative, maybe check with a database.

Thanks a lot for the prompt reply, the glyphs look great for me now!

Yeah I am using Linux and apparently the font used for these symbols is Liberation Sans.

About whether it will look better or worse for all users: It would make sense that the fonts that don't care about distinguishing HYPHEN-MINUS from MINUS-SIGN would just use HYPHEN-MINUS for both and if the font designer *does* care about this, I'd assume he/she would preview the MINUS-SIGN in a context of other math symbols (especially the MINUS-SIGN). Of course, you never know and maybe some font designer had some other idea on how MINUS-SIGN should be rendered (one that could look bad near the plus-sign). All in all, given that the MINUS-SIGN is the sign designed for this purpose, it seems reasonable it'd be the most likely to look good on average.

I know this is an old thread, but if others stumble upon, the easiest fix for me has been to reset the training data. By resetting, I can completely start the training from scratch. This has really helped me take the opening move by move.

I don't really understand why the admin are so adamant on spaced repetition in all the forum posts. I understand the benefits and that's why I am here, but ONLY doing spaced repetition means that if I haven't trained in a while, I can refresh my mind. There should be some type of "degrade" variable that automatically walks through the full lines when not trained over an extended period of time.

Reset is the easiest way to mimic such a feature

As soon as you select a looping/sorted set you need to choose a sorting method, and "random" isn't an available sorting method (as you've noticed), although there is certainly an argument that it should be!

I'm resurrecting this thread just to ask if adding a random sorting method to a looping set is on the roadmap? It would be useful!

I have mostly figured out how to configure a custom set for looping and solving tactics, but I am unsure how specifically to configure the set so that after each loop, problems will be dropped when they have been solved correctly X number of times under Y seconds per solve.

For the first, I believe I set the Correct Attempts range to X-1, choose Match Correct OR failed range, and set the failed range to a very high number?

But how do I set the second? I see the "shortest solve time"-- do I set that to 0-Y and then if I don't solve the problem in that amount of time, even if I do it correctly, it is considered failed?

Yes, both our engine and latest stockfish preferred 2.Qxf6 at lower depth, although stockfish ends up not seeing any real difference between white playing 2.Qxf6 or 2.Nxe4 at higher depth. The reason 2.Nxe4 was played here despite having worse eval was because it was the move played by the master in the original game, and playing the master defense usually provides more realistic and human like problem responses from the opponent than always choosing the best engine line which can often throw material early instead of playing the most resistant line from a human point of view. Here that hasn't happened and the problem is worse for chosing the human played line, so I've disabled this one, as it might be easy to decide not to play 1...Qf6 if you analyse the 2.Qxf6+ as it is very similar to other "winning, but plays into long endgame instead of short term win" moves that we sometimes allow to fail.