In the game Rublevsky vs Beljavsky 1995, from the course Roots of Positional Understanding, this position appears:

If the student tries 1. Nc4 then Chess Mentor (IM Jeremy Silman) comments:

"1.Nc4 is reasonable, but Black gets plenty of play with 1...a4"

Excuse me mister Silman but  1...a4?? is a blunder because of 2.Bxh7+ winning the queen!

You might consider following the course about basic tactics again.

Can you give the list of moves that wins the queen for white after a4?

Thanks

1.Nc4, a4??

2.Nxh7+, Kxh7 (or Kh8)

3. Rxd5.

Gotcha,

Qh5 would probably give more play to Black than a4 alright.

The example above shows four things:

     1. Jeremy Silman (an IM) could be human (and make an error); and

     2.  All chessplayers are human and could experience an 'oversight' or blunder; and

     3. This site provides something for everyone, even the devoted critics; and

     4. It remains a possibility - that he did that deliberately - to see who is 'on their toes' and looking for opportunities (be alert) or who is sleeping.  Those sleeping will fall into that trap in a game.  Those alert players will notice the error, most will not comment on it, merely note it and tuck it away in their minds, to play into this and see if one of their opponents swallowed that line and attempts to play it - thereby making a fast end to the game......

What a great site - something for everyone.

RetGuvvie98, don't get me wrong. My message is meant to improve Chess Mentor. This part of the lesson is clearly a mistake by Silman, and should be changed.

When after a sentence in a forum a wink is placed it means "joking". So the sentence about Silman following the basic tactics course again should not be take seriously.

Relax.

I'm so relaxed that if I relax any more, I'll be asleep.   Chess Mentor might be improved by correcting the technical error you so kindly pointed out, but maybe it is intended to be there, after all, this site does try hard to provide "something for everyone"....   and the 4th possibility I posited could be the right one - - -  that is always a potential source of errors when chessplayers are involved.   ((( devious are their little minds... soooo devious.)))

we'll fix it :) thanks!

Yep, we have corrected this response to something more appropriate now.

Wouldn't it be a good idea to run all Chess Mentor exercises through Rybka in some automated fashion to check that the "correct move" does not vary more than .5 from Rybka's (after 30-60 seconds or so on normal hardware)? Then the cases with a larger difference can be gathered and examined by a human. Of course, creating such a software is not easy.

BTW: It would be nice if there were a way to save or mark exercises either due to how interesting they are, difficulty in solving or inaccurate analysis like here.

exigentsky:

    Chess Mentor exercises are validated by humans and programs already - and the programming team is working hard on improving things all the time.  we already have a significant team reviewing all tactics and validating the tactics before approving them for public use in the tactics trainer.  Some are evaluating the high-rated - low 'solved rate' tactics as well.

whether Rybka or some other [management selected] astute program, rest assured:  things are being validated by site management.

   all that takes money of course (programmers have families to feed, too) - and site management greatly appreciates those who support their efforts with paid memberships.

Does he mean

1.Nc4 a4

2.Bxh7 Nxh7

3.Rxd5 Bxd5

A Rook and a Bishop for a Queen and a pawn