In fact, rather than call it "favorites", I should use a more generic name like "user groups", and allow each user to assign a name to each 'user group'.

With this approach, I could have:

- One user group that I call "Favorites" (maybe colored yellow). To this group, I add people that I particularly like.

- I would probably have another user group called "Myself" (maybe colored green) that has only my own user ID (so I can quickly spot my own posts).

- I could have another user group that I call "Ignorable" (colored gray). To this group, I add people whose posts I normally ignore or merely skim.

The color allows me to quickly assess each post, without actually hiding any posts. This avoids the negative impact of throwing off page numbers -- which is the main reason I don't use the ignore feature itself.

Each user group may have some additional attributes. For instance, for the "Ignorable" group, I might want to set an attribute that says "do not let these users post in my chessforum". (I might turn off that attribute when I'm running a prediction contest.)

Then, in each forum, instead of having a "Favorites Mode" which is "on" or "off", you could either select "all" (the default) or select one of your 'user groups'. Similarly, you could perform searches where the posts are limited to a specific user group.

Firstly I wish you a very merry christmas!

In his forum <MostlyAverageJoe> has advocated an alternative that would get around the problem, at the cost of more work for the chessgames.com staff. I would like to suggest a simpler fix.

You continue to post a new puzzle every 24 hours, chosen exactly as in the past, but in addition post a second position where there is nothing special to be found (a control or "dummy"). At least 19 out of 20 positions chosen at random would satisfy this requirement. The puzzle-solver would first have to decide which of the two contains the winning move (or in the case of "spoilers," the saving move), and then find the move itself.

Thanks

2009-2010 Bundesliga

Here is my earlier post to the Kotov page:

Kotov won Hastings in 1962 according to Wiki. this concurs with the PGN of the event which can be obtained from:

http://www.saund.co.uk/britbase/has...

It would be good if Chessgames.com could get these PGN downloads and incorporate them, as there are many Kotov game examples here e.g.

[Event "Hastings 6263"]
[Site "?"]
[Date "1962.??.??"]
[Round "3"]
[White "Kotov, Alexander"]
[Black "Hollis, Adrian"]
[Result "1-0"]
[ECO "D95"]
[PlyCount "121"]
[EventDate "1962.12.??"]

http://en.wikipedia.org/wiki/Hastin...

1. Do search for "any result", e.g. Endgame Explorer: Q vs QP

http://www.ippogrifoscacchi.it/torn...

Pragmatically the three differences to the scoreboard, if dummy games were inserted, is that Elena Zaiatz would have 4.5 points on a technicality and not 3.5, and Tatiana Stepovaia-Dianchenko would have 3 instead of 2, and we'd display Natalia Pogonina as having 3.5/9 instead of 3.5/7.

On one hand, one may say that these missed games represent breaking chess news, and perhaps they may even turn into a "best wishes Natalia" type thread. It wouldn't be a horrible thing to have them. And the leaderboard would show everybody has having their full 9 games.

On the other hand, the argument can be made that displaying the scores without dummy games explains what happened in this tournament better than if we inserted them. Showing Natalia with 3.5/9 gives no indication that she dropped out of the tournament; it would look as if she played horribly. Showing 3.5/7 tells the real story: she was doing average, then had to leave.

We read your post carefully (and others similar to it in the past). Our position is that we like the general premise, and are eager to incorporate something like it in upcoming features, but we doubt that we will change the puzzle-of-the-day format on the homepage.

To elaborate even more on the idea, there is an extension that applies itself very well to endgames. In fact it may be critical in the teaching of endgames effectively. For example, imagine drilling endgame theory by presenting a student with 16 simple king and pawn endings (Endgame Explorer: P vs K) say in a 4x4 grid. Given that it's White's move in each position, ask the student "Which positions are wins for White?"

This sort of ability is much different than what it takes to solve puzzles on our homepage. The student quickly learns rules of thumb like "the box of the pawn" and "king opposition" and then learns how the rules change a little bit when dealing with rook pawns, etc.

On the other hand, the old-fashioned puzzle has its own charm. It dates back to the very oldest written evidence of chess itself. We know these "What's the zinger?" style puzzles suffers from a fault in the sense that it's not exactly like playing a real chess game; but it's not supposed to be. Of course we don't want fans of Chessgames to go to tournaments and start to give up bishops on h7, rooks on g7, and queens everywhere, just because that's always the winning move on the homepage. We think most chess players understand the difference--and if not, they'll learn quick.

In short, we like the idea of "mutated puzzles", and hope to offer some sort of feature that incorporates that. It is a very powerful drilling method.