So if I'm catching up on my reading of the forum and realize I can vote, it takes extra (yet minor) work to vote and then get back to where I was. However, I imagine (hope) that it would be easy for the webpage to simply know where I was coming from when I voted and then take me back there.

Officially, the rule is that if somebody can't figure out how to properly type the move their vote doesn't count. You could consider it a test to verify the voter's overall chess aptitude. The rules explicitly state that the "x" is mandatory for all captures, so nobody has any excuse to not get it right.

In theory the World Team should perform a little bit stronger with such a heavy-handed policy, but in practice we are much more forgiving and often do consolidate the erroneous notation with the intended move.

Just so you don't get the wrong idea, this rule is not intended to be mean or to ridicule people who make notational errors. We are just limiting our liability. Just in case one day the argument is posed that the World Team made an inferior move because the arbiters failed to consolidate erroneous notation, we can rightfully point out that consolidating notational problems is not our duty.

I didn't submit today's pun, but it is inspired from my game collection, right?

Oct-12-10 <chessgames.com: <<SwitchingQ> <No, we get the right guy, just with his name misspelled>

Ohhh, I see. Whew. That's trivial. I see that the original PGN we received spelled it "Buck" so we'll stick with that. <And there is a pun here so obvious that I don't even need to say it. <<<<>>>>>>>>

Genna Sosonko's book 'Russian Silhouettes' on page 30 says "A few days later, on 28th June 1992, Misha Tal died in hospital in Moscow. The offical cause of his death was given as a haemorrhage in the oesophagus, but effectively his entire organism had ceased to function".

Hopefully a tournament page will be set up in the next day or so.

The <European Club Cup> for Men and Women will take place 16 to 24 of October 2010 in Plovdiv, Bulgaria. If you look at the teams it's like a European chess Olympiad.

http://ecc2010.chessdom.com/

Apparently Ivanchuk vs Ehlvest, 2000 shows a novelty not played, and Wang Yue shows up for games against Carlsen when searching for Wang Hao: http://www.chessgames.com/perl/ches...

Opening Explorer

Two games. One line. I click on that line: Opening Explorer

With regards to your search, avoid that problem by using the pulldown menus to get the players instead of typing them in, e.g.

http://www.chessgames.com/perl/ches...

Also you can use "EZ Search" and just type "Wang Hao vs Carlsen" like this

In your case, we just reprocessed those games and indeed the problem went away.

But you didn't ask us to fix it, you asked <Can you explain this?> The answer has very little to do with chess and a lot to do with software engineering, so if you want to stop reading this post at this point, I won't be offended. But the computer scientist in me loves this stuff, and surely other members are like-minded, so lets take a closer look:

First, understand that the Opening Explorer tries to keep track of "just what it needs to know". It doesn't strive to know every position of every game, nor does it want to. To the Opening Explorer, the only really interesting positions are the ones shared among several games, or a position belonging to only one game that defines a novelty.

When our software "opex" (the Opening Explorer indexer) gets a new game it runs through the positions (plies) in order until finally it finds what may be a novelty. E.g. after move 13.Be2 it may realize that we have no other games that contain that position. One might think that it would stop right there--but it can't. It has to go at least a few more moves to make sure that the game doesn't transpose back into well trodden ground. What's more, it wants to make sure that future games that follow this variation will be paired up properly.

So opex plods through several more ply proving to itself what should be obvious: that if 13.Be2 is unique then the positions that follow it are also unique. The software sees a series of numbers that look something like 48012, 27007, 9032, 532, 137, 122, 53, 8, 4, 1, 1, 1, 1, 1, 1, 1...

Once it hits that long string of 1's it becomes 99.99% confident that it is in unique territory, but it's not going to add records all the way into the endgame to prove it. At some point, it has to give up and stop adding new records. The number of 1's that it witnesses before it gives up is what we call the "novelty_depth" setting. We call it that because it answers the question "how deep do you index, after you find a novelty?" In other words, after witnessing a dozen 1's in a row, it stops processing.

Currently, novelty_depth is set equal to the number 12. You may wonder, what's so magic about the number 12? Nothing; it's just arbitrary. We tried various settings and settled on 12 out of experimentation. It's a number that make omissions extremely rare but at the same time doesn't bloat the database with hundreds of millions of positions that belong only to one game.

Once in a while, the very unlikely happens: those 12 ply prove to be not enough. A game we uploaded years ago ends up having a modern twin, and twin mimics the original game through all 12 of those ply and then keeps on going. And that is the answer to your question, that's what happened in the case you demonstrated.

Botvinnik vs J Guerra

G Munkhgal vs Botvinnik

G Orgil vs Botvinnik

Botvinnik vs V Papin

Botvinnik vs Peter Meier

K Mekhitarian vs Botvinnik

J Wyss vs Botvinnik