Characterizing Rating Levels

Hi Ozzie,

If you go looking through a recent version of ChessBase's Big/Mega databases, there are a number of guys with ratings in the 2200-2500 FIDE range that haven't been able to reel in the full point with K+N+B vs. K, so you're in good company!  I think the reason is that there's a trick to it (that's very hard to find OTB) as you drive the king down the side from the corner of the color opposite of the bishop's towards the mating square.  The trick is moves 6 and 7 in the sequence below.  Enjoy!

So, how many hours are we willing to spend on a device that MAY generate 1/2 point only once or less in our lifetime? 

DrawMaster, read YeOldWildman's explanation. He sums it up well. A stronger player only needs to learn the three tricky moves--that doesn't take hours. A weaker player will take hours because he doesn't know how to coordinate his knight and bishop, and that is one of the chief benefits of studying this endgame. Just as mastering Rook vs Knight has helped me become better at dominating knights in the middlegame.

(Of course, B+N and R vs N aren't one of the first ten endgames to learn!)

Now we've got onto endgames...

I've taught my son 2 rooks vs K, K+Q vs K and he has now mastered K+R vs K. What do you think I should teach him next? K+P vs K or 2 Bishops+K vs K or something completely different?

jonny, ChessKids has recommendations for kids learning chess. Of course, you're the best judge of your own child. Here's their curriculum to give you some ideas:

• The Two Rooks Checkmate
• The King and Queen Checkmate
• The King and Rook Checkmate
• Pawn Endings 1
• Pawn Endings 2
• Pawn Endings 3
• Queen Endings
• Rook Endings
• Minor Piece Endings
• Endgame Studies

Right on, likesforests!

DrawMaster:

I don't bother trying to teach this to players who can't figure out how to push the king into one of the wrong corners on their own.  It's pointless.  OTOH, I was addressing a NM who doesn't have this particular tool in his tool box.  Someone who has his skill level plus the little bit of information above will know how to do it in minutes, not hours.

Ozzie is a nice guy and I don't want him to end up as another database tragedy if the ending ever came up in a game that mattered... 

Thank you for the kind words and instruction.

I remember during one of my lessons (from a GM) that he said that he knew how to mate with bishop and knight, but only if the king started out in the "wrong" corner. He said that everybody goes to that corner because they know they are far away from the corner that must be used to mate. Pretty funny, huh?

The one thing I know about bishop and knight is how to mate given the same exact starting position as YeOldeWildman has above, except with the bishop on g5 instead of f5.

I think with both of these nuggets, it should be easy to do it. I think there are even internet flash players which will help you perfect your technique.

I found the page which has the try-it-yourself, along with a tutorial which discusses both positions above.

Tutorial: http://www.answers.com/topic/bishop-and-knight-checkmate
Try-it-yourself: http://www.chess-poster.com/chess_problems/mate_king_and_bishop_knight.htm

Excellent, likesforests! That's exactly what I needed! Many thanks.

While I've never seen the KNBvK ending in one of my a games, I have played  as a teaching exercise -- and that GM is right:  most human players do run to the non-mating corner which makes it sooooo much easier.  So do the 2-3 engines I've tried it against, though perhaps that might be a function of the tablebases or event horizon since the final denoument is something like 40 ply out there after you get the king in the non-mating corner...? 

That link above to Little Chess Parter was the toughest I've ever seen anyone defend.  I'm glad to say I can still practice what I preach...  [Blows smoke away from the barrel of his six-shooter, twirls it, and replaces it in its holster...] 

YeOldeWildman-

I think you made an excellent point about CC (turn-based) ratings here. I play just a few games at a time, and it is usual for me to spend an hour with an analysis board on each move. But I think a lot of my opponents are playing 30 or more games at a time and spending only a few minutes on each move. I figure to have a higher CC rating, overall, than equally-talented players who play more games--although my Glicko RD will be much higher, too.

But, you know, even in the days of old-time CC, there must have been some players who had only three or four postal games going at a time, while others would have dozens. (At least in the days of penny or two-penny postcards.) So, maybe that difference in time alotted has always been with CC.

Hi nimbleswitch,

Since CC chess is chess for people who have lives outside chess (and for many who don't...  ), their rating here is for them playing with the style they choose to employ here -- and there are many styles one can employ.  The place where it can get a little inconsistent is that someone can choose to change their style and suddenly they aren't playing at their rating level any more. 

Bottom line:  don't be obnoxious to  anyone during a CC game, because s/he may really be 500 points stronger if you make him/her mad at you and you suddenly get his/her full time and attention... 

try this puzzle to assess your rating

I will make a point of taking about 15 minutes to learn the B + N v K mate, before my next tournament (Western Class Championships in LA)

Yeah.... if you see less than 30 mates on the move, you are under 1200~!

(for heavyartillery's puzzle)

Well, there are so many different moves creating a checkmate-in-one available it's something of a turkey shoot, no?  It might be more an assessment of one's patience than one's overall playing strength trying to find them all... 

Since it's a puzzle and no mention of it in the call of the puzzle, there's no need to justify the initial set up which might not be reachable by legal moves from the starting position.

Personally, I don't have the patience to answer either question... 

It was a joke...

I saw a guy next to me once close out with K+B+N.  His opponent made him prove technique, which he understood and didn't mind.  Perfect technique and the other guy resigned one move before mate.  For a lesson like that one shouldn't resign, but appreciate skill of the kill and let him have it.  Winner was ticked.  Wouldn't shake his hand or say a word to him and I didn't blame him - and said so when the loser was like "So what's your problem?"

The main point I would make about the Peter Paul characterization that likesforests put forward (in post #105) is that the levels corresponds to overlapping ranges of chess.com CC ratings. e.g.

One way of using the table would be to survey the chess.com community: "Looking at this table (WITHOUT ANY NUMBERS). Where do you see yourself in this table? What is your current CC rating?"

Suppose all the contributors to this thread did just that. We would have an actual piece of the real jigsaw puzzle. It's not that easy! What if you do not meet ALL the criteria for USCF 1200, but you meet SOME of the criterial for USCF 1500? Which group should you put yourself in? Arguably, I do not meet ALL the criteria for USCF 1200. Must I go to the next lower group?

(There is also the small difficulty that chess.com surveys only collect one piece of information. The survey could be done like this:   )

this would be nice.