The theoretical draw may not be so obvious, if white starts running his g-pawn. For example after 48...♖f3 one must see that the pawn endgame is indeed not winnable. After 48...♖f3 49. ♖xf3 exf3 50. ♔e1 ♔e4 51. ♔d2 ♔d4 52. ♔c2 ♔c4 53. ♔b2 ♔d3 54. g4

click for larger view

Black can play both 54..♔e2 with mutual queening or eliminating the pawn with 54..♔e4 with getting the opposition after mutual pawn elimination.

Here's am old friend. Variation 2.Rxb2?

Thanks for putting me in my place. :-) But Olavi, the position you point to is a little different, no? There isn't the same rook trade possible.

< nok: Teichmann vs Alapin, 1908 : 69...Rf3! Maybe past masters weren't trash after all.>

The general structure e2-f3 vs. g7 is the same, these things are draws, unless there is a specific reason why not. Mamedjarov almost spoiled it with 42...e4.

<Alireza said in post match interview that no human can see 48...Rf3 was draw ... as he thought after 49.Rxf3 exf3 the pawn ending was winning for White:>

Averbakh analyzes many positions with this exact pawn structure (on second and third rank, but on different files). Since I have studied Averbakh, Cheron and others in my youth, is was obvious to me watching live that such possibilities can be there. (No false modesty.) No, I don't see engine evaluations.

Yes, I'm sure you and Timman are right about young players and endings, and it's the more striking given that there are no adjournments any more -- you have to figure it out at the board, no alternative.

But down at my level -- I've played a dozen games OTB since returning to tournaments for the first time in over a decade. Eleven of those games have been decided or decisively influenced by a blunder (meaning one side blundered and the other side failed to capitalize -- a double blunder, in other words). But I've only played one <real> ending, opposite colored bishops where I was a pawn down. And I drew without much trouble. The funny thing was, I knew there was a rule about which color squares my pawns were supposed to be on, but I couldn't remember the rule! That didn't matter, though, because it quickly became obvious that I wanted my pawns on the color that his bishop wasn't on.

I guess it is not considered cost efficient to study big endgame books cover to cover anymore. Probably even old Soviet-educated coaches no longer make their pupils go through that kind of work.

With opposite colors bishops ending, the defending side usually wants to put his pawns on the same color of his bishop, so your conclusion there makes sense..

I think that studying endgames deeply, and memorizing various types of simplified positions, is not considered cost-effective in general anymore. Its not about whether to use books or computers, its about the fact that studying openings takes too much time and energy these days.

Studying openings was never "sexy" because its mainly about memorization, but the fact is, that if you don't know your theory well, then:

1. You will waste a lot of time trying to figure out what's going on, and will get into time trouble that will seriously reduce your chances.

2. Most likely, you will not figure out the best plans (for both sides), and will end up in worse position (or at the very least, a position that your opponent understands better than you, assuming that he knows the theory there).

3. You will not have the confidence that you actually know what you are doing there.

And these days, with the amount of theory we have, and its instant availability, and with the fact that your opponent can prepare with engine and punish you for every inaccuracy, its just too much to afford, not to be well prepared in the opening, and keep it up to date.

In other words: not knowing this type of endgame, cost Mamedyarov in half a point here, but if he didn't know the Italian opening that he played here, it would have cost him much more when looking at all the games that he play.

You are absolutely right in general. But in this case one doesn't need to have memorized the pawn ending after Rf3 Rxf3 exf3, it's enough to know that many of these positions are (surprisingly) draws. To know that you only need to have had a glance at a good old book. Then he could have used his 15 minutes to think the matter over.

Across the years, I have often posted on this and completely agree; a pendant to this topic comes from a 1972 interview given by Larsen to Hugh Alexander for the latter's general work on the game.

Larsen was critical of what was then standard practice in Britain, and which I recall seeing in some of my early events here in the States, in 1972-73: adjudications.

I think, that both endgame knowledge general playing strength in this phase (these are two different concepts!) is fairly low compared to openings and middle game tactics. This applies to all generations, but to younger players in particular.

I consider myself a fairly good endgame player, especially in rook end games which I seem to grasp intuitively (combined with some of the classical positions and concepts, you just must know).

Now the simplest endgame material wise are if course pawn endgames. I thought I knew everything there: opposition, pawn breaks, corresponding fields etc. To my surprise it was exactly the pawn end games where I faired pretty badly in a recent test assessment over at Lichess. They are still much deeper, than I thought.

The wrong assessment of both Mame, Giri and Firouzja does not surprise me anymore, actually.

Regarding Teichmann -Alapin: we don't know the exact circumstances. Maybe Alapin played the move as adjournment and it was only later it was discovered this was a draw?

In any case, knowing endgames and being good at them are related, but not completely congruent concepts. The often lengthy nature of endgame combinations makes them different than say middle game mate attacks. Being able to really calculate for 10 moves in a very " linear" is a different skill set than seeing five or more relatively short branches of a middle game tactics..

I see your point, but I'm not that that is the case here.

Most likely those 2800 rated players, saw this element in past (and probably more than once), and probably saw the Teichmann-Alapin game as well.

However, it is one thing to see some rare pawns-endgame element (out of so many other endgame elements), and it is a totally different thing, to be able to envision it from a rook endgame, and use it in practice.

At least for myself, I know that I saw many pawn endgames elements (including this one I think), but I am far a way from being able to actually use them in practice.

In order to be able to use it in practice, I will need to study the element much deeper and practice it a few times.

As previously stated, the aforementioned ("weird") KPP v KP endgame, with f- and g-pawn, is a theoretical draw. Black manages to hold the position by maintaining the appropriate form of opposition. The same goes with the configuration with g- and h-pawn.

In all other (corresponding) cases White is winning. For example, the same type of endgame, with e- and f-fawn is won by White:

click for larger view

A possible continuation:

1.Kf1 Ke5 2.Kg2 Kf4 3.Kh3 and Black cannot maintain the opposition without losing the pawn:

3...Kf5 4.Kg3 Kg5 5.f4+ Kf5 6.Kf3 (followed by Kxe3)

click for larger view

A similar maneuver is not possible in the original endgame:

click for larger view

There are other draws also. For instance Kg1 pf2 pe3 - Kf5 pf3 is mutual Zugzwang, white to play only draws. That's mirrored, but a corresponding case to my mind.

The horizontally mirrored endgame is also draw:

click for larger view