Why study the endgame first?

[TexAg06]

I'm not disagreeing, just curious as to what y'all think. I was talking about this with YKW the other day and he made some interesting points I hadn't heard before, most notably that opening and endgame theory are rather finite, whereas middlegame study is virtually endless.

We've all heard titled players (old Soviet school of thought in particular) recommend to study the endgame first (after acquiring a basic opening and tactical knowledge, of course), but why is that? To me it makes sense to study the middlegame, since often times the bulk of a game takes place in the middlegame, plus if you suck at the middlegame you won't even make it to an ending. Is there something about learning the endgame that makes you a better middlegame player as well?

Sort of thinking out loud here, any thoughts would be appreciated.

[Dorilla]

I think it is better to study all areas at the same time. Also endgame is harder than middlegame. Nobody can understand RB v NN.

[John_Douglas]

I think there are definitely some endgame positions that should be known by all chess players - King and pawn endings, Q vs. Pawn, Bishop vs. rook pawn, a few of the main rook endings.

But beyond learning the basic theoretical positions, studying the endgame systematically is probably not the best time investment for most players.

The main thing is learning the right ways to think about chess - how you approach a position, how you calculate variations.

[Montrealcorp]

pretty simple imo.

studying endgame first is the ideal thing to do, tho its pretty tedious.

has to why well, how can u recognise a winning position in the middlegame if u dont know its a winning position in the ending ?

a lot of players below expert level are very weak in the ending, so u would gain an awefull lot of winning game from drawn endgame and u would draw a lot of losing endgame .

[meekrab]

There's obvious practical reasons... if you don't know the Knight and Bishop endgame, what are the odds you'll be able to force mate with 3 minutes left in an over the board game?

But it also helps by training you to avoid bad endgames. For example, knowing that a Rook pawn + wrong colored Bishop is a draw, if you have a passed Rook pawn in a middlegame, you would prefer to exchange that Bishop instead of a Knight given the choice.

[KyleJRM82]

Originally Posted by Montrealcorp

has to why well, how can u recognise a winning position in the middlegame if u dont know its a winning position in the ending ?

For a beginning player, "More material or checkmate = winz" is plenty to recognize won games with 95+% accuracy. Throw in "beware of having just an extra rook pawn" and you are up to 99% accuracy.

[Allen C]

I think it's obvious that the middlegame is far more important than the endgame. Like YNW said the scope of the middlegame is near infinite so maybe the Soviets said that so they could have something concrete to teach their students rather than admitting they have no idea how to play some position. Maybe they revered Capablanca who said the same and they didn't want to contradict him. Maybe back in the day they had plenty of books with endgame studies but not as many 1001 tactics books so they couldn't very easily recommend studying middlegame tactics.

Insightful teachers who I've read mostly encourage endgame study because it's a clear way of practicing calculation and visualization and it can reveal many chess principles with all the extraneous stuff stripped away. On the subject of learning certain endgame positions just for their own sake, several modern grandmasters say it shouldn't be a priority at all. This interview comes to mind: http://www.youtube.com/watch?v=paHdsGkP7nI (starting around 2:45)

[whiskeytown]

I'm not sure it's that useful in blitz play - I've heard endgame play is suffering amongst the younger generation because of this - and the fact you can no longer adjourn games means they get played faster as well.

I just feel like I can make better decisions in the middlegame knowing the endgame - here's two examples



I recently played a live game (Scotch but this is techincally a Ruy Lopez ending) where I strove like hell to get this position. Why? - Cause it's won for white - I can even beat Houdini with it now - unfortunately, I had to make a trade that compromised my kingside position but that's what happens when you play 500 points above your fighting weight - but my middlegame decisions were motivated by an ending I KNEW unequivocally was winning for white - (for more on this ending - go to http://www.chessvideos.tv/forum/viewtopic.php?t=9219 - he does an incredible job of explaining it)

ending number 2 - fighthing a guy 300 points above my weight class, I blew this ending in time trouble



as white, I picked up his passed pawn at the expense of my g2 pawn - but even then if I had played f4! instead of Kf5, I'd have been able to either trap his king on the h file with opposition or else outflank him to the h1 square, which would have been a draw. We all know about blocking a rook pawn by camping on the queening square, but we forget sometimes that we can have his king serve the same purpose - bet I remember better next time

Now I take positions like these, generate them in Fritz, and replay them so I don't blow them next time

here's a couple more positions I took draws on with black but they were easy wins provided I have 5 min left on the clock - (I'm usually down to about two though - doh) - actually I'm still trying to master the first one but if I can beat Houdini and Rybka and Fritz, then I know I'm golden against a human -

I think tactics and endgame play are much more important than opening study - at least till about 1800

RB

[TexAg06]

Originally Posted by Allen C

I think it's obvious that the middlegame is far more important than the endgame. Like YNW said the scope of the middlegame is near infinite so maybe the Soviets said that so they could have something concrete to teach their students rather than admitting they have no idea how to play some position. Maybe they revered Capablanca who said the same and they didn't want to contradict him. Maybe back in the day they had plenty of books with endgame studies but not as many 1001 tactics books so they couldn't very easily recommend studying middlegame tactics.

Insightful teachers who I've read mostly encourage endgame study because it's a clear way of practicing calculation and visualization and it can reveal many chess principles with all the extraneous stuff stripped away. On the subject of learning certain endgame positions just for their own sake, several modern grandmasters say it shouldn't be a priority at all. This interview comes to mind: http://www.youtube.com/watch?v=paHdsGkP7nI (starting around 2:45)

Really good post, and thank you for posting the youtube link. From about 2:45 to 7:00 is on endgames and it's really informative. After reading different stuff and talking to people, I think your second paragraph hits the nail on the head. Endgames teach raw calculation without being able to rely on intuition as much as the middlegame. And also like you said, the extraneous stuff like opening memorization isn't present, so the player has to really dig into a position and calculate accurately.

[Dorilla]

For those who are bashing endgame take a look at Carlsen. I think he owns a lot of people in that area.

[KyleJRM82]

Originally Posted by whiskeytown

There's a win for white there? That looks like the drawiest drawy draw that anyone has ever drawn.

[smilingbill]

White i'm surprised if it isn't won for Black

edit: didn't real whiskeytown's post too carefully: that's what he said

[KyleJRM82]

Oooh, that makes more sense.

If you ever have a chance and ability to win an opening like that in time trouble, it'd be pretty epic.

[bigpooch]

Originally Posted by Allen C

I think it's obvious that the middlegame is far more important than the endgame. Like YNW said the scope of the middlegame is near infinite so maybe the Soviets said that so they could have something concrete to teach their students rather than admitting they have no idea how to play some position. Maybe they revered Capablanca who said the same and they didn't want to contradict him. Maybe back in the day they had plenty of books with endgame studies but not as many 1001 tactics books so they couldn't very easily recommend studying middlegame tactics.

Insightful teachers who I've read mostly encourage endgame study because it's a clear way of practicing calculation and visualization and it can reveal many chess principles with all the extraneous stuff stripped away. On the subject of learning certain endgame positions just for their own sake, several modern grandmasters say it shouldn't be a priority at all. This interview comes to mind: http://www.youtube.com/watch?v=paHdsGkP7nI (starting around 2:45)

Knowing extremely basic endings is helpful: maybe pawn endings and rook endings at the minimum. The most important exercises should be focused on tactics and calculating lines and simple endgames or studies can help. The video seems to focus on opening theory because if anyone works hard on openings, they can really often get good or playable positions in the middlegame, but that usually applies when a player is fairly competent at analyzing. The video also suggests that it's pretty much given that 1.e4 doesn't lead to very much against a player who has simply mastered the Marshall and some other defences with 1...e5 ( e.g., the Berlin); I think Aronian said something like he would play the Marshall for a draw and the Berlin if he wanted to try to win.

[Allen C]

Originally Posted by bigpooch

The video seems to focus on opening theory because if anyone works hard on openings, they can really often get good or playable positions in the middlegame, but that usually applies when a player is fairly competent at analyzing.

Well Gustafsson began by suggesting that players study how to play typical middle game positions, calculation, and openings rather than endings in reply to a mailed in question and did a semi take back when he found out the mailer was over 2200. So that implies he thinks endgame study might be more important for higher rated players, certainly not "learn endgames first." I wish he was allowed to go on longer because I also felt like Svidler digressed a bit and was speaking to stronger players who already could conduct the rest of the game very well when talking about how important opening study is.

Of course nobody's suggesting not to learn the couple dozen important endgame ideas that come up most often and are easy to pick up quickly. I feel like this thread is more about whether to fill a high percentage of your regular chess study time on the endgame or if you're allowed to study something else before memorizing Dvoretsky.

There's also a symantical problem since positions with many pawns and two or three pieces each are position in which grandmaster shine and seem very important but I'm not sure if everyone calls them endgames. In fact, the whole classification of the three stages seem a little hazy to me. I believe the most logical way to study would be to concentrate on the most common types of positions and gradually work toward the less common, regardless of how many pieces are on the board.