Stokyo Exercises

I'm planning on adding Stokyo exercices to my study plan and was wondering how many people may have done them and what kind of suggestions you might have.

On Dan Heisman's site it is mentioned to "find a rich middlegame position" / "a fairly complicated position". A little further down he mentions that "most players are very poor at even-material evaluation."

So, would the primary criteria for a candadite position just be one that has even material and/or would it be better if the position itself be consider even too (equal or near equal computer evaluation)?

Dan also goes on to suggest a few resources to gather some good positions for this exercise. Do you have any additional suggestions/resources?

For what it's worth, I've done it with difficult tactical puzzles / endgame studies.  The principal is the same, just calculate everything you can and write it down.  Both these types of positions require you to see multiple lines.

Then you check the answer / notes to see what you noticed and didn't notice.  If I have the time / energy for it and if the main line in the solution isn't part of what I've written, I cover up the rest of the solution and think some more from that position.

To a lesser extent this can be done with an annotated game collection.  Some books give diagrams when it's a difficult position, and you'd just use the position from the diagram.  The more analysis the book gives the more useful it is.

A 0.00 or near 0 computer evaluation can be good or it can be bad.  What you want is a dynamic position with forcing moves available (or at least subject to long calculation like most knight or pawn endgames).  If it's a 0 eval in a stale position then you're just wasting effort trying to calculate your way thought it as there should be multiple correct moves.  If it's a zero evaluation in a difficult position, where you're forced to find the correct counter attack or defense, then this is useful.

Thanks for the input. Dan Heisman gives some suggestions for some resources and that is a good start too.

I was also hoping I might be able to determine a method to create some for my own games but I think the hardest part is getting a good dynamic position that isn't just immediately winning/losing. Often when I have a position that I find complex/dynamic, it many times evaluates poorly for one side when looking at it from an engine perspective.

Thanks.

I've done something similar with mates-in-two while converting the Polgar 5334 tactics book into a database for training. I need to get back to that again (only about a 1/3 of the way through it).

I've read through some of your blog posts and will try to read the rest of them soon. I kind of wonder how beneficial it is to do Stokyo on tactical positions; though maybe what you are describing is more of a Stokyo-lite.

That said, I do have some books that I think will be good for those types of exercises. I was also hoping to generate some other methods/ideas so I could maybe work on finding some of my games that had/have really good positions for the process.

I know from what I read, most suggestions are to spend an hour or two per position, filling up to two pages with thoughts, variations and evaluations. Though, I guess any structured process like this would be beneficial for improvement.

I try making whole games out of them.  It can be very tough though I found a better move than even Grischuk in a particular position.  ...Qf6! sacrificing a pawn for a strong attack was correct, but he played ...Qe8?! with equality. I still forgot some critical lines but the move was sound enough where if they appeared in a hypothetical game I'd try working out the proper continuations. 

Yeah, such exercises can be memorable.  The position in question is:

Heck, I'll post the entire game but pay special attention to move 18, it took so much out of me:

Is there any way that one or more of you could actually post a pic of your actual analysis sheet so I can get an idea of how you organize it?

Also, is the Stoyko just done one that ONE position? or do you move a move or two and start again on that position?

Here is the link to Dan Heisman's page where he talks about it; you want section 3, a little over half way down the page. http://home.comcast.net/~danheisman/Articles/Exercises.html

I haven't done one yet, but my understanding is you do one position and then you can move to a different position. In your analysis, you will likely go beyond a move or two anyway.

I'm sure there are games where you could do multiple positions from it though.

Sure I have a screenshot:

As a side comment, this method is the same one advocated by Yusupov in his 9-book chess class series.  Most of his exercises are not complete games. He does not "allow" the use of a computer (OTB only) and he asks students to study the position and make a list of candidate moves and the thought process behind eliminating or choosing each move.  In other words, write everything down that is considered, good or bad.  This is intensive, time consuming study.

but you didnt type it out as u went did u?

So you are writing down general thoughts on the position as well as the actual variations?

Type everything first before moving ahead.  Turn off the notation pane so you can't see ahead. 

Yes it's an assessment, analyze imbalances and who you think is better, by how much, and why.  Why settle for an equal line if you're better?  The goal is to look for something that leaves you at least slightly better if you're slightly better.  Deciding on candidate moves is the next step, then analyzing those. 

In the posted position the weak dark squares are key, even more vital than the isolated e-pawn in my opinion as you really wouldn't want white taking advantage of weak c5 and b6 squares.  The Be3 looks like white's most dangerous minor so far, and can easily be targeted with ...Bg4, but one has to be careful that it doesn't lose due to tactical reasons or you weaken yourself too much on the dark squares. 

The Na4 initially looks out of play, but obviously (gosh I read Hansen's Improve Your Positional Chess too much, every other word he uses is obviously!) coordinates with the Be3 very well targetting the weak dark queenside squares.  b6 is traditionally weak in the Najdorf, but here c5 is really the big point of contention as white uses it as a focal point for his operations.  The c and a pawns are weak, but white can't take advantage of it just yet.

awesome. More than one reader looks forward to seeing your stoyko exercise postings, and more than one reader recently discovered and much appreicates your blog. Thank you for putting up your quality material. 

In a certain sense, the Stoyko exercise could be a type of Holy Grail of chess exercises because essentially it is what you are (supposedly) doing in a game

I have heard it over and over that one needs to play slow games to improve...and I agree. Yet day after day I find myself playing 5 minute blitz and each and every game I miss many ideas because a blitz game like that is going to be shallow by definition.

So at least the stoyko will give us some practice doing what we will actually be doing in otb games.

"What you learn to do, you learn BY doing"

but is that in chessbase or Word or what?

The position was copied from Chessbase into Word, rest was typed into Word. 

I started on them over a year ago because to automate a skill you need to work at it for three weeks but you need to augment a thought process with new information, so after reading Shereshevsky's Endgame Strategy (recommended by Pfren and others) I'd allow mini-plans or schemes (schematic thinking) to drive analysis provoking further weaknesses, or welcome having knight vs. bishop (despite usually favoring bishops) if it results in mutual isolated d-pawn (a structure that favors knights). 

Now that I'm almost finished How to Defend in Chess I can look through some games and see if I can apply certain principles taught in that book (such as Steinitzian-Laskerian rules against making too many pawn moves without corresponding piece activity, law of economy regarding defense meaning just cover what you need to and no more, so no unnecessary weaknesses, etc., trade off most dangerous opponent's piece, etc.)

The last point is so basic that I applied it before reading the book (see above Stockyo I ripped from Bologan-Polgar.  Sometimes like Kasparov vs.Polgar at Linares, the one I didn't know had a touch move controversy or Polgar vs. Bacrot I find myself studying from her opponent's side)

Off topic but Polgar vs. Bacrot is such a beautiful game, his long unbroken pawn chain is an immortal game that should go down in history.  It was a Ruy Lopez. 

Thanks for the information,roi,Oogie,Stahl,I'm going to give this a try.Anymore links to,or books concerning,this exercise?

As for books Heisman's Improving Chess Thinker has some exercises. 

(Spoiler and shameless self-promotion alert!)

Is an example of such a protocol, it is mine.  In De Groot A Bxd5 in black's best line leads to a mutual isolated queen pawn position, so the hole to your isolated d-pawn is plugged so no major pieces can target it and white has better chances due to his piece activity after the forcing variations.  Nxc6 wins the bishop pair, but bxc6! reinforces the d5 point further and the isolated c-pawn is quite easily defended (I think Euwe mentioned that too)

  I found it odd that none of the protocols and even Heisman mentioned that in his book however. 

http://www.chess.com/forum/view/general/de-groot-c

@badger_song: I believe this exercise can be done with any position that is well annotated, in order to compare what you come up with with the annotator. Crouch is well-known for his variation-rich annotations, hence How To Defend in Chess. However, as a text it's supposed to be quite advanced. But Stoyko exercises can be done with any diagram, from any book, anywhere.  I am starting to do them with my (very elementary) tactics puzzles, just to practice the habit of looking wide. 

I did a similar exercise about a year ago, I have it in a notebook somewhere. I was just going thru oldschool games and I found one and went to a nice middlegame position in it and started looking for variations etc.

Come to find out it was one of the most complicated games ever lol. It was widely annotated but I dont remember who it was.

I can think of a way to make the Stoyko even better. Do it on opening positions THAT YOU PLAY.

for instance, say I play the Nimzo as black. Obviously I am going to see the Qc2 variation and I decide I am going to castle after Qc2. White has several popular moves after that. A3, Bd2, Nf3, e4 etc.

I could pick a line, lets say after 5.e4 I go 5...d5 6.e5 Ne4 7.Bd3 c5 8.Nge2 cxd4 9.Nxd4 Nd7 10.Bf4 Ndc5 11.0-0 Bxc3

I could find games from that position and do the Stoyko on those.

In the course of a year, if you did 10 Stoykos on each main line you faced, dont you think youd know those typical middlegame positions better?