In the extreme, why would a mass of variations which don't prove anything (e.g. which knight outpost is better and more exploitable to opponent or you) be useful to anyone?!

Tisdall would calculate some moves and link it to the actual question. E.g. Ahh Nd5 and then this might haappen - blah blah, therefore maybe Nf4 is better. I.e. the hypothesis is integrated with the testing of the hypothesis.

If you take the Kotov "system" literally, apparently a GM has three distinct independent stages:

1) Assess position

2) Form plan

3) Perform a mass of calculations

I simply don't buy into this. I think 1) and 2) need always to be aware of the exploitability element from both sides. I.e. the static and dynamic considerations should naturally be combined at all times.

Here is an analogy: You want to shave your face. Do you do this:

1) Assess where to shave

2) Plan the shaving route

3) Give the instructions for shaving using the assessment and plan to a robot who will do a mass of calculations to do the shaving for you.

Or do you:

Look at the mirror when shaving and make sure you don't cut yourself at any point. You are aware where you want to shave, and make sure you don't bleed to death in the process through severe cuts.

This is the subject of a video I'm planning, which is an extension of this one:

"An ineffective knight outpost chooser"

If one has really mastered the art of calculation like a computer with "trees of analysis" wouldn't one pick sharp battles and sharp tactical openings?!

http://www.saund.co.uk/britbase/has...

It would be good if Chessgames.com could get these PGN downloads and incorporate them, as there are many Kotov game examples here e.g.

[Event "Hastings 6263"]
[Site "?"]
[Date "1962.??.??"]
[Round "3"]
[White "Kotov, Alexander"]
[Black "Hollis, Adrian"]
[Result "1-0"]
[ECO "D95"]
[PlyCount "121"]
[EventDate "1962.12.??"]

He tied for first with Gligoric (both 6.5/9) ahead of Smyslov (6.0/9).

Maybe I have some kind of gift, but I can usually tell what works and what doesn't really quickly. Yes, some games can be won Capablanca style "I look one move ahead- the best one." In speed games, (and when I move too fast, one problem I have on the internet.) that works. I win lots of games that way.

But in a 90/30 moves game, I usually look if my opponent is doing anything with his move. So first step is: Assess opponents threat.

So once we identify what my opponent wants to do, I try to combine my plan with stopping his. See "My System". Of course, tactics predominate over plans, so we figure that out first.

If the threat is tactical in nature, then we either resolve the treat, counter it, or ignore it. See "Looking for Trouble" by Dan Heisman.If we can meet the tactical threat with combination of our plan, then do so.

If there is no threat, then it is a quiet move. Piano, if you will. Now we can assess the position after ridding ourselves of threats. Of course, if the threat is of position of nature, we are able to assess the position. Still remember that tactics are better than positional things, or at least in short term.

Now we must find a plan, or reassess our old plan. After this we break in to hardcore analysis.

Now we find candidate moves.

This is were I can find moves very quickly.

Now check for tactical errors in the moves, and then analyze a little bit in to each one, and then narrow in down a little, the go in as far as I can depth-wise, assess that position, then go backwards to look for improvements for both sides, and finally blunder-check.

Now I play my move, and hit my clock.

Kotov's approach did not work for me. I did not find his books realistic for my level of (in)competence at the board.

Except for his Essential Knowledge the same goes for Averbakh's admittedly impressive tomes on the endgame, either.

That said, I was thinking how many great players the Soviet Union had in the 1940s-1950s who now seem underappreciated and understudied... Kotov, Boleslavsky (especially), Lilienthal, Averbakh ... even lessor lights such as Lutikov. There games can be a goldmine of learning and enjoyment.

< "When you have finished analyzing all the variations and gone along all the branches of the tree of analysis you must first of all write the move down on your score sheet, before you play it." >

Very strange indeed. Almost as strange as the fact that Kotov was obviously able to play a game as late as in 2006, more than 25 years after his own death :)

Very probably, the reason for both is simply the existence of another player with the same name.

< "When you have finished analyzing all the variations and gone along all the branches of the tree of analysis you must first of all write the move down on your score sheet, before you play it." > -Kotov>

The first concept, which was well outlined (I recently discovered, though I have been thinking this way for many years) in a training video by Maurice Ashley is to NOT look at what your opponent's most recent move DOES, but rather at what it UNdoes. Every time any piece moves, it abandons its previous position, and in so doing un-protects something, or ceases to have certain moves available to it. So before you think about anything else, first look to see if your opponent's move has created any weaknesses, unprotected something, or allowed a previously impossible combination. Only after digesting all of that, THEN look to see what threats his move created against you.

Most people are so panicked about the threats that they neglect to look for the opponent's newly created weaknesses.

The second analytical technique I use which Kotov (and everyone else) completely ignores is to calculate moves in REVERSE order of likelihood and logic -- especially in sharp positions. Thus, when it's my turn to move and the position is dangerous and unbalanced, I initially very quickly consider all the WORST and most RIDICULOUS moves I could make -- just to ensure that I'm not missing some brilliancy.

So, the first moves I consider in any position are ways to sacrifice my queen for no apparent reason. Next, if one of my pieces is attacked, I ponder what would happen if I just allowed it to be captured. Then I see what would happen if I walked my king directly into danger. Then I ponder ways to give away material purposely. Etc. 99% of the time, I very quickly confirm (in a matter of seconds in a blitz game, or maybe a minute or two in a tournament game) that these terrible-seeming moves are indeed terrible. Ah, but that remaining 1% -- that's when the magic happens. You see a brilliancy that no one else even considered.

If no brilliancy manifests, then little by little I eliminate all the self-evidently bad moves until I am left with a small number of "good" candidate moves, and I can concentrate on those without any nagging feeling that I'm missing something.

You'd be surprised how often I find unexpected and disorienting replies in my games that really keep my opponents off-balance.

Re. your second point, I have a vague recollection that Kotov wrote looking at sacs was the starting point for Tolush. Satisfied there were none that worked, Tolush would begin to analyse. Possibly the same approach you use?