It is better to take a broader view: The more methods we associate with tewari the more we can apply. Of course, it is possible to call some other methods by different names like "move order analysis":) Then one needs another collective term instead of tewari though, maybe "retrospect shape and sequence analysis tools". I prefer to stick with the shorter "tewari".

It is so important that it has gotten its own strategic concept "timing".



First of all trivially: When tewari has been applied, it can or cannot have told us that something is suboptimal. It is sometimes possible to express a measure of how great the benefit it: E.g., if tewari assesses x stones to be superfluous, then the benefit is that great: x stones. Early in the game, that might be worth up to almost x handicap stones. Otherwise measuring the benefit of tewari is possible by other measurement tools like those assessing values (such as territory efficiency or mobility difference) or being analysis methods (such as local positional judgement or locally or globally applied global positional judgement). Such other tools range from simpler to much more complicated. (Details see my book.)

It's interesting that you phrase things this way. This seems to suggest the possibility that tewari is not intended as a tool to tell us if something is optimal - but can merely point out, sometimes, when something is suboptimal.

If this is the case, then there seems to be no conflict between tewari analysis and move order: move order is of great importance, and tewari can be used independently of move order to determine if something is not a good move.

In other words, it would appear that an optimal sequence will always "pass" a tewari analysis, but a sequence that passes a tewari analysis is not always optimal. Does this sound accurate?

_________________
be immersed

No, because tewari is not developed enough yet to be always applicable successfully.

"Who guards the guardians?" (or watches the watchers, or polices the police, if you prefer)

It's a rhetorical question, usually used to point out some sort of injustice perpetrated by those in positions of authority. The quote is probably originally by Seneca, though I'm not going to take the time to check Wikipedia (about to leave for work.)

(EDIT: ok, I checked anyway - it's by the poet Juvenal)

In this case he's responding to Kirby's question about how to know if your tewari analysis has come up with a correct result... basically, you can't, unless you are fairly strong and already have a decent amount of practice with tewari analysis. So "who analyzes the analyzers?"

I know the quote. I was facetiously (and rather poorly) referencing a meme.

My bad - I still haven't played either Portal game, so while I know the common meme from Portal, Portal 2's stuff hasn't trickled its way down the wall of pop culture into my brain yet.

Tewari analysis can be defined as “Breaking down a position by eliminating an equal number of stones in order to analyze the efficiency of the moves”. By doing this, players get to analyse on how to make better and more efficient moves, why some variations are better than the others etc.

This can be applied to any daily chores that you do. Try to think of something that you do, step by step, and map it out on a piece of paper. Then systematically look at each step, backward, forward, and think of the steps that can be eliminated and the steps that can be combined and after eliminating and/or combining them, you still achieve the same objective of that chore.

This was said on a blog "Falling Stones are not Heavy" in the Weiqi/Go page.
Basically saying the same thing Toge said to a point but this may make it easier to understand.



Its like knowing the wanted result and knowing how to get that result through a X sequence of moves but looking at it and seeing if it can be done by following different sequence of moves that use less moves by removing the superfluous stones / bad moves or useless moves to achieve the same or better result.

I had a teacher for a bit that said, "Look at a sequence that had a desired result and count the number of moves it took you and the number of moves it took you opponent. If it took you more moves than your opponent to achieve, how ever many more stones you have vs your opponents are superfluous stones, do it again through different sequences, moves untill you get the same or better result with the same number of moves or less than your opponent." He then went to say that even understanding it that way is wrong but that it was a simple way to put it for a beginner to understand. Another way he explained to understand it is like when you are reveiwing your game at the end and looking at the better moves that could have been played in place of bad ones. Instead of playing here you play here instead because it has this effect which is better or when you see a sequence at the end and go oh if I played here then here then here instead I get a better or same result in fewer moves or such, tewari would be seeing this during the game play and using it. Thats how it was explained to me but it sounded better than how I am trying to relay it.

I have a Pro game saved some where it shows a really good example of the use of tewari, ill look for it but it may take me a day or two to find it.

_________________
Only a child can ask questions that the wise can not answer.

I feel like everyone who's read watchmen knows that quote

_________________
"There are no limits. There are plateaus, but you must not stay there, you must go beyond them. If it kills you, it kills you. A man must constantly exceed his level." -- Bruce Lee

This is basically how I have always understood tewari. It will not tell you that you have played the right moves, but it can be a tool for discovering that you have played the wrong ones (=

Since the go board has no memory (barring superko), once you have completed the position, all transpositions can be evaluated equally (with respect to the final result). This means if you find a transposition where one player was playing reasonable "correct" moves, and the other player has made a mistake, it is likely a suboptimal position for that second player.

If you think of it like a mathematical sum...If moves all have a certain value (known or unknown)

-Start with position X, move to position Y via one of two sequences A or B

-We know original position X + B1a + W2a + B3a + W4a +B5a + W6a = Y and that X + B1b + W2b + B3b + W4b +B5b + W6b = Y

-If we can assume that all of white's moves are good when played (optimal ideally, but we're human), then these should for all intents and purposes cancel out. The value of the original position also cancels.

-We end up with B1a + B3a + B5a = B1b + B3b + B5b

-Evaluating the intermediate positions we show that when played, one of B1b, B3b, or B5b is suboptimal.

It must be true that somewhere either in an individual move or combination there is a suboptimal component in sequence A.


Of course all of this would depend on how well your assumptions made throughout this reasoning process hold (e.g. if white makes suboptimal moves as well, you can't cancel).

IMHO, even allegedly good tewari reorderings leave a lot to experience or imagination in terms of judgment. They are rarely as clean as "one side plays all reasonable moves and the other plays some obvious junk." As John Fairbairn mentions, it's often used to compare a new sequence to one that is already presumed to be even by other methods. So you already have to have some kind of reference point, such as trusting that a joseki really is a joseki. However, old josekis are falling out of favor all the time, so it's hard to be certain of even the reference point.

Click Here To Show Diagram Code

[go]$$B Old Joseki
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 4 . . . . . . . . . . . . . . . |
$$ | . . 5 . . 2 . . 6 . . . . . . . . . . |
$$ | . . . 1 . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . 3 . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |[/go]

Click Here To Show Diagram Code

[go]$$B New joseki
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 4 . . . . . . . . . . . . . . . |
$$ | . . 6 . . 2 . . . . . . . . . . . . . |
$$ | . . . 1 . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . 3 . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . 7 . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |[/go]

elsewhere

Click Here To Show Diagram Code

[go]$$W Tewari of new joseki
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 7 . . . . . . . . . . . . . . . |
$$ | . . 1 a . 3 . . . . . . . . . . . . . |
$$ | . . . 6 . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . 2 . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . 4 . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |[/go]

So here, the claim is that the tewari of the new joseki shows to being a joseki and therefore reasonable for both side, but that somehow in reponse to , is played as a diagonal rather than at 'a' which would clearly be better. This has been given as one of the reasons that black doesn't respond at the 3-3 point in the original variation as much as in the past. Of course, even though is bad, I sort of wonder about as well, so I don't find this analysis terribly convincing. But I guess some pros do, and so we are back to trusting pros.

I personally would like to hope that tewari can help me figure out for myself whether a move is good or bad or whether a result is good or bad. But actually in the end I wind up asking stronger players anyway. I think to use tewari effectively, you often have to already have an accurate sense of how much worse bad move X is than bad move Y, but that seems to be pretty hard even for fairly strong players.

I saw a lecture by Janice Kim 3p not to long ago and she was asked to judge who was ahead in a certain opening. She said it was hard to tell because both players had their stones in the wrong places!

It reminds me of a joke:

What's the difference between a Japanese joseki and a Korean joseki?

Yes, this is precisely the problem with most tewari analysis. It is faulty reasoning to say "In answer to , would be a mistake, so W is better here", because there is no guarantee that is a good move. The question needs to be more convoluted: "If W promised to play , and B agreed to reply with the (non-optimal) , who would come out ahead?"

Not that often. E.g., identifying superfluous inside stones is straightforward and counting their difference of numbers possible also for kyu players.

In the "new" joseki, black tenukis at , and in the tewari analysis, it is white who has gotten in a play elsewhere at . Does this not play a role in the evaluation (i.e., serve as justification that the suboptimal is not at a)?

_________________
Patience, grasshopper.