Miai counting/tedomari

Bass: If I compare miai values directly, and add and subtract like ordinary numbers, will I not go wrong in the situation where the three remaining endgame plays are A: a two point reverse sente, B: a 6 point gote, and C: a 5 point gote? The respective miai values are 2, 3 and 2.5, and yet, A (with the lowest miai value) is the only move.

tapir: Most probably that is because this example involves tedomari considerations with an unnaturally big size of the last given move. Deiri counting would mislead in the same way (giving 4,6,5 points as size respectively.)

Bass: I am not yet quite convinced that miai counting is as useless as deiri, it is entirely possible that I am just missing something. I wonder if Bill could enlighten me?

Bill: Neither deiri counting nor miai counting involve tedomari. Quite the opposite. They assume a rich environment of many plays ranging from the size of the play in question down to zero, in the sense that in such an environment the biggest play is best. When you have tedomari, you do not have a rich environment, but a drop in temperature, and the biggest play may not be best.

Bass: Ok, I see. Might it be possible to add some counting rules to take tedomari into account, along the lines of gote move addition having every even-numbered term negated, so that 6+5 -> 6-5 = 1, and 5+5+4+3+2 -> 5-5+4-3+2 =3 or something like that? I have not thought this out at all, but if it were possible, then the system would be much better applicable to real life go situations. It would work in this situation, at least.

Bill: Well, just use the miai values and plain arithmetic. If you make the largest play first, then the result will be 3 - 2.5 = 0.5 (The opponent gets the sente, but sente gains nothing.) If you take the reverse sente instead, the result will be 2 - 3 + 2.5 = 1.5 , which is 1 point better. :)

Note: It is conceivable that the threat of the 2 point sente is worth less than 2.5 points, and that it will not be answered. But I am assuming that the sente is currently sente (in a different sense, of course). :)

Bass: Ok, I think I get it now. The disappearing two seems to defy the rules of plain arithmetics, but I think I am just trying to put too much weight on this thing, which is not even supposed to work because of the big tedomari.

Bill: About them missing 2: 3 - Big + Big - 2.5 = 0.5 . Big = the size of the sente threat. :)