Araban wrote:
Joaz Banbeck wrote:
Araban wrote:
I do just as much on my opponent's time as I do on my own time, from reading to counting to everything.
Counting works because you have a rough idea of the temperature, and you can add 1/2 of that to your opponent's score. But reading is inefficient because you are always 1/2 ply behind.
Maybe I'm too tired, but I can't tell if there's sarcasm in this post or not.
No sarcasm was intended, and I apologize if it came out that way.
Araban wrote:
Reading on your opponent's time is certainly not inefficient just because you're 1/2 a play behind. If you read a sequence that you play the first move for and your opponent plays the second move as expected, then great - we can agree hopefully that the amount of extra reading you've done going over any variations in the sequence that both players are in was not for naught.
Certainly agreed. If the opponent chooses the variation that we have analyzed on his time, it works very well.
Araban wrote:
However, is it really for naught if your opponent plays a move that's unexpected and wasn't what you considered as you read during his time? In my opinion, absolutely not. The variations you've covered in your opponent's time, even though they were based on the assumption of a move from the opponent that never happened, serve as reference in determining whether the move he played was good or bad and allows you to make a judgement call on the local situation. If, after you've read out your opponent's unexpected move (on your own time now) and find that his move is better, than you've learned something new, whether it be a new shape or pattern or tactic or you've found a hole in your reading or you've made a silly mistake, etc. If you find that his move is worse, then it's a LOT easier to conclude that you've found the punishment for it because you've got something to compare it to - what you learned on your opponent's time.
It would be hard to argue that the intelligent player cannot find
some use for any kind of information about the game. But I'm looking for best return for my time, not just some return.
If I read N possibilities, the odds that my reading will be directly useful are 1/N. The average position has, as a wide guess, two to five reasonable possibilities. Some sequences have only one reasonable move, and others - especially early in the game - may have ten or more.
That is why direction of play analysis is so useful in the beginning of the game, when there are easily ten reasonable moves for the opponent, and attempting to read on the opponent's time leaves us operating at 1/10th efficiency. You have demonstrated how a player can get some use of the unused sequences that he has read, but I don't see that they make up a full order of magnitude.
I think that we get the best return for our time doing things other than reading. Consider direction of play analysis. If I determine the proper direction for every group on the board - and let's assume that the average game has 5 to 10 groups total - and my opponent does something totally unexpected, then maybe one or of the groups have a different direction of play. But the majority will be unchanged, and thus the majority of my analysis is still 100% useful.
The efficiency of counting follows similar logic: the majority of the board will be unchanged if my opponent does something unexpected, and thus the majority of my count will be useful. Again, this is way better than the worst-case 1/10 efficiency of reading.
Ko threats are a third possibility. I can count ko threats of a certain minimum size, his and mine, and end up with a useful piece of information. If the opponent makes an unexpected move, again the majority of my analysis is stil useful.
I'm not suggesting that you can't make some use of reading on the opponent's time, I'm arguing that many other things are more efficient.
Araban wrote:
Also, my argument only applies to local situations. ...
Ahhh..that may be why we appear to disagree at first. I have been assuming whole board for this entire thread.
Login
Register
FAQ
Search
Kirby
