return to indexE-mail me at kwoolsey@netcom.com
Perhaps my question might seem a bit silly, but how does one trainthemselves to become a better BG player? It seems like each expert player has a different set of favorite (recommended) books--but is there a general format which one should follow in terms of trying to improve their game?
Joe Carducci
Certainly not a silly question. In fact, I think the only silly question isthe one that isn't asked, since it was silly not to ask it if you wanted theanswer.
As to the best way to improve, that is an individual matter. Read all you canand watch and listen to experts. Every bit of exposure helps. And of coursebe sure to log in some playing time.
Most important, think for yourself. Don't accept anything anybody saysas gospel, regardless of how good they are. This goes both for expert humansand bots. Form your own ideas about priorities and concepts of a position.If something puzzles you, spend some time rolling it out (I mean by hand,not having a computer roll it out for you). This is how you will becomea top player.
When I was learning how to play, I took a trip to New York and kibitzed thebig chouette at the Mayfair club. Some of the best players in the worldplayed regularly in that chouette. I noted the positions where the playersmade a move which was different from my choice. Some of the moves I couldsee where I was going wrong. On others, I was not convinced.
A year later, I again visited New York and kibitzed the chouette. Again Inoted the plays on which I differed from the experts. This time I could seea few plays where I knew I was right and they were wrong, even though theywere the best players in the world. I had analyzed and understood theposition better than they had. Many of the players simply played by rote,doing what they had learned from other players. This is good, but not goodenough. You must be original and develop your own ideas about the game, andthen do the work to decide which of your ideas are correct and which onesmust be scrapped. That is how you will become a top-flight player.
Kit Woolsey
In the September issue Ron Dumont wrote, referring to the problem solving contest : � It would be fun to see the experts' results appear along with those of the readers. People not receiving top scores would know that they're in good company! �. Surely it's fun. Still, I don't think it's correct. Could be misleading , unfair (unclear to which side) and, perhaps, counterproductive - for many reasons.
First of all , readers and panelists (these two spaces are not quite separated - some , if not all, future top players/analysts will come inevitably from the space of present readers - the winner of the readers' contest is invited already to be in the panel for one month) have different purposes. The purpose of a reader when he sends his solutions is to win the contest. The purpose of each of the members of the experts' panel is different : to vote for the move that he thinks is best in the given position, which is sometimes the one that he simply feels more comfortable with. In other words, trying to win a problem solving contest is like playing in the Stock Market (with inside information [bots] being legal). The winner is not the one that knows which are the good companies , but the one that is good (or lucky) in guessing what other people think about this fact, even if they are wrong. The spirit of Ron is that panelists should just be placed along with the readers just for the fun of it only? I doubt it. He speaks of �experts' results� so, I guess that his spirit is that experts could and should be judged not only between them as a group (which is meaningless still, I think ) but �along with the readers�. Judged by who? By themselves, actually, which is unfair to the readers, since, it's according to their opinions (the panelists' ones) that the various scores are attached to the various moves (choices) in each problem. Imagine - if there was only one panelist then he would get maximum score, since he would be judged only by himself and the same would be true if the choices of all the experts were coinciding in each one problem (in the real world, in terms of probabilities, the fewer the panelists and the more they agree, the worst for the readers placement). The placement(=results) of the experts as a group reflects actually the measure of their agreement (unanimity) about the solutions. Furthermore, readers are allowed to use bots but panelists are requested (which is absolutely meaningful) to refrain from that. This , in a first look, seems unfair to the panelists, but, actually, is more complicated, if we think again who is the judge. Summing up, we have the panelists, who don't actually try to win the contest but to act as panelists, who do not use bots, so they don't know what Snowie votes for (but , each one of them knows the vote of a member of the panel -namely, himself), and readers, who try to win the contest and know what Snowie votes for. So each one (reader or panelist or bot) knows exactly the vote of one member of the panel (as long as exactly one bot is in the panel) . If there is no flaw in these thoughts, what would be the appropriate thing to do in order to have a fair competition (between) or ranking (of) the panelists, supposing that we see a meaning in this? A hypothetical answer would be �to allow experts vote two times - one as panelists and one as competitors for the contest�. This could be real fun : �I believe that play a is best , I would adopt it if this very same position occurred in one of my games, but, for the contest, I vote for play b, for I guess that the majority of the panel will vote for play b � (it must be said that for reasons of fairness, panelists should not use bots not even here -see above). Let me explain myself further : often I have a strong opinion about the solution of a given problem, estimating at the same time that the majority of the panelists will vote for another (specific) one. What should I do then? Dismiss my play in favor of what I think will be the unanimous choice , fearing that, in all probability, it will cost me points in the problem solving ranking? Or, stick to my solution, which, could be even a brilliant discovery? Worse than this, sometimes, a panelist knows that his solution will get exactly 50 out of 100 points , namely when it is simply not listed as a candidate move , by oversight perhaps (in the August issue the are two cases in point, one of mine and one of Laila Leonhardt). If he is going to be judged with a light heart by his total score, then he will face a dilemma. This I rate as my strongest argument , although , personally, I will continue to vote for a non-candidate move, if I would make it on the table.
Having said all that, let's go back and see if we really want competition (between) or ranking (of) the panelists. If panelists themselves believed that this is totally meaningless, then , perhaps, it should be of little interest for everybody. If panelists, on the contrary, believed that sure it has meaning and it is important for their reputation, they might start acting like competitors, which is something different, as I said already. In a non-perfect world (I knowingly exaggerate !) one might use even bots when nobody sees him, or ask a friend-panelist about his solutions. In no case a panelist would choose a move that would put him in the minority. He would try to be not himself, but another (a phantom-person, the majority of the others). This competition between panelists would not be fair, as well, for practical reasons. Someone can or wants to spend much more time for the problems, another not so and the second has obviously a disadvantage if he is going to be judged by this particular performance. In conclusion, I think that readers should compete but panelists should cooperate in a way, trying to shed light to the problems, each from his own point of view, sharing their reasoning, strategies, thoughts- telling when they are certain or uncertain- with the readers and the other panelists in the process, trying to teach and help bg players and create future panelists, being indifferent to any other consideration.
In any case, I am eager to hear other opinions or improvements about this matter. The subject is interesting, put for the first time (as far as I know) and the more I think about it, the more I discover hidden aspects, which I cannot mention here for lack of space.
A last-minute thought about the problem solving contest. What if there was a second round , in which the panel of experts would vote once more for the previous month's problems as well , under the light of other panelists' arguments (or, rarely, a non mentioned alternative move) that should be in common view by then? (I know, bots are stubborn).
George Klitsas
As you say, you will continue to vote for a choice you know will be unpopularif you think is is right. So will I. So will all the panelists, I'm sure.Our goal as panelists is to present our arguments for a play, not to getthe highest score. Personally I would be disappointed with myself if Iscored 800. This would indicate to me that I had not been thinking creativelyenough. I don't think there is any danger of a panelist hedging his answerbecause he doesn't think it will be the majority choice. We are all used tomaking fools of ourselves, and will gladly continue to do so. The reason Iinclude the scores of the panel members is simply because I though the readersmight be interested.
As for a second round of voting, I don't think that is a good idea (even assumingI could persuade all the panelists to vote a second time). The main idea behindthe quiz and the panel of experts is to let the experts air their views abouta position independently. Even if their views are shown to be wrong, they arestill of value to the readers who want to know how the experts think.
Kit Woolsey
Ron Dumont also answers George's letter:
I just wanted a chance to respond to George Klitsas on the topic of including the panelists' scores in the Quiz results. The intent of my request was definitely not to "judge" panelists, but rather for us readers to not get too hung up on the result scores. If a panelist gets a low score, that just means that he or she saw things in a different way from most of the other panelists. Different does not necessarily imply incorrect!
The key is to learn the most from looking at how the panelists approach each problem, and not get caught up in a "judgment" of actual scores. Displaying the panelists' scores is a way to realize that a high score is not that important.
Also, George stated that "The purpose of a reader when he sends his solutions is to win the contest". Given how the scoring system works, a reader interested in winning the contest would need to guess the consensus vote among the panelists. I doubt that this is the case. People submit what they feel to be the best solution to each problem.
Finally, I would like to thank George for providing hand rollouts in his analysis of the positions. While a limited set of rollouts may not carry statistical weight, it does provide for more insight into the positions.
Ron Dumont