| |
BGonline.org Forums
Challenges just beyond my competence
Posted By: David Rockwell
Date: Friday, 28 December 2007, at 10:54 p.m.
Here are some examples of challenges just beyond my competence that I have tackled in the past year. If there’s interest in further discussion of any of these items, I would be happy to elaborate. Believe it or not, this is my attempt at being concise. Each of these was inspired by the Scientific America article, “The Expert Mind” mentioned in an earlier post.
1. I saw a post on this site where it was mentioned that slotting often plays better than a split with an ace in the opening when there is a fourth checker on the eight point. I had not noticed this pattern previously and it wasn’t intuitively obvious to me why this would be so. So, I decided I needed investigate. I checked opening rollouts for examples of openings where the spare made a difference in the choice of response. First I tried to determine whether slotting was actually undesirable or just not as good relative to splitting in these cases. Then I analyzed subsequent moves to find significant differences in equity and more importantly variations in move choice due to the extra spare. This resulted in the discovery of a couple of themes I hadn’t noticed at first. (I’m sure I only scratched the surface of this topic). And, I should remember to slot more often in the future.
2. I’ve always been a sucker for taking hopeless cubes when I have a racing lead and one checker trapped behind a partial prime. Recognizing this weakness, I contrived a practice position where either side can reach a proper double quickly. I’ve played this position on Snowie hundreds of times. I run a Snowie evaluation after each of course. I have used this practice to improve my recognition of proper doubling and take points. I have identified key checker play decisions (e.g. run into the teeth of a quadruple shot or stay behind the prime for now) and attempted to understand when one plan is better than the other and why.
3. I was practicing bearing off against an ace point one day when I rolled 1-1 with the five and six points cleared and six checkers on the four. I moved 4-2(2) leaving no shots on the next turn and was dinged by Snowie for this mistake. 4-3(4) was preferred by a significant margin even though it left six immediate shots. I didn’t understand this. (I know I need spares on the 3 point, but felt the need for a deeper understanding than this. After all, my current level of understanding got the move wrong.) I asked a friend who is about as strong as I am and he told me, “Of course, that’s because of platitude XYZ.” So, he didn’t understand it either. In pursuing an answer to this quandary, I found other move choices I didn’t understand (with the 2, 3 & 4 to clear). I am currently compiling winning % at DMP for all combinations of checkers (a little over 200 of them) on the 2, 3 & 4 in a format where I can look at it and find the important patterns. When I succeed at that, I will check the mathematics of some sample cells so that I can understand the reason for preferring one formation over another.
4. I was practicing five point holding games one day when I missed a simple safe versus an indirect eight shot choice. I thought I understood why and tried moving the checkers around to similar positions to verify my theory. I stumbled into a position that sure looked similar to me that had an opposite conclusion. Obviously, I didn’t understand a key concept. To figure out what I didn’t understand, I used a technique which involves a fair amount of work, but has been extremely helpful. (I do this analysis for some position almost every day.) For each position, I create a spreadsheet which illustrates each of the 21 subsequent rolls, the preferred move and the equity after that move. (I use a weighted average of course.) I usually use Snowie 3 ply for the sake of time. With 2 move choices, that’s 42 data points. (This takes me about 30 minutes to compile, but a little less when I have 2 computers side by side.) When the schedule is complete, I try to group the rolls into similar categories so that I can understand the pros and cons of each move. In the example above, I did this for both positions (84 data points) and was able to conclude that the strength of my opponent’s board was the difference I was missing. It was right to leave the immediate indirect shot when a hit left good residual winning chances, otherwise play safe. I use this analysis technique for most of my blunders which I don’t understand. It yields an answer I can grasp about 75% of the time. When I spend an hour or more analyzing a position, I am likely to remember it when something similar happens at the board. 15 seconds checking an answer on Snowie is far less likely to stick.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.