Backgammon Learning

Learning

Study method

From:   Jason Lee
Address:   jleemath@gmail.com
Date:   19 January 2012
Subject:   Study method: self-made quizzes from actual matches
Forum:   BGonline.org Forums

I have a nicely burgeoning collection of matches that I've played over the board, maybe in the neighborhood of 200 matches going back about two or three years. Here's the plan ... I'm going to take positions where I've made an error bigger than some threshold. I'll print out the position. Then I'll flip over the paper and print again, this time position + bot info. XG has a very nice print function, where you can ask for position only or position + bot info. Enter a three-hole punch and a binder, and now I've got study material. Once I've got a critical mass (a few hundred, say), I can start going through them. I want a large enough amount so that I won't really "remember" the position when I go back through again later (although that's not so bad either). As I encounter a position, I decide on my plan, and then flip over to see the bot answer. If I got it right, I'll make a check mark ... if wrong, then an X. Then the problem goes back into the pile. Eventually, a problem gets seen enough times. If I'm consistently getting the problem right, I'll just throw it out. Maybe something like three straight right answers and the problem gets culled. Problems that I consistently get wrong might go into some special study pile. The theory goes like this: these are actual positions that I got wrong OTB. Something went wrong in my thinking. Maybe it was a brain fart. Maybe it involves a concept that I didn't understand very well two years ago. Maybe there's something I just don't understand, then or now. People talk about having a collection of reference positions that they study from. I haven't ever taken the time or the effort to create such a collection. This collection won't be particularly well organized at first. For now, I'm just going to focus on volume. Print out a ton of these, and start thinking about the positions. I welcome any suggestions to improve this study method -- particularly if you've done something like this yourself. I'd like to hear if people think this is a worthwhile endeavor.

Stick writes:

Your idea is for the most part good but I don't like how you're going to overload yourself with information. The key to learning is to be able to swallow bits of information at a time and internalize them for later use. If you smack off all these blunders over a certain threshold it's too much information. Here is what I suggest. Starting now, play at least one serious match a day of a decent length. In that match you will for sure make a few blunders, depending on what you define the size of that error to be. Take these positions, try to understand them, analyze them, roll them out if need be and place them in a separate folder. This folder should never have more than a dozen-ish positions in it at one time. You will want to set up variants of certain positions, roll them out. Things you still don't understand you can play out until you see for example why it's a take, where is your counterplay coming from. Once you have fully understood the position you can then move it out of that folder and make room for another. Repeat the process. If you stick 200 random positions in a folder and go through them you will end up memorizing the correct decision and not learning enough at all. At the end of the month you can have another look at all the positions you moved out of the 'blunder folder' just for recap and be confident not only in your play but more importantly what and why that decision is correct. This is time consuming. Make no mistake about it but if you want to really improve I think this is one of the best ways. Whenever I play a serious match the first thing I do after it is go through it play by play. If I find anything interesting I will keep it for the daily recaps. If I find anything I don't understand especially I will spend some time with it. People get frustrated on Gridgammon because I play a match and then they want a rematch and I tell them no, I'm going over the match we just played. Then they'll ask me two minutes later like I should be done by now. The reason I don't like using your old matches to pick out problems is that you now have no idea whatsoever what you were thinking at the time. This is why I push playing a match and going over it thoroughly immediately afterward. This way when you see an error you will know what you were thinking at the time and have a better chance of correcting that flawed thinking. In the beginning your blunder threshold will be higher and hopefully over time you'll have to lower it in order to have more positions to study. Make it whatever gets you an ample but not overwhelming amount of positions.

Did you find the information in this article useful?

Do you have any comments you'd like to add?