I know this is kind of a nothing post, but the forum seems somewhat dead, and I'm super excited to be learning backgammon at the moment and I really have no one to share that excitement with as no one I know actually plays the game.

Anyways, I've mostly been playing on backgammongalaxy.com (~1700 atm). Often I only have time for a single point game, so no cube. I started to also play against xg set in unlimited mode and noticed how much worse my performance rating seems to be in xg as compared to playing at Galaxy.

I find this very interesting as it can only be contributed to two things:
1. cube play
2. xg is so much better than my average opponent that the positions I get are much harder to play

It's point 2 that I find most interesting. In a vast majority of my games vs xg I seem to get into tricky positions and blunder a lot of these. It's probably a better resource to actually practice vs xg than play on galaxy, but it's fun to play humans.

For now, I'm just working my way through 501 Problems. Not sure what version I have. Published in 2014 if memory serves me. I have a few other books I bough a few years ago (Modern Backgammon, Backgammon Boot Camp) that I'll look at after I am done 501, but I'll probably go through 501 a few times. I think the spaced repetition will be good.

Simple question here:


The 6 is clear but why is 23/22 better than 24/23? What's the intuition as my intuition is to simply make the anchor. I know the 23 isn't a great anchor, but it's something.

His board is not that strong and he would be blocking your 6s if you anchor on 23.

One point matches on Galaxy vs. unlimited vs. XG -- I'd guess it's mostly the cube play and the fact that when gammons count, things are more complicated and there's more room for huge errors. For example, in a one point match, a big error might be giving up 5% wins, but in a money game, maybe you're giving up 5% wins and then also losing a bunch more gammons at the same time.

For the position, I think uberkuber has it right. The 23 just isn't a very good anchor here, since it's so blocked, and if White makes his 7 point, then you're totally blocked in. So it's crucial to try for more despite the extra risk (the risk isn't very high here). Note, you could get a better anchor or you could end up hitting in the outfield by virtue of the fact that you remained split.

Agree with both -- the 23 is a weak anchor and you only make it in the opening if no other decent play is available. The 22 is much better since 6s won't be blocked. The risk is still very small, so go for that.

As to your other question, it's because you're playing cubeless and gammonless games on Galaxy but using the cube and allowing gammons on XG. Take a look at XG's "Summary" window after a session and see how much the cube errors are adding to your error rate. It's probably a lot. Try playing 3-point or 5-point matches on Galaxy and your PR will shoot up and probably resemble your error rate on XG.

I knew the above question was simple and basic, but I asked it to reinforce some fundamentals. In the game I made the 23, but I see how it is blocked and why it was a small error.

Thanks everyone.



Yesterday I forced myself to get into a number of 3 point matches and you are totally right - my PR shot up. I've got a lot to learn.

You make the 23 when it's like "oh **** he's super strong and is about to pounce on me I really need to get anchored up!" That's not remotely the case here

edit: I want to delete this now, but I spent time preparing it so I'm not going to. Feel free to ignore.

I think I'll just treat this as my improving thread - to keep things self contained. If people think it would better to post new positions in new threads let me know. I don't have a strong preference.

So, I'm playing XG in an unlimited match and this position comes up:


I consider hitting lose on the 1, but decide to make the inner board point instead. I understand I'm supposed to attack a single checker, but my board is slightly better and another point would improve it further. I play 6/2 5/2 and it's a -0.1 blunder. I'd be better hitting or even leaving 3 outfield blots with 13/9 13/10. I understand 8/1* but not 13/9 13/10. I guess my play is wrong as it's too early to bury two checkers in case he escapes.

He rolls a 33, then I roll a 6/1 in the following position:


Now, I play 8/1* and now it's a mistake to hit lose on the 1 in favor of stacking up the 6 point. I played 8/1* through elimination - as I didn't like putting 4 checkers on my 6 here.

I'm struggling to understand the subtleties of these two positions.

I think I understand now: As I just finished typing this I think I get it. In position #1 after hitting lose I have a lot of covers, but in position #2 after hitting lose it is much harder to cover as I don't have a lot of builders pointed at the 1. Yeah, stacking the 6 also leaves a builder on the 8. Sigh, I guess there is value in me typing this as it forced me to work through the position differences.

I think you meant to say you rolled 5-2 (instead of 6-1) as shown on the picture.

I think you'll find as you play cubed matches that you get into much weirder positions against humans than XG. That has definitely been my experience. I often look at my position in a game and think, man, XG never has a set-up like this, where have I gone wrong?

When you play a lot against XG, you will start to notice tiny pauses between throws that indicate there are doubling calculations at work (at least I think that is what is going on) that alert you to stop and think more than you would normally.

My play against XG tends to about 2 PR higher than against humans, mostly I think because of the orthodoxy of XG's play, and the doubling hints offered as mentioned above. Coming from a PR of 10ish against humans.

You will like always find your cube PR better against XG because so many of its doubles are takes since it doubles at the first opportunity, unlike humans.

So how do you get better at the cube? It is costing me a ton of PR / equity because I am terrible at it.

Still working my way through 501 problems. I have Modern Backgammon and Backgammon Boot Camp, but I'm not sure what to work through next. 501 has some cube problems, but I'm still not getting it.

Just finished a 5 point match and when I was up 3-0 I doubled in a position I thought was a double and it cost me -0.76!! I checked it in XG and in a cash game it is a clear double, with the score 0-0 or 1-0 it is also a clear double, but at 2-0, or 3-0 it is a no double. In hindsight I can see match score is the main reason, and I feel good about the double even though match score turned it into a mistake because I'm still trying to find good double opportunities.

Anyways, that was a bit rambling, but cube is costing me a ton. As well as anytime I get into backgames - I have no clue.

Boot Camp has a lot of really good cube stuff for beginning-intermediate. Woosey's 2 cube reference position books are great, too.

Awesome. I am like 50% through 501 and I looked at the table of contents last night and I think a lot of trickiest positions I deal with are still to be dealt with. Not that I absorb every position the first time I see it.

Edit:
I have the purple cover 501 - after doing some research just now this seems to be the 2nd edition. Any value in picking up the 3rd edition is the 2nd good enough?

Mark: I got better at the cube (without costing me rating) by just playing matches against XG over and over again. With experience you get a feel for it. And also NOT playing 1-point matches! That's not going to help your cube play. By the way, I play at least 2PR better in 1 pointers, with no cube, so that would explain some of your PR difference.
Seriously, ask any good player and they'll tell you about the 6 month stretch where they played against XG 8 hours a day or something...
You're also correct that since XG is so strong you tend to get simpler positions than you would against real opponents. I remedy this somewhat by playing against XG on a weaker setting, like Expert.

The purple one is fine. The mistakes aren't a big deal. But if you really want to know what they are: http://timothychow.net/robertie/robertie501.html

So, I finished off 501 Problems. I am going to need to go through it a few more times to assimilate all of the information in it, but in the mean time I am deciding which book to read next between Modern Backgammon and Backgammon Bootcamp. Any advice as to which would be more beneficial?

As for 501 problems, I really struggled with the final chapter on bear off. I get the 8-9-12 rule, and need to practice with pip counts to be able to do the calcs faster, but this isn't my concern. My concern is that a lot of the problems come down to small percentage difference and how do you actually come up with those over the board? Also, how did Bill come up with those percentages? Roll outs? I trust them (although he has a number of pip count mistakes in this chapter which I found surprising), but how do you actually do this in practice? Is it just a lot of memorization of critical positions and then making mental adjustments from them or...?

Instead of calculating percentages, use multiplication. If you want to know whether the pip count is within 8%, take the difference in pip counts and multiple by 12.5. If the product is greater than the leader's pip count then the difference is more than 8%. For 9%, multiply by 11.1, for 12% multiply by 8.3.

Awesome, I was going to sit down one night and work out a shortcut to make this simpler, but now I don't need to as this is exactly what I wanted!

Although, that's not what I meant with respect to my previous point about the percentages he talks about and the sort of "logic" behind some of these positions. I'll give a couple of specific examples:



In this position the argument Robertie makes is:
Now, he says to memorize this position as a reference position, and I understand that, but where did he come up with the 79%? Back of the envelope reasoning that he uses in other places or just a rollout?

or..



Without writing his entire argument, I can, but I assume you have access to it:

You've got your diagrams and comments mixed up. The first position you show, with 8 checkers on both sides, is Diagram 481. It's a 4-roll position with Black on roll being 74% to win. The comment about Black being 79% to win is the comment for Diagram 480, a three-roll position where Black is 79% to win. I got the numbers by directly calculating them.

My apologies, it was late when I posted. I thought I had them right.

And thanks - feels like a lot of work to calculate these positions out so that is where my head was at when I was reading the answers - how did he come up with these numbers and am I supposed to be able to do these calculations over the board?

I've started going through Modern Backgammon now and noticed it was published in 2001. Do you feel the material is still relevant with the advances of the bots since then? Do top players think in these terms: Efficiency, Connectiveness, Non-commitment, Robustness? More importantly, should I as a beginner / intermediate?

So far the book feels excellent and regardless of the specific concepts the descriptions of the positions and thought process feels like I can learn a lot from reading it.

I double checked and indeed, I made a mistake on the first diagram, but you have answered my question.

For the second diagram, I am correct - that is the quote for position 484 about stripping 12% off.

@MarkD How were you able to get a copy of "Modern Backgammon"? Have you had it for several years already?

Because I've been trying to get one myself, but it seems very complicated to do so.

Yeah, I had an interest in backgammon a few years ago, but there was no one to play and I quickly lost interest. This time feels different as I at least have opponents at Galaxy. Whenever I get interested in something I load on a number of "the best" books on that topic, and at the time Modern Backgammon must have been considered one of the best. This was probably 6 years ago, but I never read it.

I think even 6 years ago it wasn't easy to obtain. I remember I had to order this one from Robertie's website or maybe Carol Cole's website and I am in Canada I probably had to pay shipping. I see it's out of print now.

Game 3 of a 3 game match and I'm winning 2-1. Opponent starts and rolls 6-4 -> 24/14. I roll 5-4 -> 13/9 13/8. He doubles. I take. I make a -0.217 error by taking and I'm supposed to pass?

I don't understand this as I thought these would just be automatic takes. I guess it says I'm a 55/45 dog in this situation and if I pass I'm 50/50 again... This is crazy.