Problem of the Week #23: August 9


Cash game. Center cube. Black on roll.




Part (a): Black to play 4-4.




Part (b): Black to play 5-1.

1) i'd go with bar/21 13/5* 13/9

i don't really see any other options beside bar/21 7/3*(2), but that seems bad because i think we want the bar point, and there doesn't seem to be any good 4 after 7/3*(2). plus white can easily anchor on the 5pt. at least 13/5* 13/9 gives a chance of forming some sort of prime

2) i think black just wants to run out with 22/16

edit: i guess in 1) just 13/9(2) and leaving them there might be better but the other 4 seems awkward. i dunno

#1 is bar/21 13/5* 13/9. We are fighting for our 5 point with a majority of rolls that will cover on our next roll if we are not hit. It seems pretty clear to me that this is our best move due to lack of other acceptable moves.

#2 is either 22/16 or 24/18. 22/16 seems better because if we are hit, it forces our opponent to break an anchor. It also does not force us to break our anchor on the 24 point.

In a) I play b/21 7/3(2)* 13/9 pretty quickly.

b/21 13/5* 13/9 looks "prettier" but when you go through White's responses, only 44 and 66 aren't good rolls. Every single other roll either hits or anchors. Plus you don't much want to get into a slugfest w only one home board point vs. Villain's two.

With my recommended play, a whole bunch more can go wrong for White, and you are then well-positioned to attack. And if it turns into a slugfest, at least you have two home-board points made.

Plus priming is a questionable strategy anyway with your army split in two. Attack, hope to keep Villain on the defensive (ideally on the bar), and join up your forces again when the opportunity presents itself.

In b) clearly moving the runners is correct. 22/16 to fight for the outfield looks pretty standard. 24/18 leaves you too vulnerable to disaster.

a) Glad we have this problem because I'm hopeless at this sort of thing. I can see arguments for:

1) bar/21 7/3*(2) 13/9
2) bar/21 13/5* 13/9
3) bar/21 13/5* 7/3*

1) is a more solid choice, but is solid what we want when this far behind in the race? Sure villain has two home board points to our one, but we have one anchor to his zero. Also this choice fails to knock him off the five point.

2) leaves a lot of return shots and chances for White to make the three-point anchor which will be a good anchor given our current inability to cover the nine point. It looks pretty at first sight but on further inspection seems to allow too many good rolls, without achieving enough.

3) is the nuclear option. I don't think it's any immediately looser than 2) and it achieves quite a bit more, with two in the air and White's chances of making a good anchor next roll decimated. The downside obviously is that it leaves blots all over the board, but with a good anchor in Black's board plus the bar point, we should be able to parlay ourselves into some backgame-ish situation if things go pear shaped.

I'd play bar/21 13/5* 7/3*, but I have a very good chance of being wrong.

b) I'm going to buck the trend here, I think 24/18 is clearly correct. That 1-point anchor is a liability, not an asset. This is going to develop into some sort of back game and my anchors are currently too far apart. I want to make the 4 or 3 point in addition to the 5, not the 1 point. The bar point would be OK as well. By playing 24/18 I get to slot the bar and 3 points. Yes, playing 24/18 is looser and I might get hit or even double hit. So what? My opponent has a 1 point board and I have his 5 point, so not much can go wrong. Not exactly scared of him achieving a bigger race lead.

Looking forward to the analysis and to seeing what the unifying theme is.

Maybe, but what's the rush? Just chill and await developments. I think the primary goal at the moment is to avoid getting crushed. After 24/18 villain has some killer doubles, a bunch of double-hit rolls, and a bunch of pointing hits. If things go wrong, it gets ugly real fast.

In problem 1 I stared at this for 5 full minutes and I really don't see any of these hit variations being any good. I dont hit at all.

I just like bringing down the 13 point to make the 4 prime and either slotting the 4 or playing behind with 6-2.

Even bringing in the whole stack to make the 4 point is fine, not to mention just moving the anhcor from the 18 to the 13 and slotting the 4 point.

I really cannot see any reason to hit . It is awkward and you have no board to back it up.

Problem 2 I race out with either 22-16 or 24-18. Just start playing for contact and dont touch ur front boards yet. Not sure what these positions have to do with one another but i am sure we will find out!


sheets

Some tough ones this week.

a) I don't really like anything here. It would have really been nice to make the 5 point but alas, we are forced to play bar/21.
So, my vote is to play 8/4 for the other 3 4s.

The 4 point should be useful and trading the 7 point for the 3 point and then having to leave a blot for villain does not seem like a good idea.
I can't really think of anything else that makes much sense.

Moving the checkers from the 13 to 9 point and playing 6/2 seems like a mistake if not a blunder.

b) I this position, it is important to get some more checkers into the home board and start building the prime. 9/4, 9/8 looks bad. So, I vote for 22/16.

a) I'm not very happy with 13/5*. I don't really see much upside, but there seems to be plenty of downside. You lose your midpoint anchor while creating two blots. You open yourself up to losing yet another 20-something pips in the race because of whites numerous return shots. You're behind in the construction which points towards playing safe instead of bold. It seems like a lot of risk for not much reward.

White has two checkers back, so I'm thinking about primes instead of blitzes. This takes 7/3*(2) off the table. (I think I would take 18/14(2) over 7/3*(2). I just don't think launching a blitz is really going to be a winning strategy here.) White is pretty far ahead in the race, so he's timing in a priming battle is going to be significantly worse. Therefore, I play 13/9(2). This has the added bonus of blocking the 6s for the back checker.

With the last 4, I play 6/2, I guess. I don't want to abandon either anchor in the back. At least 6/2 doesn't offer any shots to white and unstacks the heavy point. 8/4 is the only other option, but that means that all the spares are sitting on the 6 point, and that makes me uncomfortable. I don't like how it puts that checker so deep, but it's better than leaving a shot.

b) I don't know if I really want to hold onto the ace point here. I would like to move out to the bar point and cause more problems in the outfield as white brings those checkers around. But there's a whole lot of risk associated with that play since there are white checkers everywhere that would love to do two good things by hitting and slotting the bar point. I don't think I can really stomach it. So I play a little less boldly and just run the blot out to the outfield.

You bring up interesting points, but it's worth commenting that 7/3*(2) is not necessarily a "blitz". You secure a home board point, and you put Villain in the air, and those are two powerful and concrete achievements. Depending on how things go, the game can develop in all sorts of ways. You could just as easily end up trying to prime a checker that enters on the ace or deuce point.

When you look at the games of top players, often they don't fall into a nice category. It's just punch vs. counter-punch until somebody finally brings home an advantage.

In this case, I don't agree that there are *that* many ways for it to develop. If white makes the advanced anchor, he's just going to hold that point while you try to clear out your anchors from the outfield and eventually his home board. If he doesn't anchor, you're going to hit loose on the 5 point to prevent him from anchoring, and you're essentially playing a blitz that you can't finish (because too many checkers are too far away).

You're almost certainly not going to end up priming a single checker because you already hold two points that are 5 points apart (3 point to 8 point). You're not really all that likely to fill in the gaps in that prime before he can get his checkers out. You're going from a decent 3 prime to a no-prime.

In Modern Backgammon, Bill talks about being non-committal. And you're hinting at this idea by saying that their play doesn't necessarily fit in a nice category.

The problem with 7/3*(2) is that it *IS* committal. It's making a somewhat deep point which messes up your ability to make an effective prime. There's a two point gap between the 3 and 6 points, and you've created a hole at the bar point. This setting up for a blitz in the same way making the 2 point on an early 64 is a blitz.

And it's very committal relative to 13/9(2). The 4 prime play gives you the option of playing a priming game, but if white fails to anchor, you're not necessarily afraid to attack the blots in your home board, and this is much easier to do with 11 checkers on that side of the board instead of just 9.

Wait a minute... you just dumped a checker onto the 2 point, giving you 10 left... and I kept all 11 in play, since the midpoint checker will join the battle in most cases. So I have to admit I'm not quite sure what you are getting at here.

Plus Black will have 3 builders for the bar point, so it's not like you are giving up on it completely. A prime from 3-8 is still a real possibility, and extremely strong if completed, whereas the 9-point is notoriously transitory and of questionable value.

Sorry. When I was considering your play, I had only B/21, 7/3*(2) in my head, so I only saw 9 checkers, not 10. This does change my opinion about your play (it makes it worse!). White now has a huge number of return shots (11/21/51 hit the lone checker on the midpoint, 41/42/44/54 hit the blot on the 9 point, which makes a total of 12 shots), and if he gets one of those your game falls apart completely. In addition to those shots, he also has 52/53/55 which make an anchor or better, meaning that half of the return rolls do something very positive for white. That's a lot to risk for a nebulous reward. Yeah, you make an inner point, but it's a little deep and not connected to your other point, plus you're putting a gap in a developing prime.

In my play, there are 11 checkers to pull off a blitz because the slotted checker provides an opportunity to make another point. So that checker is "in play" in the blitzing option.

If white anchors, he's then going to have to clear his points carefully because you've got double coverage in the outfield. He can easily find himself leaving you with 15-18 shots to send a third checker behind your prime.

If white doesn't anchor, his rolls are going to be somewhat awkward because he won't easily be able to leap your 4-prime with the back checker. This means that white may be forced to leave a shot for you to attack as you release your back anchors and bring them around to the support and extend the prime.

(I mislabeled this position as a "priming battle" since white isn't playing the priming game with you. What I meant to say was something about how white is going to have to be the first to break his anchors, which means that you don't have to worry about your prime crunching.)

I doubt the 3-8 prime is really that likely. You don't want to be building primes from the outside in. That's just not an efficient way of doing it. It's better to have a small prime and add to it. Besides, by the time you get enough checkers around to fill in the prime completely, white will most likely be long gone. You've got four checkers tied up in your two back anchors, so you can't fill in that prime until you release one of those two anchors.

While the 9-point will not be a permanent structure in this game, it has the very powerful benefit of blocking 6s for the back checker. This is important because it puts white under more pressure.

I strongly disagree with this evaluation, so let's leave it at that.

Well I got sick of waiting so I did some rollouts of the first position, with the following conclusions:

1.) I suck at backgammon
2.) With 44 reasonable candidate moves, I have no idea how anyone is supposed to figure this stuff out. I guess you do rollouts, and then make up the reasons later.

Excellent point, this is probably the "tiebreaker".