Game 3, move 13: Forum 2-1

[dorbel]

Spoiler

We should be grateful to whaat here, because he doesn't understand the need for the "dangerous" play and keeps right on trying to get it, thus affording a learning opportunity. As ever, why is more interesting than just knowing the answer.
If you make the "safe" play here, presumably 8/7, 3/2, it is safe, but only for this turn. Next turn our position will be worse than it is now because (a) our board will be weaker (and White's may be stronger) and (b) we have more burial, so fewer checkers that can be played. The six checkers on the low points will be able to play one 2 and three 1s between them, so all but dead. Note too that we will have burned the last safe 6. The remaining 6 checkers will somehow have to combine to make their way home, through an outfield crowded with White points.
If we make the "dangerous" play, probably 15/13, 8/7 but 15/14, 8/6 is an interesting alternate that I hadn't considered, we have to say "How much do we gain from this and how big is the risk?" What we gain is mostly that we have kept our board to prepare for the hitting that is almost inevitable now or later and we haven't buried any more checkers. We have also cleared a point and given ourself the chance to make an outfield point, pretty useful if we can. The risk is that we will be hit and lose our racing lead, but we do get return shots after most hits and I believe that White will, with some rolls, choose not to hit. He can't, with the cube on 2 and a 2-0 lead take too many gammon risks, because Blue stands to gain more from gammons than he does. Further, the usual gain from hitting, where he can ship a nasty cube after a fan doesn't apply here. He just isn't going to be redoubling this any time soon.
If you still don't get it, try setting up the position on a real board and play the risky play and the safe play say 50 times each and play the game to the end. Record your results. The results may or may not indicate which is the better play, it's too small a sample to be conclusive, but what will happen is that you will gain considerable understanding of "what happens next".
Incidentally I am quite confident that the "risky" play is the way to go, but it may well be that the "safe" play is not a very big mistake. Even if the safe play was a very good alternate, I would still choose the risky play. Why? Because the safe play means that rkb's next turn is trivially easy, whereas it will very often be difficult after the risky alternative.

[close]

[vikingblood80]

Thanks for the quick reply! Just so we're on the same page, would I be right in summarising your argument as:

a)   The home board is useful if we are blitzing/attacking him. At present our home board is much more powerful. Losing one point would dramatically negate this strength. This attacking works best if he leaves his anchor.

b)   Flexibility is useful and getting hit will help our flexibility (as RKB’s board doesn't pose any problems)


However, why do we have to have it this roll?

â€" although, its not as if we have abandoned the 3 point for ever.

I guess this might be because I don't rate our attacking chances.

As I see it our attack will only really work when RKB breaks anchor, or leaves a large number of blots. I don't see RKB being kind enough to do the latter (he has a nice flexible position) and I imagine RKB would only break anchor with a hit or a double.

A hit provides tempo, thus permitting only single hit returns.


A strong board will obviously have secondary affects on RKB (i.e make RKB play more cautiously than he'd otherwise) but because we still have so much to do (i.e. bring all three points round), I don't think dancing for a few throws would worry him that much. Gammon chances will remain low as long as he maintains his anchor or we lack ammo up front


1) I characterise this as a mutual holding game
2) I believe that race advantages are more important than flexibility in most holding games, including this holding game
3)  I do not believe we have much potential, in the medium term, to attack
4)  I don't believe that the of threat of attack is significant for RKB

Hi again,

a) correct to give up the 3 point would weaken it dramatically

playing 8/6 3/2 is just too passive..it doesn't help us anyway..we only could wait rolling doubles next turn...we are in front in the race so we want to clear all our points safely and break as much contact as possible...the 15 is really useless for us.we want to clear it first. Playing the risky move with the spare on the mid we got a checker to play with or hitting back and still keeping the mid point...rbk has at least 2 checkers on the mid to play with 1 or 2 turns waiting for us not rolling doubles and leave a shot after the safe play. If we are forced already next turn after the safe play the blot on the 3 can turn out to be a disaster for us when some exchange hits will take place. It will be hard to cover and the 3rd checker on the 2 is just simply buried.The other 4 are not quite cause they are forming 2 points which help us.

All right the board helps us on attacks but it also helps us if we are able to hit one in the outfield. The chance is about 1/5 that he will stay on the bar 2 rounds.Even if he stays out just 1 turn that will help us may be clearing the next point or doing anything other useful.If he enters then at least he cant use his full roll only 4-4 and 5-5 will help him.

Well at least if he leaves 2 blots that could be enough for us to win..even gammon.

Don't underestimate the number of our returns cause he has so many points open in his homeboard.

As I mentioned rbk has two spares on the 13 and for some cases one on the 6 which he only wants to move past our anchor when he has to or to create a homeboard point. In general I would not call his position flexible.

1) true
2) depends on the amount of contact
3) yes not much cause most of our checkers are more behind, but at least some
4) at least its so significant that rbk cant afford to play an exchange of hits at the moment

we will see how the story goes 

well at least something tells me that playing the deuce with 15/13 must be right (hopefully 

[vikingblood80]

See the RO posted by Zorba?

So its definitely best playing the 2 from 15 to 13. 

[lewscannon]

I will eschew my previous vote to play it safe.

[Whaat]

Cheers for all the responses! All very helpful, and I will certainly try to manually roll this one out this evening.

I think I've almost certainly underestimated the amount of contact left in this game. But just a few final queries:

Paul, you say, that:

"8/7, 3/2... is safe, but only for this turn. Next turn our position will be worse than it is now because (a) our board will be weaker (and White's may be stronger) and (b) we have more burial, so fewer checkers that can be played."

I'm pretty sure I get the significance of a) now. However, I still unsure of the significance of b) in isolation from a). [I hope statement will make sense in a bit…]

Since there’s going to be plenty of contact now, our home board is effectively dead. So burying further doesn't really reduce the number of checkers in play (it is however a bad in relation to a).

Rightly or wrongly, I don't value any pair of our outfield points more than any other pair. If so, I would have thought that breaking any point next turn would have provided us with similar levels of flexibility as breaking a point this turn. [And as a gain â€" if you counting our rolls for the next throw, almost all of them would leave only 1 blot (as opposed to 2). For example 6s would either allow us to clear one of our points (including our anchor) or let midpoint jump home.]

If we don't have key outfield points, it appears to me (now) that the then costs just reduce to the relative damage to our home board â€" and that these costs over-power the gains from saving a blot. So if are outfield points are all of equal value â€" then I completely see where I went wrong!

However, if our outfield points are of differing value â€" and the 20 and mid points are more valuable then the 15 â€" then our position next turn (after playing safe) is much less flexible then I thought. If so, then we’ve got to clear the 15 point now while we can.

I hope this rant makes it a bit more understandably why I’m not letting this position go so quickly. Basically, if some particularly combination of outfield points are significantly more valuable then others, then I’ve completely mis-understood what or why I’m craving flexibility. And if that’s the case I going to need to roll this one our more than 50 times…

[vikingblood80]

Rightly or wrongly, I don't value any pair of our outfield points more than any other pair.

However, if our outfield points are of differing value â€" and the 20 and mid points are more valuable then the 15

At first one comment: Yes its kind of a mutual holding game, but rbk holds us more than we hold him.

2 simple questions, 2 simple final quick answers from me:

Why do have outfield points a high value for us? Because we need some (1 or 2 are may be enough) landing spots to bring our men home. A checker can't go directly from the mid to our home. (Additional : More points would harm our play and may be we would get in (big?) trouble clearing these points. I think the 8 or 9 point would work best for that purpose.)

Why do our 15 point have the least value? A point controls the range within 6 pips apart.Any other points within that range are only useful for blocking. We don't wanna block rbk instead we wanna get home.So this point is our pain in the neck and we want to get rid of it first. Its just as simple as that.

Cheers Viking!

[Whaat]

Cheers Viking!

[Zorba]

Good explanations here, I would just like to add the importance of timing here. In a one-sided holding game, with f.i. checkers on the midpoint and in the outfield to bring home against a high anchor, being ahead in the race is usually much to your advantage. There are very good chances to bring it home safely and very little chance to get beaten in the race somewhere, even if opponent rolls a high double.

In a mutual holding game, ahead in the race is a mixed blessing. It is usually still an advantage, but much less so. The problem is that being ahead in the race usually means you have worse timing than your opponent. That means that at some point, you will be forced to break useful points because you're running out of plays, while your opponent still can move some spares around, etc. This is where your opponent gets his chances to hit you, and turn the game around. Of course, a lot depends on how strong your opponent's offense is by that time.

In this particular position, we are far ahead in the race: 33 pips effectively, with 125 to go. Since we also made three deep points and an extra outfield point, our timing here is pretty bad. If we don't do anything about it, we are likely to run into trouble later, when rkb is better prepared to profit from it, and we are in worse shape because of messing up our board, or lacking flexibility, etc. "Waiting for our eventual doubles" to solve the problem is unlikely to work here, as we simple don't have enough rolls left and may in fact need two sets of doubles to get out of trouble completely!

The best way to improve your timing is usually to move your backmen up. Lacking that, breaking up a point to create to spares is a good alternative. Getting hit obviously also improves your timing, as long as the backmen aren't severely blocked. Here, we are so far ahead in the race that getting hit is not such a problem, race-wise. That is an additional reason to break the 15pt: we can afford to lose some in the race.

Another important aspect of this position is that several numbers are blocked for us (mostly threes and ones), making certain rolls play even more awkwardly. This increases the risk of having bad timing. We could be forced to break a point later with a very awkward roll, and "lucky doubles" might be blocked and turn into a nightmare (f.i. 3-3).

Last but not least, leaving two blots here is not nearly as risky as it might seem, because many hits are uncomfortable for rkb, breaking his anchor or some other point, and thus giving us plenty of return-shots. With our much stronger board and the matchscore penalizing gammon losses for rkb more than usual, this is a very serious consideration. rkb's actual 4-1 was a case in point: it turns out to be a blunder to hit with that roll.

So the conclusion is that we have bad timing here, but a huge racing advantage, a strong anchor and a much stronger board. We improve our timing with 15/13 8/7 with not that much risk. The safe play worsens our timing and messes up our board, which won't be easy to remake and buries a checker in a position where we are already inflexible.

[Whaat]

Hi Zorba. Thanks for the input, and many of the new points you raise.

However, I feel (and I’m going out on a limb here!) that the stress on flexibility and timing isn’t helpful to understanding the choice of plays â€" as opposed to the position. 

I think this is what I was trying to grapple with yesterday…

(- Btw, I understand timing â€" from your post and books â€" as either synonymous with flexibility or as means of measuring flexibility, and not really an issue in its own right. But please correct me if I’m wrong - )

“If we don't do anything about it, we are likely to run into trouble later, when rkb is better prepared to profit from it, and we are in worse shape because of messing up our board, or lacking flexibility, etc.”

“The best way to improve your timing is usually to move your backmen up. Lacking that, breaking up a point to create to spares is a good alternative.”

And for what it’s worth, I completely agree. But we could have done something about our timing and flexibility next turn. And in the absence of doubles we basically had to!

As such our gains in timing and flexibility should be seen as gains relative to our next throw (not absolute gains). With the exception of those double 3s you spotted, I don’t see there being much gain in having flexibility next turn (Rkb’s new position hasn’t made us need any more flexibility) as opposed to the turn after next.

So our decision process should focus more heavily on the other costs and benefits. And given the points raised by everyone already, the balance is pretty clear.

The benefits of the safe play are no blots (for this turn), and only 1 blot likely next turn. The costs are that we break our board. In contrast our gains after the bold play are that we get to break our least useful outfield point and we get to keep our home board intact. The costs are two blots this turn.

The gains from the safe play aren’t that impressive â€" as RKB quite often won’t hit. The costs are significant because of the amount of contact left and the affect the play has on rkb’s moments.

For the bold play, I suspect that breaking the 15 point (as opposed to a different point) wasn’t the main driver for the difference in equity. However, the benefits of keeping our home board are clear enough. In turn the costs are also slight. 

Thus â€" to me - the flexibility and timing consideration seem to wash out over the course of the two throws. However, the other costs and benefits are clearly stacked in favour of making the bold play now.

So, yeah, any chance I'm getting warm yet?!