I made the wrong decision in this position and I find it hard to grasp the thought process behind the correct play. I would be interested in hearing others' thoughts..

I don't like 18/12 13/12 it results in 3 stripped points in the outfield.
I don't see the purpose of hitting loose on the 2-pt, but it might be better than I think.
I'd be tempted to play the boring 15/8. (After that I'd be more inclined to hit loose in my board.)

I assume White owns a 2 cube here?

I probably would have made the 12, but now I think uberkuber is right that 15/8 is better. Should be easier to play next turn, though it is somewhat more risk immediately. I think hitting is too much risk and I don't want to give up the 8 point. Though if you only 1 second left on the clock, maybe you want to play for a gammon!

Sorry, I forgot to mention that the cube was in the middle. I missed a cube before the roll, but it was only a small double.

I decided to hit with 8/2* 15/14 on the premise that if White dances or rolls 13, 14, or 15, it will become a double/pass so I win straight away more than half the time, and when I get hit I still have some chances.

However I think I underestimated the size of Black's advantage after the 15/8 play. I guess the point is I will likely have chances to attack later in even more favourable circumstances..

Hitting is going to be especially bad with the cube in the center as it's going to overshoot your market too much. In other words, the double/take equity after a hit/dance sequence is going to be like 1.5, which is worth the same as if it was 1.1 since it will just go double/pass either way. So in a sense you're wasting that extra 0.4.

Or maybe a cleaner way to look at it is say have a choice of plays -- play one always leads to a cubeless win % of 80 next turn regardless of what your opponent rolls, and play two leads to a completely won game on 90% of your opponent's rolls and a completely lost game the other 10% of the rolls. Cube in the middle you want to make play one and cash the game in next turn by doubling. If you had already doubled, you choose play two. So generally speaking, if it looks like you are going to get in a strong cube next turn, you have an incentive to reduce volatility.

Sometimes you need to make a play that gives up a significant chunk of cubeless equity so you can be around the double/pass point next turn. The 5-1 position I posted in this thread is a good example:

https://forumserver.twoplustwo.com/1...40/?highlight=

I'm surprised that leaving that blot on 18 is correct. It leaves 13 shots and 4 that also make the 2 point. Seems in an even race holding game you are usually trying to minimize your opponent's shots and wait for him to give you one. But, I'm also not very good so I'm sure I'm missing something.

You could look at this almost as a pay now vs pay later situation. You could make an immediately safe play 18/12 13/12 that does not leave any direct shots. However looking at the position that gives, you almost certainly will be leaving more shots as you break your outfield points, not to mention the “safe” play creates an extra outfield point that needs to be broken creating even more potential shots during the bear in.

Further, consider what is likely to happen in the meantime if you wait to leave a shot. White likely will have covered his blot on the 2 point. He can now safely break his 8 point to do so. He might even be able to bring his spare checkers from the back into play and close his board. Note that since you have vacated the 18 point he now can safely move checkers into the outfield, assuming you continue thematically and vacate the 15 point as well. Being hit after this would obviously be disastrous.

By contrast being hit on the 18 is far from a disaster in the immediate term. Most of his hits will leave the two point blot for a return shot, or barring that shots on the 7 or 8 point. You are not entering against a closed board, and you maintain flexibility on your 13 and 8 points with an alternative like 15/8.