I was playing vs the Android Backgammon Deluxe Free app, with a 2 cube that I had accepted. The 5/3 came up, which looked like it could be my best roll (other than maybe 6/6). My first instinct was to take one of my back checkers and hit the white guy on my 12, but then I saw that I could build a 5 prime. Luckily, the 'undo' option let me game both scenarios out, with identical rolls, and even though the hit option had me first in rough shape, when he rolled out with a 6/3 and hit my blot, I eventually won by 6 or 7 pips. When I backed it up and played out the second option, I lost by 3 pips. I think hitting is better because the CPU's home board isn't very scary, and trapping 2 checkers is always preferable to trapping one. I was also way behind in the race (144 to 91), which always leads me to hit more.

Hi
i just put your position into Xg. Hitting is best because you win more gammons and protect your checker on the 8 point indirectly.(hope my explanation is ok?)
Ok, but if you remove whites straggler from the 23 point to the 22 point (1 point forward) hitting this checker ist best. You make a 5 prime and white cant anker on this point if you hit his checker on his 13 point.
Now lets remove this checker (straggler) from the 22 to the 24 point and it seems to be very close. Xg++ has hitting on the 13 point only 0,0025 ahead of making the 3 point.
Nice position imo. I dont understand your winning with 3 pips or 6-7 pips?! It doesnt make a difference imo.
I can approve Backgammon NJ for android btw

My win and/or loss were the results of the two games that played out as a result of my making each of those two moves. I played the hit option, then continued until I had one checker on the board, took note of the pip count, and backed up the game (by repeatedly hitting UNDO) to this position, then played the 5-prime option, out to an alternate conclusion, which resulted in my losing by 3 pips. Obviously the concurrent rolls and how I played them had a lot to do with the difference between winning and losing*, but this was the toughest decision I had after accepting the cube, and it appears to be at least one of the spots where

*It's worth noting, though, that even though I made a different move after backing up the game, the software dealt out the same string of rolls. This is a pretty cool function that, while encouraging the user to cheat, can be a pretty sweet academic tool.

sorry, typo in my posting above.
"Xg++ has hitting on the 13 point only 0,0025 ahead of making the 3 point."
Difference between 20/12* and 8/3 5/3 is 0,025 in Xg++

According to linuxgnubg not hitting would be a mega mega blunder. I suppose this evaluation is faulty. But let's keep it simple. Black has the much better board. Bold play means to me agressive play, and that means hitting.

Look at it this way, if you make the five prime a lot of the time they can clean up their checker and just have the one to escape. You don't have any ammo, so have no real chance of closing the board until you bring some more men around, which often gives them ample time to roll a six, escape the final man and be in good shape.

Compare this not so great scenario with what happens if you hit. Now your opponent has two men stuck behind a four prime which will often soon be a five prime. If they enter on the three they can get attacked again, if they enter on the one they have a lot more trouble escaping that man. Your men will generally also soon control the outfield, leaving white in terrible shape as they struggle to get back round.

Even worst case, where you get hit back, you're still in OK shape due to the weakness of white's board and your anchor. Small risks with large rewards warrant the hit.

If you make the 5 prime Grey will be at the edge of your prime with any 6 escaping and ahead in the race, this is not good. But if you hit you have 2 men behind a 4 prime iwth a 5th point slotted and another man within cover range.