This is a curious position which relied on a precise distribution of blots to produce an unexpected result. It is amazing it occurred in an actual practice session.
I think the key points were all covered in one post or another. I'll summarize what I think is most important here in one place. There are other factors such as cube ownership which I will ignore. By making the unusual play of breaking the four point, immediate shots are reduced at the cost of leaving three blots. Getting hit with three blots is worse than getting hit with one blot, especially since there is no double out after hit / dance with one blot. So, extra quick wins when breaking the four point are traded for extra later wins when retaining the point. And the computer tells us these are almost exactly offsetting at DMP.
However, it is a large blunder to retain the four point. And there is a lesson from this position which we may actually see again. Quick wins when the initial shot is missed are often gammon wins - over half of them. Wins after being hit are seldom gammon wins. Since gammons count, play for the quick win and collect all of those extra gammons when it works.
                     |
| ||||
| XGID=-a-EBaa--a--a-----Abbbbb--:1:-1:1:53:0:0:1:0:10 | |||||
| Blue to play 53 | |||||
| 1. | Rollout1 | 18/13 4/1* | eq: -0.3244 | |||
| ||||||
| 2. | Rollout1 | 18/10 | eq: -0.5559 (-0.2315) | |||
| ||||||
| 1 1296 Games rolled with Variance Reduction. Rolled for both No double and Double Dice Seed: 10803197 Moves and cube decisions: 3-ply | ||||||
eXtreme Gammon Version: 2.10
  |
| ||||
| XGID=-a-EBaa--a--a-----Abbbbb--:0:0:1:53:0:0:0:1:10 | |||||
| Blue to play 53 | |||||
| 1. | Rollout1 | 18/13 4/1* | eq: -0.3489 | |||
| ||||||
| 2. | Rollout1 | 18/10 | eq: -0.3643 (-0.0154) | |||
| ||||||
| 1 5184 Games rolled with Variance Reduction. Dice Seed: 24007057 Moves: 3-ply, cube decisions: XG Roller | ||||||
eXtreme Gammon Version: 2.10, MET: Rockwell-Kazaross