Here is the story with this position.
It was a friendly game, I was chatting with my opp during it, and this cube decision was discussed in great detail as I was making my choice. (I talked, the opp listened).
My thought process:
Yes, the Takepoint is what? 17% and I did feel that I had more than a 17% chance to win this game, but I also felt that Blue did have a very gammony board and if I were to get hit, which seemed very likely, I would have been gammoned.
Match score was a big issue for me as well. My take point is very low here at 5a/2a, but there are other factors that come into play, lots of pros and cons. If I were to go by race numbers alone and the TP, yes it is a take. I felt it was a very good double as well. But I had this ugly rigid board and that huge chasm to cross. Though, I did have Blue's 5 point, which was a plus for me if I did decide to take.
How does one weigh all these subtleties? I had to make a choice, and my choice was to drop that cube, I couldn't count on getting good doubles and I had way too many bad rolls. I wanted a clean, fresh, blank slate board. (if you are wondering how my board ended up like that? I rolled a few double one's as well as other ones where I just had to keep stacking).
I rolled it out to the 10k games and thankfully I chose well. If you just analyze the position using XG+, it is a take.
Michelle
                         |
| ||||
| XGID=-AbG--CBaa----a---bbBbbb--:0:0:-1:00:7:4:0:9:10 | |||||
| Blue on roll, cube action? | |||||
| Analyzed in Rollout | No double | Double/Take |
| Player Winning Chances: | 67.78% (G:19.93% B:0.90%) | 68.25% (G:20.80% B:0.96%) |
| Opponent Winning Chances: | 32.22% (G:7.32% B:0.34%) | 31.75% (G:8.50% B:0.46%) |
| Cubeless Equities | +0.392 | +1.031 |
| Cubeful Equities | ||
| No double: | +0.870 (-0.130) | ±0.004 (+0.866..+0.873) |
| Double/Take: | +1.031 (+0.031) | ±0.006 (+1.025..+1.038) |
| Double/Pass: | +1.000 | |
| Best Cube action: Double / Pass | ||
| Rollout details | ||
| 10368 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller | ||
| Double Decision confidence: | 100.0% | |
| Take Decision confidence: | 100.0% | |
| Duration: 1 hour 24 minutes | ||
eXtreme Gammon Version: 2.10, MET: Kazaross XG2