From Matt's study Note(5)

Position ID: 4HnwIBhw58EBIQ Match ID: QYkSAFAAEAAA 1. Cubeful 4-ply 24/20 13/8 Eq.: +0.3947 0.6773 0.1973 0.0060 - 0.3227 0.0526 0.0009 4-ply cubeful prune 2. Cubeful 4-ply 24/20 8/3 Eq.: +0.3536 ( -0.0411) 0.6593 0.2007 0.0068 - 0.3407 0.0556 0.0010 4-ply cubeful prune 3. Cubeful 4-ply 24/15 Eq.: +0.3046 ( -0.0901) 0.6521 0.1878 0.0077 - 0.3479 0.0747 0.0017 4-ply cubeful prune

Matt's Note: Although 13/8 is a bit stacked, it can usually be unstacked next roll (by making a point with a natural number or slotting with an awkward one). 8/3 slots a non-key point and gives us a chance to improve our back checkers with tempo.

From Matt's study Note(4)

Position ID: zG5HAETmNkMBEg Match ID: cAkAAFAAAAAA

Cube analysis
4-ply cubeless equity  +0.5986
  0.6567 0.3698 0.0201 - 0.3433 0.0982 0.0064
Cubeful equities:
1. Double, take         +0.9681
2. Double, pass         +1.0000  ( +0.0319)
3. No double            +0.6659  ( -0.3022)
Proper cube action: Double, take

From Matt's Note: John felt this was a pass, but we made a good take here. We get gammoned a lot but with two checkers semi-primed we have good counterplay if we enter.

From Matt's study Note(3)

Position ID: nwEgcEfD2zYEAA Match ID: EQELAAAAAAAA 1. Cubeful 2-ply 13/7 9/7 Eq.: +1.3494 0.9139 0.5443 0.0117 - 0.0861 0.0053 0.0000 2-ply cubeful prune [world class] 2. Cubeful 2-ply 8/6 8/2 Eq.: +1.3444 ( -0.0050) 0.9066 0.5574 0.0146 - 0.0934 0.0092 0.0000 2-ply cubeful prune [world class] 3. Cubeful 2-ply 13/7 8/6 Eq.: +1.3432 ( -0.0062) 0.9123 0.5415 0.0131 - 0.0877 0.0062 0.0000 2-ply cubeful prune [world class] 1. Cubeful 4-ply 8/6 8/2 Eq.: +1.3577 0.9101 0.5604 0.0151 - 0.0899 0.0075 0.0000 4-ply cubeful prune 2. Cubeful 4-ply 13/7 9/7 Eq.: +1.3464 ( -0.0113) 0.9159 0.5378 0.0099 - 0.0841 0.0046 0.0000 4-ply cubeful prune 3. Cubeful 4-ply 13/7 8/6 Eq.: +1.3302 ( -0.0275) 0.9107 0.5333 0.0116 - 0.0893 0.0058 0.0000 4-ply cubeful prune

XG says 8/6 8/2, and it is .037 better than 13/7 8/6, .047 better than 13/7 9/7
From Matt's Note:
My proposed play of slotting the deuce seems like the right idea here. We need to make the 2pt eventually and getting hit is not costly, and at the same time we keep pressure on the 10pt blot in case we roll a 3 next time. To me, it is hard to see but quite understandable.

From Matt's study Note(1)

Position ID: mCd4BCmwu8GAUA
Match ID: EQEOAAAAAAAA
1. Cubeful 4-ply bar/22 24/20 Eq.: +0.4450
0.6846 0.2343 0.0134 - 0.3154 0.0634 0.0012
4-ply cubeful prune
2. Cubeful 4-ply bar/22 13/9* Eq.: +0.4438 ( -0.0013)
0.6594 0.3046 0.0371 - 0.3406 0.0941 0.0046
4-ply cubeful prune
3. Cubeful 4-ply bar/22 6/2* Eq.: +0.2799 ( -0.1651)
0.6189 0.2615 0.0283 - 0.3811 0.1119 0.0052
4-ply cubeful prune

XG says 0.050 better to play 24/20. gnu 2ply does not agree, but 4ply agree.

From Matt's note:

I don’t think Mochy, Michy, or I really looked at the best play of just anchoring. That doesn’t overextend our position and creates the threat of making the 4pt next roll. We could make the 9pt after our play but it leaves 5 blots instead of 1 and doesn’t make the 20pt, and getting hit is not good, particularly if we throw a bad number from the roof.

I agree. We have a prime. If we anchor at 20pt, Opponent can't prime us, that means one-sided game which we can prime him. Hitting play let opponent to hit back and fight back decently, since board strengths are almost same.

From Matt's study Note(2)

Position ID: sNsWMAKPAQxwMw
Match ID: UYkJAAAAAAAA
1. Cubeful 4-ply 24/22 24/21 Eq.: -1.2781
0.0980 0.0079 0.0000 - 0.9020 0.5052 0.0102
4-ply cubeful prune
2. Cubeful 4-ply 13/11 13/10 Eq.: -1.3196 ( -0.0415)
0.1051 0.0044 0.0000 - 0.8949 0.5363 0.0336
4-ply cubeful prune

What XG says: 13/11 13/10 is 0.053 better than 24/22 24/21.

Matt's commented:

An interesting play. Coming off the 24 point wins less often but also gets gammoned and backgammoned less. I am not sure whether this will hold up after a rollout.

I think XG is wrong here.