I had to play 2x50 game sessions because SW doesn't allow
more than 50 games in a session. Just in case anyone would
want to reproduce the dice rolls, for the two sessions I
used 321 and 323 as seed (current month/day when I played).

Remember my experimenting with doubling immediately in
response to XG's opening 63 and a few other such moves?

Well, I wanted to expand on those and I doubled after
opening 62, 63, 64 and also after the bot's hitting me
and (single or double) slotting in it home board during
the first few rolls.

The results even surprised me: Effective +96 vs. expected -118!

It's hard to remember to double in the above mentioned cases
and I don't think I skipped any. In response, the bot beavers
half of those doubles and you are stuck with it for the rest
of the game. It took a real resolve to go on with it for 100
games.

My point was/is to ask/answer whether early cubefull rollouts
for cube and/or checker decisions are useful/meaningful at all?

For years I had a feeling that they weren't and kept arguing
that the so-called "cube skill" was mostly an elaborate fantasy
that only/gradually/minimally applied towards the end of the game.

I think I had explained in the past what the spreadsheet columns
meant and I will try to explain more when I find time.

The numbers at the botton are about who cubed during the first
4, 6, 10 rolls and later; who cubed first (SW 44, MK 55 times);
who won the starting roll (SW 54, MK 46 times); and as an amazing
coincidence both players won 50 of the 100 games. :)