Yes, of course -4 -4 with centered cube is a similar situation. Thanks for the hint with the Robertie's book. I think I've briefly read it some years ago but don't recall this section. I might want to take another look at it.
Digging deeper into the score consideration, I've made some simple calculations, and it seems that trailer's recube window when holding the 2-cube at -4 -4 or the 4-cube at -8 -8 should open when his single wins are higher than single wins of the opponent:
my.s - my single wins
opp.s - opp's single wins
2-cube at -4 -4:
MWC@2cube = my.s * .667 + opp.s * 0.333
MWC@4cube = my.s
potential redouble when:
MWC@4cube > MWC@2cube
my.s > my.s * .667 + opp.s * .333
.333 my.s > .333 opp.s
my.s > opp.s
The same goes for 4-cube at -8 -8, only the .667 and .333 values are switched to .75 and .25.
Actually, we can stretch it to all even scores where recubing kills gammons for both sides.
It seems logical, too. If the leader wins a gammon, he wins the match not matter what we do with the cube, so the single wins are what matters.
So for example, if opp wins 60% of the time, with 30% gammons included, we can already consider redoubling since we win 40% single vs 30% single.
An example:
                     |
| ||||
| XGID=----BBBBBB--------B--aAfg-:2:1:1:00:0:0:0:8:10 | |||||
| Blue on roll, cube action? | |||||
| Analyzed in XG Roller++ | No redouble | Redouble/Take |
| Player Winning Chances: | 28,41% (G:0,00% B:0,00%) | 28,46% (G:0,00% B:0,00%) |
| Opponent Winning Chances: | 71,59% (G:52,10% B:0,20%) | 71,54% (G:56,43% B:0,01%) |
| Cubeless Equities | -0,971 | -0,876 |
| Cubeful Equities | ||
| No redouble: | -0,927 (-0,050) | |
| Redouble/Take: | -0,876 | |
| Redouble/Pass: | +1,000 (+1,876) | |
| Best Cube action: Redouble / Take | ||
eXtreme Gammon Version: 2.10, MET: Kazaross XG2