i begin programming after very long time:
- Now raper alices rape together
- Other yukkuries attack them more often.
i might be not very stable you can test this version from there;
I tried your version five times, with the same yukkuri: three Marisa, two Reimu and three Alice. Two of the Alice are always rapers, and one is never a raper. Only plain food was offered, and I shook an Alice to start.
Twice, the Alice rapists went after each other and ignored the rest. Once, I shook an Alice as it was starting poo-poo, and as it was on it's back with it's anyaru in the air, it went sliding around into a Marisa and kept her pinned into a corner as the other Alice raped at the other end. The last two times, only one Alice would go rapey.
Raper Alice: 2X5 trials=10 participants; 8 participants/2 Alice-on-Alice. So, overall, it was a success in that 60% of the time one Alice was stimulated, the other Raper Alice would join.
As far as defending, the control Alice was attacked once and fought twice. Twice, the non-rapists did nothing but cry and run. Twice, at least one Marisa fought back, in addition to the one time an alice helped fight the rapers. Only one Reimu fought back out of the five trials.
Control Alice: 1X5 trials=5 participants; 2 defenses.
Marisa: 3X5 trials=15 participants; 4 defenses.
Reimu: 2X5 trials=10 participants; 1 defense.
So, it wasn't very disciplined a test, but I'd say the rapers did as you intended. I really didn't see much difference in the defenders over Eternity's version.
Going back to a post I made about rape/defense behaviors, I never have seen a strong correlation between attitude and defense. Shitheads are IME as likely to fight a raper as a good yukkuri, maybe slightly more so.
There is something I'd like to see added, though. I've seen some Marisas (not sure of others) will kill baby predators. But you don't see any killing of baby rapists.