Bug in Saber Cherry Simulator
Seems a little odd
from 
GalCiv2 Forums
GalCiv2 Forums
Hopefully Saber Cherry will notice this, sadly the homepage for the simulator doesn't have an email address.
Hi Saber,
Excellent work in making a statistical simulator, I'm also making one and was comparing them to see if they gave the same result. Sadly they don't, which could mean either one or both of us are wrong
Anyway, I noticed some strange behaviour with respect to defense. Running
java Sim 40 10 0 0 0 0 0 40 10 0 0 0 0 0
i.e. an equal battle each with 40 HP and 10 attack.
The result that comes back:
A beats B 0.4997675 of the time.
A dead heat, which is what you'd expect providing iniative was shared equally. The homepage doesn't
say how iniative is dealt with, but from the result I'm guessing it must be shared. The bonus to the attackers would be quite substantial with 10 attack and only 40HP (my simulator estimates it would be about 8%).
So good so far, but what happens if you add a single point of defense (even the wrong kind of defense)?
java Sim 40 10 0 0 0 0 1 40 10 0 0 0 0 1
And it returns
A beats B 0.6127455 of the time.
Which is quite a big shift, since both sides are still equal and if iniative is balanced you'd expect 0.5. Multiple runs give the same answer (which is what you'd expect for a statistical simulator).
Anyway, hope that helps I'll try to get mine finished and put it up soon.
Hi Saber,
Excellent work in making a statistical simulator, I'm also making one and was comparing them to see if they gave the same result. Sadly they don't, which could mean either one or both of us are wrong
Anyway, I noticed some strange behaviour with respect to defense. Running
java Sim 40 10 0 0 0 0 0 40 10 0 0 0 0 0
i.e. an equal battle each with 40 HP and 10 attack.
The result that comes back:
A beats B 0.4997675 of the time.
A dead heat, which is what you'd expect providing iniative was shared equally. The homepage doesn't
say how iniative is dealt with, but from the result I'm guessing it must be shared. The bonus to the attackers would be quite substantial with 10 attack and only 40HP (my simulator estimates it would be about 8%).
So good so far, but what happens if you add a single point of defense (even the wrong kind of defense)?
java Sim 40 10 0 0 0 0 1 40 10 0 0 0 0 1
And it returns
A beats B 0.6127455 of the time.
Which is quite a big shift, since both sides are still equal and if iniative is balanced you'd expect 0.5. Multiple runs give the same answer (which is what you'd expect for a statistical simulator).
Anyway, hope that helps
I'll try to get mine finished and put it up soon.
