Attacking - Gain in War and Offense to Send

[pathetic sheep]

random-number-between-0-and-0.5

This part is uniform?

[Bishop]

this part is uniform?

note: The random factor is not 100%

[RiffArt]

Love you Bishop.

[RiffArt]

Testing equal off to def sends would be of more interest to me.

This. If anyone's interested in collecting the data (maybe EoA if you're in the ghetto?) I'll be happy to do the maths and work out what the underlying distribution is for the random factors. Now we know there are 4 of them and how they add it should be possible to work out with just the 1 extra data point (success at equal off/def). I am assuming the same distribution for each factor - if each had a different distribution it would be messy lol

[RiffArt]

Not to necro (well the thread wasn't really dead, just a little cold), but if anyone is interested I finally got round to working this all out (assuming the four random numbers are distributed uniformly which makes sense - else why have the two extra ones?)

It's late now but I'll post some results (and triple check my working) tomorrow probably. (It does pass the test of giving 0 at 90.44%, 50% at 97% and 100% at 104.04% so I'm fairly confident I've nailed it - else I wouldn't be posting!)

Some quick results:
sending 103% gives you 99.8% probability of success
sending 102% gives you 98.5% probability of success
sending 100% gives you 87.9% probability of success

Can't get my img to work, but this should link you to an image of the graph of RatioSent vs. Probability of Success: http://pasteboard.co/22MHaOrs.jpg
(line isn't smooth as I've just joined the points linearly)

Hope it's of interest to someone.