Perrin, I remember talk of that and a discussion of whether a >99% and <1% setup should be instituted. Wasn't certain if that situation had been changed with the updates though so chose not to mention it.

Sorry if this is getting ever more off-topic, but I just wanted to reply to the Analyse Tool discussion:

It's indeed reporting the results of a simulation, not a calculation (I can't vouch if it is really 1000 attempts, but that sounds reasonable). This can be quickly verified: just click the "Analyse Again" button without changing any settings; you'll see some slight changes to the reported outcomes.

@Perrin:
I'm not entirely sure if this is what you're talking about, but a while ago I posted a bug report about the Analyse Tool reporting values which were *way* of the mark. Fizzer explained that was the result of a (broken) optimization, which didn't even run the simulation if it **expected** the result to surely be 0% or 100%. The problem (which has since been corrected I guess, but (come to think of it) haven't tested) was that for some weird luck and/or kill rate settings would incorrectly expect chances to be surely 0% or 100% (and then report them as such). To the best of my knowledge there has never been a "range" of 2% which is *deliberately* rounded to either 0% or 100%; the idea was to only not-run the simulation for results which would be 0% or 100% anyway.

Ah, found the original thread: [Discrepancy between attacks in actual game / analyzer / my own calculations](http://warlight.net/Forum/Thread.aspx?ThreadID=2335)

At the moment I'm on a computer which doesn't have Flash installed, so I can't test, but then again, I can't remember Fizzer mentioning changes to the Analyser in any of his change logs... Anyway, if you want to see for yourself if this has been fixed, go play a game with 10% offensive luck (and 75% luck modifier); if a 12 vs 1 attack is listed as a guaranteed victory, the bug is still there. Also, the 2% mentioned is actually (at least) 20%.