[12:41:50 AM] Crazy Colorz: So uh
[12:41:52 AM] Crazy Colorz: This is weird
[12:42:02 AM] Crazy Colorz: If you proc 0 stats
[12:42:06 AM] Crazy Colorz: On a level up
[12:42:10 AM] Crazy Colorz: Then it reprocs HP
[12:42:20 AM] Crazy Colorz: And if you STILL don’t get that
[12:42:23 AM] Crazy Colorz: It reprocs all stats.
[12:42:48 AM] Arch: so a lower HP growth makes shit characters slightly have a chance at more levelups?
[12:42:59 AM] Crazy Colorz: … maybe?
[12:43:21 AM] Crazy Colorz: It only does this at most 2 times though
So basically all of our calculations of averages have been wrong for the past 11 years?
Yes. Did the data always look slightly higher?
I’m pretty sure I’ve heard this mentioned on SF before. Or at least, that there were 3 attempts. The bit about alternating HP and full-stat attempts is new, though.
Either way, it makes a pretty small difference. For most units the “naive” chance of a blank level is going to be on the order of 1%, based on some quick checks.
Also, as I pointed out a long time ago, the simple “max(cap, base + rate * levels)” calculation is already subtly wrong. As a simple example, say you have a growth rate of 50% and are 1 away from your cap at level 18. The charts pretend you never capram and give you an average stat = your cap at lv 20, but in fact sometimes you fail to hit your cap and you never exceed it so that can’t be the average.
I’ve mentioned this several times before, geez!!
My DS-growths thing even cut this out to make space!
I’ve heard about this in FE6 but never heard info if this was in 7/8.
B-but you could have just re-wound the loops
How can this possibly be true? Heck, I just saw an LP where a guy leveled up 20 times in a row and got 7 levels of 0 stats. This is on a character with substantially good growths and 0 maxed stats too. I don’t think it actually re-rolls.
I’m a rebel and I do things my way
Anecdotal. Also, might have been a 0% growths run?
“it’s very unlikely for something to happen”
“look at this anecdotal evidence that has it happen a lot of times”
“it’s very unlikely to catch a pokemon legendary with a turn 1 pokeball”
“look at all these videos/reddit threads of people doing it”
“unlikely” doesn’t equate to “impossible”
Yeah, true enough. Also, his growth rates on the non-capped stats aren’t as high as I thought, so I guess it makes more sense.
Addendum: It REPROCS UNTIL IT GETS A SINGLE STAT. It does this until it rolls at most 3 times for each stat(including original proc). Also of note, on the reprocs, Afa’s drops bonus is not added.
I remember the FE6 discussion mentioning it doesn’t reproc the levels if a stat is capped, maybe that’s what happened?
EDIT: I was wrong about that, but I had the right idea.
Just to see how negligible it is, true hp growth is =
(base+afa) + (1-base)(1-str)(1-skl)(1-spd)*. . . (1-luk)base + (1-base)^2(1-str)^2. . … *base
Assuming all the stat grows are 70%! Quote generous, this turns out to be
.7 + .00015309000 + .00000003348 which is basically .70015, a .02% increase
If the stats are a bit lower, say 30%,
.3 + .0247 + .0020 = .3267 = 8.9% increase, more sizable.