True AC of Monsters

[Karui Kage]

Ignore this thread if you aren't a fan of statistics or raw data. :)

Not sure if this is the best sub-forum for it, but it might be useful for those who like to calculate the feasibility of hitting a monster of X CR. I always wondered if the values in the Bestiary were accurate in terms of Average AC, so I did my own calculations (I got bored and hey, I have the data, so why not?)

The linked sheet below includes over 1,800 monsters taken from all 1st-party Paizo sources using my Archives of Nethys database. The only ones it is missing are the last half of Bestiary 4.

Summary: Yeah, those average ACs are still pretty dead on, even given so many data points. I expected them to actually be pretty divergent, but overall, Paizo has managed to stay true to the given averages even with nearly 2,000 monsters. Impressive. Less so at early CRs, but 3+ are all pretty exact.

Fun Note: Touch ACs, on average, never go above AC 14 until CR 25, at which point they jump. Keep those touch attacks handy!

Average Monster AC Sheet

[wraithstrike]

I did this a while back by just looking at random monsters. I figured if I could blindly grab 7 monsters per CR and most of them be accurate, ignoring certain ones that I knew were outliers, that the chart would be accurate.

However the info is nice. :)

[Akerlof]

Nice info. How do the medians look, pretty close to the averages? That's a decent way to get a feel for whether the average is the "normal" or not.

[Karui Kage]

I only did a few (CR 1-5) but that looks to be the case so far.

[Cheapy]

Have you considered doing saves?

[Karui Kage]

I added saves on (new tab), though they can't be compared as easily to the suggested numbers since I believe those are meant to be base numbers (maybe?) and I really only have the grand totals. I'd have to hack some way to extrapolate the base numbers with feats, ability scores, etc. being taken out, which would take too much time.

Still, the numbers are up there now if you're curious. :)

[Cheapy]

The save totals on 1-1 are meant to be total...I'm pretty sure of that.

Thanks though! You're awesome as ever.

[Cheapy]

I suppose the tricky thing is that you would need to differentiate between all the permutations of Good / Bad saves, and calculate the averages there, to get a truly useful data set.

Totally understandable if you don't want to do that though.

[Karui Kage]

If you want to play with it, shoot me an email (asknethys at karuikage dot net) with your google account and I can add you to the users. You'd be welcome to play with new formulas to analyze the data in a separate tab.

[drbuzzard]

Very useful. Nice work. It's good to know that the provided average numbers really are accurate if you're making DPR calculations.