Stat Analysis: Progress

[Ry]

I was really astonished when I realized about six months ago that nobody had used statistical analysis (and by this I mean some pretty basic stuff like standard deviations and averages) on D&D monster stats.

With the SRD in access database form, I've been able to use a few simple pivottables and produce a very solid range of figures that describe most monsters at a given CR (special abilities, feats, class features, and spells notwithstanding).

But I'm hitting a wall - while I have decent data for monsters (sparse samples at high levels) I don't have any organized data for NPC stats. I'm not interested in doing hours upon hours of manual data entry - can anyone point me to a database of NPC statblocks?

 

[ashockney]

Rycanada,

I am really, really interested in your access database project and would also like to get access to the info and possibly help out, where possible. I started using an access database for work about a year ago, and the first thing I thought of once I started doing that was putting all the D&D monster, spells, npc's into something like this for review, assessment, improvement, etc.

 

[Ry]

I got my data from here:

http://www.andargor.com/

I actually don't have access, just excel (I have excel-fu from endless hours with it at work) but you can import tables as spreadsheets, which I did.

That database is awesome, but of course, like the SRD, there's fewer and fewer samples as you get to higher and higher CRs, or sizes outside small - large. Most importantly, though, is that it doesn't have the NPCs I would need to expand my analysis beyond the beasties.

 

[Varianor Abroad]

Now that is damn cool! My question is, how does the "standard" monster at CR X compare to certain critters like ghosts, beholders and dragons? More or less? I'm really curious about your project. As to the NPCs, could you use the DMG 3.0 tables? I wonder if someone has those somewhere? (They are probably not something that can be posted, but I bet someone has converted the table.) How do you intend to allow for the difference between core and core+splat, which anecdotally skews things?

 

[Ry]

I'd rather use 3.5 DMG tables if I can. I just don't want do do all that typing.

The stuff I'm doing right now is just standard deviations and averages, so there are "outliers" that aren't covered, but I've found those to be fairly predictable as well - and I think they cluster along the same monsters (like lower AC/higher hitpoints; lower hitpoints /very deadly attacks).

Just as an example, here's my hit points chart. For example, a manticore has 57 hp by the book, and is a CR 5 monster. The manticore has cool attacks, but we don't expect it to be particularly high or low in hp for its CR. My average says 56. Mummies are 55, again, no surprise for a CR 5 monster.

The higher the CR is, the fewer data points I have, so I'm expecting more outliers there.

 

[Ry]

Also, keep in mind that dragons are highly overrepresented in the SRD database I'm working from.

 

[ashockney]

Awesome! Thank you!!

 

[Ry]

Sticking with CR 5 hit points for a minute, our outliers are:

Orca Whale, Black, Green, and Red Young dragons, Huge Animated Objects (High)
Rasts, Ravids, Wraiths, and Unbodied (Low)

 

[jaerdaph]

Your name is Ryan - you should be out celebrating today lad, not crunching numbers!

 

[Ry]

Drink an extra green beer for me - I've got the flu.