Welcome to my Pathfinder Analysis Notebook#
A couple years ago, my Google Sheets analyses and charts came to a screeching halt. It wouldn’t load anymore. There comes a point where dozens of charts that depend on thousands of calculated rows of data that themselves depend on thousands of calculated rows that themselves… You get the idea. It was calculations all the way down.
Then I did a bunch of analysis in F# and enjoyed the process of it but burned out on trying to share it with people. Well, now I’m working on actually making it meaningfully shareable! And there’s excitement in the webpage updating so smoothly as I work on it (although getting it deployed was a bit of a headache).
What makes this different from other analyses?#
Most other analyses use averages and the GM Core / CRB guidelines for creature creation. I use every creature available on Archives of Nethys as of June 21, 2025 and make my programmatic comparisons against all creatures of the same level.
Should I have? Probably not. But a lot of times, in these discussions, people tend to get very side-tracked by debating the nuances and merits of comparing one ability against moderate AC/DC and another against high AC/DC. I also wanted to check how well official Pathfinder creatures stick to their own guidance.
Additionally, some of my analyses are permutative instead of averaged. While definitely unnecessary, I was very interested in getting as exact of results as possible rather than relying on probabilistic simulations of results. That means some of my analyses literally compute, for example, a 14d6 Fireball with two d20 saves, for an analysis of all 3 * 10^13 permutations of dice values (with some lossless shortcuts).
Should I have? Probably not. A statistical analysis would’ve more than sufficed with a sufficient accuracy and a lot less work. Still, regardless of methodology, we get to answer questions like “What is the ‘natural 1’ of a rank 10 Dragon Breath damage roll?” “What if I also factored in whether enemies saved or not?” And, “if i had to boil down my entire turn to a single d20 roll, what would it be on the d20?”
Here’s some caveats to keep in mind:
Mathfinder’s DPR critique still applies here#
My analyses do compare against every creature as of a certain date, but they are essentially boiling down to averages. Actually, they’re boiled down to 20 tiers of results (in honor of our sacred d20), but Mathfinder’s critique of averages and DPR still applies here.
You will never cast Fireball or use Vicious Swing against every single creature in the bestiary at once. You will always be in a situation where some subset of actions are optimal (casting fireball against low reflex enemies with weakness to fire), some are totally fine (swinging a greatsword at a creature with some slashing resistance), and some are ill-advised (using a precise finisher against a precision-immune enemy).
That being said, I do plan to do my best to keep this critique in mind. However, the general goal here is to compare things that are generally good to other things that are generally good across classes rather than dig deep into specific optimization niches.
Rarity and analysis#
Originally, I wanted to filter out Unique and Rare creatures in the results, but ultimately most creatures at high levels are already Rare. And while Unique creatures might be, well, unique, they still represent a design space that is part of the Pathfinder experience, especially if you’re going through Adventure Paths.
What about kineticist?#
Kineti-who?
Just kidding. I just haven’t really planned to do much analyzing of them beyond what already seems clear to me about them.
Elemental Blast: Usually on par with a weapon Strike without damage boosting class features
Examples: Earth and shortsword are d8s, air and shortbow are d6s
Overflow: Usually about as strong as a focus spell and reaches a spell slot if it is 3 actions and sustain
Examples: Retch Rust vs. Pulverizing Cascade, Rock Rampart vs. Wall of Stone
Non-overflow: Usually a little stronger than a cantrip
Examples: Aerial Boomerang (cantrip with better AoE, plus single action follow-up next round)
Counter-example: Timber Sentinel
Overflow 18+ Capstones: High rank spell slots, but limited selection.
The charts look bad on mobile#
Yup. I don’t know how to fix it yet, I’m sorry.
What else do you do?#
Also visit my Blades in the Dark (and Scum and Villainy) character sheet app if you ever need one! It has partial French support and was built with NVDA screen-reader testing!
If you’d like to continue encouraging or supporting my work, consider buying me a coffee. 💖
