Background: My first D&D was 3e. I started DMing with 3.5e and have run 3.5e, 4e, 5e, PF1, and some non-D&D games like Dungeon World, Open Legend, and a homebrew D&D-like hack for Fate Core. Dungeon World has shaped the lens through which I view PF2 and how I GM, though the effect is fairly modest, and I’d posit that
Justin Alexander has had more influence on my GMing procedure and structures.
Prior to converting to PF2, I was running a sandbox hexcrawl in my homebrew setting (that got its life in Open Legend) using 5e. See here for why we’re converting, but the TL; DR is 5e is not a great fit for what we want. Before 5e (and a few other systems), we were primarily a PF1 group using a heavily house-ruled version of PF1. I stopped running PF1 because it got to be too much, and I got tired of Golarion. We switched to Open Legend then 5e, which is where my setting started (OL) and then evolved (5e).
I was pretty cold on the PF2 playtest. I thought it was okay when I played it at Origins, but I did not like what I saw when they released the playtest materials and tried to run one of the PFS scenarios for my group. I really bounced off skill adjudication, which seemed cool in theory (assign a level to a task and derive your DC from that) but was clunky in practice. I like the simple DC ladder in the final rules, whichh makes it easy to improvise a DC.
I ran a one-shot for my group with the final rules, which went well. I knew going into it combat would probably be fine. When I ran PF1, we used the Unchained action economy. I wanted to see how exploration went and try some scenarios that caused us problems in 5e (a bunch of flying snakes TPK’d the party…). We also have a player who joined us and has only played 5e. Everyone seemed to like the one-shot.
I hew towards an older school mentality on encounter design, so I used mostly trivial and low threat encounters. Moderate threat encounters have a risk of ending the adventuring day, which makes them too risky to use regularly. We played for about four or five hours, and we got through five encounters. In addition to that, we were able to do a little urbancrawling and some dungeoncrawling. My favorite moment was when one of the PCs accidentally set off a Grimtooth’s trap and got himself flung into the spiked ceiling. Well, that and when the
zombie hulk opened up combat by pelting the party with corpses. That was awesome.
Prepping was easy, but I always skew prep-lite. Nothing in PF2 required me to go prep heavy. The official GM screen is a poor reference, but what official screen doesn’t have problems? There is a learning curve, and I think that is what all the traits do. I find them useful and like that they are there, but they are going to pose a barrier to entry.
The system itself ran very smoothly. There are very few exceptions, and the CRB is well organized as a reference (probably why it is so boring to read). I found it very easy to adjudicate improvised activities. One of the PCs wanted to use a ghoul’s corpse to bluff some ghouls in another room, and it was very easy to use the decide which basic actions applied and have it all fit naturally in the game’s action economy. Interestingly, there were no complaints about things that used to be free now costing actions (such as adjusting your grip), but I think that’s because of how largely exception free the action economy is coupled with everything’s being fitted into it.
One thing that did not go well was always revealing DCs. I had started doing that in 5e, but I think I’m going to use secret checks in PF2 as written. No matter best intentions, I think it leaks too much information in PF2 to work well given the number of checks that happen e.g., when PCs investigate traps.
After we finished the one-shot, I set to work on converting my homebrew setting. It features none of the core races from D&D and is generally more JRPG-like and less like a traditional setting, but it also pulls in other influences (the Stormlight Archive and Vance’s Dying Earth being prominent). Compared to homebrewing races in 5e, I find homebrewing ancestries much easier in PF2. The customization points are standardized, so there are lots of examples to use as benchmarks, and you can steal or repurpose them quite easily. I’d say about 25% of my ancestry feats are borrowed or modified from the core ancestries.
The campaign soft-resets and starts back up on the 5th. My players are more excited to make characters than they have been in a long time. That’s great, and what I like is we don’t have to worry about trap options or plan out builds in advance to make sure the PCs are performing at par. From what I’ve seen from theorycrafting so far, DPR between options is close enough that players can focus on concept and not be badly off.