So for several months now I've been running three different games: a more or less weekly Saturday Pathfinder 2E game, and a rotating weeknight session that jumps between D&D 3.5 and D&D 5E. In Pathfinder the group has hit 5th level, so still relatively low powered. D&D 3.5 deliberately started at level 1 and has crept up to level 4ish for most of the group now. D&D 5E rolled in at level 3 and is hitting level 5.
In each case I worked out a fairly detailed scenario/plotline to keep things focused. In Pathfinder the group is a gang of young acolytes in a local assassin's guild with strong political, patriotic ties to protecting the city itself. They face a crisis as the heir to the throne is killed, then resurrected under extremely suspicious circumstances, even as their senior leadership are taken out of action, leaving them alone to figure things out.
In the D&D 3.5 game I started with a level 1-3 zone in which I worked out a main dungeon of interest and several minor side quests. I then built it around leading in to a specific Necromancer Games module from the good old 3.5 days of Necromancer, which shall remain nameless in case any of my players are reading. The key conceit of this campaign is it is extremely sandboxy and open-ended; I don't care where the PCs go, as long as they do something of interest....I have most angles covered unless they suddenly decided to journey two hundred miles away in a random direction.
In the D&D 5E game I an running it in a different section of the same world the D&D 3.5 game is taking place, and it starts with a group of ragtag mostly monstrous heroes who work for a local investigator of an orc-dominated city; they are essentially given tough jobs that require protecting the interests of the city against the neighboring human kingdom which often mistrusts the orc-run area. The group is currently wrapping the latest investigation, into the attack and kidnap of a priestess who channels the will of a popular goddess, and it is exposing a deeper mystery of other groups who seem interested in sowing conflict between the orcs and humans. I started this campaign as a 3.5 venture for the first scenario, but then moved to 5E for the next storyline as I wanted to do exactly what this article is about: contrasting 3.5 D&D against its successors, 5E and PF2E.
Here's what I've learned now after several months of gaming:
Pathfinder 2E Remains Fun but it's Balance is Too Much
Pathfinder 2E's rigidly designed skill system is annoying. Seriously, I wish it was a broader set of skills, and not so tightly woven into the structured pathology of Pathfinder's overly balanced advancement, balanced to the point of eerie predictability. In fact, after running a level 1-20 and some smaller campaigns in PF2E, I have decided that, in contrast with the editions it is meant to replace or compete with, that it's highly structured style just isn't as flexible or fun as prior editions have been. PF2E, on occasion, has been compared to D&D 4E, and I can understand why: it's design was handled with too much emphasis on a specific play experience, and not enough feedback clearly entered during design and playtest to allow for Paizo's team to realize that there are other styles of play which their new game would not support so well (such as at my table, where I am sick and tired of calling on Society checks or generic crafting checks or Nature, Survival, etc. etc. for myriad other skills that the PCs should actually have as separate skills).
Do I still enjoy running it? Yes, particularly in Roll20, which makes it easy. But it is painfully clear that in contrast with 3rd edition and 5th edition D&D that Pathfinder 2E feels a bit more like a "sandbox playground where everything has been padded to prevent the players from escaping its confines." Moreover, my players describe PF2E as "A GM's game, for GMs who don't like uppity players." They like elements of it....such as how ancestries work, but they also sense that a lot of PF2E's design went in to removing the potential for players to design truly interest characters and unexpected synergies.
As a GM I have come to realize that combat encounters of even 1 CR more than the players can be a pain in the ass and risk unexpected deaths and TPK, it simply doesn't have the range that you can get out of D&D's editions for encounter design due to its hard focus on tight balance. I have ranted about this in prior blog posts, of course, but to give you an idea: I mostly design encounters around a CR 1-2 less than the PCs. Anything more than that is too trivial, and anything except a rare CR+1 will be too deadly with remarkable consistency.
D&D 3.5 Is Funner Now That It's No Longer The Only Game Around
Put simply: D&D 3.5's key flaws evaporate once people are playing it for fun and enjoyment and you no longer have a large player base and online presence talking about min/max game design and turning everything into an arms race. My group is having fun in a way that very much reminds me of the early fun days from 2001 to 2006. Sometime after that I feel the game hit a level of notoriety and the obsession with optimal builds began to infect everyone who played it. Now? It's just a fund game and I am enjoying a sandbox campaign with a group that is barely optimized for fighting paper bags, let alone serious stuff. I run it as a DM aimed at providing for a good time, and I don't worry too much about balance at all, a welcome reprieve from PF2E on the other game night.
One thing I realize with 3.5: I prefer the old skill system. It was flexible, a little unpredictable, and had more stuff in it that feels natural to call out for in the course of play. I am sure a great many people much prefer "perception" as a skill (or not at all in the OSR crowd) but I love the fact that Spot, Listen and Search are three different things and can reflect that one PC might be a keen eyed observer but have a hearing problem, while another PC might have bad eagle vision but can search methodically with great efficiency. Good stuff.
I don't anticipate running D&D 3.5 past level 12 or so, but who knows.
D&D 5E Feels Better to Run with 3.5 Fresh in Mind
D&D 5E is good, and running it back to back with 3.5 makes me appreciate it more. Most interestingly, sometimes I find myself using 5E as a reference point for adjudicating some moments in 3.5, to keep tings simple. Other times I find myself tempted to house rule in a few items from 3.5 to 5E, but I try to restrain myself as much as possible. Like with 3.5, I suspect that as D&D 5E goes on I may grow a bit tired of its core simplicity and lack of dynamic elements in stuff like saves and damage; but I did decide with this campaign to run it using gritty resting rules and that is going a surprisingly long way toward my feeling like the players are "tough guys in a tough world" rather than the standard 5E trope of fantasy superheroes. Still...they've only just hit their good levels, so we'll see how things go in the coming months.
Also, I don't hate the D&D 5E skill system, at all. In fact, while I still like 3.5's granularity on skills, will take the 5E skill system over PF2E's skill system any day.
After the group completes their current storyline, I am considering integrating a module, possibly Rise of the Drow, which I just snagged. We shall see.
Some Conclusions (so far)
So....it's fun running three iterations of basically the same game, and seeing how my expectations and experience in one lend to observations and changes in the other two. The real takeaway I have gotten from this experience so far has been one about how I structure and focus on campaigns. Specifically: I am not as interested in the "big story" campaigns as I once was, and the D&D 3.5 game where I basically made a sandbox for them to do whatever (including regions of different levels they can wander in to regardless of their own level) has actually been the most fun. But my structured investigation stories in the 5E game have also been a lot of fun because I took some time to lay out interesting paths of discovery and skill challenges related to the investigations. It's "pseudo-rails" in that the PCs could, like, stop investigating and go elsewhere, thus ending the module, but they had motivation and interest to proceed so it worked.
Meanwhile, the very structured big picture storyline which admittedly makes the PCs more integrated to the world and setting proved perhaps a bit too much in terms of scope and design. I realize now that I came up with a great idea, but then sort of left it as a "and so that happened," type event, without a lot to go after the main event. Luckily I proceeded to dive in to some of the smaller angles and pieces, fleshing out the game to feel more like a sandbox, but I concede it's hard to just do sandbox in PF2E because a good sandbox should allow for the PCs to get into more trouble than they can handle on occasion, and in PF2E that can quickly turn into a lethal TPK. So....we'll continue for a while on this one, but afterwards I need to think hard on whether I plan to continue with PF2E or not, because it almost....but not quite....manages to frame the sort of adventures I like to run, but just not as well as either D&D 5E or D&D 3.5, which both do it so much better.
Final conclusion.....turns out too much balance in design is not necessarily a good idea! Who knew?
Also, and this is extremely important to stress: the D&D 5E and 3.5 edition games both have a huge edge over Poor Pathfinder 2E, in that they are live games I am running in person. PF2E is online, and while the online tools make for an easier time of it, I know my lack of time to sink into enhancing the graphic elements of the experience factor against the game to some degree, as does the predilection for the overall experience to be a generally less satisfying experience than the sort where normal humans are able to see each other live and not share a single audio channel. So, I must concede that PF2E in a live environment might still be a better overall experience than I am giving it credit for. Poor Pathfinder though....I think I got about 10 levels in to the original campaign when it had to migrate to online due to the pandemic, and its more or less lingered there ever since. May need to change that soon.