There is nothing inherent in the system that "forces" a GM to write adventures that resemble official APs in any way, specifically regarding encounter difficulty.
[...]
That doesn't mean it isn't relevant to discuss the game as envisioned by Paizo. It means we need to remain able to distinguish what the rules lead to as separate from what the official application of them leads to.
Saying "PF2 forces me to TPK my characters" would be wrong. But saying "PF2 and 5E are equally deadly" would be misleading, and just as useless.
[...]
From a game design perspective, I’m coming to the conclusion that the “4 degrees of success” model inherently increases table variance, because if small changes to bonuses result in larger swings to results, different interpretations of rules/ways of playing the game (which may not appear at first glance) to impact bonuses also result in large swings in the results.
First, the math. In a 4 degrees system, if a critical hit does double damage, a +1 to hit is the equivalent of a +2 to hit in a system that doesn’t use a 4 degrees system.
This is the case even if the other system uses critical hits.
In other circumstances, if a critical success is worth more than twice a simple success (or a critical failure is more than twice as bad as a simple failure), than a +1 is worth more than a +2 in a system without degrees of success. I would argue that generally speaking, spells in PF2 fall into this 2nd category (a crit failed save is more than twice as bad as a simple failed save).
Now for the link between table variance and degrees of success. I don’t have data for this, just anecdotes.
On the PF2 forum, a poster who clearly plays the game and understands the game suggested that all else being equal, it didn’t make a difference if an attack was rolled by the player or a save by the monster. I don’t think he was wrong, but he was definitely playing the game differently from my group. We were playing on a VTT, but the DM was rolling with his physical dice. If a monster saved or critically saved, I had no idea if I was unlucky or if I was unlikely to succeed given the save targetted (given level differences, this would be case regardless of the outcome of Recall Knowledge). Obviously, that would not be the case on an attack roll.
For another example, using Recall Knowledge on a unique monster, or using Recall Knowledge based on the traces of the monster before you see it. Neither directly affects your rolls, both are likely to have a large impact on a combat.
To summarize, if small bonuses are likely to have a big impact on gameplay due to how your game is designed, typical GM decisions are also likely to have a huge impact on gameplay. This
will increase table variance, and may be the reason why people have such different experiences from playing PF2.