D&D 5E Fixing Challenge Rating

[tomedunn]

Now, the same considerations don't quite apply in 5e, but having learned the lesson "many actions make you dangerous" they A) tuned the game for battles against one monster at a time, and B) assumed that several monsters would be much more deadly, even though their total overall damage output is likely much less variable, AND probably more spread out. However, a basic numerical model cannot easily suss these things out either, so like most RPGs the numbers don't work perfectly.

It's not that more actions make you dangerous, it's that monsters in groups live longer than a single monster alone. There are benefits to having more actions, but those are smaller orders effects to what's driving the difficulty change when grouping monsters.

This is why referring to this effect as a result of the "action economy" is a bit of a misnomer. The effect correlates with changes in action economy, but it's not caused by it.

Also, while 5e's encounter building math is centered around encounters with only a single monster, that's simply a matter of presentation and not a reflection of any underlying assumptions on how the game functions (at least, none that I've come across in my research). It could easily be recentered around one monster per PC without changing the underlying math.

In fact, this is exactly what the alternative encounter building rules in XGtE do. As I show in my analysis of those rules, you can derive all of the values in that book from the math in the DMG. Presenting the math in that way doesn't remove the encounter multiplier, it just reduces the size of the error you incur for ignoring it.

The encounter build rules used in 5e aren't built around measuring how likely an encounter is to result in a TPK. They're built around estimating the average damage the PCs are likely to take relative to their maximum HP.

Two encounters with the same adjusted XP values should, on average, deal similar amounts of damage to the party. That doesn't mean their outcomes will have similar distributions, just that the averages of those distributions will be similar. An encounter with more actions will have a tighter distribution than one with fewer, and will be less likely to cause a TPK as a result, but the average strain on the party will still be similar between the two.