Reply to thread

Message: <blockquote data-quote="Li Shenron" data-source="post: 7428474" data-attributes="member: 1465">We can first set a couple of restrictions, just to make the analysis simpler:1) PHB feats only. 2) No multiclassing.I went through all the PHB subclasses features, and I don't actually see a significant difference between classes. Perhaps Wizards and Sorcerers have a slightly lighter-than-average set of subclass abilities, and Clerics and Druids have a slightly heavier-than-average set. Note that this is when considering the whole set of subclass abilities (levels 1-20) compared to a number of feats equal to the number of levels at which the subclass gains its features (e.g. all Bard's subclass features VS three feats, all Wizard's subclass features VS four feats). Some subclasses also grant bonus spells (known, prepared or added to the list), so without a choice of subclass the character will also lose those spells.When considering individual subclass features, the general situation is that there is way too much variance between their usefulness, as well as to the usefulness of feats, and that's before allowing this house rule of gaining ASI/feats instead of a subclass. Variance plus more variance means that extremely good/bad combinations can be even more extreme.There seems to be some pattern so that especially the early subclass features are strong, then those gained at intermediate levels are weaker, and the late ones are strong again. But the pattern is not necessarily the same for all classes (e.g. the last subclass features of the Sorcerer look weakest to me).But on average, it seems to me that subclass features are roughly on par with feats, so eventually the problem is that while choosing a subclass means you must accept the whole package (including some better-than-feats abilities and some worse-than-feats abilities), choosing feats means you can always pick good feats. In the course of 20 levels, there is going to be a net gain. Naturally, as most games don't reach past 15th level, the average net gain is lower than the theoretical maximum. So maybe the risk of exploitation could be kept low enough by restricting the choices a bit, so that the feats regarded as strongest just cannot be gained in this way, for example Lucky, Alert or Sharpshooter (depending on your mileage, the list could be a lot longer).</blockquote>

Verification