Do the expected values for rolled stats take into account the non-dice related rules?

[Luccan]

Rereading the PHB recently (haven't read the entire thing in years) and I noticed this bit where it discusses generating attributes


If your scores are too low, you may scrap them and roll all six scores
again. Your scores are considered too low if the sum of your
modifiers (before adjustments because of race) is 0 or lower, or if
your highest score is 13 or lower.

It's entirely reasonable, though not extremely likely, that this will happen. I'm not all that great at probability, but it strikes me that having a minimum requirement that may result in rerolling all your stats might effect the expected outcome of those stats. I ask because I've read that standard point buy and Standard array aren't too far off from what you can expect to roll... but I've heard that in the context of using this method for 5e as well, which AFAIK doesn't have this clause concerning rolled stats. So, is it accurate to say point buy and standard array are close to the most expected values for rolling in this method or not?

[Fizban]

Another poster (can't remember who, don't know where my bookmark is) said that according to their math the actual average point buy for the rolling method should be 32 points- maybe that's because of the reroll, maybe not. Either way, it doesn't change the fact that the expected stats are the elite array, which is what's used for standard NPC entries and the playtest characters.. You get to reroll if you don't have at least a +2, because they expect you to have a +2 in your main stat. You get to reroll if your total modifiers are low, because that means you have a bunch of penalties they don't expect you to have.

I say this because even if some statistical magic says that the standard point buy "should" be higher for that rolling method, it doesn't matter. Rolling is for people that want random stats to influence their characters for the entire game. But the game was not made with random stats or higher point buys in mind, and giving out more points is essentially the same thing as just increasing everyone's highest stat- the game expects +2-3, not +3-5. They never should have put rolled stats in the 3.0 PHB, and yet to this day even 5e still presents it as the first option :smallmad:

(And to finish the rant- then people make it worse by allowing ever more powerful races with higher starting bonuses, "LA buyoff," etc, and then wonder why printed CRs don't work when everyone's stats are 4+ points higher than they're supposed to be even before optimizing anything.)

Edit: okay, that took no time to find. It was bekeleven here (https://forums.giantitp.com/showsinglepost.php?p=21256134&postcount=34).

[Luccan]

Another poster (can't remember who, don't know where my bookmark is) said that according to their math the actual average point buy for the rolling method should be 32 points- maybe that's because of the reroll, maybe not. Either way, it doesn't change the fact that the expected stats are the elite array, which is what's used for standard NPC entries and the playtest characters.. You get to reroll if you don't have at least a +2, because they expect you to have a +2 in your main stat. You get to reroll if your total modifiers are low, because that means you have a bunch of penalties they don't expect you to have.

I say this because even if some statistical magic says that the standard point buy "should" be higher for that rolling method, it doesn't matter. Rolling is for people that want random stats to influence their characters for the entire game. But the game was not made with random stats or higher point buys in mind, and giving out more points is essentially the same thing as just increasing everyone's highest stat- the game expects +2-3, not +3-5. They never should have put rolled stats in the 3.0 PHB, and yet to this day even 5e still presents it as the first option :smallmad:

(And to finish the rant- then people make it worse by allowing ever more powerful races with higher starting bonuses, "LA buyoff," etc, and then wonder why printed CRs don't work when everyone's stats are 4+ points higher than they're supposed to be even before optimizing anything.)

I'm generally of the opinion that, in a more balanced game, rolled stats are fun. So long as it doesn't prevent you from building what you want, then it's fine. It doesn't always work for 3.X because there's a greater number of MAD classes. I also don't think CR is a particularly useful system the way its been implemented in either 3rd or 5th edition. Too many variations in character builds alone make listed CRs unreliable without widely different stats.