DM's fully have the right to disallow whatever they want
I place a lot of blame on the DMs. The PHB has Rule:0 ask the DM while the DMG tells the DM in several places that he or she is in charge of how the game is run at the table, which rules are used/followed, and which supplements in use.
I don't think the language of "right" is very helpful in this context. Against whom is the right held? Are we saying that players have obligations to do as the GM says?
I'm not sure talking about "blame" gets us very far either.
I agree with [MENTION=82746]HardcoreDandDGirl[/MENTION] and [MENTION=58401]doctorhook[/MENTION]: the relevant concepts are
compromise and
fun.
But, speaking as someone who didn't play 3e or 4e, it was a real shock coming back to the game... the entire culture has shifted from DM Fiat to Player Entitlement, at least going on forum culture
<snip>
in earlier editions, the DM was still the keeper of the rules, really, and I think that's a good thing for quite a few reasons. From what I've seen of 3e and 4e, the players are heavily involved in the crunch (I also appreciate that many players really like this).
Classic D&D and AD&D had virtually no PC build rules, but they had some. A fighter has to use weapons and armour; a magic-user has to choose spell load out on a daily basis; a 2nd ed AD&D thief player has to allocate skill percentiles.
I never saw GMs try and dictate weapon and armour choices, tell the MU player what spells to pick, tell the thief player how to allocate skills.
With the increase in build options - more classes, more races, more feats, etc - I think it's natural enough that players regard those options, and choices among them, as likewise being mostly up to them.
Classic D&D also has almost no action resolution mechanics, especially outside combat. (Gygax's DMG had fairly intricate social resolution mechanics, but I don't think they were widely used because somewhat opaque and scattered across multiple sections.) But where the mechanics existed, I think they were mostly used: GMs didn't regularly substitute their own judgements for players' d20 rolls, nor change the damage dice for fireball spells.
So I think it's also natural enough that, as the game developed more extensive rules for action resolution, that players assumed that the GM would apply those rules.
There are nuances, too. In classic D&D, a GM might make a secret door harder to find than 1 in 6 (and Gygax canvasses this in his DMG); or make a door harder to open than the standard STR chance. But the expectation, I think, was that the players would easily link this to the fiction (eg the GM might describe the door as very heavy or stuck; the secret door, when discovered, might be described as well-hidden).
But consider the setting of DCs for Diplomacy attempts in 3E (putting to one side the bigger issue that the Diplomacy rules are fundamentally hopeless). The GM will be influenced by all sorts of considerations of NPC backstory that the players probably don't know, and more significantly will often never know, or have any way of knowing.
This is a recipe for conflict. And it's interesting to see how the 5e designers try to diffuse this in the 5e social mechanics, via Insight checks to learn NPC motivations, thus reducing the inherently secret nature of this sort of backstory.
Every example that Player A offered was some form of special snowflake that was going to have impacts throughout the game simply by its presence.
What I don't get is when the trope emerged that the PCs are just another group of schmos. When the game started, every PC was the first/only of his/her kind - the first thief, the first ranger, the first dual-wielder, the first baby balrog, etc.
