To generalise a bit (but hopefullly not so much as to be egregiously wrong), there were two main ways of playing pre-3E D&D.my initial observations about what made earlier editions "feel" a certain way are limited by my having to rely on the rules as written, since I have very little--well, actually, non-existent experience actually playing with the rules.
Reading the rules, I came to the conclusion that 4e dispensed with the management of the mundane, which in turn meant that the attritional aspects of the game were no longer present
The first, which is largely what is promoted by Gygax in his PHB (and mostly also in his DMG, though it suggests perhaps a bit more of the second approach), is dungeon crawling. Inventory is a big part of this, although whether mundane equipment or spell load out is the more important aspect of inventory will depend on table style, party level (the higher the level, the more important the spells and the less important the gear) and probably other stuff I'm not thinking of at the moment.
In my view the greatest exponent of this style (in essay form, at least - I don't know about actual play) is Lewis Pulsipher writing in early White Dwarf. Gygax's essay at the end of his PHB (just before the appendices) makes much more sense if read by someone who has also read Pulsipher.
Key to this dungeon-crawling style is that the bulk of the content has been determined by thge GM in advance, and hence the players have an objective framework in which to (i) identify the content (via searching, using potions of treasure finding, etc), and (ii) deal with content (via fighting, avoiding, looting the valuable stuff, etc). This objective framework underpins planning, which in turn is what makes inventory key.
There are three main elements of randomness/immediacty in this style: (1) wandering monsters - and one striking feature of Gygax's DMG is the pains he goes to discuss when a GM should disregard a wandering monster check because it will spoil play; (2) reaction rolls, which (i) reward CHA-focused or charm-focused builds and (ii) can turn what would by default be opposition into something more neutral or even advantageous; (3) rolls to find secret doors, traps and the like, which can affect the sorts of intelligence players are able to gather and also their ability to actually go to various parts of the dungeon (at higher levels this 3rd factor becomes less important, as PCs have wands of secret door detection, passwall spells, etc). Gygax's DMG also talks quite a bit about how the GM can manipulate factor (3) to help ensure that the players actually get to the content the GM wants them to get to; conversely, ToH is all about amping factor (3) up to the Nth degree as the focus of play.
This style doesn't really survive the transition to seriously non-dungeon-focused play, because (a) it is unrealistic to expect a GM to generate all the content for a verimilitudinous rather than narrowly contrive setting, and (b) it is unrealistic to expect the players to gather all the intelligence and do all the planning for such a setting.
Which brings us to the second style of play of pre-WotC D&D: it's a style which uses the mechanical trappings of the first style, but not the other play procedures. In particular, it assumes that the GM is overwhelmingly in charge of deciding what the content is that the players (via their PCs) encounter and hence engage with. The Dragonlance modules are a famous and fairly early published example of this style; the whole Planescape and 2nd ed Ravenloft ouevre is a highwatermark of it in pre-WotC D&D publishing.
In this style, mechanics often become secondary: the players don't use mechanics to gather and act meaningfully upon information (because the sequence of content is in any event under GM control - a hallmark of this style is that there is a clue, but then if the players don't find the clue a NPC tells them anyway); the players don't use mechanics to gain access to the content, but rather it comes to them as the GM dictates; and the one place where it is often assumed that mechanics will be deployed - ie combat - the assumption is that the PCs will win, such that if in fact they lose the whole adventure goes off the rails or grinds to a TPK-induced halt (and there is often an overt or covert signal to the GM that s/he should fudge combat outcomes to make sure this doesn't happen).
[MENTION=16586]Campbell[/MENTION] made a good post about this playstyle in a recent thread - summing up, in this second playstyle which downplays mechanics and relies heavily on GM control over the sequence of events in play (what content is introduced, and how the players' interactions with it via their PCs ends up), the players' main job is to provide colour and enthusiastic engagement with the GM's story.
My suspicion is that quite a bit of 5e is played in more-or-less this style.
Well, the slip here is telling!A 4e fighter in the same position could pick up a tree branch and probably be sitting on the dead body of the orc king when the rest of his party finally enters the throne room, because, in 4e, the power is inside you!--er, I mean, inside the character. In a certain way, that's awesome.
4e, by making the encounter the focus of play, allows the GM to determine what content the players encounter via their PCs, but allows the players, via their PCs and via engagement with the mechanics, to determine the outcomes of those encounters. This is at the heart of its "indie"-ness, and is the point where (as per my post just above) I think 5e most radically departs. (And if "indie"-ness looks like combining elements of player authority as per the dungeon-crawl style with elements of GM-led story as per the second style of play I described: well, that's because it is! It's an attempt to allocate power across GM and players so as to get story without the GM just leading the players through something pre-written.)
5e can't really afford to put that sort of power inside the players, because once the players have principal control over the outcome of encountrs, which changes the shape of what is to come in a way that refelcts their choices and priorities, the GM's control over the "adventuring day" is forfeit; and once that happens, the intra-party balance of 5e breaks down.
I think this is the true meaning of the slogan that 5e "puts poweer back in the hands of the GM". Or of [MENTION=996]Tony Vargas[/MENTION]'s remarks that balance in 5e is on the GM, not the mechanical design.
But failing to do this is not problematic if it is assumed that the GM, in any event, is in charge of managing the outcomes of encounters and feeding from one to the next. That's not an assumption in classic dungeon-crawling play; and it's not an assumption for indie-style play (and hence, by implication, not an assumption for 4e - though the 4e modules themselves don't seem to be aware of this!); but I think it is an assumption of the default approach to 5e.Claimed Magic Items were not calculated into the math via marketing.
<snip>
See what's going on here? One group of Fighters likely takes out an at-level challenge in a round, the other group of Fighters likely takes three rounds to do it. Magic items radically change how difficult a typical combat is going to be
<snip>
Which is fine and not problematic at all provided the DMG tells you which party is at the correct power level in case someone didn't see the marketing about no magic items in the math.
Which it does not do...
If we assume that the GM is in charge of managing those outcomes and shepherding the players (via the PCs) through his/her scenario, then the fact that an encounter takes one or three rounds is just another one among myriad factors that the GM is managing, fudging around, etc. For many D&Ders being able to do that sort of thing is what they mean by "being a good GM".