Did you play AD&D? I can't remember how old you are, so hopefully that's not too offensive. If so, was a typical game really as hostile as people say it was?
That's one of those question where the answer hovers somewhere between "no, with a couple of massive caveats" and "yes, but not in the way most people think".
A lot of AD&D 1st Edition's GMing practices are pretty hardass by modern standards; however, they need to be understood in the context that the game's authors were writing for a target audience who mainly played the game in college wargaming clubs, where players would frequently transfer between groups and group sizes tended to be very large – six players per GM was considered a bare minimum, and up to a dozen player characters in a single party was by no means unheard of!
In particular, players would often bring their character sheets with them when hopping between groups, and it was considered a faux pas for a GM to reject an incoming player's existing character or request any substantive changes be made, so managing expectations could be quite challenging; even as late as 2nd Edition, the Dungeon Master's Guide contains extensive discussion of how to gracefully handle players bringing existing characters with them who aren't necessarily a good fit for the present game's tone or resource economy.
The upshot is that the culture of play these iterations of Dungeons & Dragons are targeting inherently obliges the GM to take a much firmer hand to keep things on track than a pickup game that draws players exclusively from within the GM's established friend group might – and to be sure, some GMs abused these expectations to act like petty tyrants, but some contemporary GMs do that, too.
A big part of the modern perception that 1E and 2E were extraordinarily player hostile, meanwhile, has nothing to do with the previously discussed GMing practices; rather, it emerges from the transition away from that culture of play in a slightly unexpected way.
In brief, back when D&D was mainly played by wargaming clubs, it was fashionable to run pre-written adventure modules competitively at conventions; the competition wasn't between players, but between parties, with multiple groups running the same adventure in parallel to contend for prizes. Tournament play sometimes chose its winners based on the fastest real-time completion of the module in question, or set specific objectives within the module which would award points when completed, a bit like speed-running or achievement-hunting in a video game (though neither practice existed yet at the time).
It was the survival module, however, that quickly emerged as the most popular tournament format. In a survival tournament, each player would provide or was furnished with a binder containing a fixed number of pre-generated character sheets, switching to the next character sheet in the set as each preceding character died; the winning group was the one whose last surviving character's corpse hit the dirt furthest from the dungeon entrance.
Many of 1E's most popular adventure modules, including the infamous Tomb of Horrors, were originally written as survival modules to be run at tournaments in conventions. As such, they were designed to kill off player characters both quickly and efficiently, so as to reduce the likelihood that the tournament would run overtime and get kicked out of the convention venue. When they were later cleanup and repackaged as commercial adventure modules, their text rarely bothered to explain any of this – who doesn't recognise a survival module when they see one?
The answer to that question, of course, is kids who didn't come up through the mentorship system of the college wargaming clubs, but taught themselves how to play D&D from first principles using books they bought at their local hobby stores – and when D&D's popularity unexpectedly exploded in the early 1980s, there were suddenly rather a lot of them!
These kids purchased the repackaged survival modules along with all their other D&D books; having no frame of reference, they assumed that these represented what a "standard" D&D adventure was supposed to look like – and since they weren't experienced players with whole binders full of pre-generated backup characters at their fingertips, the result was a lot of seemingly unfair total party kills, and a lot of kids concluding that the previous generation's GMs must have been objectively insane.
There is an additional amusing point of order here, which is the answer to the following two questions. I once had a discussion with someone in Gary Gygax's gaming group, who was involved in early TSR work a bit. Allow me to paraphrase my questions and his answers.
Why publish survival modules as your primary format of published adventure?
"Because that's what we had -- they were already laid out for publication. Why not publish them and make some money off it?"
Did it ever occur to you at the time that publishing adventures like these would shape the larger D&D culture's expectations of what play was supposed to look like?
"No, why would it?"
One of my favorite anecdotes about early D&D, from Blog of Holding:
"It’s hard to get that context just from reading the original Dungeons and Dragons books. If nine groups learned D&D from the books, they’d end up playing nine different games.
"Mornard told us about an early D&D tournament game – possibly in the first Gen Con in Parkside in 1978? Gary Gygax was DMing nine tournament teams successively through the same module, and whoever got the furthest in the dungeon would win. You’d expect this to take all day, and so Mike was surprised to see Gary, looking shaken, wandering through the hallways at about 2 PM. Mike bought Gary a beer and asked him what had happened – wasn’t he supposed to be DMing right now?
“It’s over!” replied a stunned Gary Gygax.
"Gary described how the first group had fared. Walking down the first staircase into the dungeon, the first rank of fighters suddenly disappeared through a black wall. There was a quiet whoosh, and a quiet thud. The players conferred, and then they sent the second rank forward, who disappeared too. The rest of the players followed.
"The same thing happened to the next tournament team, and the next. Players filed into the unknown, one after another. And they were all killed. The wall was an illusion, and behind it was a pit. Eight out of the nine groups had thrown themselves like lemmings over a cliff; only one group had thought to tap around with a ten foot pole. That group passed the first obstacle, so they won the tournament.
"Gary and his players couldn’t believe that the tournament players had been so incautious. But, to be fair, none of those tournament groups had played in Gary Gygax’s game. They had learned the rules of D&D, but they had no experience of the milieu in which the book was written. Of those nine groups that had learned D&D from a book, only one played sufficiently like Gary’s group to survive thirty seconds in his dungeon."
#ngl survival module sounds fun as fuck. maybe i gotta torture my current group a bit (via @nadaismus)
It's worth bearing in mind that tournament-style survival mode developed in the context of a version of D&D where you can create a new character and hit the ground knowing everything you need to know to effectively play them in just a couple of minutes. 5E isn't structurally terribly well-suited for the binder-full-of-backup-PCs approach, and it's definitely a recipe for disaster in 3E or Pathfinder unless your entire group consists of a very particular flavour of high-effort masochists.
It also bears mentioning that the current culture of RPGs encourages a separation of player knowledge and character knowledge. I, as a player, know that the big cat with tentacles out the back is a displacer beast, but my character doesn't, and the character that replaced the one the displacer beast killed. That separation, particularly with Survival Modules, was not the case back in the day. Characters had full knowledge shared between them, so if Dave the fighter got disintegrated by a beholder, Dave's identical twin brother now knew beholders have disintegration attacks. This is part of the reason why it was considered bad form for players to read monster books.
It's broadly untrue that the idea of separating player knowledge from character knowledge is a modern development. The practice descends to tabletop RPGs from the historical wargames they splintered off from; tabletop wargames which focus on accurately re-creating historical battles often operate on a gentleperson's agreement to refrain from acting on strategic information that your side's commanders couldn't reasonably have been aware of, or employing tactical doctrines which had not yet been developed when the re-created battle took place, and many early tabletop RPGs adopted similar conventions, to greater or lesser degrees. Heck, games like Paranoia were parodying those conventions as early as the mid 1980s! It's come in and out of fashion in mainstream RPGs over the past half-century, but it's not a recent thing.
It is, however, correct that there typically was no expectation of observing these conventions when playing survival modules in particular.
Oh, so that's where Munchkin got the idea of your identical twin turning up when you die in game.
Yeah, having your previous PC's identical cousin randomly come rocking up five minutes after you died is totally a thing that happened, largely as a response to the awkward transitional period where survival play was still in fashion, but the game's rules had become too fiddly for rolling up a new PC on the spot to be a pain-free process, so folks would just recycle their existing character sheet instead. You saw a lot of it in the 2E era!
