The Tomb of Horrors by Gary Gygax is probably the most famous dungeon in all of D&D. Part of the appeal of D&D for many gamers is the shared experience of playing common adventures, and because The Tomb of Horrors has been around for such a long time and has gained such a reputation for being a real challenge, I think it's well worth playing the module. That said, for players it can be a peculiar module. Modern D&D has largely deviated from the style of play represented by Tomb of Horrors, and in some ways, so has the system itself. For the Dungeon Master, it's often a treacherously confusing module, with an over-complicated map and unclear explanatory text. I've run the Tomb a few times, and I run it as an all-day (about 8 to 9 hours) session. This post isn't a step-by-step guide to running the module, but it contains some notes for a Dungeon Master preparing to run the Tomb, some of the problems you might run into, and how I deal with them.
WARNING: THIS POST CONTAINS SPOILERS.
For whatever reason, most D&D adventures assume the Dungeon Master is already familiar with the adventure. They're heavy on prose and description, they rarely separate atmosphere from mechanics, and they frequently make vague references to areas of the map without bothering to provide a key. They're impossible to parse at a glance.
For this reason, you must read the Tomb of Horrors in advance, and probably take notes.
I'm still working on perfecting my notation, but here's how I generally take notes (all information has been obfuscated to protect the secrecy of the module):
10. Great hall of goblets [p 193] * Illusions: DC 20 Int (Investigation) * If goblet is removed: summon Imp [MM 94] a. Goblet on pedestal: worth 1000 GP b. Secret door: DC 20 Wis (Perception) c Illusion: Blue orb at waist: crawlspace to #18 [p 194] d. Illusion: Yellow orb conceals secret one-way door from #8 [p 192]. Open with _knock_ or destroy with _disintegrate_. e. Illusion: Green orb at feet: crawlspace to #10 [p 192] f. Illusion: White orb overhead: tunnel to #21 [p 194] g. Archway south, misty: if PC enters, then 1000 GP appears in goblet
Not that unlike most D&D material, I use page numbers.
The 10. Great hall of blah is the name the book provides to the location.
The bullet points immediately under the location are global elements. For instance, all illusions in this location require a DC 20 WIS (Perception) check to overcome. There's also a monster here, but it's triggered only if blah is removed from its pedestal.
It's rare that D&D maps or books provide markers for anything, and even when they do they often treat everything with equal weight, regardless of how vital (or insignificant) an item is to the plot.
If something shouldn't be mentioned to the players until AFTER they've overcome some obstacle, then I "conceal" the item AFTER the barrier, in text.
For instance, instead of just noting that an imp may attack, I specify
If goblet is removed before I mention the imp.
This prevents me from stupidly saying something like this to my players: "There's an imp in the corner. Oh wait, no there's not. Not yet. I mean, forget I said anything. So does anyone want to grab the goblet and remove it from its pedestal?"
The same goes for all illusions in the room.
Instead of tricking myself into announcing that there are doors and tunnels, I preface them with
Illusion:, which I've already defined in my global section as a DC 20 to bypass.
I also give letters to each item in the room, and then I write those letters on the DM copy of the map. This provides quick reference when I have 4 or 5 players scurrying around the room looking at places on the map. It's a heck of a lot easier than referencing a table on page 194 that sort-of-aligns with the map way back on page 187.
The map of Tomb of Horrors is just plain stupid. It's cramped, and it arbitrarily tries to use elevation and depth when the map is obviously confined to a 2-dimensional space. For instance, room 7 in the tomb leads to a pit in room 3 (I think? the module is rarely clear on how rooms connect, so this is my best guess.) Depth in this case matters, because players can't find the tunnel connecting the rooms, and probably won't (although I've had players that do) until they end up in room 7 and find there way back out into room 3.
But why does the tunnel to room 7 have to go under room 13? Well, it doesn't. There's no reason for the tunnel to go under other rooms. The layout of the dungeon doesn't actually matter. A dungeon is a flowchart, and it would be much easier to comprehend if the map were laid out to recognise that. Room 7 could be isolated in any part of the empty space on the map, but instead the module authors continue (even in updated versions) to place the tunnel to room 7 under room 13, and the unnumbered crawl space above room 7 visually adjacent to room 13. Unless you spend an hour pondering it and another hour searching the Internet for explanations, you're VERY likely to assume that the crawlspace is a closet on the south wall of room 13, and room 7 is an extension of it.
And even if you don't, how do reveal the tunnel between room 7 and 3 without also unintentionally revealing that room 13 exists?
There are several examples like this, and it only gets more complicated the more modern your tool set. Sketching out a map on a dry erase map is one thing, but try mashing the map into a VTT.
My point is, study the map early, and make lots of notes on your copy of it for your own reference.
Finding the tomb is a major plot point, and it often takes players an hour or two to get into the tomb. I actually think this is a really neat mechanic, but I also admit it's a little awkward to run. It's fine to say "You're facing the north side of a sheer cliff, covered in overgrowth of trees and bramble. This is, you've been told, the entry to the legendary Tomb."
But then what?
As written, it seems that you're meant to have the players do perception checks, I guess literally on a square-by-square basis across the border of the map. Here's what I do instead:
There are still, potentially, a lot of awkward moments when the players are convinced they've found the entrance, and are trying to make their way through a truly dead end. I don't see the point in allowing players to spend 4 hours excavating a new entrance through a set of fake doors and a collapsed ceiling, so if your players don't figure out that they're meant to find a different entrance, then you may just have to prompt another Perception check when they camp for the night.
That leads me to the importance of tracking time.
In my weekly games, in-game time is pretty fuzzy. I often don't track it (except in dungeons), and I just acquiesce to allow players to rest when my players are low on HP or spell slots. For Tomb of Horrors, I track time for two reasons:
I track time based on the assumption that player characters spend 1 hour in each room they enter. Adventurers in search of loot, and also on guard against traps, don't just glance around a room. They probe the room carefully. So even if players breeze in and out of a room, I assume they've spent an hour at minimum. Once Perception checks and trap searches have commenced, I start considering adding another hour. Combat definitely adds an hour, not because combat takes an hour in-game, but because post-combat presumably does (searching the body, settling your nerves after a brush with near-death, cleaning your weapon, and so on.)
To track time, I use a deck of cards.
Once the correct entrance into the tomb's been found, players still need to find the entrance into the rest of the tomb. Even in updated editions, the way into the tomb from the real entry hall is astonishingly poorly written. As far as I can tell, there's one door that actually leads into the tomb (specifically, to room 8.) The problem is two-fold:
This is all compounded by the fact that the map is confusing in the first place, so you might think that the apparent lock of a connection to room 8 is a mistake, much as the connection between room 7 and 13 is an apparent mistake. In other words, on the same map, there's the same setup with the same lack of notation: room 7 isn't linked to 13, and 3 isn't linked to 8. Here's what the multiple choice looks like:
One of those is right, so as DM you have a 25% of correctly interpreting this mess. (To reiterate: the correct answer is B.)
It's extremely disappointing to me that decades after the Tomb of Horrors was originally released, the modern re-releases fail to describe the tomb accurately, much less in a way that's useful for the Dungeon Master. I'm not saying (although I'm not not saying) that the D&D authors must change the format of their adventure write-ups entirely, so that they include an easy-to-follow "rap sheet" of the dungeon, with all the flowery descriptions of the space in a visually separate text box. But surely they could at least scrub the text so that it's unquestionable what room connects to what room. Surely they could re-number the rooms, or label important objects. What's a re-release for, if not to improve the original?
In Tales from the Yawning Portal, it states that Tomb of Horrors is "a thinking person's dungeon." I think that's supposed to indicate that it's not a combat-heavy module, and that's accurate.
In practise, though, the module has an odd duality to it. The traps and surprise effects deal damage in the range of 11, 17, or 24. Even the low end of that is enough to perma-kill a low level character. Monsters in the tomb range from CR 7 to CR 18.
For whatever reason, the 5e re-release of Tomb of Horrors fails to specify a target player level. It defaults to the original advice ("players of a high level".) People on the Internet seem to agree that levels 10-14 is reasonable.
That may have been true at some point in D&D's lifespan, but in 5e a level 10 player has almost as much power as the DM has at level 1. 11 points of damage is but a tenth of their maximum HP, hardly enough to cause concern. And I don't know if you've ever pitted a CR 7 creature against four level 10 characters, but it's not a challenge for the players. The final monster (the demi-lich himself) is CR 18, but by the time players encounter him, they've probably already defeated the fake lich and gained well over 25,000 GP, and will likely immediately go buy a bunch of powerful weapons. Even without that, the CR 18 demi-lich is encountered in a tiny 10x20x25 foot room. High level characters are very likely to pass all the checks required to avoid any serious effects.
In short, this dungeon is simply not a damage-heavy adventure, unless players willfully and obstinately refuse to check for danger.
Different people's tolerances for character death varies widely, and people feel excitement and anticipation during an adventure for a variety of different reasons. To my mind, there are two ways to play this:
Reputation and urban legends aside, in the 5e era the Tomb of Horrors is primarily a challenge of investigation. What clever way can you come up with for finding a door? What ingenious string of spells can you daisy-chain together so you can [try to] extract the location of Acererak's crypt from the skeleton impersonating him? Can you endure the frustration of examining an imaginary room from all possible imaginary angles in the hopes of finding a clue about what you're supposed to do next?
A classic is a classic because it's a classic. Whether it lives up to its reputation is a complex question, because it depends on what its reputation is to you. The one thing it will always live up to is that it's the Tomb of Horrors. As a collection of rooms and traps and atmosphere, it is and always will be the Tomb of Horrors, no matter whether it killed all or none of your group's PCs.
If you run it as written, then you're at the mercy of its uneven design and re-design, but you provide your players with an accurate and exact telling of its story. As a Dungeon Master, you have a lot of knobs you can adjust to customize your group's experience. You can impose a specific player level, you can add monsters, you can make rooms larger, you can adjust trap damage, and so on.
It's up to you and your players. If they want the Tomb of Horrors to be deadly, then you can adjust it for that. If they just want the experience of the tomb itself, but definitely want to play the whole tomb, then run it as written: low damage and high player levels. I think they'll be disappointed at the ending, personally, but then again I'm a Dark Souls player and enjoy a punishing game.
I do think the tomb is worth experiencing. It's seriously fun. I try to run it as an all-day session once a year, and when I do I have a blast every time. The tomb forces players to be creative, irrational, angry, excited, confused, and ultimately proud. And they should be. It's a stupid dungeon. It's actively hostile in every way: it dis-invites players from participating, it tricks them, it lies to them, it doesn't give them adequate clues on how to proceed. Players who tackle it ought to experience the full spectrum of emotions, and then they should compare notes with other players. It's a beautiful rite D&D are entitled to, and you owe it to yourself and to your players to have the experience.