So far, we're two adventures into the Labyrinth Lord campaign. The first was Goodman Games' Dungeon Crawl Classics #0: Legends Are Made, Not Born, and the second was The Fountain of Health from Dungeon issue #39. Running an old-school D&D game has been different in a few ways from the games I'm used to.
Simplicity: I mentioned in a previous post that I really appreciate the simplicity of LL. The easy prep for me is more than just a matter of investing less of my time; every time I ran a d20 game the drudge work involved was demoralising. That meant that I'd put off the prep work to the last minute, and it would be rushed and unfinished, which didn't work as well and put me in a negative mood for the game. I actually abandoned the last two d20 campaigns I ran (Lone Wolf and Monte Cook's World of Darkness) because they were actually souring me on GMing. LL has been really liberating.
The simplicity in character options is something I had reservations about, because the group is used to games where you get to "build" your character with certain strengths. I think some of the group still aren't sold on LL's build-free way of handling characters (where one Fighter is much like another game-mechanically), but they seem to be giving it a fair shot. Hopefully the old-school ethos that characters are defined by what they do in the game and not by their powers will catch on as the campaign develops.
THAC0 & AC: Some of the group are still tripping over the THAC0 and to-hit rolls with lower-is-better AC. I even got a little confused once or twice myself. We're getting used to it again though, which isn't surprising considering the group's collective long experience with AD&D 2e.
The controversial THAC0 Defence house rule has mostly worked fine, but there was one weird moment in the first adventure where a PC fought an enemy one-on-one, which worked out as the player rolling the same d20 himself back and forward. Maybe THAC0 Defence needs a duelling exception to stop that happening, where single combat reverts to the standard system where the DM rolls attacks.
Lethality: We had one PC death in the first adventure. The second adventure was going well up till the very end, when a single bad decision got half the party killed. Not an ideal ending, but I think the group would have to admit that it was player error, and there didn't seem to be any bad feelings over it. I think the fact that it wasn't just a single PC loss this time made it seem less personal.
An unexpected benefit of the lethality comes from the fact that the players have been taking hirelings along with them to improve their odds. Having NPCs travelling with the group has opened up a little more roleplay. I think it's also helped the players feel less discouraged about the low power levels of their characters (compared to later editions), when they can see that the NPC adventurers don't do any better.
Sandboxing: I'm introducing sandboxing in small steps, because the group (me included) is used to linear sequences of adventures. So the first adventure was a straightforward "this is your mission, go and do it" thing to ease into the new game. After that, I gave the group several different adventure hooks to choose from. That drew comments about the videogamey feel, and the players approached it exactly like in a Bioware RPG; they also had a brief bit of analysis paralysis from those choices, but that didn't last long. When the PCs get more capable (round about when they graduate from the "Basic" into the "Expert" level range) I'll try opening things up further. When they start to outgrow the small town they're based in, I'll sound them out for where they want to go next, and build from there.
Experience for gold: I was surprised at how fiddly this turned out in actual play. I'm used to giving out mission-based XP, which is a simple case of picking a number to award, but gold-plus-monster XP needs a fair bit of calculation, at a point in the session where I'm usually finished GMing. The other thing that caught me off guard about it was that it makes the amount of treasure in the adventure very important for the experience award as well. In particular, the first adventure I ran was very cash-poor (which I hadn't thought to consider in advance) so I ended up multiplying the XP ad hoc, and even then it wasn't a whole lot when split between the PCs and the hirelings.
Description Over Dice: One of the biggest changes in this campaign is having the players' interaction with their environment be mostly based on the exact descriptions of their actions, instead of having skills like Perception or Search. This bit of the game style is still a work in progress for all of us to adjust to. I still instinctively feel like calling for dice rolls when it's not necessary, and I need to prompt the players occasionally to be specific about their actions. It's only to be expected that this will take time, considering how long we've been playing regular-style games but we've only had two adventures with LL.