Deconstructing My Bad Time
Let's accept the core concepts of Shadowdark. Darkness matters. It's designed to be quick and streamlined. It needs to feel like D&D. All this I'm good with.
1) A player needs to trust that they can successfully cast a spell at least one time. (Fixed in houserules.)
2) Die rolling ability scores as presented takes too long to produce an acceptable character. The 3d6 method of the past does not work in "d20 + modifier" systems. (It worked in old-school games because ability scores weren't as tied to your character's role in the party. It might've granted a slight edge to attack or damage, an XP bonus, or an extra language slot.) Shadowdark should either have a Standard Array and/or present die rolling methods that have a statistical likelihood to create characters that aren't going to be trashed (per the rules).
3) Dungeon turns and constant Initiative are immersion breaking and take too long.
4) HP and damage are out of whack. Yes, this is a compatibility issue with old D&D, but it should change. 1HD monsters shouldn't have a 50/50 chance to kill a PC with one hit in a dungeon adventure where you face swarms of them AND you can't reasonably avoid combat. (If you open a door to a 30 X 30 room and see the monsters, they see you, you can't avoid them, you can't run away because they have the same - or better - speed.)
5) The DCs of Shadowdark are mathematically wrong if players should have a reasonable chance of success for mundane tasks. An Easy DC 9 is a 60% chance of success. You only have a 37.5% chance to have rolled 12 or higher on your 3d6 ability score. So there's a slight chance that you have better than a 60% chance of success to achieve an "Easy" success. If we look at a lowly kobold's Armor Class [13] as a DC skill check, it's on the high end of a Normal check. You have a less than average chance to hit, even assuming you might have a +1 or +2 to hit.
6) Levelling up is disappointing. You get a handful of HP (average is 2-4) that don't really matter or add to survivability in a substantial way. You might get to roll on a chart that could make you 5% better at something you could already do. Assuming you're not a spellcaster, you don't unlock special capstone abilities at higher levels. (Extra attacks for a fighter, for example.)
Sorry to hear that none of your DMs knew the way to make it work. :/
1. That perfectly cromulent house rule does fix this issue, though honestly I don't think it's needed. Priests start with two spells and wizards start with three. In general the odds that you never get a spell off are comparable to or lower than you cast your one (or zero) spell for the day in B/X and the enemy makes their save. Luck tokens also substantially mitigate spell failure. Your third level Wizard knew 5 spells, had better than 50/50 odds to cast, and on average talent rolls (or if he just chose Magic Missile as one) he probably had Advantage to cast at least one of them.
2. This is a reasonably common complaint in 3d6 down the line systems. I had one player in my longterm 5 Torches Deep game need to roll something like 4 or 5 sets because he just rolled terribly a few times. My usual fix is the "mirror/flip" option. Have players roll 3d6 down the line, but they may optionally subtract every score, in order, from 21. This completely eliminates ever needing to roll a second set. It does inflate the average a bit, makes very low scores rarer and high ones substantially more common, but I don't mind that.
3) Yeah, strictly adhering to turn order when it slows the game is bad GMing. The point is to make sure the spotlight is shared and it's not just one or two talkative people constantly driving.
4. I agree with your two caveats. Which means the adventure shouldn't have swarms of monsters if the PCs have no alternative ways to deal with them, and the GM should be using the reaction rules (as Kelsey models in her APs). I can also get behind max HP at first level to give a little padding, but the death rules for SD are more forgiving than, say, baseline B/X, so going to zero HP also isn't automatic death.
5. Mundane tasks you shouldn't normally be rolling for.
SD GM quickstart page 7:
WHAT TO DO
Describe what the characters newly perceive, and then let the players respond to that with actions. Call for checks when those actions require skill. Then start again from the top!
SD GM Quickstart page 12:
CHECKS
The characters automatically succeed at what they are trained to do. Only use checks when there is time pressure and failure has dire consequences.
6. This is a matter of taste, and totally understandable if it's not to yours. Doing it this way deliberately keeps the power curve flatter and characters fragile enough at high levels that fear remains, and they can't be as cavalier about damage and danger as higher level PCs are in editions with more HP inflation.
Retreater’s experience of Shadowdark reminds me of something I read on The Alexandrian blog. Justin Alexander ran an OD&D Caverns of Thracia campaign in order to test out how the game played RAW. The 3d6-down-the-line method produced such poor characters that after several TPKs the players no longer bothered to give them unique names, instead using designations like “Bob the Dwarf III”. They measured progress in terms of how much farther they could get into the dungeon before the inevitable TPK, almost like a Roguelike computer game. It sounded excruciating to me.
While CoT is a gold classic, it is not actually designed for a 1st level party, never mind one of only 5 characters like Justin's group had. A number of their problems with the module derive from playing it much more on hard mode than it was meant to be. 2nd or 3rd level would be more appropriate, and usually a group of twice that many characters (whether all PCs or a mix of PCs and hirelings).
That was also back in 2009 and Justin deliberately disregarded a number of OSR best practices which were already known even then, in the name of trying to make his experiment fit the test conditions he arbitrarily decided to employ, at the expense of fun and playability.
It’s very important for OSR style play (including Shadowdark) to emphasize that the point of play is to solve environmental puzzles. Rooms with monsters are simply another kind of puzzle to solve.
Any party that gets into head-to-head combat with a monster group without some kind of setup to give them advantage is in a fail state. Immediate combat is going to give a coin-flip chance of a TPK.
Combat is considered a fail state in OSR, it happens when your other approaches failed. You tried to instigate a melee between the arguing lizardmen but they saw through your manipulations and attack. Or you tried to sneak around them, but they heard something came over to investigate. When they questioned you they realized you came with ill intent to rob them and they attack you. Combat is dangerous and thus to be avoided. If it erupts, you did fail some other attempt in most cases.
There are whole classes dedicated to combat. A large part of the rules are about fighting or what happens when you fight.
This combat is a fail state stuff ignores the fact that combat is a large part of what players of virtually any generation want in a fantasy game. Not 100% of the time, and not all players, certainly, but I’d hazard to guess that most want combat at some point as part of dungeoning and dragoning or shadowdarking.
"Combat is a Fail State" is a classic maxim of the OSR which was never meant to be taken 100% literally. Gus wrote a good analysis and explication of it on his blog, along with six other major OSR maxims. Some combat is expected. Trying other approaches before fighting
where possible and using smart tactics to stack the odds in your favor are also expected.
Back in the aughts and the 2010’s, a decade ago now, there was a movement in older RPGs that I was part of - the “ Old School Renaissance ” ...
alldeadgenerations.blogspot.com