To Roll or Not to Roll…

I was thinking about his recently when discussing what to do as a DM when you players roll poorly and you don’t want them all to die. What it got me thinking about specifically is that a lot of newer or younger DM’s seem to make their players roll for everything.

There are definitely times when you should not roll dice, or roll for something different than determining if something happens.

When Not to Roll

You should never roll to determine if something critical to the story happens. You may roll to determine how it happens. If the players must find the letter in the office, never have them roll to find it, have them roll to see how they find it. For example, if they roll a critical success, well whoever entered the room first found the letter sitting on the desk, easy as you please. Or you may choose to narrate that that PC found it in some clever way to make him feel smart. If they succeed normally, they find it after a short search in a desk drawer. If they fail, it takes them a long time to find it taped to the underside of the desk drawer. If they critically fail, they spent hours tearing the room apart, dumping all the shelves out, pulling all the books off the shelves, etc, only to find it sitting with all the other mail in the “outgoing” box, hidden in plain sight.

But the key here, if you make them roll at all, in a situation where they must find the letter, and how long it takes doesn’t actually matter (there are no guards to discover them or whatever), roll once to determine the narrated flavor of the scene.

You also ought not to roll when the players can succeed at something by simply trying again and there’s no penalty for failure. You’re just wasting everybody’s time and frustrating the players. Now, if there is a risk, rolling is fine. Sure, they may be able to climb over the wall eventually, but ever failure risks injury. Or they may take too long and get caught by the guards as they try to pick a lock. But if the only penalty for failure is just rolling again, don’t bother rolling in the first place.

You should also not roll for something that the PC really ought to know or be able to do, especially if there’s no risk or it’s not a stressful situation. If a PC wants to get up on a table, don’t bother rolling. If a PC is drunk as a skunk, go ahead with that Dex check to see if he stumbles. Don’t make a soldier roll to recognize ranks within his own army, but go ahead if he’s trying to read rank from another army. Don’t make a soldier roll to clean and oil his weapon, but go ahead if he’s trying to clear a jam while under fire during an ambush. Don’t make the bard roll remember the words for his favorite song, but go ahead and roll if he’s got to give the performance of his life to the goblin king or get beheaded. A Druid should not have to roll to recognize a common plant in his own environment, but ought to to determine if a strange plant in a different environment might be safe to eat.

Rolling too often and for the wrong things risks the story going off the rails in weird and frustrating ways. It can also make players feel like their PCs are getting cheated. If the barbarian wants to lift the heavy thing, I’d probably just let him. Strength is what Barbs do. If the scrawny wizard wants to try, he’s got to roll. But if you make them both roll, you may run into the weird scenario where Muscles McBrawny can’t lift a 50 pound sack, but the 40 pound Gnomish wizard can. If you, the DM, aren’t fast enough on your feet to narrate the failure properly, such as the Barb throwing his back out, or slipping and dropping the sack, then the whole situation just feels very dissonant.

Rerolling

Now, if success is not assured nor vital to the story, and rerolling presents a risk, go ahead and make players roll. But if all it takes is spamming rerolls to succeed, don’t bother rolling at all.

If they don’t have to find that letter mentioned above, but it sure would make their lives easier, make them roll. But if they fail, don’t just let them spam rerolls until they succeed. Hang a sword of Damocles over their heads. Every time they reroll they risk the guards finding them in the Lord’s study, or what have you.

In order for my players to reroll, they have to do something different. This also depends on context. I try to remember to always ask players how they are doing something before having them roll. But then, I do this for almost everything. My players know their role is to tell me what they want to do and I’ll tell them what to roll. Don’t tell me “I roll search” and roll, because then you’re only rolling once, pass or fail. Tell me “Ivar gives the room a once over, looking for clues”. If that fails, now Ivar has to do something else. “Okay, Ivar is going to search the desk.” How? He already gave the room a once over. “Okay, Ivar is going to go through each drawer, inspect the contents, and then check to see if anything is taped under the drawer.” And that is how you get a reroll. The PC is doing something meaningfully different from what failed the first time.

If another PC wants to also give the room a once over, that I’ll most likely allow, but once the desk has been searched in detail that option is off the table.

Unless…

“I’d like to have Athelstan search the desk in detail too.” No, Ivar already tried that. “Yes, but Ivar is a barbarian. Athelstan is a wizard, he may be able to recognize a formula or something that Ivar can’t.” And depending on just what they’re searching for, if that might make a meaningful difference, I’d allow it.

Of course, taking that much time almost assures the guards will find them on their routine patrols…

Rolls that Matter

Many game systems, especially those that go for “realism” can be overly crunchy with far too many die rolls that don’t mean much. For a recent game I had to use the GURPS survival rules a lot, and this was the biggest stress test I’ve put on them in quite a while. I’d forgotten how much they kind of suck. My players liked it even less. They were rolling quite a bit to simulate hiking across varied terrain in all manner of weather conditions, but the rolls didn’t feel like they meant much.

I did the math, and in order to avoid freezing to death for 10 days of hiking, each player would have had to roll Heath to almost 500 time. That’s way too much. And GURPS has no native way to compress that.

I ended up rewriting most of the survival and hiking rules to reduce everything to one roll per day per relevant factor. One roll to see how far they hiked, one roll to see if they froze to death. Optional extra rolls if they tried to forage for food.

The goal of these new rules was to reduce the number of rolls to the minimum feasible, and to put as much control of the journey into the player’s hands as possible. Which means one bad roll could screw them for a whole day or more, but they were much happier with this situation even with the higher risk.

And that’s my advice here to other DMs. You may think it’s better to roll often, especially to mitigate some problem or failure, but fewer rolls that matter more may be the better way to go. Good stories are often about overcoming failure and bad luck. Always succeeding at everything often isn’t fun. We need that risk of failure to relish the reward of success.

Don’t cheat your players out of failure.

Bad Rolls, TPK, and the DM’s Screen

So what do you do if the players just can’t roll well to save their lives?

I actively dislike letting PCs die. I game for story, and if a PC dies that the end of her story. That said, I will also not save PCs by fiat because that doesn’t serve the story either.

I may cheat with my own rolls to ensure the story flows better, but player rolls stand as they are. I may then try to modify things behind the curtain, but again, I will not use fiat to save them. Without real risk of failure, many players, at least mine, find the game less fun.

Years ago, during a teaching game no less, a series of just horrific rolls lead to my first TPK. I had to assure the players afterwards that they did nothing wrong and that what happened was an anomaly. For some reason, everyone, including the veteran player I asked to join the game to help guide the players, all rolled terribly while I rolled very well. By the end of the game I was often outright ignoring my own actual roll to try to give them some chance, but they just couldn’t roll to save their lives.

So here’s an example of something I might do in a game. Let’s say Jo has her PC check for traps. Critical fail, the trap springs. Now they’ve been grinding this dungeon for a while and are kinda beat up. Regardless of the rules, I may tell Jo roll a reflex save. She critical fails that too. Nuts.

So I roll damage. Behind my screen. It’s high. “Oof. Hey Jo, what’s Athelstan’s HP at again?”

“Um…4.”

“I see. So Athelstan accidentally triggered the foot trap while checking for it, and got shanked through the hand and arm for 5 points.”

Now here’s the secret. It doesn’t matter what my actual damage roll was. Because the players aren’t doing anything stupid, they’re just having bad luck. And I know Mark’s PC healer is within a few feet of Jo’s PC and has at least one health potion/heal spell/medical kit left, which will provide at least 1hp back. So if the party does anything useful to help Jo’s PC, he’ll be stable and not actively dying anymore, even though he’s in rough shape.

Some of my astute players may notice that I asked for Athelstan’s health before stating what the damage was. Most players will miss it. And I find tracking PC health is often more of a chore than it’s worth, but if you really want to keep everything behind the curtain, you may want to do that.

Another thing you can do, which I learned after that TPK, is that you can “fail” when necessary too and give the players a bit of luck. In a different game, the player’s ambush was going poorly. Not a bad plan, they just weren’t rolling well.

I realized that if things continued as they were, it was going to be another TPK. So I rolled. “Fuck!”

It didn’t matter what I rolled. The machinegunner had just “critically failed” and now the MG was jammed. It took three turns to clear it, but by then, with the most casualty-producing weapon out of play, the PCs were able to reverse the momentum of the fight.

So a bit of “luck” for them was able to turn things around. Now, you can’t “fail” every roll, but do it at the right time and they most likely won’t suspect a thing and you can prevent a TPK.

In Conclusion…

Don’t roll dice when the only penalty for failure is just a reroll, when the PC really should know or be able to do the thing, or when something must happen for the story to continue. And if a player insists on rolling, or you want to mask that success was pre-ordained, use the roll to determine the flavor of the event, not the actual outcome.

Part of this essay is copied from my post https://www.quora.com/What-can-a-GM-do-if-player-rolls-are-just-terrible-in-a-tabletop-RPG-GM-techniques-balance-narration-TPK-luck/answer/Murphy-Barrett