137 thoughts on “Acting on the Edge of Your Seat”

1. My group started to use the Genesys system a few months ago. I was having some issues with getting the DnD game together on a regular schedule, so I used Genesys as a fill in for the players that could make regular sessions. We’ve been having a great time with it and I think the dice have a lot to do with it. With Dnd, once you have advantage, there’s no further incentive to get more unless the GM adds a disadvantage. With Genesys, there is reason to further define a scene for boosts or to reduce difficulty.
   Each action does take slightly longer than a standard DnD action, but it has more payoff than, “Ok, you hit”. I found that the scenes are more developed as a result. Also, having the extra options for multiple successes and advantages gives more options for resolution. Just my experience so far, but I’ve really enjoyed playing Genesys.

   Reply

   • I’m not interested in another big debate about how great dice pool mechanics are for storytelling and all the wonderful options they create. It’s already been done to death on this site. This article is about pacing and designing systems that tell you the second the dice hit the table whether you succeeded or not. Genesys requires too much processing of the dice pool to fit that scheme.

     Reply

     • Sorry, I was trying to focus more on the immediate knowing that it succeeded or not more than the dice pool itself and that there is incentive to add to the story to get the added dice. I enjoy the way the dice are handled, but I would prefer a more simplified version that doesn’t use specific dice. More of a, “You have a small advantage. Add a d4 to the roll” instead of a boost dice.

       Reply

       • Alternity kind of did that, where modifiers where translated to other dice, and then you added all dice together for the final result. And depending on how well you rolled, you also knew if you had an ordinary/good/amazing success.

         Reply

2. I agree it feels better to have a target number in mind before you roll, so you can just look and immediately see “success” (or “failure”). There’s a design tension between that and the derive for detail/complexity, though. And while I mostly agree with you about telling players target numbers up front, sometimes I think it feels good to have the players not know immediately how tough a target is. A round or two of “wow, how high IS this guy’s armor class, anyway?!” followed by a successful attack, and now they know (or have a pretty good guess, anyway) and can start adapting to the situation I think can provide a quick dramatic arc in the midst of combat.

   There’s a dice mechanic a guy came up with that has some very nice properties – every +3 to the difficulty halves the chance of success. That could be cutting success from 40% to 20%, or from 0.05% to 0.025%. So it’s great for combining modifiers like “Well, the target is twice as far away as normal, so that’s 4x harder to hit — that’s plus 6 to the target number. And it’s dark, we’ll call that another 50% penalty, that’s another plus 3, so we’re at +9. So you need to roll a 10 + 9 = 19 or higher to hit.” The math geek in me loves this mechanic. But the problem is that it’s a variant on an exploding dice system. So you can roll the dice, and not know whether you succeeded or not — more rolls may be required. Sometimes it may be obvious that you’ve failed you don’t need to roll more, but that’s like the worst of both worlds – you know instantly if you’ve failed (sometimes), but except for pretty easy tasks you usually don’t know if you’ve succeeded until later.

   Reply

   • Unless you are attempting something you have literally never seen done before, it is extremely unlikely that you don’t know how close you were to succeeding.

     If you’re swinging a sword at someone, you can see how nimbly they dodged your attack. You can tell how expertly they parried your strike. You would be aware that your swing clanged harmlessly off of their armor.

     If your player throws a d20 on the table and all you do is tell them that they missed, you have not given the player enough information to understand what their character is actually experiencing. So unless you want to give a 5 minute monologue for every action that you resolve, the easiest way to communicate to your player how well or how poorly their character is doing is to let them know the target number.

     You can do a little song and dance with them if you’d prefer – you can describe the opponent nimbly ducking or you can tell them they think they almost slipped through a gap in the opponent’s armor. But that gets old quickly, and the numerical guessing game is pointless when there’s very little a player can do to affect the actual number that shows up on the die.

     There’s almost nothing that can be done after the die is rolled, and those few things that can be done usually involve a complete re-roll, which has the same 20-number spread as every other attempt.

     RPGs are about making choices, and d20 games don’t offer players enough granular control over the dice for that to really count as decision-making. Even less so when they have to play “Mother May I” with the target numbers.

     Reply

3. Something else to consider for dice pools is when the player makes their decisions and when they roll the dice (or equivalent).

   Most games ask the players to make the decisions first, which affects the DC or equivalent, then they roll the dice and either pass or fail in their action. I completely take the point that this offers a good fast pace; but it can sometimes reduce decisions to an optimization problem – simply find the best modifiers to DC to give you the best chance of success.

   Another approach is to roll the dice then let players decide what to do with them (it sounds like FFG’s RPGs do this, though I’ve not played those myself). So, to use an example from the wargame Imperial Assault, if you roll a Surge, you might choose to spend it on Stunning your enemy, or maybe instead spend it to Focus yourself instead. It does slow the pacing as you note, but the resulting choices are qualitatively different (and arguably more meaningful, as choices) than just trying to optimize the most favorable DC.

   One could even imagine a game that used a fast-paced decisions-up-front mechanism for combat, but slower-paced decisions-made-after-rolling mechanism for navigation, exploration, or crafting – letting the type of action undertaken dictate the action resolution approach, at the cost of complexity (two or more action resolution mechanisms to learn). This would be analogous to the way the excellent Total War videogames let you plan your nation’s overall strategy in a turn-based strategic map, but battles are fought on a real-time tactical map. The right tool for the right job.

   Reply

   • Yeah except the game then becomes random chance first and choice second. My games are about choices first and then the consequences of those choices. Not rolling to see what choices are allowed. Honestly, the dice need to shut up until they are needed and then get involved as little as possible.

     Reply

     • Well, every roll is – in some sense – about what choices will be allowed. The success or failure of any action always constrains or enables the next set of available actions.

       The FFG RPGs take this too far, but I’d argue the “microchoices” of surges in the miniature games (Descent and Imperial Assault) significantly enhance the tension of a roll. In particular, they offer a chance to eke out a victory on the cusp of failure – a classical dramatic motif. Because surges are always beneficial to the roller, they never snatch away an initial victory, and since they’re small choices they’re usually fast choices.

       Reply

     • Consistent with your point that “people are really good at knowing what they like, really bad at explaining why, and terrible at guessing what they might like instead.” I believe you’ve articulated what I found unsatisfying about Dragon Age (and the generic AGE System) from Green Ronin. The stunts were exciting, but for many of them, the actions followed from the mechanics.

       Also, (and this may betray me as a long-time reader) this article immediately reminded me of “Thinking Critically” (https://theangrygm.com/ask-angry-thinking-critically/) and “It’s Okay to Have Damage Rolls” (https://theangrygm.com/ask-angry-its-okay-to-have-damage-rolls/).

       Reply

4. You’re not alone, I’ve always felt this delay was an issue too. My solution was to use an “extended advantage” system on 2d6, where the only math was taking two dice, and adding those to your modifier. All the situational bonuses are done by adding dice to the pool before the roll. I haven’t started play testing yet, but I’m hoping the max-4 operations should speed up the resolution time

   Reply

5. On the ‘revealing ACs and DCs’ – I agree but don’t think ‘revealing’ is the point, it’s about ‘concealing’ or otherwise. There is genuine gameplay value in being surprised by how hard or easy a monster is to hit. Wow, my 19 didn’t hit it? No, it ducks under your swipe with lightning speed. I hit on a 13? Yeah, it’s just a huge great tub of guts that laughs off the damage you do it. That’s cool, and interesting.
   But I guess what you’re taking aim at is the idea that it should be fundamentally hidden. Over the course of a fight, players homing in on an opponent’s AC has a nice in-game correlate in the PCs learning how to fight this monster and what it takes to hit it – similar story for DCs. What is stupid is the GM shouting down players saying ‘huh, must be an AC of 20 or 21’ or similar. Pointless.
   See also: players asking each other ‘how many hit points are you down to?’ If you’re a GM who cracks down on that, you should ask yourself why.

   Reply

   • I give out AC, HP, attack bonus and DC, I highly reccomend giving it a shot. I can see your point, but AC and HP and DC represent real in-world qualities of things, and its safe to assume that expert adventurers can get a good idea of those things. A trained climber knows his rough chances of climbing a wall before he climbs it. A trained archer knows if he can resonably hit a target just by looking at it. The variables such as loose rocks, sudden gusts of wind or an unexpected juke are represented by the d20.

     Reply

     • There exist situations where concealed stats of that sort can serve a purpose, but it shouldn’t be the main method, and it depends on knowing your players. For example as depicted in the last panel of https://www.uptofourplayers.com/comic/heart-of-stone-page-28/ .

       15 is a really good damage roll in that system, and he was only shaken, it didn’t even wound him. That’s an “oh crud” moment right there. Most of the other depictions of rolls in that comic give the roll total “Vs. toughness X,” so it additionally stood out.

       Reply

6. This article reminds me of an idea I once playtested as a new DM, that I gave up as a bad idea then but now think may simply require some tweaking.

   In place of AC, monsters had armour that reduced incoming damage by a set amount. For instance, a hobgoblin reduced incoming damage by 5. This meant that an attack that did 1d6+2 damage had a 1 in 4 chance of hitting, with hit/miss and damage being the result of a single roll.

   While this admittedly doesn’t work for balancing faster, lighter attacks, it was interesting to see some players resorting to unusual tactics when their characters weren’t capable of hitting with their standard attack.

   As I type this, I suspect this may be a better ability for a boss monster rather than a central mechanic.

   Even if you all hate it, I hope this provides food for thought.

   Reply

7. I’m loving the “psychology first” approach to game design! Another thing to consider would be degree of success being built into the roll… If I know that I need to roll a 12, and the die comes up 18–not only do I immediately know that I succeeded, I instinctively know I succeeded really well. Or rolled a 3 and failed badly.

   Now, not only do you have the quick emotional pacing to the outcome–you have a direct link from “hold breath” to the degree of consequence… Which adds to the anticipation of the gm’s reveal, and sets up the next pacing curve as those consequences incite the next choice.

   …or even to keep the emotional pacing up through the damage roll. It’s a let down to roll high to hit and low to damage… But, simple example, if you have +1 to damage for every point that you exceed the target #, then rolling high guarantees that you can’t drain all the emotional energy during the damage roll.

   Fascinating to follow the quick moves in the player’s consciousness as the action unfolds, because the pattern doesn’t let up until the scene is resolved. The immediate payoff of the roll is integrated in the player’s mind, but then instantly followed by tension/anticipation as the degree of the roll creates the next bit of uncertainty. They’re never off the hook until the climax, and the action finally falls.

   Reply

8. Something else to consider is how this hypothetical system would affect group DC checks, like evading a fireball. With a party that has varied Dex mods the number to roll that the DM would have to declare could be different for each player, adding time to the deliberation phase.

   Not naysaying the idea, just offering further design considerations. I’m with you on the need to manage the pacing from a mechancal standpoint. At the very least it’s a fun thought experiment that I’ll enjoy pondering for awhile. Perhaps their are assumptions that are limiting the ability to come to an answer. I wonder if we would need to rethink the physical dice themselves as well, perhaps the ztandsta D20 style dice won’t work in this case

   Reply

   • If the fireball were one attack roll against a target defense of those caught in be blast, though, it’d still be quick resoluton

     Reply

     • But then it becomes very binary if the targets have similar defense values.
       Imagine using fireball on a large group and killing either all of them or none of them, and never anything inbetween. It just feels wrong.
       Same problem with passive checks. All or nothing is less interesting, and if anyone does succeed it’s always the same guy every time.

       Reply

9. Me and my roleplaying buddies are taking turns at DMing (about 4 sessions each) and there’s a bunch of stuff I want to try, including just plain revealing target numbers (especially AC). I considered pre-substracting the attack bonuses for the same reasons you stated. There are a lot more stuff that I want to try to make the players seem more competent too, which may or may not be wise to do, but hey, if I’m going to break it, I might as well go all the way and see what works and what doesn’t.

   Reply

10. If we don’t ask someone to roll dice for an action that’s basically guaranteed, like “roll to climb stairs,” what benefit are we getting out of asking a character who’s an apparent fighting expert to roll dice to determine whether or not they hit something? Is it really so difficult for a person to use a stabby thing to make contact with a house-sized monster that we need to roll to determine whether they succeed?

    What’s the reason we wouldn’t: make damage more random, (there’s a chance to only deal 1 damage); reduce damage dealt to high-level creatures via an armor stat or similar; and throw out to-hit rolling?

    Reply

    • If the house sized monster is covered in armor plates, then it isn’t a matter of ‘to hit’ exactly, but can you effectively hit it and hurt it when it is trying to eat your wizard’s face? I try to make my narration reflect ‘it is agile’ versus ‘it is tiny’ versus ‘it is armored to all heck, and your hits are glancing off it’, even if all of those are represented in game as ‘did you make AC 16? No? You miss’.

      It also means that if the characters are taking a melee weapon to a static target with no real reason to hurry, I’m not asking them to roll for hits. (Or damage unless it is time sensitive — if you can damage an object, and do enough damage that you can break it before a reasonable person would give up, I’m going to say you broke it eventually.)

      Reply

      • Thanks for the reply. The fact that all of those different “flavors” of missing a target aren’t reflected in the mechanics feels like a red flag.

        I’m also still not seeing what advantage we’re getting from possibly missing, as opposed to possibly dealing negligible damage.

        Reply

        • Let’s be clear here, missing doesn’t feel good. Failing doesn’t feel good. But this is also one of those cases where players – and even GMs – can’t be trusted to offer valid feedback because no one likes to miss or lose. It will always feel bad. So will dealing negligible damage. But the advantage to hit rolls offer is that they are in-line with every other mechanic in the game. The game is a game of choice, resolve the action, live with the results. Which is why universal action resolution mechanics became so popular. Why do the same thing seventeen different ways. Damage rolls are the odd-thing-out. It is the only time in the game we follow a choice to act and an action adjudication with another, different roll based on completely different rules and dice, to see if the success was actually a success enough. THAT’S why, if you actually want to economize, DAMAGE ROLLS need to go.

          Reply

          • Agreed we want at most one roll per outcome, and most outcomes are pass/fail so it’s tempting to make all rolls pass/fail.

            If you want, you can just have a pass/fail “fighting” skill that you’d use just like climb or like persuade in a multi-stage social encounter. The player can describe an approach like, “I hide behind the drapes and burst out upon the orcs as they settle in for tea, slashing at them with my sword!” and they roll the die and the orcs are cut to ribbons or not.

            But if you want a tactical war-game inside your RPG and as part of that war-game, you build actions with integer-valued outcome (e.g. an attack, the outcome of which is to do X damage), you’ll have to accept some complexity and departures from your non-tactical non-war RPG layer. And you are better off having the uncertainty as close as possible to the final output of the action – in this case the damage roll.

            Consider that having it in a to-hit roll doesn’t actually fix the problem of a successful hit being essentially negligible in terms of damage – it just means it’s not a matter of chance if it is.

            And if HP is disclosed, you wind up with Alice knowing “if I hit, I’ll do 10 damage, if Bob hits, he’ll do 14; it has 13 HP left, so my hit is pointless if Bob is going to hit anyways, which he probably will”. Alice’s action has no impact on the probability of success of Bob’s action. But if Alice instead does 0-X damage and Bob does 0-Y, a good roll from Alice means Bob doesn’t have to roll as high. Even if he does, and Alice knows rationally that her hit did not help, Alice is still more likely to feel like she contributed to the team success, because it *could* have helped.

            You can get rid of the dilemma by avoiding integer-valued outcomes and boiling every outcome down to pass/fail (usually by saying “it takes N hits to eliminate an enemy”), which has other design trade-offs.

            IMO the consistent design principle to follow is that you want the die roll as close to the outcome as possible, not that you want all rolls to be pass/fail. If you have a continuously-valued outcome, you’re usually better off avoiding a bimodal distribution of that outcome (e.g. X% chance of 0 damage, 100-X% chance of Y damage).

            But we agree that to-hit roll followed by damage-roll is the worst of both worlds, and if you’re just tweaking d20 it’s way harder to eliminate the to-hit roll than the damage roll, due to the pass/fail mechanic of AC.

            Reply

            • Wow. You certainly wrote a lot of words. And we’re way beyond the scope of this article. And also, this is a false dilemma. These are the ways various RPGs have handled expanding the scope of the tactical game, not the ONLY ways. You need more imagination.

          • Hm, sorry to hear you say that.

            Well, still interested to see what you propose in the future.

            Reply

            • Eh. I’m sorry. I shouldn’t shut the conversation down. Carry on. That wasn’t cool of me.

          • I agree that in games with a universal resolution mechanic it is the damage rolls that need to go. Static damage values provide plenty of tactical options if you look at combat outside of the context of a single combat round against one target. Going back to Joshua’s example: if Alice hits and Bob misses then the enemy will be able to be killed by her in subsequent turns. She could attack a different target, or try to achieve another objective if the combat has one. Alice could perform an action that would improve Bob’s chance of hitting the enemy (an Aid action or disabling the target in some way). This actually lets the players use more tactical thinking than to just attack a target as much and as fast as possible to maximize the damage done.

            Alternately the players may know the AC of a target but not its exact total HP. This would still satisfy the pacing curve since Alice knows the outcome after a roll is a hit or miss, but she doesn’t know if the consequences include the target being killed until told by the GM.

            Reply

          • “But the advantage to hit rolls offer is that they are in-line with every other mechanic in the game. The game is a game of choice, resolve the action, live with the results.”

            I agree with the second sentence, but the game doesn’t have a single universal action resolution mechanic, it has two: saving throws. We have action initiator rolls against target (attack rolls), and target rolls to defend against action initiator (saving throws). What if saving throws was the primary action resolution mechanic and the onus was thus on the defender to get out of the way? If dodging attacks as a dex save was a reaction, and armor just provided damage reduction it might set up some interesting choices while defending, while also removing the to-hit roll, and also not inventing new core mechanics. It would certainly change the pacing curve.

            I imagine it would feel more like a fencing match – I present my attack, you have to do something about it or else be hit, then you do something back, I have to do something about it or else be hit, and so on. Parrying could be an actual thing as a weapon ability – all shields would have it, daggers probably, etc., and it just gives you an extra reaction for the sole purpose of “dodging” an attack. I dunno…just spitballing. The whole question is certainly an interesting design challenge.

            Reply

    • To be fair, an expert fighter is not “just making contact” He is ducking milliseconds before the creatures tree sized club connects, jumping off it’s knee and stabbing between it’s third and forth rib all while avoiding getting stomped.

      Reply

    • I agree: Eliminate to-hit rolling. Designers can still get effects around armor vs evasion and all that by using damage reduction, two HP pools, or similar. You can differentiate “precise” characters vs heavy-hitters by tagging attacks as “precise” or not and having different forms of damage reduction or HP pools apply.

      The cost of having two-phase attack resolution is really high, and it can slow some games to a crawl as people swap dice or figure out what next set of mods apply to their damage roll.

      Reply

      • Which can be eliminated if players just roll damage dice at the same time as the attack roll. If the attack hits, add your damage modifiers and go. If the attack misses, ignore the damage dice.

        Reply

        • That’s a good point, and would eliminate or at least front-load the dice swapping time. I don’t think it solves the mental overhead involved in applying two sets of modifiers.

          Since I don’t see a good reason to have a two-phase attack, to me it feels like just asking the players to hack around the design.

          Reply

        • That can crush player morale, rolling max damage along with a miss just feels horrible, like the dice are taunting you.

          Reply

        • Or a fixed amount of damage for most attacks, like most video games

          Reply

        • Did you not read the part of the article where Angry spells out the exact disadvantages of to-hit rolling, or do you just disagree with him?

          Reply

          • Huh. I missed that part. Where did Angry spell out the exact disadvantages of to-hit rolling specifically?

            Reply

          • *the exact disadvantages of to-hit rolls followed by damage rolls the way they’re implemented in DnD 5e

            “Honestly, even damage rolls suck for the same reason. Because after the moment of highest tension, there’s another die roll and that die roll could turn a lucky hit into a pointless graze that might as well have been a miss.”

            Can’t reply to Angry’s comment below, but there you go.

            Reply

            • You don’t think, apart from the pacing issue common to ALL resolution the way I’m saying here, that it’s actually the damage roll that I’m calling out? Not to-hit rolls?

        • I’ve always kind of liked the “damage is the amount by which you exceeded the target” mechanism. You’re not rolling to hit. You’re rolling to see if you do any damage to your foe this time.

          Reply

    • I’m not arguing with the mechanics you’re suggesting per se, but the way AC works does create some interesting situations. Some enemies, like zombies, have abysmal AC, but are difficult to kill because they have a lot of health. Other enemies might be heavily armored and difficult to hurt with simple attack roles, but they have only average dex, making attacks that require saves, such as acid splash or fireball, more viable. High AC also makes securing advantage more important.

      The contrast between these different types of enemies can help to vary the effective tactics from battle to battle, giving the PCs a reason to use different techniques than they might otherwise.

      Also, I would argue that requiring a roll to hit is very believable. Sure, anyone can hit a zombie; that’s why they have 8 AC. But how about a deft and agile assassin, equally if not more experienced than the PCs? What about the naturally armored dragon, whose thick scales easily deflect any attack that doesn’t come from just the right angle, whose snapping jaws and beating wings make it difficult just to get within swords reach of?

      Reply

      • “The contrast between these different types of enemies can help to vary the effective tactics from battle to battle, giving the PCs a reason to use different techniques than they might otherwise”

        I see what you’re saying in an ideal hypothetical system, but if we’re talking about 5e, would you elaborate on what specific decisions a given fighter PC would make when confronted with a zombie vs a dragon?

        Rolling to hit is realistic in some scenarios, yes, but A) what’s the ratio of slowish monsters to not slow ones and B) are we giving up too much for the sake of realism?

        Reply

      • I agree with what has been said elsewhere that there’s ways of forcing the players to change their tactics against the enemies you describe without requiring a to-hit roll. People have been fine before with having both the agility of an assassin and the armored scales of a dragon be modeled as AC, so why not model them both as different kinds of damage reduction instead. The damage reduction from the assassin’s agility is countered by attacks with the “Precise” property, and the damage reduction from the dragon scales is countered by attacks with the “Piercing” property The realism of being able to miss a target occurs from damage rolls that are too low to bypass the target’s damage reduction. The way an enemy’s defenses work is only important in that it is consistent and supports the action resolution mechanic.

        Reply

      • I think the way that HP tends to get represented/described could solve this issue for you. I tend/try to think of hit points not as how many cuts and bruises you can take, but it terms of how much experience/ability you have to afford blunders/take hits.

        In D&D I tend to think of it in terms of “take hits and continue fighting”, but that’s mostly because AC reflects your ability to Dodge/block and because the stat you add to HP is Constitution.

        Someone with more experience (levels) in fighting knows better how to mitigate the consequences of an enemy getting inside their defenses. A character doesn’t really get truly hit until they have been worn down and no longer have the stamina (HP) to deflect or roll with the hit. Or perhaps that’s just how long their experience helps them push their luck. Eventually it runs out. You can only parry/deflect/dodge so many attacks before one finally hits you right.

        That is why a level 5 fighter can take more than a level 1 fighter.

        Of course, even this breaks down in D&D 5E once you get really into the specifics, but my point is just that if you can shift your thinking away from HP = how many cuts I can sustain then it makes a lot more sense to do away with a to-hit roll, and to accept that its possible that the super-fast rogue can be hit almost every round. The rogue’s speed reduces the difference in the target number vs the roll, for example, if you wanted to stick with AC vs attack roll. If you used the difference as damage then you still get a different feel. When you hit a fast creature your attacks would likely do less damage. When you hit a heavily armored creature you do less damage. When you hit a slow creature with lots of health its easy to do more damage, they just shrug it off easier.

        Reply

    • Look at Into the Odd. It has armor as damage reduction and eliminates rolling to hit (rolling for damage instead). By all accounts combat, is brutal, quick, unbelievably tense, and best avoided. That is one among several good reasons to give the game a try. I think you can get the pdf for like five bucks or free.

      I’m slowly moving toward something like this myself, because I run for kids and all the rounds without hits really cut into their patience.

      Reply

    • Because climbing the stairs and trying to hit someone in combat aren’t comparable actions. The equivalent comparisons are climbing the stairs vs. trying to hit a stationary target outside of combat. You wouldn’t roll for either of those. If you were trying to climb a set of slick stairs on a ship tossing in a dangerous storm, or trying to stab a person in combat you would roll to determine the outcome.

      There are ways to determine success and make damage random in a single roll.

      Reply

11. This article is another turning point moment in Angryverse lore when Angry takes a definitive stand on a right and wrong way to run your game. “Yeah. I’m going on record. If you think DCs and ACs should be secret, you’re just bad at this. End of story.”

    I’m glad this is now a definitie thing.

    The arguments for open DCs, ACs, and HPs has been made before by Angry. This is “player vs character knowledge.” The player needs to know the DC, AC, HP because the character knows things about the world that can only be expressed in these terms.

    Your highly expert martial knows how tough a foe is and how effective their attacks are against the foe. That concept in game is expressed in AC and HP. It’s what they learned in fighter school and in their career.

    Your highly expert rogue knows if a lock/trap is very much like or nothing like those they have open/disarmed before in thief school. That concept in game is DC.

    Knowing these things is necessary for the players to make valid decisions about their response to the description of the situation. If a DM denies this information, the player responses are made with assumptions. Each encounter might as well be an involuntary game of Russian roulette. All you have is agony between the squeeze of the trigger and the result…and your game slows down at what should be an exciting moment. Go dolphin or get out from behind the screen.

    Reply

    • I would say that the only caveat to all DCs and ACs should be open might be cases where the DM wants to keep the secrecy, like if an NPC bought a lie or you successfully detected a trap. Otherwise, I am in complete agreement. Lately I’ve been very open with what the target number my PCs have to hit is, and it’s wonderfully refreshing.

      Reply

      • Yeah. Of course. If the players aren’t going to know if they succeeded, they shouldn’t know what the DC is. But, those are the exception, not the rule. And, frankly, even then, GMs WAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAYYYYYYYYYYYYY overvalue that kind of secrecy. It’s part of the brain damage that is “fear of metagaming.”

        Reply

    • Well said!

      Reply

    • How do you handle something like searching for traps? The door to the treasure room is heavy oak with iron banding; any barbarian can tell it’s going to take a DC 20 check to open it, likewise any rogue can tell that fancy schmancy lock is going to take a DC 25 to pick. But what about the DC 15 pressure plate trap in front of the door, or the DC 22 needle trap hidden in the lock mechanism?
      In particular, if the rogue says “I don’t trust this door, I want to check it for traps before I try for that lock,” then rolls an 18 (with modifiers)?

      Reply

      • You mean all those decisions where the players aren’t making a decision so they shouldn’t be rolling? If you’re searching for traps, the tense moment isn’t when you’re searching, it’s after you decide there’s no traps and you open the door.

        Reply

      • Dealing with detecting:

        First, I use heavy foreshadowing that traps are in play because “traps suck.” Angry wrote that.

        Second, I know the passive perception of all characters, and TELL them there’s a trap if their passive exceeds the detect DC. Somebody usually has high wisdom…

        Third, if they check for a trap or to disable a trap and there’s no immediate pressure (combat, etc), they roll no worse than their passive score on the first attempt. After that, the success is automatic if they accept another die in the time pool.

        All else failed, I use the ‘click’ rule. You hear a click, and your named reaction either advantages, disadvantages, or has no effect on your save to avoid the effect of the trap.

        Not all traps do damage. Some are just annoying time wasters, such as a symbol of hypnosis, WIS save or incapacitated for 1d4 rounds. All it does is put another die in the time pool…which leads to plot-progressive random encounters.

        Reply

    • Counterpoint: there’s narrative value in the player not having an EXACT value for their DC, even if their character should know how tough the lock/climb/door/critter is. I know Angry’s opinion on Mercer, but in the games I’ve played/run, I find that breaking things down into rough quarters using Merceresque language – “the critter’s hale/injured/bloodied/on its last legs”, or “You’ve seen a lot of locks in your time, and this one looks trivial/straightforward/pretty tough/masterfully built” both grounds them in the narrative more than “it’s a DC 17 lock, gimme a lockpick roll if you want” or “the critter’s got 22 AC and 236HP” and gives them as much information as they want/need to have to make an informed decision.

      There’s usually class abilities, feats, or equivalent stuff meant to give you more precise/exact information, and also little reason a character would be able to hone in with scientific precision on exactly what they need. My table enjoys it more when things are just a little smudgy, though that’s probably just picking nits, really.

      Reply

      • I agree there is potential narrative value.

        But values must be placed against other values. I value speed of resolution over narration during the (non-boss) monster’s turn, AKA “waiting around to die.”

        I value narration over speed of resolution during the player’s turn. It’s a spotlight time for the players. So the fastest way to get to a(n) (in)glorious moment is knowing that they hit or missed, followed by narration of the effect as I simultaneously mark off monster hit points/flip over the dead, and/or call off the next initiative number.

        I like the narrative quarters for outside of combat DC shorthand, but the mechanic relies on numbers.

        Reply

        • I think Yurei has a point though. I have a question that might need clarification, following from Angry’s articles and Yureis point.

          If we want to reveal a target number to help clarify things before decisions for players to have information…when do we do so? In the last part of the article above, the example given was after the player had decided to attack, so it does not serve that purpose. It does however, serve the purpose that was explained which is quicker dice rolls. Which is cool.

          Reviewing the art of Narration article, I noticed from a quick skim – might be wrong – that target numbers aren’t in the scene setting examples provided.

          I can imagine in combats which are longer this might be softened because they are extended, so in the 2nd round you might know enough to make the decision. But in cases like NCNI encounters where one roll could do something, I’m having trouble imagining how to describe the situation. For example…

          DM says : “The assassin is about to make it across the bridge, and before he goes he slashes at a cart freeing a horse which barrels down the street. To calm this horse it requires a dc 23 animal handling check, which requires a roll of 9 for the druid.”

          I don’t know if that’s exactly what was being espoused so I just wanted to clarify. Could an example of providing DC’s before a decision be shown? Did I misunderstand?

          Reply

          • I expect it would be part of the negotiation phase, when the player suggests a course of action the DM can respond by saying how difficult that action would be to accomplish.

            So in your example:
            DM: “The assassin slashes at a cart, freeing a horse which barrels down the street. What do you do?”
            Druid: “Can I calm the horse down to stop it rampaging?”
            DM: “You can try, but that would be a DC 23 Wisdom check. With your skill in Animal Handling, you’d need to roll a 9 to succeed. Do you attempt it?”

            Reply

      • You could always compromise by using standard descriptors that your players understand.

        My players know that “bloodied” means half hp or less, and that “heavily injured” means one more hit will kill them.
        The PHB also provides common DC shorthand with “Easy”/”Medium”/”Hard” difficulties being set at 10/15/20, respectively.

        If your narration is consistent enough, you can give your players the exact values without explicitly stating the numbers every time.

        Reply

        • Yeah what I think about the first reply approach though is that for everything that the GM intends to be an approach or something with a default approach the players have to prompt the GM for the information, which I don’t like.

          “How about fighting the zombies” “AC 16” “How about that other thing?” “DC 20”

          I think the Easy/Average/Difficult division is one that definitely works considering narration and providing information.

          But it makes the players store that division as information themselves. It becomes more if the GM forecasts a consequence.
          If the effect is the players having to keep a list, I was thinking why not provide an actual list?

          Set the scene, provide a paper with information, players can review it while deliberating, then declare and roll. I feel I sort of like the idea of providing a list or paper with information for some scenes. I was thinking it helps players with things they might miss as well.

          Anybody else think providing an actual paper is over the top? Any thoughts? I think I sort of like the idea.

          Reply

12. As I read through this thinking about dice mechanics with satisfying pacing curves the one that kept constantly being brought to my mind was Monte Cooks Numenera mechanic. Its a D20 system where when the player decides to take an action the DC is announced based on the level of difficulty rated between 1 and 10 where the DC is equal to the level of difficulty times 3. Then the player can add skills and other situational benefits to lower that level of difficulty effectively lowering the DC. Once all the abilities have been applied the player rolls a single D20 and if its above the target you succeed.

    If you havent checked it out take a look because while the game as a whole has some other issues it does have a very satisfying dice mechanic and also does some cool things with its exp system.

    Reply

13. Angry, I think you have helped me decide why I dislike the FFG dice pool system so much. When my group first tried it I couldn’t understand why the dice mechanics felt so terrible to me. I thought at first that it was just the symbols, which are pretty non-intuitive and take some time to figure out. But after a few sessions when I had learned pretty well what the symbols were, for some reason I still hated the mechanics.

    I think it was the two problems you talk about here:

    First, it takes a long time to put together the pool, decide if you have anything that can adjust it further, make the roll, match up the cancelling symbols to see what you have left that’s good, and then decide what you are going to spend your net advantage on, or have the GM decide how to screw you over with bad results. It takes a LOT of time, even for players who don’t have to look at the tables to see what the dice symbols mean anymore.

    The second was the feeling of too much randomness. Yes I can add dice to my pools and try to mitigate bad dice with abilities, but it still felt a lot like the dice were in control, not me the player. Having a more experienced character that can throw more good dice at a problem helps with that, but it still FEELS very random to me, because it’s all dice rather than set targets you can aim for. Then you have the dice requiring the GM to come up with some reason why your attack failed, or why you didn’t manage to fix the hyperdrive, and in the worst cases you roll a despair but a lot of advantage, and the GM has to figure out the up side to a roll you failed in the worst way possible. The dice just have too much power.

    Some people in my group like the system, but I have a near-instinctual aversion to it. It doesn’t feel right to me for what should be a fast-moving space opera system where the heroes can have setbacks but always succeed when it counts.

    Reply

14. Incidentally, this is a good explanation why the writers of “Stranger Things”, rather than not understanding how D&D works, and “getting it wrong” in the first episode, actually displayed their understanding of narrative pacing, and took liberty with the actual rules of D&D in order to create a scene that had a lot more impact for the audience (as long as they could ignore their inner nerd voice screaming “That’s not how Fireball works!!!!”).

    Reply

    • Ditto for the Community D&D episode.

      “I’m going to throw my sword at the amulet.”

      “That’s a tough shot, you’ll need an 18 or higher to hit.”

      Reply

    • Because television writers ARE more likely to understand good narrative pacing than your average DM. There is a certain logic to this theory.

      Reply

15. As i was reading this I realized we’ve been doing math first in my D&D games basically forever, because i’ve always been open about target numbers and my players usually take a second to do the math and say “okay I need a twelve” before they roll, because it feels better.

    So math first feels better than math second, at least for my group, which is fairly good at math and usually remembers all their modifiers.

    Reply

16. First things first: I totally agree that the players should be able to recognize the result of their action after rolling the dice, and that the least amount of effort on it, the best. In fact, i’ve being toying with the “Conquest of Nerath” board game by WoTC, where you roll a die (a d6, d8, d10, d12 or d20), and you have a success on a roll of 6 or more. Your skill proficiency could indicate what kind of die you roll. The system is still bare-boned, but I think it has potential.

    On the rising tension problem, I think that there is another component on this equation: the weight of the outcome. In my experience, most GMs ask for rolls to actions with little-to-none consequences. They ask a climb test to go up a small cliff outside a combat, to search for something in a room that is not hidden etc. But this is just bad GMing. When this become a problem is when its ingrained in the rules.

    I will try to give an example: in D&D (let’s say, 5e), a 1st-level character attack has much more weight in to it. A single blow from a fighter can take it out a foe (let’s say, a goblin) thus making the rest of the combat easier for everyone else, so failure means that that enemy will continue to menace the party. Likewise, the monster attack can easily take out a rogue or a wizard, so everyone is afraid when the GM rolls its attack.

    But at higher-levels? A fighter needs to land 3-5 blows to take it out a foe, or the monster does so little damage that the party can wait for the combat to end to chug on potions. So each attack has less weight, and the action looses its tension.

    That’s, by the way, what got me out of 4e – you had to take a lot of actions with little weight each, specially after you used all your encounter powers and had to spam your at-wills.

    So I think this is something to considerate as well: what is the weight of the action? Is this consequence severe enought to deserve a dice roll? Is the result exciting, or its just bookkeeping?

    As aways, thanks for the inspiration, and sorry for the bad english

    Reply

    • Brilliant. Weight of action is an extremely good concept.

      This article got me rethinking damage – and I wonder if the weight of the action in combat could be pushed over to damage. (thinking while I type mode): Suppose (using D&D) that instead of dealing 1d8+Str or 2d6+Str, etc, a weapon did a straight 4, 6, or 8 (whatever the average is), and roll the damage die. Instead of doing additional damage, the roll could indicate a status effect. The target would vary by weapon….

      For example: roll to hit, then damage! As damage is set, the immediacy of the action and resolution work it out quickly (as mentioned in the article), roll the damage dice. On a ‘hit’ deal an extra 3 damage, or cool specialty items could maybe knock down an opponent, impose disadvantage, etc.

      Or maybe just use straight damage it that sounds too kooky. I might experiment with this next session.

      Reply

      • A simple way of doing this is every attack does average damage. That sword (1d8, average damage 4) wielded by a 16 Str fighter (+3 mod) does (4+3) 7 damage on a normal hit. For every point you roll over AC, you gain a +1 to damage, up to your weapon maximum (11 in this case). A critical can just add either an extra d8, or an extra 4pts, or an extra 8 points.

        Reply

        • Ooo! Have you play tested this idea?

          Reply

          • I haven’t play tested it. I literally just came up with it on the fly reading the above comments. But it is not a new idea. After I posted, I ended up googling it to see if anyone had play tested it. I think it would work well, as long as the DM/Players can add fast on their feet.

            Reply

        • Your calculation for an average amount of damage is a good idea, but the rest of the math seems a little complicated for smooth pacing.

          What if damage was a set amount (as you suggest) and when you roll to hit you also roll, say a d4? If the hit is an (immediate) success, you multiply the (average) damage by the amount shown on the d4. It’s easy to see if you hit and if the damage is high (and the math is easier).

          I’m usually not a fan of set damage for every attack because that becomes boring.

          Reply

    • Darn, someone already came up with my brilliant shower thought in an actual game? Haha jk. After reading through the article, it popped into my head that you could use different dice sizes to determine success chance. I went the opposite way. Small dice are better. Highest 3 numbers are a hit let’s say (so 17+ on a d20, down to 50% chance on a d6 with 4+). You start with a d12 or something. Disadvantage? You’re bumped to a d20. Relavent skill? D10. Night vision in low light combat on top of that skill? To the d8 you roll. Granted, this makes the d20 your enemy, and the lowly d6 your best friend which is a big flip in normal thought processes. the system could still have +1 through +3? Modifiers to die rolls too. There’s a whole mess of unanswered design stuff left like how does AC work, but that’s my food for thought.

      Reply

      • There are loads of mathematically interesting reasons to use “roll under” systems, including decreasing dice sizes to increase success chance while decreasing randomness, thus simulating the increased precision of being reliably better at things without changing the target numbers (1 is the lowest value on every size of dice, which is great for roll under systems and terribly frustrating for weapon damage systems).

        It’s just too bad the very concept of roll under systems is psychologically repulsive. 🙁

        Reply

17. Having the roll be the final step can be really good because it gives that moment of truth where everyone is on the edge of their seat (and not trying to do mental math while very excited and tense). The players know that the GM doesn’t know what the outcome will be; everyone finds out together.

    But it takes discipline. The math-before approach has a challenge of combining GM-side target numbers / modifiers and character-side modifiers. You can’t have the GM apply the character modifiers for the player because that starts to erase the mental link between their choices and their chances. And the players usually don’t know the target number for the task. So the player has to tell the GM the mods, the GM has to tell the player the target number; at some point someone adds this up to see what is needed, and then the die roll happens. But when put into practice at the table, players often find it faster to just roll and add their mods onto the roll and tell the GM the final result (one less communication).

    Dungeon World tries to streamline further by eliminating target numbers, so that the GM doesn’t really have to communicate much of anything to the player. The player rolls and the outcome (6-, 7-9, 10+) is obvious to everyone. The drawback is that this really limits the space for differentiating the difficulty of tasks, so you can’t allow too many modifiers or else there starts to be very little the players can’t do, and failure is unlikely and feels more like RNG screw when it happens. But if you limit the modifiers too much you get back to your basic coin-flip RPG.

    Exploding dice mechanics (used sparingly!) are great for rolls where every step of the roll is a committed result in and of itself. It’s exciting to go “12 damage! No wait, more! No, even more!!” They are terrible for rolls with pass/fail outcomes because the intermediate results are meaningless: they just delay discovery of whether the player passes or fail. A rule of thumb: if the mechanic isn’t essentially applying the outcome of the action twice (or more), don’t use exploding dice.

    I agree roll-to-hit then roll-for-damage sucks. Consider assigning attacks/weapons a fixed mean damage (if modifying an existing system with a damage roll) or better yet eliminating the to-hit roll and only having damage rolls (if making a homebrew system).

    Reply

18. I’m in the boat with all the players who want to know the “secret” numbers like AC and DC, but knows their GM won’t reveal them because of the “tension” the mystery adds.

    Traveller is a system that I think is at least trying to solve this, since it puts focus on public understanding of the difficulty system. Everyone at the table should know you need a 6 for a “Routine” task and a 12 for a “Very Hard” task or whatever, and the books mention that the GM should be telling the players that what they want to do is considered “Routine” or whatever.

    I also think that a side benefit of streamlining RPGs this is just speed. A three-session adventure could boil down to two or maybe even one if there’s less math in the way.

    Reply

19. Revealing ACs makes sense, but I don’t support your call to reveal DCs, at least until after the effects happen.

    I could support a case where there’s no reason to conceal an AC, at least after the first swing (characters can see the armor, and see how an opponent reacts, etc., so they could deduce the AC of a given opponent as long as something has happened). I support revealing DCs is fine for obvious tasks (“This is a DC 15 climb.” or “You examine the lock, and it looks pretty complex – i.e. DC 20 to pick.” or even “The portcullis is massive and made of thick iron. You’d need a 25 to lift it.”).

    However, revealing DCs for non-obvious tasks is suspect, unless you just don’t let your players roll for, e.g., searches. If I tried to tell my players “you can ask to make a search, but I will roll the dice, and you will never know what was rolled” I’d have some grumpy players – they like to roll dice. It took some effort to even train them to wait to roll for some kind of skill check until they described an intent and approach, instead of yelling “Perception check” and throwing a D20 at the dice tower (which can lead to a fascinating level of physical comedy, but that’s an aside). I think the process of declare search –> roll –> announce modified roll –> DM reveals what you’ve found is a good compromise. The players get to roll dice, and I don’t have to tell them “don’t bother rolling, there are no traps on this door.” or worse, “If you get a 15, you find the simple trap, but you need a 20 to spot the hidden needle trap inside the lock.”

    Reply

    • I don’t see an issue with revealing most DCs, save for very specific stuff like insight or perception checks, though really, I think there an argument could probably be made whether the PCs should make those checks at all, or if the DM should always make them secretly. Especially for insight where you don’t have to make a check at all to determine if someone is telling the truth.

      Reply

20. I recall playing a game (possibly Star Frontiers) many decades ago. One of the things I liked was that each weapon had a fixed maximum damage value, but the amount of damage an attack with that weapon did was related to how good of an attack you made. The game used percentile dice with low numbers being better (if I recall correctly), but a low value of 1-5 was always maximum damage, but you might get full damage from 1-10, then you might get 75% of the damage with a 11-30, 50% damage with a 31-50, 25% damage with a 51-70 and no damage above 70. The benefit was that you knew the target ranges up front and one percentile die roll gave you immediate knowledge of a hit and how much damage was done. You chose the weapon and actions that made sense (based on range, cover, etc.) and you knew how much potential damage you could do, but you had to perform well enough (die roll) to get the full effect. Deciding to drop prone, for example, had an impact on your target ranges for firing your rifle. The system then used appropriate armor (if any) to offset damage.

    Bringing that to a generic D20 system would be difficult, but not impossible. Bringing it to Pathfinder, 5e, etc. might fundamentally change the way we think about equipment and combat resolution, but as an example (leaving armor as adjusting AC instead of damage absorption), instead of a weapon doing 1D8+Str damage, it would do 8+Str maximum damage. Then when facing a monster where you had to roll a 9 or better to hit, if you rolled a perfect hit (20) you did 12 damage, an above average hit (17-19) you did 9 damage, an average hit (11-16) would be 6 damage and a poor hit (9-10) would do 3 damage. If we wanted to bring armor in as damage absorption (as opposed to AC), then it gets weirder because the DC to hit would be changed and the game would have to be recalculated for different armor types, natural armor, dexterity based defenses, etc.
    Of course there are all kinds of fiddly bits that would have to be worked out such as feats that change crit ranges, what we do with magical attacks, etc.

    As an aside, in my 13th Age game, I always let the players know the target AC (or PD or MD) before any die roll that they make. It works well and we avoid the “Does a 17 hit?” question, so the question mark of the attack becomes elation when they see the number that comes up a hit and they move right on to damage dice.

    Reply

21. Interesting musings for sure. I’m planning to run my next game in Savage Worlds, which uses a (mostly) static target number system with modifiers applied to the roll, and I’m toying with the idea of applying the modifiers to the target number instead just to get the math out of the way before rolling the dice.

    Of course, Savage Worlds also has exploding dice. I’ll be interested to see how the psychology of that works out at the table – I can see the case for the additional delay messing with the pacing, but I can also see the case for the extra time taken to roll an exploding die again just ratcheting up the tension and making it even more satisfying when the action is finally resolved.

    Intuitively I’m expecting the second to win out there (people tend to enjoy exploding dice in general) but we’ll see.

    Reply

22. One way to combine the narrative pacing of roll-under target systems, the ease of roll-over target systems, while maintaining the visibility of the player’s influence on the outcome, might be to have proficiencies tell you how they affect the DC instead of how they affect the roll. For example, if the rule was that an easy DC was always 8, Medium was always 10, hard was always 12, etc. and instead of proficiency giving you +2 on your roll, it decreased the difficulty of the task by one step.

    “I’m going to try to climb the cliff.”

    “Ok, it’s a hard climb, so you need a 12 or higher.”

    “I’m also trained in Athletics, so that should be Easy for me.”

    *roll*

    “11! That’s a success!”

    Ultimately, it is just the “GM does the math first” solution in different clothing. But if the difficulty scale is hard-coded and presented to the players in the player-facing part of the rules… I don’t know, I guess then you’d still get GMs who insist on not telling the players the difficulty of the task. But it seems like a step in the right direction.

    Reply

    • I think Numenera does something like that.

      Reply

23. “Yeah. I’m going on record. If you think DCs and ACs should be secret, you’re just bad at this. End of story.”

    Thank you! I have been saying that for years, but everytime I mention it online I get shouted down like some sort ranting lunatic

    Reply

    • Yeah, saying the same thing I’ve been saying for years and agreeing with a dude named The Angry GM. That’ll prove to the world that you’re no ranting lunatic.

      Reply

    • You need to follow up with, “You can run your game any wrong way you want, then.” That’ll set them right, those people who said something wrong on the internet!

      Don’t feed the trolls. You said it. You feel better about you. That’s sufficient.

      Ranting, yes. Lunatic, no.

      Reply

24. What about a dice pool system where there’s a binary “if 1 or more 6, then success” or similar, so that you instantly see if you have succeeded? (Like the Time Pool mechanic, really)

    Then you could modify due to skill and circumstance by adding or subtracting dice, or change the value you need to roll higher than, but you is still instantly know the result simply by observing at least one high enough result, or none.

    Reply

    • This.
      I strongly suspect that the best dice-rolling mechanic would more closely resemble World of Darkness than anything else. It’s not perfect by any means, but searching for any number over a set value is one of the fastest resolution methods that exists (most nWoD rolls only need 1 success, and even those that need more can cause tense moments of frantically counting all your dice for just one more success). It would be even faster if only the maximum value indicated a success, e.g. a 6 on d6, which incidentally is the easiest dice size to buy in bulk).

      Incidentally, I hated in 2e when you had to delay the dice roll just so you can invert all the numbers and apply them to values that they don’t narratively correlate to. There comes a point where front-loading the math before the roll isn’t worth the extra time/hassle it takes overall to do so.
      Plus, it’s just so psychologically pleasing for the dice roll to represent the strength/skill/luck being applied and for the DC to represent the difficulty to be overcome.

      Reply

25. We’ve been play Stars Without Numbers recently, and one of my favorite things about it is the way it does its saves.

    Combat is done with a compatible-with-D&D d20 system, except the saves are a predetermined “roll this number or higher to pass.” Effect happens, DM says which of the three saves to use, dice hits the the table.

    Having the Big Bad open the fight with a party-wide cone of cold, and the table responding with 4 fails and a pass feels better than a bunch of number and a pause while the DM checks them all.

    Reply

    • That is exactly why I approve of 5e returning to the old saving throws system.
      4e had good reason to believe that the one casting the spell should be the one who rolls the dice, but AoE effects just go so much smoother when the targets can all roll their saves simultaneously.

      Reply

26. I like the idea of rolling 1d20 to hit a target number. In particular, I like the idea of rolling below a number to hit, so for example the unmodified target is 5, you have a bonus of 4, so you need 9 or below and you have 9 in 20 odds of success? For especially difficult tasks, you would have negative targets, which are bad, but you could relabel those as “impossibility points”. Then the task is impossible unless your bonus is high enough, and your odds of success are your bonus minus the “impossibility points”. There is subtraction, but it happens in the “rising tension” part of the pacing curve.

    I was thinking about the difference between video games and RPGs, and I think part of it is that your odds of success are really determined long before you act, when you select your stats and gear. What if you had several different attacks with different odds to hit, but some of them are more costly, like in Dark Souls? E.g. you can either attack normally, or you can do a careful attack for better odds of hitting, but then you can’t attack next turn. Then the target number would be important in the “select an action” phase, not just the resolution phase. (A fighter in D&D, for example, has no reason to know his odds of hitting because he’s just going to attack anyway. In video games, you have several attacks available, and you know roughly how likely they each are to succeed, and you can choose between them in the moment of attacking).

    Reply

    • “What if you had several different attacks with different odds to hit, but some of them are more costly, like in Dark Souls? ”

      I agree with this. A choice of three basic attacks would add to the game. One would be the default, then the others would be a careful attack (more likely to hit with lower damage), while the last would be a wild swing (unlikely to hit, but massive damage potential).

      I saw this ruled on a magic item in 5e, where the weilder could choose to attack with disadvantage to reroll all 1’s and 2’s on damage, or to attack with advantage with the drawback of rerolling all 5’s and 6’s on damage (the weapon damage was d6 based).

      Naturally, if the user already had disadvantage, they couldn’t add disadvantage voluntarily.

      I’m a fan of this concept for all weapons, but would be tempted to add a modifier instead of the advantage/disadvantage, be that a fixed ±5 or a ±1d6 to the attack roll. There’s a thrill to the desperate gambit of a wild attack when backed into a corner, and pinging small amounts of damage on the big bad is much more satisfying than constant misses.

      Reply

      • Then why have the middle of the road one? I mean, it’s boring. Why not have just have a floating thingy that gets assigned to either attack or damage. Like a stance or something. And you could tie it into different effects for different weapons. Hell, why not have it float between attack, damage, and defense.

        It is always extremely dangerous to put a middle of the road non-choice between the actual choices because people are risk averse and lost averse and they will fixate on what they are losing whenever they choose any other option. So most people will just use their middle of the road attack. Look how rare it was for MOST – not all, I don’t want to hear your one anecdote about the player who was better about the tactics – for MOST players who took feats like Power Attack to actually use them.

        I am all for this, by the way. In fact, you’re stealing the idea from me. And if you were under NDA, I could prove it. But, that middle of the road thing is the best way to kill a meaningful choice.

        Reply

        • A fair point, well made. I think I was intending to have the two additional attacks be more situationally useful. As highlighted, this was not a wise, carefully thought out decision.

          Having never played pathfinder, I can assure you that I have never seen a player use the Power Attack ability. You were technically correct, the best kind of correct. In your experience, does it matter how extreme the options are before the middle of the road non-choice becomes a useful option?

          I genuinely didn’t realise I was parrotting your advice. In that case, I’m delighted to find that some of your material has actually lodged in my brain.

          Are you implying that I could get a sneak peak at the Angry RPG if I were argumentitive enough to prod and willing to sign a NDA? That’s tempting, but I’ll just wait and hope that the rest of the system is as interesting as stances and tension pools.

          Reply

        • “Look how rare it was for MOST – not all, I don’t want to hear your one anecdote about the player who was better about the tactics – for MOST players who took feats like Power Attack to actually use them.”

          I think there is quite a lot to say about player’s aversion to risk, and how to design good risk vs reward options alongside middle of the road ones.

          I totally get Power Attack doesn’t see play a lot. However, in my games, I see Barbarian’s Reckless Attack used quite a lot. Assuming this is not only true for my table (I would be curious to know), why does Reckless Attack works and not Power Attack?

          First element of answer is, I think, how meaningful the effect is. Power Attack gives a small damage bonus in exchange of a small attack penalty, so why bother? Reckless Attack, however, gives you Advantage on your attack roll, in exchange of giving advantage to all attacks against you until your next turn. That’s quite big of a reward for quite a substantial risk.

          I also think it has also something to do with the time frame in which risk and reward occur: trading later risk for immediate reward is probably more attractive the opposite. In the first case, the advantage is immediate and seems guaranteed, and the downside is hypothetical (if you kill the enemy first, he won’t benefit of Advantage on his next attack against you). In the later, you take guaranteed risk (a lower chance to hit) before an hypothetical reward (if you hit, you do bonus damage).

          Reply

      • My system of choice right now, Shadow of the Demon Lord, has a pretty simple way of doing what you’re talking about. Anyone can make a basic attack. Anyone can also make a more advanced attack, to knock the target prone, keep yourself guarded, disorient them, throw off their attack of opportunities etc etc, but you just apply negative modifiers for each additional effect the attack has. My players, for the most part, haven’t figured this part out yet, but the NPCs have been employing the advanced attacks.

        Reply

      • Some 5e feats have an attack level option (-5 to hit = +10 damage). Wonder if that couldn’t become the mechanic. It would nerf the feats, but there could be a corresponding buff (-5 gets +20 damage, -2 gets +10).

        Optional rule. A -1 penalty to hit adds +2 damage.

        Reply

        • I’ve come to realise that 5e’s power attack feats are not fit for purpose anyway, as they actually favor weak, light weapons over the slow, powerful weapons they supposedly specialise in.

          Flat penalty to hit penalises high-damage attacks.
          Flat bonus to damage and unlimited usage both favor multiple hits.
          It doesn’t help that the melee version also grants an occasional bonus attack (that also happens to be chance-on-hit, and therefore favors multiple hits), and the ranged version also removes the biggest penalty of hand crossbows (disadvantage beyond 30ft range).

          Limited-use (per round, or otherwise) or costly (action or other resource) bonuses to a single attack/hit would favor fewer, stronger hits.
          For an example, see Stealth giving Advantage for Sneak Attack (because mechanically, Rogues feel more like slow, heavy-hitters than great-weapon Barbarians do).

          An idea I’ve had for some time now, but been unsure how to introduce (Feat, Fighting Style, or baseline), is to spend an Attack (similar to how grapple/shove works in 5e action economy) to perform a “Backswing”, giving a massive bonus to your next attack with that weapon (I’m thinking Advantage + Double Damage, because Advantage alone would actually be a net loss over just attacking twice). This would go a long way to closing the gap between two-weapon fighting and heavy-weapon fighting (I know other people say the opposite, but in my experience dual-wielding is considerably superior due to the vast abundance of flat damage bonuses in 5e).

          Reply

27. Interesting stuff, thanks Angry.

    This thing of “deliberating” first by building a dicepool that you ascribe to Fantasy Flight Games is also the case with Burning Wheel, and is one of my favourite things about that game. I haven’t actually played it (other aspects of it really don’t appeal to me) but I’ve watched a lot of streams of it, and the resolution mechanic just strikes me as deliriously fun.

    Reply

    • Yeah, and as someone who has run a couple of different Burning Wheel systems, I’m going to tell you where it fails. It mixes up the deliberation and the decision in its conflict resolution. Most of the decisions about how players act and even what actions they can take are dictated by the conflict resolution system and the party’s deliberations through them. So, the players don’t actually choose their fictional actions until AFTER a bunch of deliberation based on the dice mechanics. The actions follow from the mechanics, not the other way around. Torchbearer was absolutely egregious about this. I don’t doubt, though, that it’s a fun system to WATCH other people play given the way it functioned. But playing and running it… just hurt. It just hurt.

      Reply

28. While most people could probably get used to any reasonable time to resolve an action, I think that people who enjoy immersion and fantasy as one of their core engagements really need a short loop on action and outcome. You’ve said yourself that time rolling dice is time not being immersed in the world and the characters. It comes down to who you’re creating the RPG for.

    For that reason, the only thing that appeals to me about dice pool systems is the possibility of side outcomes. For instance, if my wizard shoots a lightning bolt at a goblin, a d20 or d100 roll alone is fine to tell me if I hit it, but if I might also want to roll a d6 at the same time to see if that bolt also jumps to a nearby goblin. You could also tie a dice pool to negative outcomes, like the side dice could tell you your efforts to break down a door also alert enemies you bypassed a couple rooms ago.

    Reply

    • Your lightning bolt example reminds me of some magic weapons I saw a while back that had bonus effects occur if max damage was rolled.
      Even when the bonus effects were just extra damage, it still felt like a fun mechanic, especially if the bonus damage itself could also activate the bonus damage, causing theoretically infinite damage bursts of infinitely small probability. 🙂

      Reply

29. I think you already have the basis for your system in your time pool.

    To succeed, you only need to roll a one on any single die. Every action roll starts with the same number of dice; let’s say 4.

    The character’s actions, abilities, skills, equipment, etc. add or subtract dice from the pool.

    The task’s difficulty determines the size of the dice being rolled from d4 for a very easy task to a d12 for a very difficult task.

    Have something like advantage? You succeed on a one or two. Disadvantage? You need to roll a one on two dice.

    The system is simple, no math, results are instant.

    Reply

    • Ok, even I can see the math doesn’t work here. It’s elegence proved too attractive.

      Reply

30. I agree that generally speaking if players are rolling to hit, damage rolls should get the axe. You’re right.

    The way I see it, the problem with using the same mechanic to pick a lock and hit a guy is that you don’t have a scenario where 6-12 characters are all taking turns trying to pick each other’s locks for 2-5 rounds. At the macro level those are completely different use cases, so why WOULDN’T we benefit from doing those 2 different ways?

    Yes, universal mechanics are nice and you can brag to all your cool game design friends about how you managed to make the hitting stuff and the talking stuff and the skilling stuff all use the same die, but what does it add to the game aside from being mildly more approachable for the players’ first handful of sessions – and aren’t we giving up a lot to get there?

    Reply

31. Don’t variable damage rolls, along with variable hit points, act to “even out” the randomness of the combat system? That is, they act to dampen success, take the sting out of failure, and help to ensure that the side that you expect to win is the one that wins the combat?

    The side with the higher hit points won’t be as likely to be killed by a string of lucky to-hit rolls because you have to have lucky damage rolls too, and the side with the lower hit points isn’t going to win just because they got missed more often, because the one or two times they are hit the damage rolls might be on the high end. Having extra rolls acts to average things out a bit.

    If you take out the variable damage rolls, you have made to-hit rolls more important, and therefore made the outcome of the fight more dependent on getting good to-hit rolls and less dependent on having high hit points or high damage rolls.

    Reply

    • It seems to me like you’re trying to apply the statistics of an infinite set to what amounts to a small number of rolls. Most combats last 4-5 rounds, have 8-10 entities involved, and each entity makes 2-3 rolls per round for a usual total of 64-150 rolls (only half of which the players are making). In that small a set, any significant deviation from the standard is going to be felt. Furthermore, most D&D parties won’t get wiped out by a single encounter so long as it was build to be within the normal difficulty range.

      Reply

      • Smaller samples only INcrease the relative effect of random chance.
        With an infinite sample, the statistically superior team would eventually trend towards victory, but with a small enough sample a few unlucky rolls can change everything.

        Reply

32. Suggestion:
    Creatures have hp, defence, and armour. HP works similarly to normal D&D. Weapon damage is static. When you roll to hit, you roll your attack vs their defence like a normal ability check (however that would work with maths first or after). When you hit you get to keep rolling until you miss or kill them. Armour acts as damage reduction to the whole attack, not each individual damage instance, and defence is normally entirely based off dodging ability. Heavy armour will usually reduce defence for greater damage reduction.

    This brings things more in line with the current skill rules, makes armour and dodging feel different without different rolls, and, while it does increase the lag between the action and the success, I feel as if it increases the tension, or at least doesn’t reduce it too much. The additional rolls could be worth the extra time because they add some randomness to the damage dealing, as players seem to like that in my experience.

    Thoughts?

    Reply

    • Isn’t that just an even worse version of rolling for damage?
      After you hit you then have to keep rolling a bunch more times until you miss (repeating all of the to-hit math every time), and then multiply the number of hits by the damage value?

      If you do decide to go this route, it becomes even more important to do the math before rolling, so at least you can roll quickly in succession and not have to keep checking whether each value hits.

      Reply

33. Delta’s D&D hotspot has done a few articles on the best way to deal with D20 rolls that are worth a look (I’d give a link but I’m on a phone and lazy). He goes at it from a different angle, basically looking for the most mentally efficient way to resolve actions, and brings in some scientific papers to back him up. He doesn’t worry about whether the calculation is before or after the roll, instead focusing on how addition is more mentally efficient than subtraction. Also he plays 0D&D, so the numbers are very different. But with those caveats in mind, it’s still worth a look.

    I’ve been running a system that expects you to calculate the difficulty before the roll for a couple sessions, and the issue I’ve had is that my players are used to just rolling immediately as or before they really decide what they’re doing (no skill system and they still want to “use diplomacy on the goblin” the poor devils). Hopefully they’ll catch the vision soon.

    I have noticed that rolling first almost seems reflexive for a lot of people, to the point where I’m not sure if it’s learned behavior or due to some sort of natural impulse, something like “I’m not sure what to do, the dice decide what happens next,I’ll start rolling then figure it out from there.” I guess I’m saying there’s a psychological barrier between the method you hint at and the way people play that may cause some major issues.

    Reply

    • Hypothetically, having to do the math first could help to get players into the mindset of stating their approach before rolling?

      Reply

34. It occurs to me this is an advantage (one of few) of playing online D&D over tabletop; you have your macros, formulas already made, so when it’s time to roll you just click a button and BAM result! Cheers or boos all around.

    It’s a good patch over this problem if you’re going to be playing D&D, but if you’re designing a new game from the ground up, then the best ideal to shoot for is to have your mechanics elegant enough that you don’t need a ‘patch’ like that in the first place. Looking forward to what you come up with, Angry.

    Reply

35. Best. Article. Ever.

    Nothing more to say.

    Reply

36. I started keeping the DCs secret in Pathfinder for pacing reasons: it gave me time to calculate to figure all of the situational modifiers while the players were rolling the dice and adding up the result. Additionally, it helps avoid leaking whether there is something actually there when the PCs choose to actively look for it. I like the idea giving the players immediate feedback on the result, but I also like keeping the game moving and avoiding inappropriate metagaming by the players. Now, the former concern is moot because I’ve switched systems to 5e, but the latter still seems like an issue. Is there a good way to have all the things?

    Reply

    • Personally, my approach to this dilemma is to work my best to adjudicate dice rolls results the fastest I can. I won’t give exact DCs to my players before they roll (so they don’t get bogged down trying to figure out the exact result they need), but 80% of the time, I will be able to tell them immediately if they failed or succeed. Level 1 fighter against a AC 13 mercenary? Roll 1-5, that’s a miss, roll more than 10, that’s clearly a hit. With practice and by knowing the stats of your PCs, you end up getting a sense of what dice roll is needed in any given situation, and only ask for precise modifiers in the 30% cases where it’s not clear. And as players don’t know the exact DCs, they won’t notice if I make a small mistake here an then (anyway, if I take a guess in order to adjudicate faster, I tend to advantage players, so it won’t feel unfair).

      However, I give them a sense of difficulty by telling them my guidelines in terms of DC: an easy task for a skilled character has a DC around 10; an average task, 15; etc. But I present those as approximations (even if in reality, I try to stick to those exact numbers), for the same reasons I just explained. My GMing style assume numbers in RPG are here mostly to give a rough sense of how likely any action is to succeed, and that if it would take more than five seconds to figure out the result of a roll, it’s my job to make an educated guess about it.

      Reply

      • That’s pretty much how I handled it in Pathfinder. If the PCs rolled really well, there wasn’t a reason to continue adding up the modifiers. If they rolled poorly, same. It’s the leaking of hidden information that I’m probably interested in how to handle.

        For example, suppose the party is exploring a dungeon and comes across a locked chest. The rogue wants to examine it for traps. If you don’t tell them a DC, they’ll know the chest is okay. If you do, they’ll assume it is trapped. I wonder if the solution to that problem is to give them a fake DC. The same would go for if they’re actively looking out for ambushes, calling for a Perception check of some sort—or really for any kind of check like that.

        Reply

37. This is the best reason to have two different stats listed on a character sheet for the same ability or whatever.

    I use empty bubbles to fill in as the players customize their characters, and then underneath each bubble is the number they’ll need to roll equal to or higher than to succeed on a d20. It ends up with the visual advantage of players seeing their characters “get more power” as the bubbles fill up on the page, while also showing them at a glance what the bonus actually does when rolling the dice.

    I combine that with a system of placing tokens on their character sheet to represent d6’s used to calculate damage. The tokens are physical objects, and so are the dice, so players get a tangible correlation there, and the tokens are designed to be moved around for tactical choices.

    Once I did all of that, I turned every obstacle into something with variable “health” rather than variable DCs. I do what you do and tell my players the “health” (aka progress) needed to complete the task or kill the enemy. I keep strict track of turns to add complications over time, so players know that the amount of progress they make really matters.

    They get rising tension with each obstacle. They know they CAN finish pretty much any task, given enough turns. When they decide to roll the dice, the resolution is simple, but I have still struggled with the amount of time it takes to add up dice for damage. I’ve seen on people’s faces the loss of tension/interest in the moment between rolling damage and realizing they don’t know whether they’ve gotten enough damage, so they’ll have to add it up. The tense moment has turned into a task to be done.

    Reply

38. You talk like these “bs” articles are a waste of our time.

    I, for one, love them.

    While they may be more abstract and less to the point, I feel like I get a whole lot more “game knowledge” from them.

    This blog has 3 qualities that set it apart from everything else I’ve found.

    While most content out there is just tips and tricks that are situational, you have three distinct types of article which all contribute to a very structured learning process.

    1- Concept Articles, that teach us to think critically about GMing, empowering readers to make their own informed decisions as a GM. (Like what you wrote on adjudication and writing adventures)

    2- Crunchy Articles, which bring easily applicable ideas and solutions to our games and show us the value of making good prep work for your adventures. (like the monster building series or the critical path building you did in the megadungeon series)

    3- Abstract Rambling articles. In which we get to apply that critical thinking and come along for the ride as you tackle a new dillema. I think of them as concept articles that havent “reached their conclusions”.

    And I think its invaluable to us to read such articles specially because of that. It stimulates us to develop our own perspectives when tackling similar decisions. I also supect it must very useful for you as well, as a content creator.

    So please, keep them up.

    Reply

    • Can agree on the game knowledge part. Knowing why what’s there works or doesn’t is as valuable as a full article about one mechanic.

      Angry calls them “random BS”, I call them “analysis”

      Reply

39. Brilliant work Angry. I love that your years of experience and well-tuned intuition are paying off in such good insights. I’ve only recently become interested in game design and I have SO MUCH to learn and absorb before I get to the level you are now.

    I’m gonna work on the question you posed, but I’m guessing either the comments section will be closed or you’ll have posted a follow-up before I figure out an answer I’m happy with. ^^”

    Either way, thanks for taking me to school!

    Reply

40. Your comment on Borderlands got me thinking… If the time between a player’s input and the result is a problem, then why does Dark Souls works so well? I mean, you press R1, then your character does a big wind up animation, THEN attacks, THEN you see the result. My guess is that it has to do with the fact that the action is not resolved yet, that in Dark Souls the person defending against the attack still has choices to make when they see the opponent winding up their attack: Do I wait until last moment and roll? Do I try to parry? Do I go for a counter? Do I just raise my shield and block? Backstep? Even after the ‘I attack’ decision there are still choices to be made. So I guess that the real issues is that the moment action should be resolved is right after no one can do absolutely anything about it, when there are no more choices to be made. Which I know you may scream at me for being semantic and pedantic, but I’m willing to take my chances.

    For what it’s worth, HARP has a one roll combat system that, whether you like the charts of Rolemaster or not, maybe is worth looking into. Basically you have a Combat skill(say 80), and a Passive defense(say 20), and so does your opponent. You can use you entire bonus to attack or allocate points to boost your defense. So you can decide to use 40 +d100 to attack and save a passive defense of 60(40+20).
    Then you attack and compare it to your opponent defense(who also has the same options), if the number is positive, you did damage, the higher the number, the higher the damage done(it is determined in a critical chart). That’s the one system I know that resolves the entire action in a single die roll, although it does have the problem that the GM must read the outcome in a chart. But you can take the good and leave the bad. There is also a fan made version that I never got to try that actually allowed the player and GMs to allocate the defense AFTER the die roll. So basically in the other example if you got attacked by a 90, you can decide to sacrifice a 70 points(plus the other 20) to block the entire damage. It was called System of Parry and Riposte.

    Reply

    • The moral is that the time between the action being finalized and the result being revealed is *tension*. Increasing tension is okay, but if the tension doesn’t pay off people get upset and impatient.

      So slow isn’t always bad. Slow is *hefty*, meaty. A slow attack should stagger the enemy and do large chunks of damage. Then it becomes a game of finding an opportunity to use that heavy attack.

      Dark Souls gets away with being slow and tense by making the game rocket tag, and putting ridiculous effort into making every hit feel fair. A heavy attack with the Dragon’s Tooth can easily do over half of an enemy’s health and almost always knocks them to the f$&%ing floor, with a huge hitbox and lots of screen shake to sell the impact. Even smaller weapons have consistent stagger and crippling damage in that game. But if you mess up you can be killed yourself within a couple seconds.

      If Borderlands was using slow animations and delays, but didn’t use sufficient recoil and the bullets were narrow and the enemies barely noticed being hit, guns would feel weak and sluggish compared to what the animations were promising the player.

      Which is partly why I like crit confirmations in Pathfinder. That nat 20 is a moment where everyone holds their breath and intently watches the next roll, and the slowness has a real chance to pay off. There’s a small feeling of “screwjob” to failing confirmations and I don’t know how to elegantly avoid that, but the weight of a successful crit feels great.

      Reply

41. Reminds me of when I thought about a Fixed Damage variation that dealt damage in 1/4ths depending on how “well” you hit (an attack that is AC or 1 less id 25% dmg, an attack that goes waaay over is 100%)

    It might not make the game faster but it does eliminate the stupid part of “Yeah, you caught the goblin dead ln and dealt 3 out of 40 possible damage”

    Reply

42. The only time I couldn’t see revealing DC is secret doors. You could save failure on finding traps means you set it off (search for traps could be treated like a another save); but secret doors just stay there. They are the most static element I can think of and giving a DC reveals something is there. The DM could roll secretly but that functions the same as hiding DC, same as having fixed search results for characters and keeping the DC secret.

    Could set difficulty by wall type to a level to be 100% positive there are no secret doors. Which means if they hit it, they can move on. Then you couldn’t ever set a DC on a secret door higher for that wall type.

    Secret doors are something where even its existence isn’t known and there is no immediate feedback like for other elements.

    Reply

    • I don’t give out DCs on secret doors. I should have said that up front. I give out DCs on things that have known and generally binary consequences. You will or will not pick the lock, make the jump, climb the wall. Secret doors are a special case.

      You could succeed and find one that is there, or not succeed and not find one that is there. But how do you succeed and not find one and be sure there isn’t one there? DM tells you,”You’ve pretty sure if there was something hidden, you would have found it”???

      And how long do you and your players want to play, “Find the hidden thing?” They just want a resolution.

      Perhaps a resolution would be a tradeoff? Offer to add 5 to their first search roll results in exchange for a die added to the tension pool. Then they will be able to decide when to call it quits, and the swingy d20 won’t be the reason the session sucked.

      Reply

43. I will again recommend RedMarkets Its resolution mechanic is one roll with 2 colored d10 for everything and success is evident at a glance. At the same time various game mechanics especially damage rolls are determined by the individual d10 values. So you have the damage rolled but it doesn’t take up time.

    Recently I was hacking Eclipse Phase RPG. I was going to use dice pools for damage rolls to “realistically” model armor. You showed me how wrong that was.

    Now for the hopefully useful stuff. EP uses Blackjack d100 (roll as high as you can under target).
    Because of that it already has “the moment when all that’s left is to roll”. All the modifiers are applied before to the target before.
    But it is a game of intrigue, spies, sometimes horror. There is a lot of factors unknown to the players and characters. Because of that I thought it would be useful to try spelling out the known factors as the modifier to the Target and unknown factors as degrees of success needed. Example of a Perception check for Karen the Scout with a Perception of 70. There is something in the fog around her. The fog is a known factors so it applies -20 to Target known to the player. But unknown to the player the Stalker in the fog is wearing invisibility cloak requiring 3 Degrees of Success to spot him. Counting a degree of success as a first number on d100 we see that Karen needs to roll between 30 and 50 to spot the Stalker.
    Players know that higher means better so they can try to influence the dice to aim for their target (with Will points/hero points analogue).
    The unknown degrees of success needed adds to the tension of the situation and feeds the player’s paranoia.

    What do you think?

    Reply

44. I’m still having trouble with the premise that all the math needs to be done beforehand so that you know “before the die hits the table” that you succeed or fail in order to maintain tension.

    I’ve played many games that use “Roll to Hit + Roll to Defend” mechanics, and that dramatic tension still occurs during the rolls. That moment of “damn the GM rolled well, so I gotta make this”, adds to that dramatic tension in a good way.

    Yes, there is a little simple math, but the breath holding moments are still there, they just happen at a slightly different point in time than what you have described.

    Having recently switched our group to from such a system, (original GM is taking a break, new GM runs D&D 5th) the players dislike the feeling that their fate is down to a static number (AC). It takes away that sense of agency that their fate is in their hands, even if the fate is REALLY in the hands of a D20. The fact that they are rolling it FEELS good.

    So, I think building the dramatic tension is less about when the “math part” occurs, as long as the game helps to ensure the tension is there.

    Reply

    • I’ve noticed this sometimes too.
      One of my players mentioned explicitly that she disliked the lack of agency in dodging an attack, you just have to stand there and take it with your static AC value.
      I’ve also noticed some DMs using contested checks for everything, even when it doesn’t make sense (how exactly does one contest a persuasion check to convince them you’re telling the truth?), so I do think a good system should have sensible guidelines on what is and isn’t contested.

      Personally, I do hate that taking the Dodge Action in 5e basically feels like skipping my turn.
      I do agree that passive dodging could easily be a roll, although depending on the system it might slow things down too much.
      I have thought extensively about the additional hooks it could bring to the mechanics though, not least of which just from having another roll to hang advantage on. You could differentiate between different forms of AC by having armour be a passive score (essentially a minimum) and dodging/parrying/etc be an active roll. You could be denied the active roll when paralysed/blinded/etc, or have effects that ignore armour but can still be dodged (essentially the return of Flat-Footed/Touch AC without having to track 4 different AC values). You could use feats/etc to substitute other skills for your roll. You could have some kind of total amount of dodging per round, possibly expending attacks or reactions to parry or avoid additional hits.

      So many possibilities!

      Reply

      • When I mentioned Dark Souls on a comment above, it’s precisely for the fact that in that game Defense is not a passive skill. The problem in D&D is that defense is just a passive bonus with no choices in actual combat. Something like choosing between dodging(medium dificulty, avoid all damage if succesful, take 50% extra damage if not), parrying (hard difficulty, get a riposte for extra damage if success, get damaged if failure), and blocking(easy difficulty, but get a penalty for attacking next round) could provide some extra options to defense. Of course, resigning attacking next round shouldn’t be an option unless some extreme circumstances.

        This is just an example of trying to emulate part of Dark Souls, but any system that provides choices with different risk/benefits should work. The price of course being the extra complexity that people may or may not want.

        Reply

        • Or what if only the PC’s rolled? They make to-hit rolls to hit the static AC of the bad guys, and when they get attacked they make their defense rolls to avoid the static “HIT” stat of the bad guys. Kinda like in Dungeon World where the GM doesn’t roll any dice.

          It might make the players FEEL better, since they are always the ones using the random number generators. It might also open up more choices for the players in combat.

          But I guess it doesn’t get any closer to knowing the result when the dice hit the table, because the PC’s still have all their modifiers they need to add…

          Reply

          • I’ve tried that in a game and it didn’t really work. I think the problem was players enjoy getting mad at foes when they manage to hit them, but they don’t enjoy feeling lousy about themselves when they screw up and manage to get hit.

            It’s not logical, but Angry has taught me if you design a game expecting human beings to be logical then you are all in for a bad time.

            Reply

45. This article reminds me of why I dislike the “proper” use of stealth in 5e (inasmuch as 5e actually describes the proper use of stealth anyway).

    Roll to “enter stealth mode”, then nothing happens immediately, then keep your rolled “stealth value” static as you explore indefinitely, then if a creature might notice you it checks it’s passive perception against your previously-rolled stealth value (hope you still remember what it is), then if it actively searches for intruders it checks it’s active perception against that same value, repeat until you stop trying to be stealthy.
    Talk about holding your breath between rolling and resolution! You practically have to write down your roll and come back to it later!

    (Not to mention the infamous “I rolled a 1… I stop trying to be stealthy, and then I try being stealthy again.”)

    Reply

    • The advantage to using stealth in this way is that you avoid ‘rolling to failure’ as described here:
      https://thealexandrian.net/wordpress/38798/roleplaying-games/gm-dont-list-2-rolling-to-failure

      (no idea how this site deals with links)

      Basically, the more rolls you require, the more likely you are to see a failure, which disincentivizes the use of stealth, since if the enemy is mathematically guaranteed to find you you’re better off just walking up to the gate with you’re whole party together than getting caught after splitting up.

      The issue is, exactly as you point out, that if you use that rule as written you won’t find out if you succeeded until you either A) finish the entire sneak or B) find the guard with a good enough passive perception to beat you’re roll(if you allow the guards all to use active perception, you’re basically rolling to failure again, but in reverse). Also the side issue of cheesing the system to improve your roll.

      So if you want to encourage stealth in your game (and I think it’s worthwhile) and still make sure the success/failure is obvious as soon after the roll as possible (I also think that that is worthwhile) I would suggest treating the entire scouting/sneaking attempt as one challenge and figure out the DC accordingly (say the passive perception of the guards). Then if, while sneaking around, the player does something especially risky (decides to turn toward the heavily guarded vault, tries to kill one of the guards, etc.) let her know the danger there and ask for another roll. If they don’t do anything especially risky, just let the scene play out without them getting caught, and let your players know that’s how it’s going to go.

      Tl,dr: treating the whole string of sneaking attempts as one big sneak attempt works fine as long as you treat the whole string of sneaking attempts as one big sneak attempt, but if you try to treat it as both one and many sneak attempts, it sucks.

      Reply

46. Also, Damn I thought by the title of this article that you’d be giving advice about improvisation and acting.
    Those are my biggest flaws as a DM. That and a lack of foresight, which is a terrible combination. 🙁

    Reply

47. Wow, that explains me a lot. We switched from D&D to GURPS several years ago, and I wondered why the pacing of combat feels better, while it actually takes much more time (advanced tactical space combat with velocity, module-targeting and other crap). The key thing is – you must do all your math BEFORE rolling (though I must admit complicated math is ok for our group). The other reason – everyone gets only one decision per turn.

    Reply

48. Hello, I’m a bit late but as a designer for my own rules I ‘ve experimented quite a few ideas over the last 4 years and you might be interested by my results.

    1) I tried no damage rolls, the problem: after a few of games my monsters had to deal enough damage to hurt characters specialised in defense but that amount of damage was fatal for the less armored ones.

    2) My monsters don’t roll their attack, it’s the players who roll their defense vs my monster attacks. I tell them the value they need to hit and usually they do the math to figure out what is required on th dice to succeed. This way they can do things to improve their defense and it raises the tension when they get hit by a potentially lethal attack.

    3) Players throw a D10 to resolve actions, 1 if autofail, 0 is autosuccess, when rolling 1 or 0 I make them roll again to see if it’s a special or critical success/failure. Whatever happens they have at least 10% chance to succeed or fail. I used to have a table spelling whatever happenned on the second roll but it took too long to check and broke the flow, I’ve settled for the same special effect for 4 results and 1 critical effect for 1 result, the critical one happens 1 or 2 times per 10hour session.

    Now my thought and insights:

    If you want fixed damage then armor can’t lower the damage, like D&D does.
    Also you will want to play with high values if you want some distinction between strong and weak hitters that is significant and the possibility of lots of minor modifiers that add up without too much imbalance. That’s for any roll and that’s why D20 might be better than D10 which I use.

    About automatic success and failures, my gut tells me that 20% would feel better to add tension to the dice rolls. However more than 5% specials is too much.

    As of today I use a damage system with dices, like 2d10 dmg, fighting specialists can improve their rolls by changing one die result as they like to improve (when hitting an ennemy) or lower (when hitting themselves or an ally) their damage. So 2d10 averages between 11-20 rather than 2-20. Only my boss monsters use that rule. Whatever damage is rolled the right always gets hurt for a least 1 Hp which is about 5% of a normally tough character.

    Reply

49. tl;dr Technology doesn’t satisfy our need to throw dice and limit maths.

    Maths is my main criticism of live games (and betting in poker) and why I prefer rolling dice on Roll20.

    I click one thing on my character sheet, and it determines all the results.

    If I were to be academic about it, I would say that decreasing the number of dice rolls limits the complexity of the game. As our tastes evolve, we want more complex games. And there is a visceral sensation provided by throwing dice that electronic dice-throwers don’t replicate. What we need then is a way to have increasingly complex dice rolls (that eliminate maths) that also satisfy the need to throw the dice.

    A few years ago, a company made a bluetooth die that would sync with your ipad. You roll a d6 and it displays the results. I see no reason why something like this couldn’t be utilized with a digital character sheet on an ipad. You highlight the Ability/Skill/Weapon, roll your requisite bluetooth die and the iPad tells you the results of the roll. No human maths involved.

    I’m not trained enough to know how complicated that kind of app engineering would be, nor how expensive bluetooth dice would need to be to be properly balanced, but I have money and would give it a try if someone developed such a solution.

    Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.