I strongly prefer games where the rolling is exactly the same for any character, whether PC or NPC. Essentially, I want the resolution mechanics to remain identical regardless of which side of the screen the roller is sitting. D&D largely does this.

The OP sort of begs a follow-up question: do you prefer if there is always some sort of defense roll? Or attack roll vs a set defense value for the opponent? I have seen both done well. I sort of like Runequest's attack roll and then parry roll. But I also made Rolemaster work pretty well, and here you have attack roll plus attack bonus minus defense bonus (refer to table with armor type).

Quote from: Trond on August 24, 2022, 01:29:00 PM
The OP sort of begs a follow-up question: do you prefer if there is always some sort of defense roll? Or attack roll vs a set defense value for the opponent? I have seen both done well. I sort of like Runequest's attack roll and then parry roll. But I also made Rolemaster work pretty well, and here you have attack roll plus attack bonus minus defense bonus (refer to table with armor type).

For me, no.  If it's a 'basic attack' and the only consequence is 'damage' there's really no need to impact speed of play adding an additional step.  Since the point of a defense step is, for me, primarily psychological, I can see different groups having difference preferences.  For myself, if an attack includes something more substantial (like being knocked unconscious), I like giving that to the defender.  Deciding what counts as more substantial is up to debate. 

For us, regular attacks (including sneak attacks), and critical hits (including damage directly to wounds) are 100% determined by the attacker's rolls.  Potential action denial (like being stunned) or attacks that give you a chance for a greater or lesser effect (inflict wounds, save for half), etc, go to the defender to roll.  In part, since the defender has to track that stuff anyway, we don't lose table-time by letting them make the roll. 

Quote from: Visitor Q on August 24, 2022, 11:45:44 AM

Quote from: BoxCrayonTales on August 24, 2022, 11:36:52 AM
In my observations, I've noticed that players like rolling dice to give them the illusion of control. This is why so many systems have more dice rolls in combat than they really need: e.g. roll to hit, roll to dodge, roll to damage, roll to soak. These rolls can be streamlined to just the players rolling to hit and to avoid being hit, which gives players the illusion of control and avoids any harebrained impetus to add more rolls to make them feel in control but is more streamlined than using opposed rolls for everything.

For example:
Player: I roll to hit the goblin with my sword. *rolls dice*
GM: *compares result* You make a glancing blow to his arm. He lashes out with a spear. Roll to defend.
Player: I raise my shield. *rolls dice*
GM: *compares result* The spear bounces off your shield.

I don't entirely disagree but also in fairness adding extra dice steps can help the immersion of the game.

Knowing precisely that your character failed to parry but only your dwarven-made plate saved you from damage is a different experience to being told that in the aggregate the orc's attack failed to inflict any damage.

And in my observation combat is a horrible slog that takes forever and the details of which are completely forgotten by the start of the next round.

There's a strong reason why I preferred playing Neverwinter Nights servers over playing in person. The combat slog is one of those reasons.

In terms of slogs though, I have overall found that D&D's separate hit and damage rolls are often faster than margin of success based single roll systems.

Compare (is it higher than defense number?) is always going to be faster than a subtract attack from defense to get damage (particularly when at least one system I've seen then applied a multiplier to the result to reflect different types of weapon).

The biggest issue with modern D&D I find is too much hp bloat and the subsequently larger damage rolls (both in number of dice and modifiers). There's a noticeable difference between 1d8+1 vs 18 hp... and 1d8+2d6+5 vs. 58 hp in terms of math complexity.

Players should roll for anything done by their PCs except certain unconscious checks like perception. For combat rolls by NPCs and monsters, like GhostNinja said, the GM should roll. In my book, those rolls should be in front of the players to avoid temptation to the GM to fudge a roll, and to let players confirm rolls for themselves. Outside of combat, the GM should roll behind a screen occasionally and in front of players occasionally to stoke disquiet, uncertainty, and fear. Every once in a great while, the GM should roll a die behind a screen, closely inspect the die result, pretend to consult a book or chart, look at the players, and ask them what they are going to do next, not answering any questions from the players as to what that was about.

Quote from: Chris24601 on August 24, 2022, 08:43:37 PM
In terms of slogs though, I have overall found that D&D's separate hit and damage rolls are often faster than margin of success based single roll systems.

Compare (is it higher than defense number?) is always going to be faster than a subtract attack from defense to get damage (particularly when at least one system I've seen then applied a multiplier to the result to reflect different types of weapon).

The biggest issue with modern D&D I find is too much hp bloat and the subsequently larger damage rolls (both in number of dice and modifiers). There's a noticeable difference between 1d8+1 vs 18 hp... and 1d8+2d6+5 vs. 58 hp in terms of math complexity.

Yeah, I can see where the expedience factor comes in. I personally prefer abstract rolls rather than separate hit and damage rolls because the former avoids the whiff effect where you hit but deal minimal damage and allows greater leeway in terms of describing the attack. I don't like things like minimum damage or redundant defense rolls or whatever that devs feel compelled to include. E.g. "you swing the battle axe but only make a glancing blow" is generally impossible in most systems because of mechanics like minimum damage and multipliers.

One of the mechanics I like in 13th Age is that they add a cumulative bonus to your attacks every subsequent round of combat to headoff the whiff effect.

Looks like preferences are all over the place, which I suppose is a good thing for budding game designers.

So a follow up question:  in my current project I'm far enough along to be locked into my chosen mechanic, which is opposed rolls, all rolls are single d6, margin of success or failure matters, and the totals (die + talent + skill + mods) will be at most in the low teens for beginning characters and mid 20s for experienced one.  I get that this might require a little extra time or attention (though I'm hoping not).  Would you expect any other difficulties with this mechanic?  And any suggestions for handling those difficulties whilst keeping the mechanic?

And please carry on with the original topic as well - it's been really helpful so far.

General rule for me is players roll for their actions, almost exclusively when real risk is involved. Combat is the classic example. Pickpocketing, bluffing, etc., fall into this category as well. If there are positive and negative outcomes, and there is significance for either or both, I want them rolling.

I tend to roll for the environment, the mundane, or when not knowing the result would add more tension. Wandering monsters, random monster tables, NPC reaction, etc. I roll for monster combat, although I do like game mechanics where players can roll a defend roll instead—no preference either way though.

Given the above, I typically scrub random search and perception type rolls from my table. We can almost always roleplay those out. This is where player skill is developed and tangible results are achieved.

Quote from: Mishihari on August 26, 2022, 04:59:03 AM

So a follow up question:  in my current project I'm far enough along to be locked into my chosen mechanic, which is opposed rolls, all rolls are single d6, margin of success or failure matters, and the totals (die + talent + skill + mods) will be at most in the low teens for beginning characters and mid 20s for experienced one.  I get that this might require a little extra time or attention (though I'm hoping not).  Would you expect any other difficulties with this mechanic?  And any suggestions for handling those difficulties whilst keeping the mechanic?

A single d6 with those kinds of mods (treating everything but the die as essentially a mod for this discussion)?  You've got to be really careful with your mods.  As opposed rolls, that's essentially a 2d6 system in other garb (which is not a bad choice), but the scaling will sometimes chop of the upper or lower ranges of the 2d6.  There's only so much scaling you can get before it starts to get that treadmill feel, but if the design isn't meant to scale very much, that may not be an issue.

Note that being careful with your mods doesn't necessarily mean being stingy all the time.  It probably is being stingy with most true situational mods, unless they are meant to be overwhelming.  You've really only got room for +1 and +2 in the vast majority of cases, and some things that systems give mods for wouldn't even justify a +1, I'd think. The main problem there is stacking mods, where several minor things barely justified as a +1 end up as +3 or +4, when if you thought about them as a group, they might barely be +2.  The same thing can happen with negative mods, though of course having some negative mods gives you more room for stacking, if the positive and negatives often cancel out.

With skill and talent, being careful can go the other way--again, depending on design intent.  If you want certain talents or skills to be overwhelming when fully developed, then that's a better place for the spread between the upper teens and mid-twenties on the totals.

I personally would probably find the mechanic a little off-putting aesthetically.  The mods are too much of the total compared to the dice roll to hit my preferences, which invokes for me a CRT table in a wargame.  The game is about stacking as many mods as you can get.  Note that Toon has a similar mechanic with much lower numbers.  It invokes the cartoon feel because the real mods in Toon are few and far between and over the top.  I've said before that if I were to run a Star Wars game, I'd use Toon as the engine--because me running a Star Wars game would be a parody.  Of course, that might just me and my gaming background and preference speaking.  There's no inherent reason why your mechanic can't be used in a more serious game, but I suspect that whatever tone is embedded into the description of the mods and their effects is going to be baked in.

Quote from: Mishihari on August 26, 2022, 04:59:03 AM
So a follow up question:  in my current project I'm far enough along to be locked into my chosen mechanic, which is opposed rolls, all rolls are single d6, margin of success or failure matters, and the totals (die + talent + skill + mods) will be at most in the low teens for beginning characters and mid 20s for experienced one.  I get that this might require a little extra time or attention (though I'm hoping not).  Would you expect any other difficulties with this mechanic?  And any suggestions for handling those difficulties whilst keeping the mechanic?

If I'm following correctly we could have a situation where one player is rolling 1d6+12 (13-18) and another character is rolling 1d6+14 (15-20).  You take those rolls, subtract the smaller from the larger, then compare to margin of success.  So if Player A rolled 18 and Player B rolled 20 (they both rolled 6) player B won, and if the margin of success if 2 or less they get a 'better success'. 

In terms of resolution, we have to assume that both players are responsible for calculating their bonus, so the 'time steps' come from rolling, adding, comparing. 

Mathematically, it would be virtually equivalent to set a TN of 3+ the target's bonus.  Ie, Player B has a +14 bonus; the TN against Player B would thus be 17.  Player A must roll a 5+ on the d6 with their +12 bonus. 

It should be apparent that if the bonus between participants is more than 6, the lower skilled character cannot succeed against the higher skilled character.  With bonuses ranging from teens to twenties, it may require extreme discipline on when bonuses apply to ensure that there's any point in rolling. 

If one character has a bonus of:
+1, that character wins 58%, loses 28% and ties 14%
+2, that character wins 72%, loses 17% and ties 11%
+3, that character wins 83%, loses 8%, and ties 8%
+4, that character wins 91%, loses 3%, and ties 3%
+5, that character wins 97%, loses 0%, and ties 3%
+6, that character wins 100%

Those numbers appear very punishing if there is usually a discrepancy between total net bonus. 

Quote from: Mishihari on August 26, 2022, 04:59:03 AMSo a follow up question:  in my current project I'm far enough along to be locked into my chosen mechanic, which is opposed rolls, all rolls are single d6, margin of success or failure matters, and the totals (die + talent + skill + mods) will be at most in the low teens for beginning characters and mid 20s for experienced one.  I get that this might require a little extra time or attention (though I'm hoping not).  Would you expect any other difficulties with this mechanic?  And any suggestions for handling those difficulties whilst keeping the mechanic?

Chiming in to support deadDMwalking's point above: To me single opposed d6s on either side narrow the margin of "interesting variability" very strictly. When as little as a +2 edge vs. an opponent, or a difficulty factor, means you're going to win 7 out of 10 rolls, the system has very little room to make rolls interesting: they'll either be far too easy, far too difficult, or essentially a toss-up.

If I were working on this system I might steal from (what else?) The Riddle of Steel and introduce an element of allocating commitment. In combat or other active conflicts, this means that you have to divide your action value total between offense and defense, and the trick is determining how low you can risk setting your defense in order to get an effective offense. This also helps because a conflict allows for both multiple rolls (allowing for extended odds management) and a finite limit (usually fatigue will decrease both combatants' scores until one can no longer effectively resist the other).

For non-conflict action rolls, I might suggest that the PCs have to divide focus between quality of result and speed of result, where you can try to do something well but potentially too slowly, or quickly but potentially too sloppily (you could use other areas of focus, as well, like quantity, originality, future carry-forward bonuses, etc.).

Quote from: Steven Mitchell on August 26, 2022, 09:06:01 AM

Quote from: Mishihari on August 26, 2022, 04:59:03 AM

So a follow up question:  in my current project I'm far enough along to be locked into my chosen mechanic, which is opposed rolls, all rolls are single d6, margin of success or failure matters, and the totals (die + talent + skill + mods) will be at most in the low teens for beginning characters and mid 20s for experienced one.  I get that this might require a little extra time or attention (though I'm hoping not).  Would you expect any other difficulties with this mechanic?  And any suggestions for handling those difficulties whilst keeping the mechanic?

A single d6 with those kinds of mods (treating everything but the die as essentially a mod for this discussion)?  You've got to be really careful with your mods.  As opposed rolls, that's essentially a 2d6 system in other garb (which is not a bad choice), but the scaling will sometimes chop of the upper or lower ranges of the 2d6.  There's only so much scaling you can get before it starts to get that treadmill feel, but if the design isn't meant to scale very much, that may not be an issue.

Note that being careful with your mods doesn't necessarily mean being stingy all the time.  It probably is being stingy with most true situational mods, unless they are meant to be overwhelming.  You've really only got room for +1 and +2 in the vast majority of cases, and some things that systems give mods for wouldn't even justify a +1, I'd think. The main problem there is stacking mods, where several minor things barely justified as a +1 end up as +3 or +4, when if you thought about them as a group, they might barely be +2.  The same thing can happen with negative mods, though of course having some negative mods gives you more room for stacking, if the positive and negatives often cancel out.

With skill and talent, being careful can go the other way--again, depending on design intent.  If you want certain talents or skills to be overwhelming when fully developed, then that's a better place for the spread between the upper teens and mid-twenties on the totals.

I personally would probably find the mechanic a little off-putting aesthetically.  The mods are too much of the total compared to the dice roll to hit my preferences, which invokes for me a CRT table in a wargame.  The game is about stacking as many mods as you can get.  Note that Toon has a similar mechanic with much lower numbers.  It invokes the cartoon feel because the real mods in Toon are few and far between and over the top.  I've said before that if I were to run a Star Wars game, I'd use Toon as the engine--because me running a Star Wars game would be a parody.  Of course, that might just me and my gaming background and preference speaking.  There's no inherent reason why your mechanic can't be used in a more serious game, but I suspect that whatever tone is embedded into the description of the mods and their effects is going to be baked in.

Your point on situational mods is well taken and has been a challenge.  Taken proportionally, a +1 on a d6 is about the same as a +3 on a d20.  There are a lot of situations that seem to merit a mod, but less than a +1 on a d6.  My approach thus far has been to ignore them.  This fits overall with my goal of trying to keep the system simple, but it pains me to ignore things that should make some difference.

Stacking has been an issue too.  The most serious example is the stealth skill, which has a list of probably 30 or so situational mods that I feel should rate at least a +1.  I need to find some way to cut this down, first because I don't want the total of situational mods to be so much larger than everything else, and second because this is just more mods than anyone is going to want to consider for a single roll.

As for the aesthetics, well, it's intentional.  I've always felt that of skill, situation, and luck, luck should have the least effect on the outcome.  It's definitely a bit experimental – I don't know of a great many other games that have tried this approach.  My hope is that players will realize that luck will not be enough to save them if outmatched on a particular type of skill check, and will look for different approaches rather than just continuing the same attack and hoping the monster runs out of health first.  A "different approach" might be looking at the environment to find a way to create another favorable mod, using a different combat maneuver (there are about 20 just for melee), or changing the type of conflict; frex if you're outmatched in a weapon melee, you could go for a grapple, open some space and switch to missile attack, run and initiate a chase, pop off a spell, get even more distance and hide, or use athletics move somewhere your enemy can't follow.  The goal is to provide enough meaningful choices that even a simple standup melee fight has some depth to it.

Quote from: deadDMwalking on August 26, 2022, 09:49:59 AM

Quote from: Mishihari on August 26, 2022, 04:59:03 AM
So a follow up question:  in my current project I'm far enough along to be locked into my chosen mechanic, which is opposed rolls, all rolls are single d6, margin of success or failure matters, and the totals (die + talent + skill + mods) will be at most in the low teens for beginning characters and mid 20s for experienced one.  I get that this might require a little extra time or attention (though I'm hoping not).  Would you expect any other difficulties with this mechanic?  And any suggestions for handling those difficulties whilst keeping the mechanic?

If I'm following correctly we could have a situation where one player is rolling 1d6+12 (13-18) and another character is rolling 1d6+14 (15-20).  You take those rolls, subtract the smaller from the larger, then compare to margin of success.  So if Player A rolled 18 and Player B rolled 20 (they both rolled 6) player B won, and if the margin of success if 2 or less they get a 'better success'. 

In terms of resolution, we have to assume that both players are responsible for calculating their bonus, so the 'time steps' come from rolling, adding, comparing. 

Mathematically, it would be virtually equivalent to set a TN of 3+ the target's bonus.  Ie, Player B has a +14 bonus; the TN against Player B would thus be 17.  Player A must roll a 5+ on the d6 with their +12 bonus. 

It should be apparent that if the bonus between participants is more than 6, the lower skilled character cannot succeed against the higher skilled character.  With bonuses ranging from teens to twenties, it may require extreme discipline on when bonuses apply to ensure that there's any point in rolling. 

If one character has a bonus of:
+1, that character wins 58%, loses 28% and ties 14%
+2, that character wins 72%, loses 17% and ties 11%
+3, that character wins 83%, loses 8%, and ties 8%
+4, that character wins 91%, loses 3%, and ties 3%
+5, that character wins 97%, loses 0%, and ties 3%
+6, that character wins 100%

Those numbers appear very punishing if there is usually a discrepancy between total net bonus. 

Your point that using such a small randomizer could result in punishing discrepancies is well taken, though I have tried to design the system such that doesn't happen.  I'll describe it here, and if anyone sees a flaw I'd love to hear what it is so I can make corrections.

You have the gist of the system mostly correct, but I'd like to provide an example to flesh it out a bit.  This is the simplest attack, using the melee action "strike."  The checks are

Attacker:  d6 + weapon skill + attack talent + situational mods
Defender:  d6 + weapon skill + defense talent + armor + situational mods

To resolve, find the margin, which is the attacker total minus the defender total, and if it is at least one, the defender takes damage equal to the margin.

One takeaway from this is that a roll is always needed, as the margin determines damage.  If one is lucky he might be able to keep health loss to a minimum even if he is outmatched otherwise.

A second takeaway is that the most of the components of the check con be computed and placed on a character sheet ahead of time to speed things up during play.  This becomes more important for other maneuvers that have more components to the check, but even in this case a player can add his PC's attack talent and sword skill and write "Sword strike +6" on his character sheet.  He then only needs to add a roll and any situationals that apply (there aren't many).

As I mentioned to Steve, my hope is that a player that finds himself at a great disadvantage in a specific opposed check will change approaches.  The game is skill rather than level based, so if one's opponent outmatches him in melee, it doesn't necessarily follow that he will also outmatch him in grapple, chase, magic, or whatever.

I'll also mention the skill progression, as I think it's relevant
0   Untrained
4    Competent
10  Expert
20  Master

In a simple melee using only strike actions, and all other things being equal, a competent warrior will clobber an untrained person and an expert will clobber the merely competent, as your table shows.  I think that's realistic and I'm okay with that.  It's much like D&D in that a 5th level pc will generally greatly outmatch a 1st level one.  Of course in an adventure, all other things are not equal, and I'm hoping that the meat of the combat strategy will be finding ways to use the strengths of one's PC against the weaknesses of his enemies.

All that said, I really need to get this far enough along that I can have some friends try it.  It looks fun to me, but the only way to know for sure is to actually play the game

Quote from: Mishihari on August 30, 2022, 02:43:27 AM
Stacking has been an issue too.  The most serious example is the stealth skill, which has a list of probably 30 or so situational mods that I feel should rate at least a +1.  I need to find some way to cut this down, first because I don't want the total of situational mods to be so much larger than everything else, and second because this is just more mods than anyone is going to want to consider for a single roll.

To minimize the number of modifiers, you can put them into categories and specify that you simply use the highest in that category.  For example, you could have concealment bonuses that range from +1 (soft cover) to +4 (invisible).  If you're invisible AND have soft cover, that would just be a +4 (not a +5).  Ideally you can combine your 30 existing bonuses into something like 3 categories, which could still be POTENTIALLY overwhelming, but with most bonuses just a +1 or +2, hopefully not! 

Quote from: Mishihari on August 30, 2022, 03:48:52 AM
You have the gist of the system mostly correct, but I'd like to provide an example to flesh it out a bit.  This is the simplest attack, using the melee action "strike."  The checks are

Attacker:  d6 + weapon skill + attack talent + situational mods
Defender:  d6 + weapon skill + defense talent + armor + situational mods

To resolve, find the margin, which is the attacker total minus the defender total, and if it is at least one, the defender takes damage equal to the margin.

If the Attacker has 3 sources of bonus, but the Defender has 4 sources of bonus, that's likely to give Defenders a pretty significant advantage.  One thing you could consider is treating melee as both players attacking each other; this means that the defender COULD wound the attacker; and if both characters spend their actions fighting each other, you could potentially deal damage twice in the same round.  I would recommend giving the attacker the advantage; people will think it is weird if they do more damage when someone attacks them versus when they attack someone else...  But if you don't give enough chance of success for attacks combat will be grindy.