I've been wanting to design my own rpg for a while, tailored to my own tastes. To what end? I don't know yet, so starting with some generic ideas. Something that starts as a baseline to springboard other ideas and concepts, with the baseline being somewhere around realistic while being playable.

Enough rambling the first thing I'm considering is a core mechanic or basic resolution system and dice math.

Linear math. Roll a dice and try to beat the target number.
-d20, osr does this has a linear probability adjusts in 5% increments. Has a pretty wide range can be kind of swingy. Also means characters must progress farther to be more reliable.
-d10, interlock system is a good example, same as above but a smaller range of numbers. It seems by shrinking the range you end up more often as highly unlikely to fail for competent characters or highly unlikely to succeed if you suck at something.
D100, the most transparent in terms of success chance. Usually a roll under tends to put a cap on progression.

Bell curve,
2d6 or 3d6, like gurps or Cepheus. Savage worlds I think also falls into this category. Characters seem like they'll reliably succeed or fail depending on the target number but still allow for exceptional results.

Dice pool, exponential curve.
Usually d6s, but I've seen d8s and d10. Can set the difficulty on successes needed and what numbers count as a success. Can really obscure probability for less math savvy players. This curve makes each additional die increase the chance of success dramatically, though at the expense of needing a bunch of dice.

Anyway I'm talking out load thoughts on resolution mechanics?

Quote from: Ocule on October 17, 2022, 05:32:50 PM
I've been wanting to design my own rpg for a while, tailored to my own tastes. To what end? I don't know yet, so starting with some generic ideas. Something that starts as a baseline to springboard other ideas and concepts, with the baseline being somewhere around realistic while being playable.

At the risk of getting off topic, I think you want to get more details, however tentative, about "realistic while being playable" before you start with mechanics.  Mechanics that work and are playable are determined in part by the game model.  The model drives the math.  Then the math drives the mechanics. All informed at each step by what is working, what seems playable and reasonable, etc.  But starting with the mechanics is most likely to either paint your design into a corner or leave you holding onto something that isn't working way too long.

This is not a theoretical concern.  I've got 2 designs I'm working on.  The first one started ignoring everything I said above, and in fact was much like what you described.  The only difference was that I was artificially limiting myself to certain mechanics so as to be partially compatible with an existing game.  In fact, it is an attempt to make a much more playable game while using the heart of the same mechanics.  I had two serious roadblocks:  A.) I was trying to do too much with the game.  Once I really did a retrench and analyzed why it wasn't working, I realized that I was trying to design 2 or 3 different games at once, some features of which were incompatible with each other.  B.) Some of what I was trying to do did not work with those existing mechanics.    Thus I'm confronted with either abandoning some of my goals or the mechanics.

In my second design, I had learned both lessons.  It has gone much better.  I had something my groups were able to start trying out in a 1/4 of the time compared to the first one, and those "tests" were more enjoyable for those groups, gave more information, and led directly to actionable improvements for later iterations. Whereas, during the initial tests of the first one, the actionable item was often something akin to "throw out this sub system entirely, and go back to the drawing board."

If you'd like to discuss that more, I'd be happy to.  Though I think it might be off-topic for this forum.  The game design forum might be a better choice.

I've played and taught a wide range of games, but not designed one, so take this with a grain of salt.

There's something in the gambling center of the lizard brain that gets stimulated when the roll immediately shows the success/failure result. I think roll, add/subtract mods, compare modded result to target number is popular due to inertia mainly. Don't get me wrong, it's perfectly functional, but post-roll arithmetic can dampen the thrill of the roll. The thrill is preserved if you reverse it to be: determine base target number, add/subtract mods, roll and compare. d100 and dice pool systems tend to do it this way, but it could work with any mechanic.

Also, roll under does not necessarily put a cap on progression. If you use negative modifiers and allow skills to go past 100, then d100 roll under still works fine.

All game models put an effective gap on progression.  Some are just more obvious and strict about it than others.  Just because a mechanic can mathematically be extended, doesn't mean that it will continue to work in all the ways it needs to if so extended.

I think the question you want to ask yourself is: do I really want to design a new RPG, or do I want to design a variant RPG.

In other words, if you take an existing system and impose heavy house-rules it's not really a new RPG. Most of the clones on the market aren't new RPGs, but are old RPGs built around a new setting with some occasional changes. A lot of existing RPGs have some sort of SRD so that you can start with a d20-based game, a FATE-based game, a BRP-based game, or whatever, and build on that framework.

Building a RPG totally from scratch is a lot of work and requires a LOT of playtesting. Often, a game mechanic which sounds great in my head turns out to be mediocre when I actually give it a test-drive. Starting from scratch with attributes, dice modifiers, skill or class lists, and so on, can take a lot of time and creativity.

Quote from: Ocule on October 17, 2022, 05:32:50 PM
I've been wanting to design my own rpg for a while, tailored to my own tastes. To what end? I don't know yet, so starting with some generic ideas. Something that starts as a baseline to springboard other ideas and concepts, with the baseline being somewhere around realistic while being playable.

Enough rambling the first thing I'm considering is a core mechanic or basic resolution system and dice math.

Linear math. Roll a dice and try to beat the target number.
-d20, osr does this has a linear probability adjusts in 5% increments. Has a pretty wide range can be kind of swingy. Also means characters must progress farther to be more reliable.
-d10, interlock system is a good example, same as above but a smaller range of numbers. It seems by shrinking the range you end up more often as highly unlikely to fail for competent characters or highly unlikely to succeed if you suck at something.
D100, the most transparent in terms of success chance. Usually a roll under tends to put a cap on progression.

Bell curve,
2d6 or 3d6, like gurps or Cepheus. Savage worlds I think also falls into this category. Characters seem like they'll reliably succeed or fail depending on the target number but still allow for exceptional results.

Dice pool, exponential curve.
Usually d6s, but I've seen d8s and d10. Can set the difficulty on successes needed and what numbers count as a success. Can really obscure probability for less math savvy players. This curve makes each additional die increase the chance of success dramatically, though at the expense of needing a bunch of dice.

Anyway I'm talking out load thoughts on resolution mechanics?

1d6 is also an option.  Standard Roll, is 1d6.  You have a few options for Target Numbers, and there is a meaningful difference between their different odds of success.  You can also pull in Advantage / Disadvantage, and use a 2d6 Roll.  This is the path I have chosen.

Before you settle on a task resolution mechanic one of the things you should define first is what are you looking for that isn't already done by other games out there? What are you trying to get out of this system specifically? What sort of things about existing games don't you like that you want changed? Etc.

Starting with task resolution is not necessarily a bad thing, but it probably won't get you far unless you want to experiment with a specific task resolution for some reason. And even then, it'll probably just turn into a game design exercise, rather than an actual game. But keeping in mind what your greater design goals are will give you direction, and get you beyond just fiddling with some numbers trying to figure out how players will make rolls in this game.

That being said, sometimes playing around with different mechanics and ability layouts to see want you come up with can give you an initial insight into how games are configured (what works, what doesn't, etc) that might help you later on once you're ready to move on to a more fleshed out design. But you have to be ready to cut things out if they don't serve your greater design goals, and avoid getting married to ideas that you come up with in a vacuum, or that may conflict with or over complicate the things you truly want out of the system.

One option for dice you missed is "roll multiple dice, use best." Savage Worlds uses this and there's also the versions from LUG Star Trek and the Silohette system (Jovian Chronicles, Heavy Gear). Variations include adding bonuses to the result (skill level in LUG, your attribute in Silohette... each reflecting their setting in whether skill or inborn ability should be more important to a reliable outcome), add a wild die/dice explosions (LUG used a wild die that if it came up 6 added to your next highest or a 1 resulted in complications, Savage Worlds has every dice rolled have the potential to explode), or multiple results adding (Silohette has each 6 rolled on a check add +1 to the result... so a 6, 5, 3, 6 counts as a 7 (6+1).

Savage Worlds use results in people not being guaranteed a success no matter how high their skills because sometimes both dice end up being 3's or less and spending a Benny only gets you a reroll with no guarantee the dice will cooperate. This works rather well for pulp-style games where even the best heroes flub it at times and unskilled types occasionally pull of miraculous results (case in point, someone with a d4 skill and a -4 penalty managed something like a 19 check result a couple weeks back... managing to diffuse a situation with suspicious natives that would have broken out into combat).

For LUGTREK the system used multiple d6's for attributes and static numbers for skills meant that the higher your attribute the more reliable your skill performance became, but you really needed your skill to be high enough to hit the target number -6 if you wanted to be good at something. A high Intelligence but only 2 ranks in Propulsion Engineering meant a TN 9 engine repair was beyond you without extreme luck or a bit of metacurrency.

Silohette by contrast had attributes as the steady value (-3 to +3) and skills as 1 or more d6s based on skill level, meaning talent was very important for hitting basic thresholds (as befits the anime robot pilot genre) while increasing in skill meant both level of performance both smoothed out (more likely to get a six on one of the dice) and allowed a chance of something even more (multiple 6s, each adding another +1) and in a system based on margin of success on opposed rolls and needing to reach a threshold to do meaningful damage... every +1 counted.

Different dice systems give very different feels and finding the right one for the genre you're emulating is important. It's why I invariably create bespoke systems for anything I run rather than trying to hammer a generic system into a genre it's not meant for.

Savage Worlds is great for a pulp setting (I'm adoring it for Star Wars... even if we're finding that refluffed fantasy or modern numbers work better than the sci-fi ones), but I'd never use it for something like Star Trek because it's way too swingy.

For that I'd probably do a hybrid of Silohette and LUGTrek (Silohette dice addition with LUGTrek's roll for attribute with static skill and slot in something closer to WEG's attributes (i.e. mechanical, knowledge, technical, perception) so 3/4 of the skills don't fall under the Intelligence attribute).

I'm working on addressing this exact question for a system of my own, and I think that decision is best informed by having a pre-understanding of the range of modifiers you want.

For example, OSR D&D usually keeps the modifier range on something like an attack roll in roughly the -5 to +10 range (though for most of the game it's more like -3 to +5), which works pretty well on a D20, with target difficulties usually somewhere between 8 and 25. One of the major features I want in my system is all skills being affected by multiple attributes as well as skill ranks (I also want a simple "die+modifiers over target" system. That means that I'm looking at modifiers which would have to start no lower than around +6 to +8, and easily spiral upward into the teens or 20s as characters progress. If I keep it on a D20, that means that as the game goes on the game will get noticeably less "swingy", i.e., the dice will matter less and the modifiers more. I could go that route, but it will lead to a game where a low-level character is not much threat to a high-level one. Alternatively I could go with 2d20, and push things back into the "swingy" category, or even go up to a d100, increase the modifiers even further, and wind up with something more similar to a MERPS/"Against the Darkmaster" type system.

Well, what kind of mechanics are you thinking on building on top of it?  Specifically, what are the mechanics which are going to be hogging player attention during combat? During roleplay?

If you're thinking of doing something in an existing system, then you should probably spend a moment reverse engineering that existing system and try to figure out if you can further optimize it by changing the core mechanic. For example, I generally think that D20 games could better be played with D12s because the D12 favors more easily memorized arithmetic, and an 8.3% bump to a skill is an improvement most players can feel.

For my own game (see link to my core mechanic) I knew I wanted to implement a FIFO stack of declared actions instead of an initiative roll. This meant I was going to have a lot of player attention going to micromanaging their Action Points and manipulating or strategizing their use of the Stack. The FIFO stack, however, uses a lot of player memory in the short term. It's relatively robust because it's decentralized--all players who have an action in the stack will have an idea who should go next--but they can't stop to do a percentile roll while thinking about where their action is in the stack.

This was why I opted for a step die pool. Players can generally remember where their action is in the stack while fishing for 1d8 and 2d10.  Moving the crunch of the core mechanic off the arithmetic and onto the dice is an intentional decision to adapt how it hits the players' brain.

I've been thinking a lot on this... that maybe my approach is wrong and I need to kind of sus out and maybe starting with building on a setting or style first.

So far my own desires in a game are
-combat should be quick and brutal. Between 1-3 rounds
- a gritty gameplay that also does not assume characters are incompetent, characters should be enjoyable out of creation.
- mechanics should be intuitive and easy to understand, forcing players to strategize if they want to survive
-realistic in a sense that with any of the abstract systems you can still look at it and go okay I can see that happening
- handle a fairly wide scale to account for monsters with inhuman stats,
-monsters are actually fucking monsters
-assumed setting usually modern or near future sci-fi. Love me some grit and horror

Quote from: Steven Mitchell on October 17, 2022, 07:05:28 PM
All game models put an effective gap on progression.  Some are just more obvious and strict about it than others.  Just because a mechanic can mathematically be extended, doesn't mean that it will continue to work in all the ways it needs to if so extended.

There are game models that don't put any caps on progression.

Quote from: gatorized on November 02, 2022, 09:11:13 PM

Quote from: Steven Mitchell on October 17, 2022, 07:05:28 PM
All game models put an effective gap on progression.  Some are just more obvious and strict about it than others.  Just because a mechanic can mathematically be extended, doesn't mean that it will continue to work in all the ways it needs to if so extended.

There are game models that don't put any caps on progression.

Some are not as obvious as others, but they all have caps.  Name one.

Edit:  To be fair and clear, a game model is not simply the math and mechanics, but also what it simulates, describes, etc.  Which is why I said originally an "effective gap on progression". 

Quote from: Steven Mitchell on November 02, 2022, 09:46:05 PM

Quote from: gatorized on November 02, 2022, 09:11:13 PM

Quote from: Steven Mitchell on October 17, 2022, 07:05:28 PM
All game models put an effective gap on progression.  Some are just more obvious and strict about it than others.  Just because a mechanic can mathematically be extended, doesn't mean that it will continue to work in all the ways it needs to if so extended.

There are game models that don't put any caps on progression.

Some are not as obvious as others, but they all have caps.  Name one.

Edit:  To be fair and clear, a game model is not simply the math and mechanics, but also what it simulates, describes, etc.  Which is why I said originally an "effective gap on progression".

Prowlers and Paragons, for example.

Quote from: rytrasmi on October 17, 2022, 06:47:43 PM
There's something in the gambling center of the lizard brain that gets stimulated when the roll immediately shows the success/failure result. I think roll, add/subtract mods, compare modded result to target number is popular due to inertia mainly. Don't get me wrong, it's perfectly functional, but post-roll arithmetic can dampen the thrill of the roll.

I've seen this same thing. Our homebrew system is opposed single die rolls (usually of different sized dice), with players meeting or beating the GM roll. I've seen my players play five different mechanisms and the mechanisms without math in the roll are the ones that capture their attention the most. They really like the opposed rolls, because even if you roll a 2 or a 3, the opponent could roll even less. (I like opposed rolls because it means as GM I get to roll a lot too.)

In Traveller, people in our group pay a little attention to normal skills rolls (2d6 + bonus >= 8 standard target). But people really got into it when the GM used chase rules and we had 2d6 + bonus versus his 2d6 roll. That actually generated excitement. (We were using Foundry VTT, so it retained the no-math aspect since Foundry was doing the math and the results were instantaneous.)