Ah, my mistake. So the die roll is 2d10, discard higher result, and on a result of 9, reroll and add, repeating the last step as many times as applicable.

A little random at the top, but it fits with what you're going for.

Quote from: Daddy Warpig;503927Fundamentally speaking, the Log chart inspires exactly the wrong approach to Torg.

And now I understand what you're going for.

Quote from: Daddy Warpig;503927Mechanics tell players and GM's what is considered important in play. They focus the players' attentions on specific things. By this, they influence what people think of at the table.

Indeed they do, which is why I never ended up playing TORG, or even thinking about it, the way you describe. So disregard any 'system needs' comments I may have made, as you definitely have a handle on it.

You've even changed my mind on dead weight to a degree and got me thinking: Doesn't the 'exploding die' outstrip the 3 degrees of success generated by dividing by 3 rather frequently? In other words, it seems like an exploding die will pretty much guarantee an Incredible Success. So maybe the degrees should be calculated in a more logarithmic fashion, such as > 4, > 8, > 16 instead of > 1, > 3, > 6.

Quote from: chaosvoyager;504118Indeed they do, which is why I never ended up playing TORG, or even thinking about it, the way you describe.

You mean it never quite clicked as an Action-Movie game? That's too bad. It worked for me and my friends, because I did all the V&M stuff, and they never had to worry about it.

From the beginning, I focused on the Action-Movie elements, as GM. Now that I'm building a system of my own, I can ensure that the mechanics are focused on those elements as well.

(Analogy: Having the V&M in Torg is like dating a supermodel, whose bodyguard punches you in the gut every time you kiss her. Sure, it's great, because hey, Gisele Bundchen and kissing. But wouldn't it be better to date a supermodel, whose bodyguard didn't punch you? All the kissing, none of the gut-hurt. Building my own system means I can ensure the latter.)

Quote from: chaosvoyager;504118Doesn't the 'exploding die' outstrip the 3 degrees of success generated by dividing by 3 rather frequently? In other words, it seems like an exploding die will pretty much guarantee an Incredible Success.

Well, I've run the numbers. I'll give the analysis first, and save the probabilities for the end.

Let's grind some math. When the DN and the Total Skill are exactly equal, you need a minimum of +1 to succeed. 0 or lower, and you fail. Failure happens 55% of the time.

The General Success Chart is as follows:

0 SL = Failure
1 SL = Barely succeeded
2-3 SL = Success
4 SL = Superior Success

When DN = Total Skill, to Barely Succeed you need 1 SL, which happens 45% of the time. 55% of the time you fail.

To get a Success, you need a result minimum result of 4, which happens 21% of the time. 1 time out of 5, you just succeed.

To get a Superior Success, you need a minimum result of 10, which happens 1% of the time. (1% to get the 9, if you do, you're guaranteed at least a +1 on the reroll, making it a 10).

When the Total Skill is equal to the DN, you get a Superior Success 1% of the time. (Side effect? Unskilled people can't ever get a Superior Success, even if their Attribute Base is equal to the DN and they roll a 9. Just succeeding is all they can aspire to.)

Here's how the probabilities break down:


Result % Chance this or better
1 1SL 45%
2 36%
3 28%
4 2 SL 21%
5 15%
6 10%
7 3 SL 6%
8 3%
9 (1%)
10 4 SL .9%
11 .8%
12 .7%
13 5 SL .6%
14 .5%
15 .4%
16 6 SL .3%

Increasing your skill by 3 plusses, only gives you (on average) 1 SL more. So, if your skill exceeds the DN by 3, you get a Superior Success 6% of the time. If it exceeds it by 6, you get a Superior Success 21% of the time.

And so forth.

The heavily 0-weighted bell curve, and the fact that hot rerolls happen 1% of the time, means super-results are fairly rare. Not impossible, but not gamebreaking.

(Or, at least, that's what my math-fu is telling me. Other analysis welcome.)

Quote from: Daddy Warpig;504136You mean it never quite clicked as an Action-Movie game? That's too bad. It worked for me and my friends, because I did all the V&M stuff, and they never had to worry about it.

From the beginning, I focused on the Action-Movie elements, as GM. Now that I'm building a system of my own, I can ensure that the mechanics are focused on those elements as well.

Actually, if we're going for pure Action Movie feel, don't bother with the concrete measurements at all. If a GM wants to know the exact measurements, put them in the "Behind the Scenes" section. But as far as the players are concerned, all they need to know is a general description of how hard it will be: Easy even for normal players, Average, Difficult even for the experts, etc. That, combined with knowledge of how talented they are will be enough for them.

Quote from: Daddy Warpig;504136(Analogy: Having the V&M in Torg is like dating a supermodel, whose bodyguard punches you in the gut every time you kiss her. Sure, it's great, because hey, Gisele Bundchen and kissing. But wouldn't it be better to date a supermodel, whose bodyguard didn't punch you? All the kissing, none of the gut-hurt. Building my own system means I can ensure the latter.)

Personally I don't consider pulling out a calculator to be the equivalent of being punched by a bodyguard, but even so, the problem is solved by having a good enough PR guy to get you in good with the bodyguard so he doesn't punch you in the first place...

Quote from: Daddy Warpig;504136Well, I've run the numbers. I'll give the analysis first, and save the probabilities for the end.

Let's grind some math. When the DN and the Total Skill are exactly equal, you need a minimum of +1 to succeed. 0 or lower, and you fail. Failure happens 55% of the time.

The General Success Chart is as follows:

0 SL = Failure
1 SL = Barely succeeded
2-3 SL = Success
4 SL = Superior Success

When DN = Total Skill, to Barely Succeed you need 1 SL, which happens 45% of the time. 55% of the time you fail.

Simple enough. I would phrase it as "to have any success, barely or otherwise," but that by itself isn't crippling.

Quote from: Daddy Warpig;504136To get a Success, you need a result minimum result of 4, which happens 21% of the time. 1 time out of 5, you just succeed.

In light of the definitions of success, I wouldn't call 2 SL "just succeed." That sounds too much like the Barely Succeeded option of 1 SL. I would call it "succeed without complications" or something like that.

Quote from: Daddy Warpig;504136To get a Superior Success, you need a minimum result of 10, which happens 1% of the time. (1% to get the 9, if you do, you're guaranteed at least a +1 on the reroll, making it a 10).

Next nitpick. You only need the 9, but with the hole in the chart, you won't get an exact 9. That's what you're trying to say, but what you did say is slightly inaccurate.

Quote from: Daddy Warpig;504136When the Total Skill is equal to the DN, you get a Superior Success 1% of the time. (Side effect? Unskilled people can't ever get a Superior Success, even if their Attribute Base is equal to the DN and they roll a 9. Just succeeding is all they can aspire to.)

The "even" here is confusing. It implies that no unskilled person can ever get an Attribute Base higher than the DN, which is not true. Removing it conveys the essential idea and doesn't create confusion.

Quote from: Daddy Warpig;504136Here's how the probabilities break down:

Code Select Expand


Result % Chance this or better
1 1SL 45%
2 36%
3 28%
4 2 SL 21%
5 15%
6 10%
7 3 SL 6%
8 3%
9 (1%)
10 4 SL .9%
11 .8%
12 .7%
13 5 SL .6%
14 .5%
15 .4%
16 6 SL .3%

Increasing your skill by 3 plusses, only gives you (on average) 1 SL more. So, if your skill exceeds the DN by 3, you get a Superior Success 6% of the time. If it exceeds it by 6, you get a Superior Success 21% of the time.

And so forth.

True, and the in-between values can be figured out as well.

Quote from: Daddy Warpig;504136The heavily 0-weighted bell curve, and the fact that hot rerolls happen 1% of the time, means super-results are fairly rare. Not impossible, but not gamebreaking.

(Or, at least, that's what my math-fu is telling me. Other analysis welcome.)

I've been playing with the (d10-d10) with rerolls on either die when it hits 10. That has a 9.09% chance of a positive reroll, which is more in line with what the original TORG had as a re-roll chance, but may be too unbalancing for what you are trying to do.

I still think it's easier to take 2d10 directly and add it to either a base DN or one based on the target, adding 11 to convert it to the same probability base.

Here's an odd effect of the probability charts which I didn't anticipate, but which works for me.

If your skill is substantially lower than a DN, say more than 9 points, each plus you gain is a negligible increase in the odds to barely Succeed: .1%.

As you get close and closer to the DN (raising the skill through experience), your increases are larger: 1%, 2%, 3%.

When you're in the near vicinity of the DN (-4), you're getting 7%, 8%, 9%.

Compare this with a system where each +1 is exactly 1% increase. (Like a straight roll-under percentile system.) Advancement is exactly even.

With this die method (and, I'd assume, other bell curve systems), each plus is actually more valuable, the closer you get to the DN. When you get very close, From -2 to +2, you progress from a 28% chance of success to a 36% chance, then 45%, 55%, 64%.

When you're very close to the DN, you move from "usually fail" to "usually succeed" very quickly. At a much lower skill level, gaining 5 plusses may only have increased your odds by .5%, here those same points increases the odds by 36%.

But, once you're surpassed the DN, each additional + does less and less to increase your odds of success, in relative terms.

What does this mean?

In game terms, high DN's are steep challenges. Success is nearly impossible.

Once you're close to mastering them, a little more skill makes a big difference.

Then, when you've already surpassed them, each point in skill makes less and less difference. Success becomes nearly assured.

This progression exactly mirrors the ideas underlying the 5-Stage campaign model. At each stage, there are low skilled opponents, easy to defeat, comparable opponents which may be more or less challenging, and highly skilled opponents, nearly impossible to beat.

Once you advance a stage, the comparable opponents of last stage are now easily defeated, the nearly-impossible opponents are now comparable challenges. This is a good experience curve, it makes the players feel like they are progressing.

D&D implements this through levels, here's it's implemented in relative skill plusses and the Dice Method.

EDIT: It also mirrors real-world skill mastery. When learning a new skill, people struggle for a very long time, seemingly to no avail. They don't get perceptibly better. Then, once a critical threshold is crossed, they quickly move from "that's impossible!" to "that's routine."

After that more study does't make a big difference in mastering that task, it becomes a process of periodic discoveries, small additional increases in skill. When this is achieved, most people are working towards mastering another, more difficult part of their field of study, repeating the process again.

Example, in mathematics: Arithmetic->multiplication/division->geometry->algebra->calculus->advanced calc.

This is mirrored in the descriptions of the skill levels I gave earlier and the "relative individual skill levels".

I really like this effect. It works in terms of game balance, rewards players for gaining experience, and mirrors real-world phenomena. That's a win, in my book. YMMV.

I think we have an interesting start here. The main thing that still bothers me is that "equal skill vs. difficulty = 45% success" mechanic. For me, this is an affront on verisimilitude; if my skill is equal to the task (literally), I should succeed at least as often as I fail. It's not the punch from the bodyguard, but for me it's like having to fill out the Storm Knight Registration Act forms every time I want to date the supermodel.

I don't want to rehash the "should matching equal success" debate from the other boards, but I do want to propose a way that a straight 2d10 can solve it. If we go from the hot/cold die model to a straight 2d10 rolling bonus d10s for a natural 20 and penalty d10s for a natural 2, we have a topologically equivalent die rolling system. But with the simple change of adding 10 to your difficulty numbers instead of 11, it allows a skill of 8 to succeed against average difficulty tasks 55% of the time instead of 45%, making it "equal to the task" in most cases. (The base difficulty would be 8+10, or 18, which 8+2d10 beats 55% of the time.)

Quote from: Rabbitball;504487The main thing that still bothers me is that "equal skill vs. difficulty = 45% success" mechanic. For me, this is an affront on verisimilitude; if my skill is equal to the task (literally), I should succeed at least as often as I fail.

Here's the disconnect. In Destiny, you only roll when:

1.) The outcome is important to the adventure.

2.) Adverse conditions apply.

3.) The DN is higher than your Total Skill. (Really, this is a subset of 2, but I broke it out for specificity.)

or

4.) The player wants to roll anyway. (To get more than just a Bare Success.)

The rest of the time, DN's equal to your skill mean you score a Bare Success. (This is equivalent to the "take 10" rule, but integrated with other, similar circumstances.)

So that 45% success rate isn't for standing in a gun range, plinking at targets, or sitting in the office, lazily looking for a form you misplaced.

It's for "I must shoot this thing now!" and "There's no light in here, and the grues are getting closer. Where's my damn dagger?"

This is a cinematic action movie game. Don't make players bounce dice if the outcome doesn't matter. Don't make them roll if it's a casual exercise of the skill.

Roll out the dice when the shit's flying towards the fan, when the enemy's bootsteps are at the door, or when the roof falls in.

That's what turns something you can easily achieve into something you'll stand a good chance of failing. That's why the 45% is apt. It represents the circumstances under which people will be making checks at all.

(Also, I implemented this deliberately because in combat, ties go to the defender. To hit someone, you need not to equal their skill, but to beat it. That just makes sense.)

But, let's assume you reject the above logic and want to change the percentage from 45% to 55%. You don't have to use a different dice mechanic.

1 change implements it: 1 SL is a result of 0-3, 2 SL 4-6, and so forth.

Same result, no need to change the dice mechanic.

Which die mechanic I like. It's colorful and unique. Using this mechanic also avoids an even bigger problem:

"My skill is 8. The DN is 18. But I succeed half the time? How does that work?"

45% Success rate against a DN equal your skill is a far better solution than 55% success rate against a DN 10 points higher than your skill. It's a far easier sale, with regards to verisimilitude.

In my opinion. YMMV.

Quote from: Rabbitball;504417Actually, if we're going for pure Action Movie feel, don't bother with the concrete measurements at all.

I disagree. The point of the game's mechanics is to easily enable the vivd depiction of the action. Concrete measurements are an important part of that.

"Their car is speeding towards you at 60mph."

"The grenade lands less than a foot away."

"The giant easily lifts the two-ton car and tosses it your way."

All those are evocative descriptions. The distance or speed of things immediately triggers a response.

Grenade 1 ft away = "Danger!"

"Giant can throw a car" = "Badass! Run!"

And so forth. Concrete measurements, while also being valuable to people who chose to use miniatures or other figures, are great ways of quantifying what's going on.

You just don't need a V&M chart to use them. Just say "50 feet" or "2 ton car". Forget the V&M, just describe what happens in concrete, understandable terms.

Quote from: Daddy Warpig;504136You mean it never quite clicked as an Action-Movie game? That's too bad.

It was, but it did teach me a lot about how to speed up this kind of game design.

Quote from: Daddy Warpig;504136Here's how the probabilities break down:

Code Select Expand


Result % Chance this or better
1 1SL 45%
2 36%
3 28%
4 2 SL 21%
5 15%
6 10%
7 3 SL 6%
8 3%
9 (1%)
10 4 SL .9%
11 .8%
12 .7%
13 5 SL .6%
14 .5%
15 .4%
16 6 SL .3%

I think there's a problem, or perhaps I once again don't understand how you're generating the results.

While getting a Hot 9 is indeed a 1 in 100, you actually have the same exact chance of getting any value between Hot 10 and 18 afterwards, which is 1 in 10. In other words, you have the exact same chance of rolling a Hot 10 as you do a Hot 18.

And the only reason this does not apply to a Hot 19 as well is because if you are treating the 0 as a 10, you actually have TWO ways of generating a value of 19: an initial roll of 9 + reroll 10, AND an initial roll of 9 + reroll of 9 + reroll of 1. And this meas that you are actually TWICE as likely to get a Hot 19 than anything between Hot 10 and Hot 18.

And to be honest, I still say the division by 3 could and should be eliminated. But if this division is only being used to calculate SLs, then it may still make sense to stick with it. So what are the raw values being applied to again?

Finally, open ended FATE rolls are easy: Just have the player roll additional dice whenever all their dice come up the same result (all +s, -s, or 0s). In fact, I even let them choose how many extra dice (within reason) they want to roll when this happens.

Here's an Anydice link to the probabilities for upto 8dF.

Quote from: chaosvoyager;505221While getting a Hot 9 is indeed a 1 in 100, you actually have the same exact chance of getting any value between Hot 10 and 18 afterwards, which is 1 in 10. In other words, you have the exact same chance of rolling a Hot 10 as you do a Hot 18.

Right so far.

Quote from: chaosvoyager;505221And the only reason this does not apply to a Hot 19 as well is because if you are treating the 0 as a 10, you actually have TWO ways of generating a value of 19: an initial roll of 9 + reroll 10, AND an initial roll of 9 + reroll of 9 + reroll of 1.

You don't get a reroll for rolling a 9. You get a reroll for Maxing the Die.

On the first 2d10 roll, the maximum the die can be is 9. (Because a "10 and 9" is a 9 and 2 10's are 0.) You Max the Die by getting a 9.

When you roll again, you're rolling a number from 1 to 10. The maximum the die can be is 10. You reroll when you Max the Die.

That is fairly straightforward. I admit, if you think "reroll on 9" it's confusing. But it isn't "reroll on 9", it's "Reroll on Maxes".

Maxed the Die? Reroll and add again.

Quote from: chaosvoyager;505221But if this division is only being used to calculate SLs, then it may still make sense to stick with it. So what are the raw values being applied to again?

The result, the amount by which the Skill Check exceeded the DN, is used to calculate Success Levels. That's it. Everything is done according to the Success Levels you generate.

EDIT: Though I'm not sure that would make a difference.

Quote from: Daddy Warpig;505280Maxed the Die? Reroll and add again.

So on the initial roll you discard the maxed dice, but on a reroll you reroll the maxed die. It's a little convoluted because it requires two different mechanics, but not that bad.

I suppose you're OK with incongruities between certain numbers, such as a result of 19 also being a result of 20. It bugs me a bit, and I actually fixed the issue with Shadowrun by treating the 6s as 0s, but it's nothing I can't get past. It's simply not pretty.

Quote from: Daddy Warpig;505280The result, the amount by which the Skill Check exceeded the DN, is used to calculate Success Levels. That's it. Everything is done according to the Success Levels you generate.

Then honestly, I think you should find a way to avoid the division.

Currently, getting an SL>0 is 1000 in 1000, an SL>2 is close enough to a 50 in 1000, and an SL>6 is 1 in 1000. I would try to find a dice rolling method that models these probabilities instead of requiring division by 3. Also, rounding up is less intuitive than rounding down, because while rounding up requires you to change the whole value, rounding down just requires you to discard the decimals.

Quote from: chaosvoyager;506050So on the initial roll you discard the maxed dice, but on a reroll you reroll the maxed die.

Sorry? That's not how I see it.

Roll 2d10. If they tie, you discard both and take a 0.

If they don't tie, discard the higher of the two. Then, depending on whether the kept die was hot or cold, you add or subtract the number on the kept die.

So, when looking at the kept die, the highest it can ever be is a 9. That is a Maxed Die (not dice), as it simply cannot ever be higher.

9 is the Max. Rolling it, you Maxed the die.

Maxing the Die earns a reroll. When you reroll, it can be from 1-10, and rolling a 10 is Maxing that die.

Those are the same mechanic: Max the Die, earn a reroll.

Let us suppose we had mechanic which was a d10, minus some number that varied according to circumstances. On a straight d10, 10 is the Max. You reroll. On d10-1, 9 is the Max. You reroll. On a d10-5, 5 is the Max. You reroll.

That's not three separate mechanics, its the same mechanic. Roll the Max, get a reroll. The number which equals a Max differs, but the rule is the same. Roll a Max, get a reroll.

Savage Worlds, the same. d4, Ace on 4. d6, Ace on 6. d8, Ace on 8. Max the Die ("Ace"), get a reroll. That you max on a 4, 6, 8, or whatever isn't 3 separate mechanics, its the same mechanic.

The rule is: Max a Die, earn a reroll. Sometimes the Max is 9, sometimes 10. Either way, the Max is the Max.

Quote from: chaosvoyager;506050Then honestly, I think you should find a way to avoid the division.

But that's exactly the opposite of what you said last time:

Quote from: chaosvoyager;505221But if this division is only being used to calculate SLs, then it may still make sense to stick with it.

People can change their minds, its like a divine law or something, what I would like to know is why you changed yours? Explaining your reasoning will help me see your point.

Quote from: chaosvoyager;506050Currently, getting an SL>0 is 1000 in 1000,

I don't understand what you mean. The chances of getting 1 SL vary based on what your Total Skill is and what the Difficulty is. The chances of getting 2 SL and higher also vary according to those two numbers.

There is some intricate probabilities that go on within the SL mechanic. I discussed the sudden curve towards success as the Total Skill approaches then exceeds the DN.

Much lower, each additional plus of skill makes a small difference in the % chance of success. Closer, it makes a bigger difference in that chance. As it gets within 1-2 points, the chance of shoots up.

What also happens with a higher Success Level is the same thing. As the Total Skill gets closer to the DN, you get a big boost in succeeding at all, but a much smaller boost in gaining 2 SL and a very small boost in gaining 3 SL.

Here's how it works:

Skill equal to DN = 45% chance 1 SL. 21% chance 2 SL. 1% chance 4 SL.
Skill exceeds DN by 1. = 55% chance 1 SL. 28% chance 2 SL. 1% chance 4 SL.

So, the chances of getting 1 SL increased by 10%, 2 SL by 7%, and 4 SL by not at all. You got much better at barely succeeding, but not much better at succeeding, and not at all better at succeeding brilliantly. This is deliberate.

We can get a clearer picture by presenting a composite probability. On an given roll, what are the chances of getting a failure, 1 SL, 2 SL, and 4 SL or greater? (Done right, they should total to 100%.)

Skill equals DN. 55% fail. 24% 1 SL. 20% 2 SL. 1% 4 SL+.
Skill > DN by 1. 45% fail. 27% 1 SL. 27% 2 SL. 1% 4 SL+.

By exceeding the DN by 1 point of skill, your chances of Barely Succeeding and Success are now equal. You're just as likely to get one or the other. Again, this is desirable.

Skill > DN by 2. 36% Fail. 28% 1 SL. 33% 2 SL. 3% 4 SL+.
Skill > DN by 3. 28% Fail. 27% 1 SL. 39% 2 SL. 6% 4 SL+.
Skill > DN by 4. 21% Fail. 24% 1 SL. 45% 2 SL. 10% 4 SL+.

You've gotten to the point where failure seldom occurs. Nearly half the time, you just succeed. And 1/10 times, you succeed brilliantly.

Skill > DN by 5. 15% Fail. 21% 1 SL. 49% 2 SL. 15% 4 SL+.
Skill > DN by 6. 10% Fail. 18% 1 SL. 51% 2 SL. 21% 4 SL+.
Skill > DN by 7. 6% Fail. 15% 1 SL. 51% 2 SL. 28% 4 SL+.

You fail rarely. Most of the time you succeed (not barely succeed, succeed). And over a quarter of the time, you succeed brilliantly. You've mastered this level of challenge.

This is progression is desirable, and only possible by taking the bell curve along with the "Rule of 3" SL mechanic. It magnifies the effects of mastering a task.

So far as I see it, the 2d10 hot/cold mechanic + Rule of # provide the following benefits:

• It's fairly unique.
• It's colorful (quite literally).
• It provides for "Awesome!" moments on a reroll.
• Models how real world skill-mastery develops.
• Leashes munchkins.
• Is, comparably, fairly simple.

What other method will provide me with all the above benefits?

Quote from: chaosvoyager;506050Also, rounding up is less intuitive than rounding down,

I'm not saying you're wrong, but the problem appears to be minimal. Again, Savage Worlds uses the same mechanic to calculate raises and it seems to work fine in play. At least, of all the criticisms of SW, this wasn't one of them.

That's not playtesting of my mechanic, but it is playesting of a similar mechanic, and it has been received well, from what I can tell.

That hurts my brain. I would never use that die mechanic in a game.

Quote from: danbuter;506232That hurts my brain. I would never use that die mechanic in a game.

Really?

Roll 2d10. Keep lowest.

Equals are 0. If it's a red die, add the number. If it's blue die, subtract.

That's brain-bending?

I'm not trying to convince you, I honestly don't understand why that's complicated. It worked for the Bab5 RPG.

Feel free to tell me why it's too complicated for use in a game.

Because I know I'd have to explain it 10 times every night we tried to use it in a game. Most of my friends aren't interested in learning "innovative" dice mechanics.

Quote from: danbuter;506238Because I know I'd have to explain it 10 times every night we tried to use it in a game. Most of my friends aren't interested in learning "innovative" dice mechanics.

So what kind of die mechanics do you prefer?