Have some sacrilege! Tinkering: skill/combat rolls -> dicepools

[Surgo]

I imagine this is going to seem like blasphemy to many of you. I was working on my Shadowrun rules adaptation and had some difficulty coming up with a way that players could quickly adapt content from books into HERO. Some concepts just didn't translate well. So to move this along, I decided to turn how HERO combat and skill rolls work from 3d6 roll-under into dicepools. As a reminder, in a dicepool like Shadowrun you roll a number of d6's equal to the size of your pool and every 5 or 6 counts as a "hit". Furthermore, you can "buy" a hit by spending 4 dice from your pool.

 

The basics: Your dicepool size is equal to your CHAR/5, plus any skill bonuses. A basic test requires one hit.

 

The probability: Basic skill rolls in HERO are 3d6 roll under your (CHAR/5) + any skill bonuses. At the most basic level (Characteristic of 10, skill purchased) this gives you a 50% chance of making the roll. Under a dicepool of two dice, you have a 55.5% chance of getting your one hit. Not identical, but close.

 

What about when we have bonuses? Let's say we've bought up +4 to the roll (or 4 dice in a dicepool system). We have a 90% chance of success with the standard HERO method. With the dice pool system you can just buy a hit, so you succeed 100% of the time.

 

What about harder checks? Let's say the check is for something hard so you take a -4 on the roll. We're back to 50% chance under the hero system. A -4 in dicepool land means removing 4 dice from the pool, so once again we're back where we started.

 

Honestly, this isn't all that interesting. The probability curve is a little skewed but...it looks similar enough if you squint at it. So why bother? Aside from making it easier to adapt Shadowrun content, I didn't like how opposed tests were working out. Also you get more variance at higher levels of dice.

 

My Shadowrun rules adaptation involve a lot of opposed tests. I've even extended something like the combat-as-skills rules from APG2, so combat represents opposed rolls instead of a roll against a DCV. A lot of powers have a custom advantage that succeeding on a skill roll by a certain amount gives them +5 extra character points in the power. I didn't want +1 to skill rolls to represent a raw +5 points to the power, that's a bit excessive and at high enough skill values you're just going to shift how many extra points you get by the results of that 3d6 roll. The use of dicepools means that an extra die only "counts" for 5/3 extra points in a power on average; the extra points are thus compressed down compared to vanilla HERO rolls by a factor of 3. There's still plenty of room for randomness though (and you get more variance in results with more dice, which I want). Both tests and opposed tests aren't any easier or harder (in a rough sense) by using dicepools instead of vanilla HERO rolls, but you get less impact on your powers from that extra +1.

 

Okay, but who cares? You could just divide how much you beat the roll with by 3 and have something that looks roughly the same. Well, like I said above, using dicepools makes it easier to lift content straight out of the SR4 books. Also I think my players will just find it easier to count. I'm finishing up a long D&D campaign with them and while I have multiple engineers as players, it's at the end of a long work day and nobody's really good at math at that point. Counting hits is easy, handling opposed tests is trivial, and there's no weird addition or subtraction involved.

 

Ultimately the only reason to do this is if you're a filthy blasphemer like me and like dice polls or rolling piles of d6s. Personally I love rolling piles of d6s. It's great fun. Or you want some more variance in your results.

 

Aside: Test Difficulty

 

In general, circumstances should add or remove from the dicepool instead of modifying the number of hits required. Some tasks are going to take more than one hit though. Shadowrun 4th edition recommends the following thresholds:

 

Easy: 1 hit

Average: 2 hits

Hard: 3 hits

Extreme: 4 hits

 

I'd stick with that.

 

Aside: Buying Skills

 

Familiarity is gone. Purchasing the skill gets you the characteristic roll for it (no rolling it without the purchase). After that it's a standard +1 for however many character points the book says.

[Hugh Neilson]

I imagine this is going to seem like blasphemy to many of you. I was working on my Shadowrun rules adaptation and had some difficulty coming up with a way that players could quickly adapt content from books into HERO. Some concepts just didn't translate well. So to move this along, I decided to turn how HERO combat and skill rolls work from 3d6 roll-under into dicepools. As a reminder, in a dicepool like Shadowrun you roll a number of d6's equal to the size of your pool and every 5 or 6 counts as a "hit". Furthermore, you can "buy" a hit by spending 4 dice from your pool.

 

The basics: Your dicepool size is equal to your CHAR/5, plus any skill bonuses. A basic test requires one hit.

I think I am with you so far. So my character with the skill at 11- gives me two dice, at least one of which has to come up 5 or 6.

 

The probability: Basic skill rolls in HERO are 3d6 roll under your (CHAR/5) + any skill bonuses. At the most basic level (Characteristic of 10, skill purchased) this gives you a 50% chance of making the roll. Under a dicepool of two dice, you have a 55.5% chance of getting your one hit. Not identical, but close.

Your math is wrong. Under the standard Hero approach. I need to roll 11-. That is a 62.5% probability of success, not a 50% probability (that would be a 10-). With 2d6, there are 36 possible rolls, of which 20 will have at least one 5 or 6, so 55.56% success. Pretty close, but success is a bit more, not slightly less, likely under Hero rules.

 

What about when we have bonuses? Let's say we've bought up +4 to the roll (or 4 dice in a dicepool system). We have a 90% chance of success with the standard HERO method. With the dice pool system you can just buy a hit, so you succeed 100% of the time.

Again, the math is wrong. At +4, the roll is 15-, which is 95.37% likely. If I get the dice pool system, I need +2 to the roll, or an 18 or higher characteristic, to hit 4 dice. 13- would succeed 83.8% of the time, so Hero rules leave a 1 in 6 chance of failure at the minimum point the dice pool will provide guaranteed success.

 

What about harder checks? Let's say the check is for something hard so you take a -4 on the roll. We're back to 50% chance under the hero system. A -4 in dicepool land means removing 4 dice from the pool, so once again we're back where we started.

We’re back to the starting point of getting the Hero math wrong, but as long as we remove 1 die from the pool for each penalty, we’ll get the same difference between Hero success odds and Dice Pool odds.

 

Honestly, this isn't all that interesting. The probability curve is a little skewed but...it looks similar enough if you squint at it. So why bother? Aside from making it easier to adapt Shadowrun content, I didn't like how opposed tests were working out. Also you get more variance at higher levels of dice.

I’m not sure how it became so much easier to adapt Shadowrun content, if it’s that easy to equate the two models. One benefit I see, if one wants it, is that we get an “automatic success on tasks which are pretty much routine” – basically a “take 13”.

 

My Shadowrun rules adaptation involve a lot of opposed tests. I've even extended something like the combat-as-skills rules from APG2, so combat represents opposed rolls instead of a roll against a DCV. A lot of powers have a custom advantage that succeeding on a skill roll by a certain amount gives them +5 extra character points in the power. I didn't want +1 to skill rolls to represent a raw +5 points to the power, that's a bit excessive and at high enough skill values you're just going to shift how many extra points you get by the results of that 3d6 roll. The use of dicepools means that an extra die only "counts" for 5/3 extra points in a power on average; the extra points are thus compressed down compared to vanilla HERO rolls by a factor of 3.

So you need to change to a dice pool because your custom advantage provides too big a power enhancement for each 1 point the roll succeeds by, did I get that right? You can’t just change the custom advantage to add +2 CP for every point the roll succeeds by (that’s 6/3, so a little more) or +1.5 CP for each point of success (4.5/3)?

 

There's still plenty of room for randomness though (and you get more variance in results with more dice, which I want). Both tests and opposed tests aren't any easier or harder (in a rough sense) by using dicepools instead of vanilla HERO rolls, but you get less impact on your powers from that extra +1.

Actually, you get less randomness with more dice as they trend more towards the average. Rolling 2d6, I will get a 5 or 6 on both 4 times in 36, or over 10% of the time. Roll 12d6 and see how often you get more than 4 5s or 6s. Getting all 5s and 6s? You will be rolling for a LONG time.

 

Meanwhile, as noted above, we move from “1 in 6 rolls fails” to “all rolls succeed” at the Hero 13-/4 dice pool level

 

Okay, but who cares? You could just divide how much you beat the roll with by 3 and have something that looks roughly the same. Well, like I said above, using dicepools makes it easier to lift content straight out of the SR4 books.

It would be even easier if you just played Shadowrun, wouldn’t it? I’m not clear where you see a benefit of moving to Hero if you prefer the Shadowrun mechanics.

 

Okay Ultimately the only reason to do this is if you're a filthy blasphemer like me and like dice polls or rolling piles of d6s. Personally I love rolling piles of d6s. It's great fun. Or you want some more variance in your results.

No one who plays Champions can complain about rolling piles of d6s.

 

Too many quoted blocks - have to break this to multiple posts.

[Hugh Neilson]

Okay In general, circumstances should add or remove from the dicepool instead of modifying the number of hits required. Some tasks are going to take more than one hit though. Shadowrun 4th edition recommends the following thresholds:

 

Easy: 1 hit

Average: 2 hits

Hard: 3 hits

Extreme: 4 hits

 

I'd stick with that.

OK, here it goes off the rails for me. I would suggest an “easy” task in Hero is not a base skill roll. A person can be qualified to work in a profession with a Familiarity, which is an 8- roll, so clearly there is a bonus if one is undertaking an easy task.

 

Getting 2 hits with that basic skill under the dice pool model? You have an 11.1% chance. However, if one looks at p 58, an “easy” task should have a bonus of +1 to +3, so an 8- skill roll can succeed at an “easy” task (2 point bonus in the middle) half the time.

 

An average task is the baseline roll. A difficult one has a penalty of -1 to -3, “extreme” is -3 to -5 and “sheer folly” is -5 or worse.

 

Based on the Skill Roll Table in 6e v1 p 56, someone with a 12- roll (3d6 in the pool) finds routine tasks easy and more difficult tasks well within his abilities. To me, he should be able to succeed in an average task pretty routinely, and have a shot at hard tasks. So how will that work out in your model?

 

Well, he will succeed in an Easy task (roll at least one 5 or 6) 152/216 times, so 70.37% of the time. Frankly, that does not feel all that Easy.

 

How does he do on an Average task? 56 out of 216 combinations of 3d6 will have at least 1 5 or 6 – so he succeeds on an average task 25.93% of the time. He does not feel very skilled to me. He probably needs 5d6 (14-) to feel like he has a decent shot at achieving an Average task, but he would be 90.74% likely to succeed in such a task under Hero rules.

 

A Hard task? He has 8 chances in 216, so 3.7% success. Wow, that does not feel like a task which is “well within his abilities”. In Hero, he has a 10- shot at success with a -2 penalty, so 50% likely. Now, at 9d6 in a dice pool, he can be virtually guaranteed of success (one of them has to come up 5 or 6, and he can trade out the rest). But that’s an 18- skill in Hero, so he would be 16- (98.15% chance of success) with a -2 penalty. Gut feel, I am thinking he needs 6 or 7 dice in the pool to feel good about his odds – he will need to either roll 3 5’s or 6s, or roll 2 and take 4 dice away. So we are up to 15- or 16-.

 

An Extreme task? Dice pool means no way. Hero says 7-, so a 25.93% chance of success.

 

Above, you say you are looking for more variance. However, your model is definitely providing less variance once we move to the “x Successes” model.

 

I think it would be better to choose between penalties/bonuses reducing/increasing the pool and penalties/bonuses increasing/reducing the number of required successes.

 

Aside: Buying Skills

 

Familiarity is gone. Purchasing the skill gets you the characteristic roll for it (no rolling it without the purchase). After that it's a standard +1 for however many character points the book says.

OK, so Everyman skills are gone – no one can have odds less than 55.5% to perform an Easy task, and if you did not buy the Climb skill to be at least 55.5% likely to be able to pull yourself up on a rope (which I would suggest is an Easy task, with a +2 or so bonus, in Hero), you are utterly incapable of the task.

 

This also ignores how often the Hero means of that 8- Climb Check succeeding is “extra time”. You need a 10- to climb to the top of the rope in 1 phase? You can spend a minute and have a 74.07% chance.

 

END RESULT

 

I think the dice pool mechanic could work, but I think an Average task should be a single success with 2d6 55% of the time. Maybe that means Familiarity should be enough for an Easy task, so you get 1d6 at an 8- and succeed in an Easy task 1/3 of the time.

 

That brings an 11- to 4d6 – an easy task is automatic. The step up from Easy to Average is 1 to 3 dice, so call that 2d6 removed. We are now back to 2d6 for an average task if you have an 11- roll. Take another 2d6 away to make it Difficult, and another 2 (or more) for Sheer Folly, and we are getting somewhere.

 

While that would fix the odds a bit, it bears noting that, especially with the “4 dice = 1 success” rule, the volatility falls, rather than rising. Enough dice to feel reasonably confident at one step of difficulty is an autosuccess at the next level of difficulty down. In Hero, an 11- becomes a 13- if the task is one step more difficult, so more likely, but still not guaranteed.

 

Evaluation of the dice pool mathematically would be a ton of work that I’m not inclined to undertake. What are the odds of, say, 7d6 generating 3 5s or 6s (which would be success at an extreme task since the remaining 4 dice can simply be traded in)? Way better than the odds of rolling 3 5s and 6s on 3d6, since any 3 of the 7 can come up 5 or 6.

 

My guess is that there are benchmarks in Shaowrun for a poor, average, good, great, phenomenal skill – how do those align with the benchmarks in Hero?

[dsatow]

A long time ago, in a town not too far away, I was thinking how can I homogenize the mechanics of HERO a little more.  This was around 4th ed.  This was the root of what I came up with.  Its not great, but possibly workable.  Beyond thinking up the idea, I did not explore it farther.

 

Treat skills like normal damage:

 

Treat all characteristics like Strength in that 5 points gives you 1d6.  Treat difficulty as Def in 2 x rank difficulty = Def.  A difficulty of 1 would be Def 2.  A difficulty of 3 would be Def 6.  Normal human scale would be from 1-5.  Super scale would be 6-10.  Skills would be "extra dice" bought for 2 points per die.  "STUN" would be the average time or rolls to complete.  If you exceed the Def in Body, you succeed easily regardless of the amount of "Stun" left.  Any bonuses would be +1d6 to the damage roll.  Complementary skills would not give bonuses to the roll but rather subtract from the Stun of the difficulty.  Thus, if you only have a familiarity with something (about 1 or 2 dice), you are not going to really help. 

 

So a person with a 20 INT and science +2d6 would roll 6d6 for that science.  They have a problem of difficulty 4 (8 Def).  The GM thinks its gonna take about 5 rolls to complete or about 60-75 "Stun".

[Durzan Malakim]

Treat skills like normal damage:

 

I like this variant because it would make skill checks an ongoing challenge that multiple people can cooperate on rather than a one-phase pass-fail challenge. How would you assign STUN to difficult tasks? Is something like developing a unified field theory in physics a 6 DEF, 100 STUN task or a 6 DEF 1000 STUN task?

[doccowie]

Always delighted to discuss rule tweaks. I like the unified approach - it lets you differentiate between tasks that are just too hard for any normal person to do, and those that anyone can achieve with time.

 

Obviously you set the hardness and "stun" based on what you want as a GM

Unified Field Theory in physics? OK, not solved by the smartest people in the world so far. Let's assume they have 20INT, and +3d6 skill.

Do we think it's just really hard, but once you get it it's pretty simple? Like say - relativity? OK Hardness 10 STUN 10. A few really good rolls from a really smart person and it's cracked!

Or it's tough, and there's just so much to it. Like, say - quantum physics? So, we're looking at Hardness 7 STUN 1000. Lots of people can chip away over time, and you're probably going to get there in the end.

 

Obviously can integrate modifiers for time, once you decide on the base time for the skill check add a Hardness for every step down the time scale!

[dsatow]

I like this variant because it would make skill checks an ongoing challenge that multiple people can cooperate on rather than a one-phase pass-fail challenge. How would you assign STUN to difficult tasks? Is something like developing a unified field theory in physics a 6 DEF, 100 STUN task or a 6 DEF 1000 STUN task?

It would depend on what you thought the average number of rolls would take to complete.  Lets say an average scientist was doing 6d6.  That would be 15 STUN each hit.  So if you though it would take 5 hits, that would be a 75 point Stun task.  

 

Again, I did not think it thoroughly through, but I am glad you liked the idea.

[doccowie]

Ah, I was thinking of it being similar to attacking an object - you want to smash that unified field theory, not just STUN it. I should have said BOD 10 :- )

 

So with Hardness 10 even a good  scientist (15INT, +3d6 Skill = 6d6) will probably never crack it, especially if you assume that the period of time per skill roll is, say, 1 year - 30 rolls over a career, chance of getting 11BOD on the roll more than once very small.

Once you get up to 7d6, or even 8d6 then you begin to stand a chance, and if you were Dr Destroyer (30INT, +4-5d6 Skill) you'd be chipping away at it every other year, probably solving it within your lifetime.

And that's something really tough!

 

Or, if you assumed it was something that lots of people could work towards, maybe make it Hardness 7 and BOD 100, those 15INT +2d6 scientists are beginning to contribute, especially if you let them Coordinate. A few labs with 20 really smart +6/7d6 scientists will solve it in a single generation!

[dsatow]

I'd probably use BODY for breakthroughs.  Do enough Body and enter a new theory of the science.