Some Questions

Discussion on game mastering Earthdawn. May contain spoilers; caution is recommended!
Tattered Rags
Posts:374
Joined:Mon Nov 28, 2016 12:04 am
Re: Some Questions

Post by Tattered Rags » Thu Oct 18, 2018 10:06 pm

Slimcreeper wrote:
Thu Oct 18, 2018 9:34 pm
It niggles me less than a D20 system where the roll pretty quickly becomes subordinate to the modifier. I think 5th DnD is better about that, but I still prefer the Step system. And it niggles me less than the entire Shadowrun system (despite the fact that I listen to two AP Shadowrun podcasts pretty religiously!)
Ugh. Don't get me started on linear probability curves in my rpgs.

Huh. I love the Shadowrun system...aside from almost dying if I accidentally knock the dice on myself.
Last edited by Tattered Rags on Fri Oct 19, 2018 10:34 am, edited 1 time in total.
Adventure I'm running:
Under the Stars

Adventure GM post-mortem:
Under the Stars Postmortem

Slimcreeper
Posts:1061
Joined:Mon Nov 28, 2016 11:44 pm

Re: Some Questions

Post by Slimcreeper » Fri Oct 19, 2018 4:00 am

I earnestly promise to not get you started on linear probability curves.

I bet I would like playing Shadowrun if someone else were playing it, but I'd need some handholding for sure.

Lursi
Posts:207
Joined:Sun Jun 17, 2018 2:09 pm
Location:Münster

Re: Some Questions

Post by Lursi » Tue Oct 23, 2018 5:25 pm

Recently I saw something pretty neat in terms of dice mechanics:
Splittermond, a german RPG
Per default you have 2d10 plus modifier. The higher the better, a critical failure occurs at an unmodified dice combination of 1&1 or 1&2.

Now the feature: you can either „Play Safe“ and throw only one die or, take up to four dice.
While the former can not produce critical fails, the latter sums up the highest two dice and has a much increased chance of producing a critical fail, in which case nothing is added.
Of all things I lost, sanity I held dearest.

Jaracove
Posts:119
Joined:Mon Nov 28, 2016 7:58 am

Re: Some Questions

Post by Jaracove » Fri Nov 02, 2018 10:09 pm

Mataxes wrote:
Fri Sep 14, 2018 4:24 pm
Understanding that I haven't really dug into the underlying probabilities, I think the difference is about the same. I know the basics of Savage Worlds, and they have a similar kind of stepped-die system with rerolls. This can do odd things with certain DNs depending on what you're rolling.

So if those kind of hiccups don't bother you, Earthdawn's system shouldn't give you any problems.
I guess it all depends on how pronounced the 'odd things' are. I have OCD and an obssessive personality so a lot of things I just can't ignore. Trust me, it's a pain.

Anyone have any idea just how odd these dice rolls are?

User avatar
Mataxes
Posts:745
Joined:Sat Nov 26, 2016 10:39 pm
Location:The Great Library
Contact:

Re: Some Questions

Post by Mataxes » Sat Nov 03, 2018 2:24 am

Jaracove wrote:
Fri Nov 02, 2018 10:09 pm
I guess it all depends on how pronounced the 'odd things' are. I have OCD and an obssessive personality so a lot of things I just can't ignore. Trust me, it's a pain.

Anyone have any idea just how odd these dice rolls are?
The main one I'm thinking of is with single-die steps. For example, the odds of success are the same for DN 6 and DN 7 with a Step 4. If you roll a 6, the die explodes and guarantees a minimum result of 7. (Old-style Shadowrun had this problem as well -- a modifier that pushed the target number from 6 to 7 didn't matter.)

Going from Step 7 to Step 8 is weird. The odds of success on DN 12, for example, is only about 8.3% with Step 7, but 21.3% with Step 8 (a difference of about 13%).

Going to Step 9... is only 24.9% (about a 3.5% increase).

But as the DNs get higher, the edge goes to Step 7 for a bit (this is the chance of rolling DN X or better):

DN 19:
Step 7 - 4.17%
Step 8 - 3.94%

DN 20:
Step 7 - 3.47%
Step 8 - 3.01%

DN 21:
Step 7 - 2.78%
Step 8 - 2.26%

DN 22:
Step 7 - 2.08%
Step 8 - 1.70%

For most of these you're looking at a really small differences. For DN 19, for example, Step 7 will succeed where Step 8 fails about one roll in 435. All of these are less than 1% difference

The weird stuff happens at the margins (mostly)*, and is only noticeable when you crunch the numbers. While -- statistically -- you've got a slight advantage with certain steps against certain DNs, it only matters one roll in hundreds.

* Going from Step 7 to Step 8 is, as I said before, weird, and unavoidable because of the way one die works versus multiple dice.
Josh Harrison - josh@fasagames.com
Earthdawn Developer, Forum Admin

Personal Website: www.loremerchant.com

Jaracove
Posts:119
Joined:Mon Nov 28, 2016 7:58 am

Re: Some Questions

Post by Jaracove » Sat Nov 03, 2018 9:26 am

This is brilliant Mataxes, thank you.

"Going from Step 7 to Step 8 is weird. The odds of success on DN 12, for example, is only about 8.3% with Step 7, but 21.3% with Step 8 (a difference of about 13%)."

To clarify, when you say it only happens one roll in hundreds with the margins (even if it happened one roll in 99 - I'm fine with those odds), does that apply to the above number crunching as well?

User avatar
Mataxes
Posts:745
Joined:Sat Nov 26, 2016 10:39 pm
Location:The Great Library
Contact:

Re: Some Questions

Post by Mataxes » Sat Nov 03, 2018 5:59 pm

If you're talking about Step 7 to Step 8?

It's... complicated, but on average you are much better off increasing to Step 8.

In general, you are better off increasing your Step, especially once you get into multi-die Steps. Especially in the "meat" of the Step's DN.

I ran a bunch of numbers on the original (ED1) Step System back in the day, and hauled those spreadsheets out of archive to take a look. (These example numbers still hold, since the dice haven't changed, but I'll be honest I didn't double check my figures.) For the following DNs, here are the increased chances of success with a Step 8 as compared to a Step 7 (e.g., you have an 8% better chance of success against a DN 11 with Step 8 than you do with Step 7):

* DN 2: 5.56%
* DN 3: 13.89%
* DN 4: 16.67%
* DN 5: 16.67%
* DN 6: 13.89%
* DN 7: 8.33%
* DN 8: 5.56%
* DN 9: 4.63%
* DN 10: 5.56%
* DN 11: 8.33%
* DN 12: 12.96%
* DN 13: 8.33%
* DN 14: 5.32%
* DN 15: 3.01%
* DN 16: 1.39%

You have a weird little hiccup around DN 12 because, coincidentally, the bonus dice for both Step 7 and Step 8 kind of cluster around there. Once you get above this range, the odds of success are within half a percent, sometimes favoring Step 7, sometimes Step 8 -- but the odds of success against DNs 24+ are 1% or less for both steps anyway.

And this example is the weirdest part of the overall probability table (which makes sense, because you're going from one die to two dice -- changing the fundamental shape of the probability curve). As you increase the Steps, you'll get this little 'bubble' of increased probability around where the DN equals the new Step value, but it gets relatively smaller compared to the equivalent bubble for the previous Step.

I haven't done this degree of in-depth analysis for the ED4 version of the system, but in general, the more dice you have, and the larger the dice, the more the bubbles and hiccups of the individual dice cancel or balance each other out. And, aside from a few places at the margins where the difference is marginal at best, you are always better off increasing your Step.
Josh Harrison - josh@fasagames.com
Earthdawn Developer, Forum Admin

Personal Website: www.loremerchant.com

Lursi
Posts:207
Joined:Sun Jun 17, 2018 2:09 pm
Location:Münster

Re: Some Questions

Post by Lursi » Sun Nov 04, 2018 1:33 pm

I told my players:
D4 is step 3
D6 is step 4
D8 is step 5
D10 is step 6
d12 is step 7
D20 is step 11

Now go and make all resulting steps in a combination that you like.

If you want Step 15 as 5D4 instead of D20 +D6, be my guest.

I know that it changes probability but that can be a design clue. Against a difficulty number of 10 the former has about 88% success chance while the latter has only 72%.

Of course, hitting a harder enemy with a DN of 20 is almost impossible with the former and has a 22% chance with the latter.

So knowing the purpose of the roll lets you control the risk profile of the roll.
Of all things I lost, sanity I held dearest.

User avatar
Mataxes
Posts:745
Joined:Sat Nov 26, 2016 10:39 pm
Location:The Great Library
Contact:

Re: Some Questions

Post by Mataxes » Sun Nov 04, 2018 6:55 pm

That's... certainly a thing.

If you want to build your own step tables, more power to you.

But everybody at the same table should be using the same chart. And I really don't think you should be able to change the dice you use for a given Step from roll to roll.*

Those things play havoc with consistency (and, as a result fairness). Humans are good at seeing patterns (even--perhaps especially--when they aren't there), but actually bad at figuring probability. This is why casinos make money.

* (Not that you necessarily said you did the latter, but it wasn't clear.)
Josh Harrison - josh@fasagames.com
Earthdawn Developer, Forum Admin

Personal Website: www.loremerchant.com

User avatar
Mataxes
Posts:745
Joined:Sat Nov 26, 2016 10:39 pm
Location:The Great Library
Contact:

Re: Some Questions

Post by Mataxes » Sun Nov 04, 2018 7:09 pm

Lursi wrote:
Sun Nov 04, 2018 1:33 pm
If you want Step 15 as 5D4 instead of D20 +D6, be my guest.

I know that it changes probability but that can be a design clue. Against a difficulty number of 10 the former has about 88% success chance while the latter has only 72%.

Of course, hitting a harder enemy with a DN of 20 is almost impossible with the former and has a 22% chance with the latter.

So knowing the purpose of the roll lets you control the risk profile of the roll.
According to Anydice:
5d4 has a 90.5% chance of success against DN 10
27.91% chance of success against DN 20.

(Select "at least" -- this shows the chance of getting a result "at least" equal to the number.)

d20+d6 has a 74.58% chance of success against DN 10
25.93% chance of success against DN 20.


So... even against a DN 20, 5d4 has a slight edge, and is far from "almost impossible." (Remember what I said about people being bad at figuring probability?)
Josh Harrison - josh@fasagames.com
Earthdawn Developer, Forum Admin

Personal Website: www.loremerchant.com

Post Reply