Help needed with custom roll

Hi. From what my searching turned out I’m either too weak at searching this forum or this functionality will require some subscription-based features. I’d like to know which is the case. :) I’m trying to create a custom roll that does the following: 1) 3d20 is rolled and sorted 2) the highest-scoring die result is reduced by an ability score, so we have now: d20-ability; d20; d20 3) then the middle result gets reduced in the same manner (this gives two possibilities, we either have [d20-ability; d20-ability; d20] or [d20-ability-ability; d20; d20]) 4) this set of results gets compared against an attribute with each die equal or lower than that attribute being counted as a success 5) (optional) the number of points that the middle die differs from the attribute is returned along with the number of successes Example: ability score of 5 and attribute score of 12 1) roll is 19; 15; 11 2) roll is modified to 14; 15; 11 3) roll gets modified again and nets 9; 15; 11 4) two successes are found (9 and 11) 5) the middle die (11) beats the attribute score 12 by 1, so the final result is “2 successes, difference 1” Another example: ability score of 3, attribute score of 11 1) roll is 19, 15, 11 2) roll is modified to 16, 15, 11 3) roll is modified to 16, 12, 11 4) one success is found (11) 5) the middle die (12) fell short of the attribute score 11 by 1, so the final result is “1 success, difference -1” So, am I out of luck, trying to code something like that with the dice rolling engine alone? Is at least some part of this achievable? Thanks for any input. ;)

This would need the API to get everything automatically. You could roll 3d20 and sort them highest to lowest with /r 3d20sd , but you can't then subtract from two of the three rolls. You really  can't subtract from one roll, sort again, and then subtract from another (potentially the same) roll. You can compare the results to a target number, but that's not useful if you're not correctly subtracting the modifier from the appropriate dice. And even if you did have the target number comparison working, the system wouldn't report how much you beat the target or fell short by.

Interesting dice mechanics... Which system is this for?

@Brian — It’s as I thought. Thanks for the answer. :) @Silvyre — it’s a modification of mine for the rolling system of two Polish RPGs: Monastyr and Neuroshima (the latter being better known internationally due to the board game Neuroshima Hex). The original system was more elegant (3d20, subtract the ability from any die then compare the roll to the attribute lowered by the test difficulty; you need two successes to pass the test), but it undervalued abilities. It was quite normal to fail a test despite having obscenely high abilities paired with average attributes, so investing in abilities was a bit pointless. A popular fan-mod is to allow to subtract the ability twice (from up to two dice), but I found that it overemphasizes abilities (a moderately high ability makes it too easy to never fumble a test). So I made my version, where you subtract the ability twice, but one of those has to be from the highest, least useful die. After simulating it on a couple millions of rolls I found that the resulting distribution reflected my expectations best. Unfortunately, gameplay-wise, this isn’t very convenient to do and often ends up with situations like „wait, which die can I modify now?”. But I still prefer this to rewriting the whole system with a new mechanic…

Asthner K. said: @Silvyre — it’s a modification of mine for the rolling system of two Polish RPGs: Monastyr and Neuroshima (the latter being better known internationally due to the board game Neuroshima Hex). The original system was more elegant (3d20, subtract the ability from any die then compare the roll to the attribute lowered by the test difficulty; you need two successes to pass the test), but it undervalued abilities. It was quite normal to fail a test despite having obscenely high abilities paired with average attributes, so investing in abilities was a bit pointless. A popular fan-mod is to allow to subtract the ability twice (from up to two dice), but I found that it overemphasizes abilities (a moderately high ability makes it too easy to never fumble a test). So I made my version, where you subtract the ability twice, but one of those has to be from the highest, least useful die. After simulating it on a couple millions of rolls I found that the resulting distribution reflected my expectations best. Unfortunately, gameplay-wise, this isn’t very convenient to do and often ends up with situations like „wait, which die can I modify now?”. But I still prefer this to rewriting the whole system with a new mechanic… That's very cool! Like Brian said, completely automating this would require an API Script. I wrote up a macro that uses Roll Highlighting in an attempt reduce the amount of arithmetic needed to determine successes. The macro includes a chart that describes whether or not a die's Roll Highlighting indicates a success. For example; I know blue dice are always successful and red dice are always failures, so I can immediately tell that I have at least one success and one failure . My median (middle) die is yellow, so I'll reference the chart. The chart indicates that a median yellow dice is a failure . So, I've determined that I've rolled one success and two failures without really having to do any mental math. Here's an example of where some math is needed: My blue dice is one success . My median die is green, so I'll reference the chart. The chart indicates that I must check whether 15+5≤19 is true or false. It is false, which means the median die is a failure . Our highest die is yellow; since we have already determined that 15+5>19 is true (because we know 15+5≤19 is false) we have also determined that our highest die is a success . So, I've determined that I rolled two successes and one failure, only having to solve one inequality. Here's the macro , if you're interested.