Here is a metascript solution... but the explanation wasn't exactly clear, so I had to make some assumptions. Assumptions 1) In the target roll equation, 0 represents the "Melee Weapons" skill, and 4 represents the value of the STRENGTH attribute. 2) Successes are counted across both d20s (that is, each die is individually compared to the skill rank and to the total target number; the number of successes generated from each die are added together for the total number of successes. 3) Critical successes (a 1) and critical failures (a 20) do not affect the overall number of successes. 4) There is ... apparently... no way to reference the skill rank or the attribute by attribute retrieval syntax from the sheet; that is, if you could refer to... @{CharacterName|MeleeWeapons} ...it would be better than hard-coding... 4 ...into the command line, since you'll have to change every instance of that number when the character trains another rank. If there is a way to reference the skill and/or attribute from the sheet, then I would use it in my final command line (it would alter slightly the solution, below). Solution For all of those assumptions, here is how you can do this with the MetascriptToolbox: !&{template:fallout_skills} [[[[10]][Melee Weapons] + [[4]][STRENGTH])]] {{assist=assist}} {{playerName=Jona Tán}} {{skillName=Melee Weapons}} {{strength=STRENGTH}} {{skillValue=$[[0]]}} {{target=$[[2]]}} {{attributeValue=$[[1]]}} {{num_rolls=2}} {{roll1=[[1d20cs1cf20]]}} {{roll2=[[1d20cs1cf20]]}} {{successes={&r}{&if $[[3]] <= $[[0]]}2{&elseif $[[3]] <= $[[2]]}1{&else}0{&end} + {&if $[[4]] <= $[[0]]}2{&elseif $[[4]] <= $[[2]]}1{&else}0{&end}{&/r} }} {{tag=}} {&simple} Explanation Here's a quick explanation so you know why I made the choices I did. They'll match the assumptions, above, so if one of the assumptions is incorrect, hopefully this section will tell you where to go to change it. 1) I rewrote your target roll to be: [[[[10]][Melee Weapons] + [[4]][STRENGTH])]] This closely associates the label "Melee Weapons" with the part of the roll that seemed associated with that base number, and the label "STRENGTH" with the portion of the roll that seemed to derive from that attribute. Further, I wrapped the skill rank number (in this case, the 10) in its own inline roll brackets, and the attribute rank (in this case, the 4) in its own inline roll brackets. That lets us refer to these numbers by their roll indices elsewhere in the command, and will minimize the number of places you have to change this command should the character train up (or should you want to copy this command line as a starting point for another skill). Finally, but importantly, I altered your roll to change the skill rank from a value of 0 to be a value of 10... this was so that I had a number I could actually roll beneath in order to test the logic of what I was trying to produce. Once you've validated that this command line works for you (that is, that it produces the correct number of successes for each roll outcome), you can change this 10 back to a 0 (or to whatever the skill rank is) to apply it in your game. Again, because of the way this command line is structured, that should be the only place you have to change it. 2) I moved the roll between template parts just to eliminate the need for ".computed" syntax 3) Success - in the successes field, I run an IF check for both roll $[[3]] and roll $[[4]]. These are the d20 rolls. They each get compared against roll $[[0]] (the skill rank) to see if they should earn 2 successes. If not, they get compared against roll $[[2]] (the overall target number -- the attribute value takes up roll $[[1]], so the target number is $[[2]]) to see if they should earn 1 success. If neither of these cases pass, 0 success are earned for our d20 roll. Whatever numbers are produced by those two IF checks are wrapped in... {&r} ... {&/r} ...which are metascript tags representing inline roll brackets. Basically, this lets us disguise from the Roll20 parsers the final roll totaling the number of successes until those successes have been derived from the IF blocks . In other words, if we used normal inline roll brackets, here, the roll would try to resolve the first time Roll20 saw the message, before the metascripts had a chance to do their work and evaluate the number of successes.
