> While reading through the rules last night, I came across a feature of the
> contest mechanic that doesn't sit right. Any time that you have a tie of
> the success level (i.e. both the actor and the opponent succeed), it is
> statistically better to have the *lower* target number. The same holds
true
> for the mastery special case for both the actor and the opponent failing.
>
> The character with the rating of 12 stands a better chance of getting a
> success roll less than a success roll of one with a 16. If the latter
> character rolls a 13, 14 or 15, he *cannot* win if the former gets a
success
> of any number. This completely slews the chance of a minor victory in
favor
> of the lower-rated character. It doesn't feel right to penalize a
character
> for being better.
Yes, assuming both roll a success, the lower roll *does* have a better chance of rolling lower, but it has to be compared against that initial chance of rolling a failure.
Your 12-vrs-16 example:
Chance of Crit/Success/Failure/Fumble: TN 12: 1/ 2-12/ 13-19/ 20 or 5%/ 55%/ 35%/ 5% TN16: 1/ 2-16/ 17-19/ 20 or 5%/ 75%/ 15%/ 5%
The TN 16 guy has a 20% better chance of winning right off the top (Getting a Crit/Success when TN 12 rolls a Failure/Fumble). Eliminating rolling Criticals or Ties, the TN12 has a ~60% chance of rolling lower. But his *overall* chances of winning are only ~20-25% for a Simple contest and ~40-45% for an extended one (according to David Dunham's HW Calculator).
RR
> I can understand the reasoning behind going with "the lower wins". It
makes
> it consistent that you want to roll low. I considered, briefly, that it
> should be the higher number wins, but this then breaks the simple group
> contest. The best resolution to fix this, I think, is to instead of
> comparing the die roll itself, compare the differences between the die
roll
> and the target number.
Powered by hypermail