Reply to thread

Message: <blockquote data-quote="Jens" data-source="post: 109410" data-attributes="member: 605">I once calculated this with explanations, but now I can't find it so you only get the conclusion. It should be easy to remember:To maximize expected damage, use Power Attack to equalize your average damage per hit and the number of sides on the d20 that will hit.This also works with multiple attacks, just use the average number of sides on the d20 that will hit. Let's for example say that, without using PA, your 2 attacks will hit the target on rolls of 6+ (15 sides) and 11+ (10 sides), so that you will hit on an average of 12.5 sides if the d20. If your average damage is 8.5, you'll equalize the two by PA'ing by two points. This will increase your expected damage by 3.76% since (10.5*10.5)/(12.5*8.5) = 1.0376... Not a whole lot <img src="https://cdn.jsdelivr.net/joypixels/assets/8.0/png/unicode/64/1f641.png" class="smilie smilie--emoji" loading="lazy" width="64" height="64" alt=":(" title="Frown    :("  data-smilie="3"data-shortname=":(" />Qualification: This is how to get the maximum *internal* solution without considering the limits. So if you have a really poor chance to hit, PA'ing fully may be better (running may be better yet, but that's a different matter).Why? 1. Expected damage is proportional to average damage per hit and to the chance to hit. 2. The largest rectangle with a certain circumfrence is a square.Atroomis' example: +21/+16/+11 vs. AC 20, on average doing 4.5 point of damage on a hit. Without PA, the average to hit number of sides is 17 (ignoring limits). (17-4.5)/2 = 6.25, or about 6. So he should PA by 6 points, exactly Atroomis' other calculation.Crits: 'Normal' crits don't affect this once they're accounted as part of average damage per hit. For example, the average damage from a hit by a longsword counting crits is (4.5+S)*1.1 with S being the '+' bonus from strength etc. Special crits effects (butst weapons etc.) are more troublesome and will change things a little (you should PA a little less, I think).</blockquote>

Verification