Talk:Testing spell power coefficients

Cumulative Rounding?
In the "Rounding" section Tejon says that non-integer damage amounts are accumulated, rather than being fed to an RNG. If a spell was supposed to do 100.2, exactly every fifth occurence would hit for 101.

Are there published test results to confirm this? If so, is the accumulation saved per caster, per target, per caster+spell, caster+target, or caster+spell+target?

The following log entries would seem to indicate that it is not caster+spell+target:

0:03'21.406 Portal Guardian #2 suffers 1043 Nature damage from Erdluf Insect Swarm. #1486  0:03'23.422 Portal Guardian #2 suffers 1043 Nature damage from Erdluf Insect Swarm. #1509  0:03'25.391 Portal Guardian #2 suffers 1042 Nature damage from Erdluf Insect Swarm. #1532  0:03'27.375 Portal Guardian #2 suffers 1042 Nature damage from Erdluf Insect Swarm. #1564  0:03'29.391 Portal Guardian #2 suffers 1043 Nature damage from Erdluf Insect Swarm. #1591

In order to get consecutive ticks at the high value, you'd need the floating point number to be greater than 1042.5. In order to get consecutive ticks at the low value, you'd need the floating point number to be less than 1042.5. If there is an accumulation, it must have been "bumped" in some way. Presumably the bump was caused by either other actions of the caster, or other things done to the target.

Erdluf (talk) 20:04, 21 May 2009 (UTC)

Just tested with lifebloom, cast on myself. No other casts in progress. Ticks were 234 or 235, with a sequence of

4,5,5,5,4,4,5,5.

This is not consistent with accumulation. Erdluf (talk) 01:32, 23 May 2009 (UTC)