Well like I said, if this were true I should have seen it hit me for around 56. I had two absorbs one at 95, the other at 94. But the highest critical against me was 46. Which is why I doubt this scenario. (Although it's possible the sample size wasn't big enough.) I'll probably run a similar test sometime to prove how the absorb works though~
Kirschy, did you do any testing without any PDT on to see how the values compared? Would be interesting to see if those absorb values changed at all.
Also to double check about shadow mantle and absorb conflicting, you could go try to fight something routinely hitting you for 0 without mantle to see if you can ever get a message for absorbing 0. It might just be an issue of not having a message as opposed to not proccing.
Also would be interested to see how Phalanx affects the absorbed damage in any way.
floating point takes more memory to store than an integer and floating point math is processor expensive compared to integer math. it makes good sense to truncate numbers as it allows them to do more calculations on the same hardware. look at their calculations of haste, mdt, pdt, etc. none are the actual %'s represented on the item, but fractions of 256 (or possibly a multiple thereof) which is a special number in computer science.