Originally Posted by Setia

I did some more tests for the interaction between Spellpower/Ice Shards and the CSD. It looks like the old (advantageous) formula is back.

On a test with 36 crits, 171 hits, I had a crit multiplier of 236,35%.

This means the "critmultiplier after talents*1.03" is out, because it would give a 231,75% multiplier, which is outside my 95% confidence range of my experiment (lower bound: 232,6%).

And my results are really near the expected 236,25% multiplier of the "((base crit multiplier*1.03)-1)*talent multiplier+1", so it's probably the one now used.