Disc priests. Halo vs Cascade vs DS

[Genzen Han]

Is there a way on the logs to see how much DA a specific spell procced? I just took a quick look at my logs, and it seems like my Cascade is doing about 63k heals per cast, whilst my PWS does 70k, however, the Cascade could also be proccing DAs. I could just multiply the Cascade's Crit percentage by its total healing, then add them together, right? The only problem is that crits are worth more than 100% extra, since Mastery increases the size of the DA (meaning a spell healing for 50k is going to do more than a 50k DA, right?).

I'm bad at math and could be looking at this all wrong, but if my crit percentage was 15%, and 63k+15%=72.450, then doesn't this make my Cascade a smidge better than my PWS? (there's also Multistrike on Cascade, but I don't know if that's already been factored into its healing on the logs).

[Myllior]

Is there a way on the logs to see how much DA a specific spell procced?

There is no easy way; if you wanted a 100% accurate number of its effective healing by spell, you'd need to trawl through the events tab of a log and note where Divine Aegis is being generated and then consumed. However, you can approximately calculate the DA produced. To do so, you take the raw critical healing (including critical multistrikes) of the spell and multiply by (1 + Mastery %) to get the raw DA produced by the spell. In the event that you don't know the Mastery percentage of the Priest, you can estimate it by dividing the raw DA of the log by the raw critical healing (including critical multistrikes) of the log, which would give you the (1 + Mastery %) factor. If you then wish to know the effective healing done by that DA, you can estimate it by assuming that its overhealing was equal to the average overhealing for all DA in the log; i.e. Effective DA = (Raw DA)*(1 - DA Overheal %).

This is only an approximation, since you can't exactly deal with things like the legendary ring buff (which I've heard DA douple-dips in; haven't tested it myself though), but it's likely to be reasonably close to the actual number.