http://www.fflogs.com/reports/Dr4kq1vwGg2dnHA7/#view=rankings
I really do not understand how DPS(e) is cut in half here? Even after accounting for overkill? Is there something funny going on, or am I misunderstanding something?
Any help would be much appreciated.
DPS(a) stands for the optimal DPS you could have reached given the numbers in the parse if you had been active dps’ing all the time. It’s not an indicator that excludes overkill or anything like that as far as I know.
The report shows you’ve been active dps’ing 52.79% of the parse’s length.
1,290.3 (DPS[a]) * 52.79 /100 (% active) => 681.14; aka your actual dps: DPS(e)
See also: Ast's DPS(e) is being calculated incorrectly in A6s
That seems quite impossible, as yesterday my DPS(a) was 1290, and today, it was ~1800 on a longer cuff of the sun (savage) run, meaning typically you would expect possible DPS to be lower, not higher. Also, if you move over to the damage done tab, you can toggle overkill, and it affects both the DPS(a) and DPS(e) numbers. Therefore, it clearly does exclude overkill to some degree, and if it was an optimal number it would not have any overkill to exclude (or very little, definitely not 100 DPS).
Although you may be correct that it’s not any actual number that I did, the DPS(a) is within 130 DPS of what ACT told me I did. ACT also added up the parse to be very close to 4.4 million total damage, but 1164.89 of that was mine in the ACT parse, whereas only 681.14 was in DPS(e) here. I also spent a considerable amount of time looking at other healer parses, and a similar sort of ridiculously high disparity between DPS(a) and DPS(e) seems to be happening in cuff, that I’m not really seeing on jobs from other roles, where typically the two numbers are within 100-200 DPS or less of each other. Here, for reference, is what ACT shows me:
for that very same encounter.
Now, I’m well aware of the reasons to “expect to see lower numbers than ACT,” including different DoT calculation, exclusion of overkill, and the inclusion of downtime, but those three things combine should not lead to a parse that is half, right? I’m not really sure it’s mathematically possible for half of my DPS in cuff to be overkill/weird ACT simulation differences. Now, after looking through all of this, it appears that fflogs did credit me with fewer DoT applications and ticks than act, but a not very significant amount (like 300 potency), and it seems very hard to match up the numbers even taking that into account.
If that is the way DPS(a) is calculated though, that seems rather silly, since DoT lengths and GCDs do not line up neatly with enemy deaths and fight length in such a way that if I spent nearly double the time actively dpsing it would give me that number. For instance to avoid overkill, I would need to use my GCD near the death of the enemy on a weaker potency, non-DoT spell for damage.
I appreciate the input, but maybe the similarity between DPS(a) and what ACT said my EncDPS was for that very same encounter, coupled with the much lower DPS(e) assigned by fflogs, is what had me confused. But in that case, how is there ~500 DPS difference between those two? Even given a difference in DoT calculation/cutting overkill/downtime, that’s very high, like ~71% my total DPS according to fflogs.
EDIT: reposted the parse on an image service that doesn’t auto-prune
I think DPS(a) is simply calculated from DPS(e) taking in account the activeness %; the fact it’s close to the actual DPS showed up in ACT being a pure coincidence. Imo it just acts as an indicator, so for healer indeed it’s not really relevant.
I’m not sure if i’m actually right about that though, that’s definitly up to @Kihra to confirm.
That huge difference between your DPS(e) and what’s showed in ACT is odd though indeed (as opposed to DPS(a) vs DPS(e)) ; those values should definitly be close. (+/- 50 in general); which actually seems to be the case for all your others party member at first sight.
Did you try breaking down your damage by skill in ACT and compare it to what’s showed on FFLogs see if a skill damage isn’t bugged?
That could also apply to damage on specific mobs?
But that’s all I’ve got for ideas atm, up to Kihra to investigate probably after.
DPS(a), or “Active Time DPS”, is a familiar concept to WoW players, but seems to be largely unknown to FFXIV players. The Recount add-on in WoW uses DPS(a), and the Skada add-on in WoW uses DPS(e).
DPS(a) is your total Damage Done divided by your time spent actively DPSing. It isn’t a number to pay much attention to, really, other than as a way of noting how much DPS you did during periods where you had time to actively DPS. It can also be handy if you died to see how much DPS you were doing before you died and were taken out of the fight.
DPS(e) is the number you are actually ranked on. DPS(e), or Effective DPS, is just the total damage done divided by the fight length, and is how much “damage per second” you actually did over the course of the fight.
DPS(a) and DPS(e) will have a disparity any time you spend large amounts of time not dealing damage,
I should add that I’m probably just going to bury Active % and DPS(a) in tooltips, rename DPS(e) to just DPS, and then people will stop being confused by there being two DPS numbers. 
Alright, that clears things up. Thanks!