Forging efficiency drop? (in General)

I've noticed that recently the amount of % increase I'm getting per BA cycle is dropping - when I started this job, I was generally getting between a .068 and .07% increase per forge cycle. That number has dropped to as low as .064 and I've not seen anything close to the .07% range in a while - is my forging efficiency decreasing as my score drops or something?

What are you forging?
If it's the + (I suspect it is given the low increase you're talking about) then your efficiency will drop as the NW gain increases. I think that's even the case in-between advances...?

Forging a pair of HGs from

+15
NW: $9,794,097

to

+16
NW: $17,794,832

so that shouldn't be the case. Oh, and I just dipped down to .063%. Granted, I know that's not a large difference between the two numbers, but when I'm trying to get to 100.00% in such small increases, every little decimal number means quite a bit...

But the real question:

Is 8M NW worth an extra 3% missile damage when your already doing over 1M a hit?

HGs, Dr. Popsicle, not BGs

I think "in-between" may be the magic word in here...

Then I stand corrected.
But is the extra +1 on the HGs worth the 8M NW and loss of all that xp you would get from fighting?

for someone planning on having a giant UC minion (NightStrike), it makes lots and lots of sense...

I suspect you're seeing the randomness the hardcore forgers talk about...

Right, but before I was getting between .068 and .07 - now the highest I've recorded this morning was .065 and the lowest was .063 - that indicates that something's different than when I started on these a week ago. Maybe I was getting super lucky on the randomness then and not so much now, but it doesn't appear that way to me.

Oddbird is definitely on to something here.

I'm currently forging the x on my SoD and the efficiency has dropped by about 7% compared to what it was a couple of days ago. (0.960 vs 1.035)

It has been my experience that the normal 'randomness' at the forge runs +/- 5% from the mean. I've had once cycle this time around that was -11% from the previous average.

Furthermore, the single least efficient cycle in my previous bouts of forging was more efficient than anything I've done today.

And yes my character was in a clan with similar, if not identical, mpr. And my BA cost is constant.

Disturbing this trend is...

forging has an intrinsic 10% randomness, applied to each of the four RPM increases, so most of the time you only notice the 5% or so you mentioned. Also, the round-off in the display of RPM because more noticeable at higher +'s.

I think the intrinsic randomness does not explain fully the results I am seeing.
Case A:
Based on my original data set (about 15 to 20 cycles) the average forge efficiency on the x of my SoD was approximately 1.035 * BA Cost.

Case B:
After 9 cycles today the efficiency is approximately 0.980 * BA Cost.

Since my BA cost and the upgrade cost are fixed the variation must be cause by some other factor.

It is also interesting to note that in Case B not one single cycle was as productivity of the average for Case A

It has been a while since I did stats so I won't try for a statistical significance test here but it certainly seems to satisfy the duck standard of proof.