Q:
Promoting items in MF return
Yes, mat promotion rates are still random but with a large enough of promotions it does converge on an average rate of roughly 1/3rd of the amount you put it, other than promotion to T6. That means 50 in gives you roughly 17 and 250 in gives you roughly 84.
I promote numerous times nightly and get within 10% of these averages 95% of the time.
RIP City of Heroes
As Behellagh says, the variance is extreme.
Let me repost some data I posted a while ago. This is for promoting kitten into T2-T5. I’ve never collected data on T5 to T6.
You get on average roughly 1/3rd of the lower tier mat. So 50 in yields 16-17 out while 250 is 83-84. ON AVERAGE
I can’t say that enough. ON AVERAGE
The range for 250 mats in is 40 to 200 mats out. There’s been times I’ve gotten over 650 back from 5 stacks of 250, other times I got less than 300. But ON AVERAGE I get in the low 400s or roughly 1/3rd.
I have data an 175, 250 mat in promotions. Now binning the samples into buckets 10 wide …
40-49: 25
50-59: 20
60-69: 32
70-79: 27
80-89: 10
90-99: 8
100-109: 4
110-119: 10
120-129: 15
130-139: 7
140-149: 1
150-159: 4
160-169: 1
170-179: 4
180-189: 2
190-200: 5
Median (half samples below/above this number): 73
1st 20%: 40-53
2nd 20%: 54-66
3rd 20%: 67-80
4th 20%: 80-121
5th 20%: 121-199
Mean (numerical average): 86.75
1 sigma range (68.2% of the time): 51-127
2 sigma range (95.4% of the time): 42-190
What you notice is the median is significantly less than the mean/average. That means if you will see more output below the average and that can seem like RNG is not in your favor, that you are getting unlucky. But if you look at the distribution you see two “bumps”, one around 50-79 and a smaller second bump around 110-129. It doesn’t take many returns above the median to pull your current session’s average up to the “historical” one.
That’s why you have to look at all the promotions you are doing over time and not focus on individual ones, or even a few session’s. You have to trust that the Law of Large Numbers, that with enough trials the results will converge on the projected average, will prevail in the end.
It’s also the reason I don’t bother chasing the “big” profit per skill shard ones if the % of profit is below 30-40%. The variance in promotion output is too high. Right now the highest profit mat promotion according to GW2Shinies is green wood planks to soft wood planks at 2.5 gold per spirit shard which is fantastic. However it’s less than a 4% profit. They use an average of roughly 86.33 which means if you get only 83 from the promotion you don’t even break even. However platinum ingots, the next most profitable at 1.73 gold, has a 24% profit and you will sill make a few coppers even if the promotion returned 70.
TL;DR – play the averages but take % profit into account which mats you promote if you are doing this simply for the money.
RIP City of Heroes