Another disappointing mod

i have to say i'm impressed here, a lot of data analysts on the road B)
But you lost me approx. on the 2nd line of EduardoCadavs reply :#

@EduardoCadav Thank you for the thoughtful response. I took more than enough math, but never statistics and for engineering not social sciences.

They could always just open source it for us and the question would be answered definitively.

"cannonfodder_iv;c-1888475" wrote:
@Mephisto_style How big of a sample size do you need though? I have 500+ rolls tracked and speed procs at a 20% rate over that sample. I would have expected 25% as you said. It's possible that I'm on a bad trend (expecting 125, got 100), not a stats person to know how big of a sample I need and too lazy to figure it out.

This is also almost exclusively speed sets. I wouldn't be surprised if speed secondaries roll lower on speed sets than on other sets in order to equalize the value of each mod set over time. Meaning that the 10% base speed increase can be made up over time with non speed sets. This would be awesome, if true, because of the benefits of other primary sets (especially offense and health).

This got me thinking of how I might track mods and mod rolls. It seems to me that you'd have to track chance to proc and chance to roll a bit differently.

• Mod 1 secondaries: Speed, Offense, Offense %, Health
  
• Mod 2 secondaries: Protection, Potency, Offense, Tenacity
  

First, proc. Would this be per mod as such:

• Speed Procs: 1/2
  
• Offense Procs: 2/2
  
• Defense Procs: 0/2
  
• etc.
  

Or per secondary slot as such (I think this way is probably right):

• Speed Procs: 1/1 (rolled on the first chance, removed from the possibility for the next 3 chances)
  
• Offense Procs: 1/2 (mod 1), 1/3 (mod 2)
  
• Offense % Procs: 1/3
  
• Health: 1/4
  
• Protection: 1/1
  
• Defense: 0/8
  
• etc.
  

???

Then rolls... For rolls, I think you'd need rolls and chances to roll side by side.

• Mod 1 rolls 2 Offense, 2 Health
  
• Mod 2 rolls 1 Offense, 3 Tenacity
  

Would be tracked as such:

• Speed Rolls: 0/4
  
• Offense Rolls: 3/8
  
• Health Rolls: 2/4
  
• Tenacity Rolls: 3/4
  
• All others present: 0/4
  
• Defense: 0/0 (and all others not present)
  

???

Is any of this right? If it is then I think I'll pass on tracking mod rolls. Seems like a lot of work. :/

Edit for formatting.

@jhbuchholz

"cannonfodder_iv;c-1889159" wrote:
@EduardoCadav Thank you for the thoughtful response. I took more than enough math, but never statistics and for engineering not social sciences.

They could always just open source it for us and the question would be answered definitively.

It is actually a really easy set of data collection to track with your rolls. You could and should do it in a spreadsheet if so inclined. It requires more meticulousness and persistence than depth of thought. You are simply trying to nullify the hypothesis: the chance of rolling a speed upgrade on a grey mod is not 25% and if we keep upgrading those grey containing mods as your n approaches infinity if your % speed upgrades approaches 1/4 then you disproved your hypothesis.

Caveats:
Only roll 1 type, slot, and primary. Those would be potentially confounding factors.

Ideally you would want all of the same secondaries too.

But then, there would still be a really really easy way to solve this problem. Simply ask CG Carrie to find out and assume she is not lying.

@jhbuchholz
No,
Choose your mods.
Record all of the secondaries and primaries in a spreadsheet for those mods.
Begin your TOTAL data collection.
This data collection is way too easy to not record everything as to potentially answer more questions later.

"Mephisto_style;c-1889198" wrote:
jhbuchholz
No,
Choose your mods.
Record all of the secondaries and primaries in a spreadsheet for those mods.
Begin your TOTAL data collection.
This data collection is way too easy to not record everything as to potentially answer more questions later.

I don't follow. What exactly are you saying no to? The Proc stats? The Roll stats? The Roll vs. Roll Chance stats? It seems to me that if you don't track all of this your numbers will be off.