TL;DRs if you dont wanna math:

• 1st vaal infuser never corrupts (or at least never did in over 314 tracked trials, plus another 70 or so untracked)
• Chance to arrive at 30% quality uncorrupted: ~11.6% +- margin of error of about 3.5%
• Vaal infusers have a 5% chance to corrupt PER quality over 20 on the item when you apply. Meaning 10% chance to corrupt a 22% quality item, 35% chance to corrupt a 27% quality, etc.
• Both vaal infusers & scrap (& whetstones and etchers) have a ~20% chance to apply +2% quality to an ilvl 82 item instead of +1%
• Graph of results of 314 items having infusers applied until it corrupts or hits 30%

Vaal infusers, if you didn't already know, drop from temple. They increase the quality of an item PAST 20% (to a max of 30%) and have a chance to corrupt that item (NOT krangle it, just add the corrupted tag so you can no longer modify it (except for modifications you can make to corrupted items like architects orb, etc)). Infusers also, like armour scraps or whetstones, have a chance to apply +>1% quality instead of +1%. I believe this is based on ilvl so this chance will vary.

What we DONT have good info on is:

• Whats the chance of corruption?
• How likely is it that you can get to 30% without corrupting your item?
• How many infusers do you need?
• What's the chance to get +2% quality?

So. Settle in kids. I did multiple experiments. Starting out with some vile robes,some akoyan spears, some ancestral tiaras, and some siege crossbows!

My method:

1. quality everything up to 20%
2. Apply VIs until 30% or corrupt
3. Note how many VIs, ending quality %, and special checkmark if we got to 30% without corrupting.

So my entries on my notepad looked like

4 6

4 4

6 8

8 10 ✅,

etc

Worth noting, it can (and will) corrupt AND get to 30% too, which is always a bummer!

Final results:

• Total items tested: 314
• Vaal Infusers applied: 1,734
• Quality added: 2095% [which does mean that 361 of them added 2%, more on that later]
• Final 30% uncorrupt items: 35
• Final percentage of "wins": 35/314 = 11.15%

Theory: Vaal infusers work exactly the same way as whetstones & scraps & arcanist's etchers. I confirmed this by applying vaal infusers to very low ilvl items and you'd see +4% sometimes from one infuser. What I DONT know is if the ratios are exactly the same. I wasn't tracking scraps/whetstones the whole time, and only did on my final run of 144 ancestral tiaras:

Scrap Findings:

• Scraps used to quality up 144 tiaras from 0 (i double checked that none started above zero): 2443
• Quality % gained: 2880

• of scraps that rolled 2%: 2880-2443 = 437

• Percentage estimate for scraps which hit 2% on ilvl 82 items: 18.91% (see below for details, it isn't just 437/2443)

We might think we'd look at 437/2443 to get the ratio, but that's unfortunately wrong. BECAUSE, if a scrap is applied at 19%, it does not have the chance to roll 2%, and therefore shouldn't be counted. However, how many did we roll at 19%? Well, this requires a bit of recursive math, but we can ballpark it. If none of the final ones rolled 2%, then that means 144 out of our 2443 applied were guaranteed 1%ers, which means 437 out of 2299 hit 2% rolls. If ALL of our final rolls somehow were from 18%->20%, then that means 437 out of 2443 hit 2%. So somewhere between 17.89% and 19.01% are hitting 2% (factual). This means we can just assume about 18.45% of our final rolls were final rolls that hit a 2%. This brings us to a final tally of not 2443 which could have hit 2%, but we take away 91.55% of our 144 (which were 19% before hitting 20%), which is 132, giving us a proposed 2311 which COULD hit 2%, of which we know factually 437 did. Final results: 437/2311 = 18.91% chance to hit 2% quality on ilvl 82 item

Vaal Infuser Findings:

The reason I wanted to track EACH piece of data, rather than just a final ratio, is because of the primary question many of us have believed, yet not known specifics of: "Does corruption apply at a higher rate at higher qualities?". Turns out, yes, yes it does.

• Theory 1: The FIRST vaal infuser (adding to a 20% quality item) is ZERO chance to corrupt. That's right. You're safe to add one vaal infuser to EVERY ITEM YOU OWN. Based on this graph, you can see there are zero items that ended their run after 1 infuser was applied.

• Theory 2: Based on the above fact, as well as a random guess, I believe there is evidence to show that the corruption chance is +5% per quality above 20 on the item when you apply VI. Meaning it's 0% to corrupt a 20% quality item, 10% chance to corrupt a 22% quality item, if you're at 29%, that's a 45% chance to corrupt, etc. Presuming I did the math right (I am a math teacher but still make plenty of mistakes.)

Expected graph if 5% per quality above 20 is correct

Actual graph of final results

Spreadsheet with Data: https://docs.google.com/spreadsheets/d/1mqAY4r5iElCj1F_U7Y0p1qOEUJ0sK8_S9TwfPIGhlE0/edit?usp=sharing

• Theory 3: Using the same strategy as we did on scraps, it seems that VIs have a 21.13% chance to add 2% quality. However, this has a margin of error for a 95% confidence interval of about +-2%, and the scraps number has a MOE of about +-1.5%, meaning it's entirely possible (and I would argue probably) that VI's and scraps have the exact same chance to add 2% on use, and it's in the ballpark of 20% for ilvl 82 items.