It’s kind of being a jerk if you are demanding cites to prove something that isn’t relevant to the disagreement between you.
Here’s a cite that shows 6000 is about the number of PTO challenges per year to validity:
https://www.uspto.gov/sites/default/files/documents/AIA Statistics_March2017.pdf
That also gives statistics on the outcomes.
This shows the number of patents filed per year in the US:
It also shows how many patents actually issue as valid per year.
[automerge]1588467742[/automerge]
The stuff in parentheses just means they didn’t pay attention to the first few months of IPRs, because there were too few to reach a conclusion. The IPR procedure (and the other procedures mentioned in that document) are pretty new, starting in the last decade.
As for the rest, what happens is if you think a patent has invalid claims (again, patent claims are the only thing that matters), you can go to the USPTO, pay a bunch of money, and show them why. If they think that you may have a point, a procedure takes place between you and the patent owner, that ends in a trial in front of a panel of 3 administrative law judges, where each side tries to prove their point, on a claim-by-claim basis.
So if a patent has 100 claims, and you decide to invoke this procedure on 10 claims, then in that best month you would guess that 5 or so claims survived and the rest were invalid (out Of the 10 of course). In that worst month, somewhere between 0 or 1 claim surviving would be the bet.
On average, looks like something like 3 or 4 out of 10 claims survive.