No horseshoe rules will be needed as we have a correct guess, estimation, calculation, prediction, or whatever method used to arrive at an entry. I wouldn't lean too heavily on Mr. Keen's results, as that would suggest his batch of 100 FETs are an accurate representative of what's in the world and I don't think that was meant to be a takeaway. It's simply a matter of sourcing.
There were 25 quad sets found and that means @dmnCrawler is our winner!
On fig's farm, you pick your own so here are the actual measurements. The blue and green blocks are obviously the matched sets (some are even perfect matches). I just need one reference from the second column to pull the set. The samples in pink are OOS**. If you catch any errors, be a sweetie and please point them out.
**I've got my Phase 90 breadboard almost built and will be testing those pink nightmares to determine how low can you go with the gate-source cutoff voltage.
Thanks to all of you for playing along. I think I'll have an extra quad set or two leftover, so if you have trouble sourcing for a build, let me know
Thanks always to @PedalPCB for all their patience, hard work, and groovy stuff.
Oh! I have another batch of FETs from a group buy to measure, so stay tuned, but for now...go build something!
I just noticed that if I slide set 13 down one spot, it becomes a perfect quad. I'll make that adjustment. It doesn't change the count...but makes a great set perfect.
Actually, the math is a quite a bit more complicated than just a percentage. Here's an example of a similar problem for those that like a little math. Birthday Problem This, actually, has real world applicability in cryptography (I was a network security programmer for many years).