Quarterly update, a bit (4 flights) late but with significant methodological improvements.

https://docs.google.com/spreadsheets...it?usp=sharing

The sheet now has data back to 2017-05-20 except for upgrade results which I only started on 2018-01-01. Right now I have a rolling EQD just over 18k and have been upgraded 83% of the time (when the plane in question had a J cabin).

Minor notes:
1. I changed the upgrade result column to use numeric values, on the theory that someday I might want to do quantitative analysis on it. Positive numbers indicate when an upgrade cleared, in minutes before flight time. The zeros are "battlefield upgrades" i.e. upgraded after already seated in Y. The 10s are "gate dances" cleared off the airport list. Negative numbers indicate failure; the number is how many slots away I was from clearing off the airport list. If I was 4th on the list and 2 cleared, that's a -2 (but that could also mean 8th on the list with 6 clearing). Null (blank) cells indicate either a)the plane didn't have a J cabin so no upgrade was possible or b)the flight was before the first of the year when I started keeping track of whether I got upgraded. The success percentage leaves the nulls out.
2. The "fare purchased" and "fare credited" columns seem redundant unless you look closely, in which case they seem confusing. But it's really just to check random refaring, usually because of IRROPS. Often if I have a flight canceled the rebooking ends up credited as a full Y fare...this lets me see that.
3. The "Max EQM/EQD" field is just the simple ratio 100,000/12,000 ...in other words, the EQM/EQD you'd have if you got the bare minimum of both. It's only there to provide a reference for the cell next door which is the YTD actual ratio, which you want to be below the max.
4. The link is live, so you'll see results up to the last time I updated it even if you click this months from now.

Anyway, as with last quarter, I'm really happy with these results.