Upgrade Pro BETA

How we predict your upgrade odds — and how often we're right.

Upgrade Pro turns real airline data — waitlists, fare buckets, seat maps — into a calibrated probability you'll clear the upgrade list. Here's exactly how it works — and how we hold ourselves accountable, hits and misses alike.

WHAT YOU'LL SEE

Every flight, scored like this.

Schedule, aircraft, open premium seats against cabin size, fare — and the upgrade score: a calibrated probability with its honest range, plus one-tap booking on the airline's own site. And where we hold real data, you'll see things nobody else shows you:

READING THE ROW BELOW

10 of 20 premium seats open now (~6 expected by departure); you project around #3; modeled from typical elite demand on this route (no waitlist yet) — a real contender — reasonable to book, unwise to count on.

FlightScheduleAircraftSeatsPriceUpgrade score
UA 1525 10:24 AM → 6:05 PM
4h 41m · nonstop
Airbus A321neo 10 / 20PZ 2 · PN 4 $486
57% Contender band 47–67%
Book on United ↗

Then, five days at a glance.

Picking the wrong day is the most expensive upgrade mistake there is. So every search opens with five days side by side — morning, afternoon, and evening for each — and the best window gets flagged. Move your trip half a day and your odds can double. Now you can see that before you book.

MONJun 22
34%Afternoon
TUEJun 23
41%Afternoon
WEDJun 24
57%Afternoon
BEST
THUJun 25
38%Afternoon
FRIJun 26
26%Afternoon

Illustrative example with sample numbers — run a search to see the odds for your status.

GRADED, NOT CLAIMED

We check our own homework.

Anyone can put "95% accurate" on a landing page. We don't. Instead, after a tracked flight departs, we go back and grade the prediction: did that passenger, at that position, actually clear? Hits and misses both count. Here's how that's going:

Tens of thousands
of flights behind the model
Built on availability data from more than 23,000 real flights. Not survey answers. Not simulations.
0.91
ranking score (ROC-AUC)
Pick any passenger who cleared and any who didn't. Our model ranked the cleared one higher 91% of the time. A model with no idea scores 0.50 — a coin flip.
0.112
Brier score — how far our probabilities land from reality
Lower is better. A coin flip scores 0.25. Guessing the historical average for every flight scores 0.226. Ours is 0.112 — about half the error of guessing.

Put another way: when we call a flight Strong, it clears about 2.5× as often as the average flight. When we call it a long shot, believe us — those clear about 1 in 18 times.

And our model keeps improving the more you use the site — every search and every tracked flight becomes new training data for the next version.

These figures come from our audit table and update as flights are verified — a prediction only counts once we've seen both the queue before departure and the outcome after it.

OUR STORY

Built by a road warrior, for road warriors.

I spend a serious share of my life in seat 12C, hoping it turns into 2A. I also spent years building machine-learning models for a living.

So I got tired of folk wisdom and gate-agent tea-leaf reading, and built a calibrated model to tell me which of my own flights were actually worth booking. It changed how I book — sharing it felt obvious.

Upgrade Pro tells you the truth the way I wanted to hear it: a probability with honest error bars, not a marketing number.

THE FOUR BANDS

What a verdict actually promises.

Every prediction lands in one of four confidence bands, set by the model's calibrated estimate. The ranges below are the odds you can expect when you see each label on a flight.

Strong 65–100%

Book it expecting the upgrade. Still not a guarantee — roughly 1 in 4 will miss.

Contender 40–65%

Leans your way: reasonable to book, unwise to count on.

Underdog 15–40%

Possible, not probable — the odds are genuinely against you.

Long shot 0–15%

Don't count on it clearing. Book this flight for the schedule, not the seat.

0255075100%
Where we're weakest — honestly
EARLY DAYS

This product is young.

Upgrade Pro is in its early stages. Coverage deepens airline by airline, the model retrains as every tracked flight becomes new training data, and verified outcomes accumulate slowly — a flight only counts once we have both a standby queue before departure and a confirmed result after it. Expect rough edges, expect visible improvement — and when something is estimated, expect us to label it instead of dressing it up.

If something looks wrong, that feedback genuinely shapes what gets built next.

What we watch, airline by airline

ALASKA · ATMOS

The upgrade list, before you ask the gate agent.

WL

The actual published list. You know the drill: empty first-class seats, a gate agent who says "we'll call your name," and no idea where you stand. Within 3 days of departure, Alaska rows show the airline's own queue — open First seats, how many are waiting, how many already cleared.

4 / 16 · 12 on upgrade list · 8 cleared

Which way it's moving. When we've captured a flight twice, you see the direction: 9 → 4 open since Tue. Both numbers are real airline availability.

WHAT WE CAN'T SEEBeyond ~3 days out, Alaska doesn't publish the list. Those rows use the calibrated model, and they're labeled that way.
UNITED · MILEAGEPLUS

PZ and PN, by name.

PZ

The buckets you actually clear through. If you hold PlusPoints, one number decides whether they're worth burning: is PZ open? We show PZ and PN on captured rows, spelled out — PZ 0 means that bucket is closed right now, and you just saved yourself the points.

10 / 20 · PZ 2 · PN 4
96

The 96-hour window. United releases complimentary elite upgrades at the 96-hour mark. Our capture timing and odds are built around it.

WHAT WE CAN'T SEEUnited blocks most automated reads. When a capture doesn't land, the row is modeled and labeled Estimated — we tell you which is which.
MORE CARRIERS

Delta & American

Full calibrated odds today. Delta and American flyers get the same probability engine, confidence bands, and honest data labels on every search.

And we're always working on new features. Coverage deepens airline by airline — what you see on Alaska and United is where every carrier is headed.