The Mini Out-Cornered My M40i. It Lost by Seven Seconds Anyway.

Part 4.4 of my AI-coached track weekends. The virtual race against Rich's Mini was close. This one isn't — and that's exactly why it's interesting.

Two facts from the same Sunday at Road Atlanta, both true:

Fact one: at six of the seven corners the data can measure, Rich's 2,350-lb Mini Cooper S matched or beat my Z4 M40i's cornering speed — and at four of them it was outright faster through the apex, in a car with barely half the horsepower.

Fact two: my M40i beat it by 7.318 seconds a lap.

This post is about how both of those can be true at once — which turns out to be the clearest answer I've ever seen to the oldest argument in any paddock: where does lap time actually come from?

Where seven seconds lives

The AI split the lap into zones and asked, bucket by bucket, where the M40i's margin was earned:

Where	Time gained by the M40i	Share of the 7.318s
Acceleration and straights	4.152s	56.7%
Braking zones	2.942s	40.2%
The corners themselves (apex speed)	0.224s	3.1%

Read that bottom line again. Across an entire lap of Road Atlanta, the 382-horsepower car's advantage in the corners was two tenths of a second — a rounding error. And the braking-zone bucket is less about braking skill than it sounds: most of that 2.9 seconds is the M40i arriving at each braking zone carrying 15–20 mph more speed, banked on the straight before. Time-at-distance math credits the zone; the horsepower did the work.

Nearly the whole seven seconds is one thing: what happens when both drivers flatten the right pedal. On the back straight alone — corner exit to the chicane braking point — the M40i takes 3.9 seconds out of the Mini, over half the total margin in one stretch of road. The speed ladder tells it plainly: at corner exit the two cars are even; 600 meters later the BMW is +22 mph; by the end of the straight, +23 and still pulling.

Same corners, two solutions

The hole (Turn 7) — the slowest corner on the track, and the Mini's masterpiece: 48.8 mph minimum to the M40i's 44.1 — 4.7 mph faster through the tightest point of the circuit, in the car that can least afford to slow down and most needs to carry everything. The M40i still wins the zone by 1.9 seconds — but every bit of that is the approach, arriving 20 mph faster off the run from Turn 5. The corner itself belongs entirely to the Mini.

The 10a/b chicane — the most complete portrait of the whole matchup in one corner. The M40i arrives at 136.3 mph, the Mini at 120. Then the Mini out-brakes it (−1.03g peak to −0.67g), out-grips it (1.38g lateral to 1.13g), and carries a higher minimum through the flick (55.2 to 53.1 mph). It does everything better — braking, grip, apex — and still loses two seconds in the zone, because the racetrack pays out on arrival speed and the BMW shows up with more of it every single time.

The corner in question: the M40i chasing a GT4 through the 10a/10b section — arriving with the speed that pays the bills.

The first Esses dip — the footnote that made me laugh: it's the only corner where the M40i carries a higher minimum than the Mini (60.5 to 58.0). Which is also the only corner where my bone-stock E86 out-cornered the Mini in the virtual race. Two different Z4s, two different analyses, same single corner. Apparently BMW engineered the Z4 chassis for one specific dip at Road Atlanta, and I own two of them.

Turn 12 — the only zone the Mini wins outright on the whole lap (−0.12s): the fast downhill right onto the front straight, where direction-change agility is the entry ticket and the straight that follows isn't long enough for the M40i to collect its usual rent before the line.

The M40i dropping down the hill through Turn 12 — the one zone on the lap where the Mini takes time from the BMW.

You can only steal in one direction

Here's the part worth taking home from this whole exercise.

If I'm the M40i driver reading this data, there's free lap time on the table: the Mini's minimum speeds are proof those corners can be taken faster — same track, same day, on less tire and less budget. Commitment is a skill, and skills transfer. Every mph of the Mini's apex speed is something I can go practice.

If Rich is the Mini driver reading this data... there is nothing to practice. He's already doing his half of the equation better than I am. The 7.3 seconds he's missing costs about $50,000 and comes with a turbocharged straight-six attached. The power car's driver can steal from the momentum car's driver, but not the other way around. That asymmetry is the whole case for why momentum cars make better teachers: they hand you nothing, so everything you gain is yours. And it's the case for why power cars make comfortable liars: mine flattered me down every straight while quietly losing six of seven corners to a Mini.

It's also — full circle — the argument I made for the E86 project in the last post. The soft, slow, stock car is the honest one. Ask me again after the coilovers go in.

The nerd footnote: both laps are real telemetry from the same Sunday — the Mini's AiM at 10Hz, the M40i's Harry's LapTimer rig at 25Hz — aligned by the same distance-normalized method as part 4.3, self-checked against the true lap-time margin (it lands on 7.318s exactly). One fairness note: the two loggers measure g differently, and the difference biases against the Mini — so where the Mini's braking and cornering g beat the M40i's above, the real gap is, if anything, bigger.

Austin is going to settle something

In October, Zach and I take his Mach 1 to COTA — one 480-hp power car, two drivers, plus the public F1 telemetry from the same corners. After this post, I know exactly what I'm looking for in our data: not who's faster on the straights (the car decides that), but who's braver in the corners — because that's the only column of the spreadsheet the driver actually owns.

Race me: my standing lap times and how to challenge them are at the end of part 4.3. Send your 10Hz+ data — Harry's LapTimer .hlptrl, or CSVs from Porsche PTPA, AiM Race Studio, RaceBox — to upload@trackdays.ai.