Car or driver?

One of the shockers at the last Lemons race was that the winning car won by so may laps. Usually the winners are separated by 1 lap, not 9. Is the “Shake and Break” E30 vehicle that much better than the rest of the field or do they have better drivers? It’s hard to answer this question without having the same driver in both vehicles. But I’ll try anyway.


The weather on Saturday was pretty consistent, and with ~110 cars there wasn’t that much traffic. That means that lap times should be a reasonably accurate indicator of performance. But performance is a combination of driver and vehicle. To separate the two, I segmented the laps based on the driver. I don’t know which driver was in when, but every driver change involves a long-ish pit stop, so I simply called each driver change a new driver. I’ve included the fastest lap and the median lap below (in seconds, not minutes:seconds). Laps longer than 300 seconds were removed because those represent either a pit stop or some extended full-course yellow or even red flag.

Given the similarities between the Shake and Brake #1 and #4 drivers, I think they are the same person (Anthony Zwain).

  • Shake #1: fastest 201.687, median 210.80
  • Shake #2: fastest 213.709, median 221.25
  • Shake #3: fastest 210.30, median 216.64
  • Shake #4: fastest 204.18, median 209.54

Here are Eyesore’s times. I think they ran 4 different drivers.

  • Eyesore #1: fastest 222.95, median 229.68
  • Eyesore #2: fastest 221.10, median 230.74
  • Eyesore #3: fastest 209.66, median 216.56
  • Eyesore #4: fastest 210.09, median 215.60

Rather than comparing laps between cars, let’s look at the variation within each team focusing on the median. The Shake #1/#4 driver laps in the 209-210 range. The #3 driver is typically ~6 seconds behind, and the #2 driver is another ~5 seconds behind. There’s a heck of a lot of variation between the drivers! The Eyesore times show that the #1 and #2 drivers are pretty similar to each other and ~14 seconds behind the #3 and #4 drivers. That monstrous gap represents a 6% difference in lap times, or about 2 laps per stint. The Shake #1/4 driver is around 3-5% faster than the #2 & #3 drivers, or 1-2 laps per stint. After some sketchy mathematics, I come to the following conclusion: if you remove the fastest driver from Shake and remove either of the slowest drivers from Eyesore, the result would be Eyesore winning by ~1-2 laps.


Sunday was a little damp early, then rainy mid-day, then drying, then dry. A wet race is very different from a dry race. It’s more about the driver than the car. Let’s look at what happened in the rain. This time, we’ll look at the graphs at SpeedHive. There was a red flag in the middle of Saturday, which you can see as the big spike in lap times near lap 60. This was preceded by some full course yellow. Note that the peaks of Eyesore (orange) and Shake (white) aren’t aligned because the graph is based on laps, not time of day. By the time the red flag occurred, Shake was already well ahead of Eyesore. Look towards the right side of the graph and you can see the baseline swells. This was when it was really wet. During this period the lap time differences are really interesting. At the start, the Shake driver is lapping much faster than the Eyesore driver, maybe 15-20 seconds. But then they pit and the next drivers are doing very similar times. If there’s a big difference between these cars, it’s not apparent from the lap times.


I don’t think the Shake & Break E30 is anything special. If it was, you would have seen that at the last 2 Sonoma races. However, in those 2 races, they were slower than Eyesore. The reality is that Shake brought in an exceptional driver and ran a clean race. There’s a huge difference among drivers on the same team and when you add the rain, there’s more than enough variation to explain the 9 lap difference.

4 thoughts on “Car or driver?

  1. or…. shake and break #1 and eyesore #3,#4 are equivalent drivers and the others are where they are. The car is ~n sec faster due to power. power is negated in the rain; therefore rain lap times between “fast eyesore” and “fast S&B” are the same. is that a viable alternate theory? Optimum Lap says Mx-5 w/ 130% power adder would be 5sec faster on old T-hill, so 9sec on 5-miler? i feel we could use one more data point –> video of main straight in the dry; comparative lap times at Sonoma; something like that… that’d allow us to separate out alien driver theory from alien car theory.

    As an aside 6-8 months or so ago i saw a craigslist ad for a 325i E30 M20 sleeper engine that a team had used in Lemons. if i recall – 3.0L sleeper w/ stoker crank, rods, ~10:1 compression, hot cam, megasquirt… – but looked crappy from outside. they said it put ~230hp. stock is ~170. curiously that’s a 35% power adder… it was back when i was contemplating another car…


      1. i was there for that event. it was a blast. it’d be awesome if they did that again! we had coleman drive the first hamster and he was something like 5sec faster then the fastest of us on T-hill… if i recall he set the fastest lap for the weekend in the thing. we weren’t anywhere close. ah… good times.

        i bet if you ask someone would do a stint or partial stint… why not? if you were willing to give up a seat you could probably get just about anyone to do at least some laps in the thing. that’d actually be pretty cool. and you could capture all their data!


      2. I’ve actually had a couple of hotshoes drive my cars. Pete Pressley (Cerveza) drove the Miata at Laguna Seca. Pablo Marx (Risky Whisky) drove the Yaris at Buttonwillow. I also had Roger Eagleton (he’s a professional) drive the Miata at Thunderhill in the highly unusual Middle Hill configuration. Global rally driver (including WRC) Matt Shinnors drove the Yaris at Laguna Seca. All told, there have been 34 drivers in my 2 cars.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s