I've known Brett Doar since we built a LeMons racer together back in 2008. Since then Brett has made his career out of similarly suspicious engineering, especially making large, rambling Rube Goldberg machines.

He was part of the team that made the OK Go Rube Goldberg machine, made one for the Colbert Report, and now helped make the colossal Athlete Machine for Red Bull.

Here's how one person engineers our favorite form of chaos.

Jalopnik So, what is all this? How did it all start?

Brett Doar: It's an athlete-mediated chain reaction machine. A Rube Goldberg machine. Actually a bunch of little Rube Goldberg machines linked together by a bunch of athletes. The site was about five acres of what used to be El Toro Marine Base. Our hangar was where they've shot a lot of stuff for Top Gear. Back before they built the base, in the 40s or so, it was world's largest lima bean field.

I think they had been in the planning stages for a year or longer, but we got the go-ahead in early September, and we began the actual building in mid-September. Adam Sadowsky, the president of Syynlabs, mentioned it to me in August. He started assembling a team and we had an all hands meeting in september- we had a general idea of what athletes were going to be involved, and then just breaking the whole thing down into sections, and figuring out who was going to be responsible for them. Shortly after that my friend Brett Phillips and I started just building stuff.

The Man Behind The World's Biggest Rube Goldberg Machines Explains How You Control ChaosS

It usually starts with a shopping trip — Matt Samsel and I just jumped in the car and drove to a sporting goods store and were like "how about this?" "I dunno, throw it in the cart." So we got to the checkout counter with a fishing pole, some barbells, various balls, a machete, some fishing nets, an inflatable kiddie pool, some rifle ammunition, a bunch of airsoft pellets and paintballs, a couple of skateboards, a kid's archery set, some of those exercise trampolines and I think a badminton set. The guy behind us goes "You guys are gonna have a weird weekend".

There's a lot to be said for just being open to what hits you — you're in Home Depot getting lumber and suddenly it's like "this is a really cheap chainsaw". Throw it in the cart! We'll return it if we don't use it! (we aren't gonna return it).

Jalopnik How long did you spend designing the overall machine? Who did what?

Brett Doar: Usually the design happens in tandem with the building. You start out with a general idea and refine it while you're building it. You're building something that has never existed before, and even if you have a pretty clear idea of how things interact together, you don't really know if it's going to work until you've got it built. There are plenty of things that theoretically should work, but theory is one thing and practice is a whole other thing. You're also dealing with a lot of slop in the system — you're trying to work with very wide tolerances at first, and then trying to isolate problems one by one. It's hard, for me at least, to do that on paper. Ultimately design is prediction, and this stuff is hard to predict.

The Man Behind The World's Biggest Rube Goldberg Machines Explains How You Control ChaosS

It's hard, at least for me, to predict what will come up when you're trying to, say, drop a machete to cut a rope, which releases a contraption to fling a soccer ball 30 feet to hit the corner of an 80 pound soccer goal that is just barely balancing on its edge, and then figuring out how that is going to make a lever drop 15 feet above that.

Then, in this case, we're outside, and it's an airfield so it's very flat and so there is a lot of wind. The sun changes position — so in the morning there's a little bit of dew on things, and it's cool, and in the afternoon things heat up, so wood swells and shrinks. Plastic gets a little softer. Stuff that slid very easily in the morning is binding now in the afternoon.

With the middle part (the part with the sander), the original plan was that this beam was going to be knocked and rotate around, and it was going to have a razor blade at the tip, and that was going to pop a series of balloons. And it was happening inside, which I thought was good because that would be a more controllable environment, but it turned out that with these huge doors open on both ends the hangar just channelled the wind through the place, and the balloons where just getting thrashed, and it became clear that this was never going to work. So it was like "well we gotta get from here to there and we need something to do this" — in this case I needed to yank a pin out very quickly from a hole in the handle of a screwdriver that was holding it in this tube, and the screwdriver fell about 5 feet and landed in a funnel where it connected a circuit that turned on the belt sander.

And so I'm looking around and I see these C-stands, stands that hold lights and stuff, and I'd ordered them but never ended up using them, and I started playing around, and figured out that I could arrange things in a very precarious (but repeatable!!) way that, when it was just tapped, it caused the entire thing to just collapse, and in the process could yank the pin out. I'm really proud of that part because it was the part that I became really worried about, and in the end I had two production assistants running it that had never seen it before the shoot, and they were doing so well that I was able to totally hand it off to them, I didn't even want to touch it, it was like Bob Dylan giving "All Along the Watchtower" to Hendrix. I wish I could remember those guys (the PAs) names, because they were awesome.

But I guess the point is, you can spend lots of time designing, and for me at least, it all ends up being improvised at the end anyway. In the beginning there's always designerly kinds of talk- "we should make it out of machined aluminum! we should make these giant gears that mesh together!", but in the end it's about "what can you build?" And it's supposed to be a kluge anyway- it's in the freakin' name! Kluges are about improvisation more than planning.

I should say that that's my approach, though- I think better with my hands. Matt Samsel is the guy who was in charge of that beautiful derrick and the airplane, and the stuff leading up to Lolo Jones and the hurdles, and it seemed like he knew exactly what he was going to do long before he did it, and it all worked pretty much exactly as he expected it to, with some minor adjustments that needed to be made. And the NASA guys (Oscar Murrillo, David Boyle, James Flynn, and Jake Shaeffer), who built the bowling ball light bulb smashy thing and the rubber band propeller on the wire thing, seemed to have worked a lot of stuff out beforehand, but I'm not sure.

Jalopnik When designing, do you make scale models? Or do you use math and other arcane magic?

Brett Doar: My main design tools are an abacus, a hunk of ambergris, and a dowsing rod. The dowsing rod is very important.

A scale model really isn't going to help actually figure out what is or isn't going to work. You might sort of place blocks around a drawing of the site to figure out how big something needs to be, how much distance you need to cover, or just get an idea of what type of interaction needs to happen, but it really doesn't work in the actual design of the thing.

Jalopnik How did you pick who would be involved, driver, cyclist, runner-wise? Was that decided from the start, and did you work around them?

Brett Doar: That was decided by Red Bull, and it tended to shift a bit even while we were building. There was at least one part that was designed for a specific athlete who was going to be involved, then not, then it was on again and ultimately he canceled just because he was going to be on the other side of the planet. That was really frustrating until I realized, "Oh, yeah, this isn't the Justice League". Danny Macaskill and Robbie Maddison are not hanging out with Lolo Jones at Red Bull HQ playing pool and waiting for red phone to ring. They're out there in far flung places doing their thing, and it takes a lot to get them all in one place. And even so, we really only had them for one day.

One of the things Kenny Abney, one of our really experienced builders, said really nails it for me: "the particular skills we have, lots of people have them. We're not getting hired to do this because of those skills. The skill we're getting hired for is flexibility". And so you figure out a way around those things.

The Man Behind The World's Biggest Rube Goldberg Machines Explains How You Control ChaosS

I think those kind of changes are often useful, because you look at that thing that you built and you say "how good is this really? Is this good enough to keep somewhere else, or do we just trash it?" A lot of things ultimately get trashed.

Jalopnik How much freedom did Red Bull give you? any hard guidelines or things you had to work in or keep out?

Brett Doar: One of the guidelines that I really respected was that they didn't want their logo all over the place. In one case I wanted to use this little rolling cooler, the kind that are in convenience stores, that looks like a red bull can, and I wanted to roll it down a ramp, and I floated the idea by someone with Red Bull, and they said "Maybe, but don't make it look like a red bull can". The only way anyone will know what it is is if it looks like a Red Bull can! So we lost it. And it wasn't that great of an idea anyway.

I honestly can't think of anything where they really said "do not go here". I don't think they would have wanted us to use guns, but that really wouldn't have played well on video anyway. I wanted to do something with a wrecking ball, but that was nixed more because of the logistics of getting a crane out there didn't justify the payoff. And obviously we weren't going to show the product in a bad light.

I think we, the designers and builders, and the client were really on the same page during the build, and I think it says a lot about them. But they wanted a big machine, so I guess we gave that to them and they're happy!

The same goes with the athletes — you're putting them inside the machine, and they have to interact with it in a specific way, and they could get hurt if you're not careful, and that's their living. I really expected somebody to show up to set and go "I'm not going near that", and having to completely do something else, but they were completely amenable to everything.

Jalopnik Putting athletes and their careers in grave danger. Got it. What specific parts did you build, and what was your favorite?

Brett Doar: I was sort of supervising the beginning up to the release of the drift car, and offering suggestions and filling holes when need be, but I was directly responsible for the "soccer ball catapult to chainsaw" part that released Danny Macaskill, the "rolling dumbell to belt sander" part that Ryan Sheckler sets off, and the "Truck crash through the bucket wall to drop tank release" between Bryce Menzies and Rhys Millen. I had a lot of help building that stuff though — Brett Phillips built a lot of the chainsaw part, as well as a lot of the middle part and the chainsaw was his idea. The whole Bryce- to-Rhys part needed pro riggers involved, and David Paris and Chris Stockton really killed that, along with Kenny Abney, who was really important in terms of keeping that kind of stuff safe.

I think that part (with the drop tank) was probably my favorite for the sheer kinetic ballet of it. The truck smashes through the wall of buckets, and that releases a barrel that rolls down and knocks into what is essentially a hammer- it's a section of five inch steel pipe that's filled with sand and then welded shut, stuck on the end of a long steel shaft. It weighs about 90lbs, and it's balanced really delicately so that the barrel just needs to nudge it. It smashes into a sort of teeter totter, launching this "rocket" — I don't know why I decided on a rocket, it just seemed like that's what it should be. The rocket shoots up on a cable, and trips a lever which releases that thing that looks like a bomb — it's actually a drop tank from a WWII era Corsair fighter (Matt Samsel has 3 of these for some reason). The drop tank swings and trips a "cripple knee", which is supporting the rear differential of the drift car. The cripple knee collapses and drops the car, and off he goes.

The Man Behind The World's Biggest Rube Goldberg Machines Explains How You Control ChaosS

That part to me feels the most elegant, and I'm not sure why. I think there's a good combination of brute force and delicacy about it, and the scale is right, and there's a good balance of fast and slow. I liked it so much I put my initials on the rocket!

Jalopnik What was the hardest to pull off?

Brett Doar: It's always happens that the night before, you've tested and tested and retested everything, and you're ready to go, and then you show up in the morning and nothing works. And once you've gotten things squared away again, the thing you were having all the problems with is working beautifully, and the thing that has always worked perfectly is suddenly having problems.

One of the hardest things was the part I just mentioned, the part with the drop tank. We had the wall of buckets braced really well in that archway, but a storm was rolling in on the shoot day (which is why the clouds looked so spectacular), and the wind shifted, and the wall of buckets, which had been parallel to the wind, suddenly became a sail, and just kept blowing over. Then suddenly, the lever that the rocket hit suddenly was having problems. Luckily the guys running that section, David and Chris, are amazingly capable guys. Chris' nickname is "Fixit", and it's for good reason.

Jalopnik How dangerous was this thing, really?

Brett Doar: The day before the shoot, they brought an ambulance on set and a garbage truck was parked beside it, and we were joking about how those were your choices of how you left if you got injured.

You can probably hurt yourself pretty bad on any movie set. Even without the machine we had forklifts, heavy stuff hanging from the ceiling, a camera car driving around pretty quick. But yeah, we had 90lb hammers primed to drop.

I brought a steel helmet to set, because I thought it would be funny, and pretty soon I realized there's a sledgehammer just barely balanced right above my head, and I think "I'm gonna put my helmet on".

With the soccer ball catapult, there's a machete on the end of an eight foot arm that cuts the rope to release it. So that's a bad place to be when that comes down. At one point I was working on the trigger, and the machete was down, and at some point I go "what's poking me in the chest?" and I realized that I was about to impale myself on the machete. So you say "Okay, I'm gonna put the sheath on that, and not lean into it".

Ultimately we know pretty well what we're working with and what those things can do, so we always build in safeties- the chainsaw has an automatic kill switch so even if it does go off accidentally, it won't do anything past a certain point. And you create procedures and checklists. You want to approach it like a pilot- there's a detailed list and most of it is really stupid, but it's always the thing you assume you did that you forgot to do. If you're careful and methodical about it, it's not that dangerous, even though there is a lot of potential for danger.

I think the most dangerous thing was actually Robbie Maddison's jump(s), but he hit it every single time. I think he does that when he goes to the store to buy milk.

Jalopnik Any final thoughts, hot stock tips, anything else?

Brett Doar: There were a lot of people involved in this, and I want them to get credit- We have a full list of everyone involved on the syynlabs website.