Smash Lab and the Scientific Method

Recently, a new show has appeared on “The Discovery Channel” called “Smash Lab”. The premise of the show is simple: four smart people – an engineer, an “ideas guy”, a scientist, and a designer – get together and try to approach real-world problems with innovative technology. For instance, using carbon-fiber wrap to hurricane-proof a mobile home, or using softened concrete to slow down a run-away highway vehicle. I’m impressed by all of the problems so far posed; the hurricane-proofing of mobile homes was particularly interesting, given how many of these homes are in the path of hurricanes along the coastal United States.

However, several things about the show have bothered me, as a scientist. The pilot episode exemplifies my concerns, though I will say that I’ve become more impressed with the discipline of the team as the show has progressed. In the first episode, the team tries to slow down run-away cars and buses using an easy-crumble concrete. This concrete is already used on some airport runways to slow down and stop run-away planes. The concrete crumbles more easily, absorbing the energy of a collision gradually to reduce the instantaneous forces on passengers inside the vehicle.

The team tried several ideas, including changing normal concrete road barriers into softened barriers, or putting a surface of the crumbly concrete in the median so that a vehicle crossing the median slows and stops before hitting on-coming traffic on the other side of the highway. Despite their effort, the ideas were all miserable failures. That’s fine – science is about failure, and learning from failure, to achieve success. However, the team seemed to declare victory at the end. Why? Well, they all failed to stop a car. However, they were able to slow-down a bus crossing the median on the crumbly surface.

What bothers me about this is that the team redefined success at the end of the experiment, allowing them to declare a victory. Cars are far more common on U.S. highways than buses, and so therefore the probability of a fatal accident from a run-away car is far higher than from a bus. The failure of the experiment to stop a car without killing the passengers, or without actually stopping the car, amounts to a failure in the bulk of the cases in which this technology could be tested. Succeeding in stopping a bus is good, and dramatic, but hardly more important than stopping a car.

The team should have admitted defeat. In fact, they’ve done so in a later episode when they failed to fully protect a mobile home from a hurricane, or even when they failed to save a car on railroad tracks from a deadly train strike. That’s good science. But in the pilot, the lead episode of the show, they redefined success at the end so they had something good to claim.

Failing to meet the goals of an experiment can be just as useful as finding success. Perhaps collapsable concrete is a bad technology, one which cannot or should not be adapted to solve the problem of a run-away car. Declaring some measure of victory, using a smaller part of the overall problem as the basis of the claim, can prolong the life of a failed technology. It would be better to spend innovation on a better idea than try to pump up a bad one. That’s an important aspect of science.

I’m growing to like Smash Lab. The earthquake-proofing of a house with ball-bearings was swank, something I think should be adapted to normal house construction. But the concrete barrier should have been something abandoned by the team as a bad idea. It’s a shame they couldn’t proudly accept defeat, instead of redefining success just to achieve it.

Pant, Pant

Well, we went from the summer of nuts, to the winter of insane, to the fast-paced late-winter of WTF. I’ve been cutting non-essential things out of my life as a result, including this blog. Oh well – last to the table, first out the door. Today, however, I’ve been forcing myself to rest more than usual. I’m between waves right now, a good time to catch some extra sleep, call friends and family, and do things that are off the core list.

Since BaBar switched from being a “B” factor to being a “b” factory, my colleagues and I have been working our asses off to catch as much good physics as possible. A whiplash change of physics program, well-motivated or not, doesn’t mean that all the hard work of the last 10 years just continues to work without intervention. Needless to say, a small change in collider energy has needed a lot of love from the collaboration to stay on the rails. So far, so good.

Last week, BaBar had it’s first “analysis school”. Students, post-docs, and faculty became lecturers to about 70 new or new-ish BaBar members, mostly students. I didn’t recognize about half the people in the room, which felt good. Event in the last year of running, and the last few years of data analysis, BaBar continues to attract the best and brightest students. Their enthusiasm held out over six days of lectures, from 9 am to 6 pm each day, with computer and physics exercises. There were great lectures on how to do an analysis, how to avoid statistics pitfalls, how to understand detector systems, and even how to do back-of-the-envelope calculations. It was great. I wish BaBar had done this 5 years ago.

Jodi’s been traveling a LOT. She’s on her way back today (see below), only to rest for a day before leaving for a big dark matter conference at UCLA on Tuesday. She and I will play plane tag for the next month; I’ll be in Washington, she’ll be in the midwest, and on and on. We’ll have a short break together at the end of March before heading hard and fast into the summer conference cycle.

Whew. I’ve taken some time today to clean up the cottage, spend time with the cats, and even reflect on the fact that I am now an uncle-in-law (and God-parent) to two newborn twin boys. I can’t wait to meet them!

This week is the Babar Collaboration meeting. So much to do!

Meanwhile, let’s keep an eye on Jodi: