Video breakdown: What does this crash say about lane-splitting?

This video of a lane-splitting crash has been viewed a quarter of a million times. Some see it as an "unavoidable" crash that shows that lane-splitting is unsafe. Was this incident really unavoidable?

Motorcyclist and YouTube user Austin Kim posted the video of his painful but thankfully not life-threatening crash in the spring. I’m not sure where, exactly, he was when he captured his own accident on a helmet-mounted GoPro, but the first responders came from the Los Angeles County Fire Department.

When Kim posted his video he summarized it as “Getting hit by car crossing double yellow.” Many of the thousands of comments are from motorcyclists who, like the OP himself, seem to put 100 percent of the blame for this crash on the car driver. Other commenters — mostly from outside California — noted that this “unavoidable” crash was a perfect illustration of why they’d never want to lane-split.

But was it really unavoidable? Does it prove that lane-splitting is inherently unsafe? A "Crash or Don't Crash" frame-by-frame analysis of the incident shows that, as usual, this crash happened because two guys did something stupid at the same time, in the same place.

crash video

T minus 1.0 second

First, what the rider does right: He’s riding a Kawasaki Ninja 300. A reasonable choice for a rider with less than a year’s experience. It’s a nice, narrow bike that’s well suited to slipping through traffic. If you watch to the end, you’ll see he’s also wearing “all the gear”; a full-face helmet (the GoPro was mounted on chin bar), Dainese jacket, and Sedici motorcycle gloves; I can’t tell what he’s wearing for trousers. He’s lane-splitting between lanes one and two, which is usually considered the best choice.

What we don’t know: The video starts exactly one second before impact. So we don’t know anything about the rider’s behavior, or traffic conditions, leading up to the accident. More importantly for my purposes, we don’t know if the Ford driver’s done anything to telegraph his desire to move to the right. The Ford has a fairly dark rear window tint, but we can still make out the driver’s outline in the GoPro video. The human eye could see more. Did the rider fail to notice or respond to a shoulder check from the driver?

The Ford is drifting right slowly at the beginning of the video. That could indicate a desire on the part of the driver to change lanes, but in the first few frames, the variation in his lane position is within the standard deviation, it could just be lane drift. The Ford driver’s not doing anything particularly alarming, but there are reasons why the rider should already be paying attention to that particular car.

What the motorcyclist is already doing wrong: Comparing the position of the BMW to the rider’s right from frames 1/5th of a second apart, it’s clear that the auto traffic is barely moving. The Kawasaki’s speedometer indicates 40 miles per hour.

The rider’s speed may have been appropriate two minutes ago, or even five seconds ago, but right now, when the traffic column is barely moving, a closing speed of 35 mph is too fast. Most of the videos I analyze start in Don’t Crash mode and serve to illustrate a moment where the rider inadvertently flicked the switch to Crash. This guy’s already in Crash mode, and all we can look for is a moment where he could flick it back into Don’t Crash.

Back in my Motorcyclist magazine days, I commuted from San Diego into Hollywood every time I went to work. So I’ve spent thousands of miles lane-splitting in traffic like this. I learned that freeway traffic that’s flowing at a reasonable speed, say above 35 mph — especially if it’s flowing at the same speed across all lanes — tends to be predictable. Bumper-to-bumper traffic that’s completely stopped is also predictable, if only because cars can’t cut across your bow unless they have a space to move into.

crash video

T minus 0.65 second

The rider’s outfitted his Ninja with a cool set of adjustable levers. Too bad that he doesn’t realize the lever on the right is also the Don’t Crash switch, while there’s still time to use it.

There are two reasons he should brake: First, his closing speed is simply too fast; and second, drivers get frustrated in stop-and-go traffic. As a motorcyclist lane-splitting under such conditions, you need to be especially ready for sudden, impulsive actions.

Although the BMW to the rider’s right does not show brake lights, it’s barely moving (if at all). Now, compare the traffic density in lane one to lane two; the Ford driver’s probably doing that himself. The HOV lane he’s in is bumper-to-bumper, while lane two has gaps between cars.

When I was lane-splitting, I liked to pass between cars traveling abreast of each other, because I knew neither one was likely to suddenly change lanes. I even liked riding beside semi-trucks because although they’re wider and leave a smaller path, car drivers almost never make an absent-minded lane change towards a vehicle big enough to crush them.

This is a novice/average rider’s last chance to flick the Don’t Crash switch. Even though the Ford hasn’t done anything alarming yet, the speed differential alone should have set the motorcyclist’s spider sense tingling. But there’s also this: The Ninja rider should be aware that the lane to his right is poised to move ever so slightly faster in the next second or two.

crash video

T minus 0.5 second

We’re now half a second, and about 25 feet, from impact. The speedo still shows 40.

The BMW’s lane position is good; that driver may have seen the motorcycle coming and made a little space. But, the BMW is hardly moving, so that driver can’t shift much more to the right, even if he wants to.

So far, we’ve seen the Ford drift about one foot to the right. The approaching motorcyclist has a lot to process right now. He needs to look at the Ford’s right front tire; does it indicate the driver’s actually steering to the right? (In the same way that you can spot car movement by looking at the top of the wheel, the best way to spot a steering input is to watch the gap between the front of the tire and the wheel arch.)

Does this look like a car that would have working turn signals? (Probably.) What about the driver’s body language? Has he glanced over to the right? Is the driver’s right-side wing mirror properly positioned? (The motorcyclist should be able to see the driver in that mirror.)

A lot to think about, eh? If the guy on the Ninja doesn’t have the bandwidth to process all that information, he’s going too fast.

The rider’s responded to the Ford’s rightward drift by making a slight move to his right. Neither the speedo nor the video suggest any effort to brake. The rider may still think the Ford’s just drifting in the HOV lane. But by now, the Ford’s trajectory is distinctly threatening. If the car driver bothered to signal, or if the motorcyclist had started scrubbing speed a few car lengths back, this would merely be infuriating.

crash video

T minus 0.4 seconds

“Oh yes you are, you...” Double yellows be damned, the Ford’s moving over from lane one to lane two. Usually, even a pretty rude lane change still leaves enough space for a motorcycle between vehicles (maybe not comfortable room, but room.) Not this time.

We’re about 0.4 seconds from impact. Even if the rider was only going 20 mph, it’s unlikely that he could completely avoid the Ford’s rear bumper. So, he’s in Crash mode; the only question now is "How hard?"

Since the distance between the Ford and BMW can’t possibly accommodate a bike, the last Don’t Crash outcome would have been passing the Ford on the left, over by the Jersey barrier. We can see clear space in the Ninja’s left mirror, so an emergency lane change, direction left, would have been a possibility, if — only two letters, but such a big word, if, eh? — he’d started adjusting his closing speed a few car lengths back.

If you look carefully, you can just see the rider’s right index finger is properly positioned for normal riding. The problem is, he’s not squeezing it. He’s still going 40 mph. At that speed, even the world’s best supermoto racer on hot slicks couldn’t make that move stick.

crash video


When it touches the Ford, the left handlebar violently countersteers the motorcycle. That tucks the front and it falls away from the car. Speedo still shows 40 mph. Dude’s riding a Ninja, but he clearly doesn’t have a ninja’s reactions. Honestly, even if he’d been traveling 20 mph, he might well have hit that Ford (though he probably would have been able to dust himself off and continue home.)

crash video

T plus 1 second

Now desperate to avoid impact with concrete, the rider reaches his hands out in the Superman position, to see if he can fly. He can’t. Luckily, surrounding traffic was moving slowly and he was not run over.

Conclusion: Lane-splitting done right

The video continues for 15 minutes, ending with the rider being carried off by paramedics. The motorcyclist was not legally at fault. I suppose he’d better hope the Ford driver had good insurance, because as soon as a bystander tells him 911 has been called, he plaintively moans, “I can’t afford the ambulance, dude.”

By pointing out that while not legally at fault he still shares some responsibility for his crash, I’m not blaming the victim. All I’m saying is, it is better to get home than to be legally in the right in the hospital. Later in the video he asks, more than once, whether his GoPro was on. The poster has other less-viewed videos on his channel that are kind of lane-splitting porn. So I think it’s possible he was riding unsafely in the hopes of getting a popular video. (He was successful!)

On a track they say, “When in doubt, gas it.” But in traffic, most accidents can be avoided by giving yourself more time and distance. For practical purposes, that means that the Crash or Don’t Crash switch you should use most often is your front brake.

For a while, California posted guidelines on lane-splitting but later took them down. As other states have considered legalizing it, proposed legislation has set out to define “safe” lane-splitting. The Washington state Senate just approved a bill to allow lane-splitting at speeds up to 25 mph, with a maximum speed differential of 10 mph. That bill now moves to the state’s Transportation Committee and will presumably get a vote in Washington House of Representatives.

Personally, I’m not sure there should be a specific rule.

In my experience, a number of conditions influence safe lane-splitting speed and there are several things a rider should keep in mind:

  • Riders need to be aware of upcoming exits or freeway splits. If you’re splitting between a HOV and a normal lane, watch for HOV exits because they concentrate lane-change activity.
  • Riders need to be aware of traffic conditions that are likely to provoke jealousy among cagers. Any time you’re in between two lanes traveling at different speeds, you need to be on high alert.
  • Being between two vehicles is better than being between a car and a gap, due to the risk of a sudden lane change.
  • Not all lanes are created equal. According to CalTrans, the standard freeway lane width is 12 feet, but lanes can range from 15 feet to as little as 10 feet. That’s a factor in determining a safe speed differential.
  • It’s a lot better if you can see into cars you’re about to pass. That’s more difficult at night of if cars have dark tints.
  • Car drivers that have just been passed by a lane-splitting motorcycle are more attentive to following bikes. The best lane-splitting position is several car lengths behind another rider, so if a rider catches up to you while lane splitting, let him through. That said, don’t let anyone else’s pace dictate your own.

Those guidelines helped me ride thousands of commuting miles, lane-splitting daily, and still stay in Don't Crash mode.

comments powered by Disqus