How do you respond to a New York Times piece that illustrates, in depth, the essential un-usability of your hot new $100,000 high-performance electric sedan? If you're Elon Musk, Tesla Motors CEO, you do what any good tech executive does when he wants to appeal to an audience of geeks: You dump a lot of data.
In a long blog post today, Musk attempts to rebut New York Times auto reporter John Broder's weekend piece about the disastrous D.C.-to-Connecticut trip he took in Tesla Motor's new Great Electric Hope, the Model S. It's a classic of the form — accusatory, self-righteous, melodramatic, only vaguely successful at raising questions about Broder's piece.
Musk dumps five graphs and a map, knowing his audience is pre-disposed to trust "raw data," regardless of the source. And understanding that many people who read it didn't read, or don't remember, Broder's original piece, he focuses on the things he knows he can dispute, and rebuts claims that are never actually made. A quick Twitter search shows that the presence of "data," unexamined, was enough to convince most people (and especially the tech community from which Musk comes) that Musk has fully refuted Broder's article, even if most of that data doesn't correspond to anything Broder specifically claimed.
Broder's article presents a fairly damning picture of the Model S: in an attempt to take a reasonably untaxing road trip up the northeast corridor, under normal conditions of use, Broder finds that he needs to put the car on cruise control at 55 m.p.h. and turn off the heat to avoid sapping the car's battery. After spending the night in Connecticut, Broder wakes to find the battery had lost two-thirds of its estimated range. At one point, the car shuts down completely and needs to be picked up in a flatbed.
"[T]he Model S battery never ran out of energy at any time, including when Broder called the flatbed truck," Musk writes. Broder never said it did, simply that the car shut itself down and refused to turn on again until it had been charged more. (Our friends at Jalopnik spoke with Broder's flatbed towing company, which says its records "indicate the car's battery pack was completely drained.")
Musk writes, "At the point in time that he claims to have turned the temperature down, he in fact turned the temperature up to 74 F. In fact, Broder never specified exactly when he turned the temperature down. About 20 miles after the point Musk identified on the temperature graph, there's a precipitous drop. For about 50 miles, by Musk's own graph, the car was running along with a cabin temperature around 64 degrees.
Did Broder set the battery up to fail? "Despite narrowly making each leg, he charged less and less each time. Why would anyone do that?" Musk asks. But again, he's not addressing the original piece. Broder was clear that he had charged "less and less." (In a followup post, Broder wrote about one charge: "I added 185 miles of range at Milford, knowing that I wouldn't need 242 or 265 miles before recharging the next morning.")
Misdirections aside, Musk's piece does three things. One, it notes a clear error in the graphic accompanying the piece, which contains a self-contradictory caption.
Two, it seems to dispute a claim Broder makes about advice from Tesla: Musk writes that, prior to the final leg of his trip, Broder "disconnected the charge cable... expressly against the advice of Tesla personnel." Discussing that charge in the article itself, Broder writes that "after an hour [Tesla's experts] cleared me to resume the trip," i.e., he disconnected with the blessing of Tesla personnel. That's a real dispute of a fact asserted in the piece, though there's no obvious reason to believe Musk's account over Broder's.
And three, it challenges certain figures in Broder's piece—places in the article where the raw data from his surveillance software contradicts numbers claimed by Broder: "Cruise control was never set to 54 mph as claimed in the article [...] Broder in fact drove at speeds from 65 mph to 81 mph for a majority of the trip [...] The charge time on his second stop was 47 mins [...] not 58 mins as stated in the graphic attached to his article." (The latter claim is particularly important, because, Musk claims, the extra nine minutes would have made it "virtually impossible to run out of energy for the remainder of his stated journey.")
To take those claims seriously you have to trust Musk's "raw data," taken from a car that Broder's article specifically charges with inaccurate readings—data that, as Jalopnik shows, contradicts not just Broder's claims but the data of the tow truck company that picked up the Model S. Maybe Broder lied about setting the cruise control on 54 when he set it on 60; maybe the cruise control was calibrated wrong.
Musk also included a graph zooming in on an odd interval when the car traveled "[a]pproximately 0.6 miles" at erratic low speed, caused by "[c]ontinuous back-and-forth driving in a small, 100 space parking lot." Musk presents this as evidence that Broder was intentionally killing the battery: "driving around in circles ... trying to get the Model S to stop." Broder told New York magazine he was "circling the parking lot ... looking for the unmarked and unlighted Supercharger port."
Ultimately Musk only vaguely addresses the real concerns that Broder brings up—that the car takes too much of a penalty in cold weather; that its computer displays inaccurate readings; that its battery can drop nearly dead overnight. The nine-minute difference in charge time can be attributed to lot of things that aren't Broder lying—he might have mistimed; the data could be wrong—but the loss of two-thirds of battery due to a "computer glitch" can't be attributed to much else than a basic failure on the part of the Model S. When you're defending the performance of your $100,000 supercar by putting a hot red arrow to flag where the driver momentarily went "over 80 MPH" on an interstate drive, you're rather painfully missing the point.
Broder—if that's even his real name [GRAPH SHOWING IT'S NOT]—has said he'll respond this afternoon. It almost doesn't matter. Enough people will have heard about Musk's post to turn the question of whether or not the Model S is a shitty car into a "controversy." And Musk will have succeeded in burying the original issues under his mound of raw data.