Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

theatlanticwire.com

Post reply on HN
Print view View original

131–140 of 262 posts

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#131

Post by DanBC »

OGinparadise wrote: post #35
DanBC wrote: post #27

Earlier quoted context omitted.

When you say you are limping along at 45 mph you shouldn't be driving at over 50 mph. > 45 mph or 52mph, it's the same crap to me. Would that work with traffic police?

Would that work with traffic police? who cares, but usually yes, at least in US (you go fast, you slow, fast...and police are looking for the worst offender) . My point is: a car that has to be driven with a gazillion, million, billion specific instructions or else it's a brick, maybe it isn't for me or many others. "Your car died and you're in the middle of nowhere because you went 5mph over the recommended by Tesla…

> My point is: a car that has to be driven with a gazillion, million, billion specific instructions or else it's a brick,

You know that cars need to be filled with gas, right? And that using diesel in a regular (or vice - versa) is a big problem? And they need you to check the oil, and top up the oil? And that not topping up the oil is a big problem, right?

> "Your car died and you're in the middle of nowhere because you went 5mph over the recommended by Tesla speed,"

Well, no. "Your car died because you didn't charge it. You didn't charge it even though you knew you had a journey of X miles to go, and you knew you didn't have enough charge, and you were given strong advice (from the car, and from the car makers) that you hadn't charged it enough". Plug the car in - that's not a gazzillion million billion specific instructions. It's just one. Plug it in, let it charge.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#132

Post by DanBC »

OGinparadise wrote: post #71
zevyoura wrote: post #57

Earlier quoted context omitted.

You're arguing a straw man point; while the speed difference may have contributed to the discharge, I don't believe anyone's arguing it made the difference.

Not really, if you look at Tesla's blog post, they even highlighted that went at 80mph for a while or once. Why did they do that? To prove that he didn't write a book (a story wouldn't fit it) and detail every speed he drove or because they speed matters?

They did it to prove that he lied about the speed he was driving.

He claimed to be driving at one speed. He was driving not at that speed, and sometimes significantly faster.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#133

Post by killerpopiller »

halviti wrote: post #109
lawdawg wrote: post #106

Earlier quoted context omitted.

Did you (or 99% of HN) read the NYT article? "When I parked the car, its computer said I had 90 miles of range, twice the 46 miles back to Milford. It was a different story at 8:30 the next morning. The thermometer read 10 degrees and the display showed 25 miles of remaining range" "I called Tesla in California, and the official I woke up said I needed to “condition” the battery pack to restore the lost energy." Look…

Nobody is lying, but if you get in your car and it tells you that you can only you can only go 25 miles, do you really think it's ok to drive over twice that?

it actually does work. Cars usually have a fuel reserve and thus range up to 50km (5 litres) when it already shows 0km.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#134

Post by correctifier »

halviti wrote: post #109
lawdawg wrote: post #106

Earlier quoted context omitted.

Did you (or 99% of HN) read the NYT article? "When I parked the car, its computer said I had 90 miles of range, twice the 46 miles back to Milford. It was a different story at 8:30 the next morning. The thermometer read 10 degrees and the display showed 25 miles of remaining range" "I called Tesla in California, and the official I woke up said I needed to “condition” the battery pack to restore the lost energy." Look…

Nobody is lying, but if you get in your car and it tells you that you can only you can only go 25 miles, do you really think it's ok to drive over twice that?

Well remember that the car "lost" 60 miles of range overnight. If the Tesla people told him that charging it for a bit would restore the range it is conceivable that he would think that the whole 60 miles would come back. Remember that not everyone has a detailed understanding of how cold temperatures affect lithium ion batteries.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#135

Post by knowaveragejoe »

lawdawg wrote: post #91
onethumb wrote: post #37

I thought the most convincing part of Elon's post was his 3rd bullet point, which relies on no data outside of the article, and which this Atlantic Wire journalist ignores entirely: "In his article, Broder claims that “the car fell short of its projected range on the final leg.” Then he bizarrely states that the screen showed “Est. remaining range: 32 miles” and the car traveled “51 miles," contradicting his own stat…

Actually it did when the car reported 90 the night before and a call to Tesla assured him that a quick charge would "re-calibrate" the system because of a software bug.

Seeing as no-one has heard the phone call, it's hard to say what exactly the Tesla rep assured him of.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#136

Post by danso »

onethumb wrote: post #37

I thought the most convincing part of Elon's post was his 3rd bullet point, which relies on no data outside of the article, and which this Atlantic Wire journalist ignores entirely: "In his article, Broder claims that “the car fell short of its projected range on the final leg.” Then he bizarrely states that the screen showed “Est. remaining range: 32 miles” and the car traveled “51 miles," contradicting his own stat…

> The car said it would go 32. It went 51. Broder claimed this "fell short". Clearly it didn't.

Your statement is on point only if you don't consider what was in the actual graphic:

http://www.nytimes.com/imagepages/2013/02/10/automobiles/10t...

An excerpt:

"Driving a 2013 Model S with an 85-kilowatt-hour battery pack, which has an EPA range rating of 265 miles...the car barely reached the next Supercharger, 206 miles away in Milford, Conn. The next day, in 10 degree weather, the car fell short of its projected range..."

So when you read the statement in context, and you actually look at the graphic, you realize that the "fell short of its projected range" is a claim made in the summary of the graphic, in which it is clear that, because of space reasons, the journey is described as day by day segments.

So the NYT does not assert that the Model S's final leg was just the 32 miles left on the car. That "fell short of its projected range" is referring to the 265 EPA that the car normally would've had had it not been so cold out.

Should the NYT have been more verbose? Well, maybe. But I guess they also expect readers to actually look at the infographic, the purpose of which is to show greater detail.

It's quite misleading for Elon Musk to harp on a summary-statement in a different context.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#137

Post by ceejayoz »

sk5t wrote: post #88
ceejayoz wrote: post #46

Earlier quoted context omitted.

Also, that's the heater setting, not the cabin temperature. I can set my car's internal heat setting to 50 degrees after getting it nice and toasty, and as long as I don't turn the A/C on it's going to stay warm for quite some time.

Quite a few newer cars don't draw a distinction between running the heater or A/C: you just set a desired temperature, and the climate control system figures out what to do based on internal and exterior conditions. (Whether this is a good approach for a car's climate control system is another question, however.)

I doubt the Tesla does that. If it did, Musk would've said something along the lines of "setting the cabin temp to 50 kicked on the A/C".

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#138

Post by rhizome »

taligent wrote: post #48
grecy wrote: post #31

Earlier quoted context omitted.

> Given that his job was to write an accurate review Why do you assume that? His job was to attract readership and hits for the NYT. Nothing does that better than controversy.

The New York Times is not some Murdoch trashy paper that deals with gossip and controversy. The integrity of their work is important because that is what is keeping the paper alive whilst the others are falling by the way side. Maybe you should actually try reading it sometime.

Pace your sense of the NY Times motivations, I offer Judy Miller:

http://en.wikipedia.org/wiki/Judith_Miller#New_York_Times_ca...

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#139

Post by rayiner »

naftaliharris wrote: post #49
adastra wrote: post #24

The "Norwich Charge" data point is all you need to understand what happened here. Notice how vague Broder is in his original account: "After making arrangements to recharge at the Norwich station, I located the proper adapter in the trunk, plugged in and walked to the only warm place nearby, Butch’s Luncheonette and Breakfast Club, an establishment (smoking allowed) where only members can buy a cup of coffee or a pla…

> Elon might have been better off concentrating on this point more, rather than some of the smaller inaccuracies. I agree. Often when you're arguing, you have a really strong argument and a few weaker ones. You'd think that making both the strong argument and the weaker ones would be the best case, (ie, that the strength of your case equals the sum of the strength of its components). In fact, in public discourse ofte…

> In fact, in public discourse often the opposite is true.

This isn't just true in public discourse by the way. It's true in proposal/grant writing, legal argumentation, or even just arguing with your parents. You should always lead with your strongest arguments, and think very carefully before including any weak ones. Humans are just bad at doing weighted analysis of persuasive arguments. Say you have an argument, with three points, with weights: 0.9, 0.05, and 0.05. Your opponent destroying your two weak arguments is going to convince the reader or decision maker far more than the 0.1 combined weight of the arguments.

Re: Elon Musk's Data Doesn't Back Up His Claims of New York Times Fakery

#140

Post by slantyyz »

specialist wrote: post #53
davidiach wrote: post #20

Earlier quoted context omitted.

I think the truth is somewhere in the middle. Broder might not have wanted to "be evil". He might be like the guy who doesn't actually want to hurt the company he is working for but still spends all his day playing games at work. Given that his job was to write an accurate review, the fact that he didn't really want to wait until the car is fully charged and was bored driving the car at normal speed and so on, we can…

As someone who's tried, with varied success, to work with reporters on technical stuff, I don't subscribe to "the somewhere in the middle" view. To us geeks, often, stuff is either true or not true, with no room for maybe. To reporters, they're both trying to understand what's happening without direct knowledge as well as convey that in a way a layperson can understand. Unfortunately, in that game of telephone, a lot…

>> What's great (rare) about this story is there's published data

I've only seen precompiled analyses, including the charts and graphs from Tesla's blog.

There is no real data per se - Wired requested the raw logs several times and have been declined by Tesla (http://www.wired.com/autopia/2013/02/tesla-logs-nytimes/ - last sentence).

And I don't get how there's this perception here on HN that Tesla's data is ironclad.

The logging routines were probably written by humans, so there could be bugs in them. Logs are usually written to text files, which can be edited. The people interpreting the data could be making mistakes with the analysis before it even gets to Musk (look at all the posts about the problems with Excel on HN in the past couple of days).

Tesla's a startup. Startups can be a little chaotic. Their procedures and controls with respect to their data may not be as robust as, say a credit card company's, yet there's this big presumption that the data is all good and pure.

I would give Tesla the benefit of the doubt, since it would be very stupid to doctor up the data, but let's not all kid ourselves and not assume that Tesla's not going to shade their "analysis" in a way that is favorable to their argument. It's not much different than what the NYT reporter and the NYT are doing.

Post reply on HN