Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[KarenRei]

Troy, thanks for your work. Here’s what I see as the problem.

Early quarter forecasts are by their nature less accurate and have a greater unknown. That’s understood. What’s not understood is that your early quarter forecasts seem to have a consistent low bias. And it’s in opposition to guidance.

Do you have a plan, or idea, how to improve your early quarter prediction accuracy? At least so that you miss high, just as often as you miss low?

This is a legitimate point. Early-season forecasting and late-season modeling based on data reporting are very different things.

 

[lklundin]

Congratulations, Dennis - you're our TSLAQ Fraud of the Day (as well as having severe Munchausen syndrome)! Oh, and Washington Post? Check your freaking sources better next time. :Þ

Also, by pretending to be one this despicable person is making life harder for actual veteran amputees.

Btw, here is a tweet of mine to him, before he decided to take his fraud somewhere else:

 

[KarenRei]

Do you mean 12 LR packs for Semi, not 4?

There were some earlier diagrams (don't have a link at the moment) that showed 4x battery packs in Semi. Which if accurate could be 4x200kWh (e.g. potentially Roadster battery packs). 4 packs also corresponds with 4 drive units, e.g. one pack per drive unit (with power balancing), for added redundancy.

But there's little hard data to go on here.

 

[The Accountant]

Semi will have a slow ramp. Fleet buyers are very cautious. They test new equipment for 18 months or more before purchasing in volume. Tesla can build handfuls of Semis in the corner of an existing factory or buy/lease a building in someplace like Lathrop.

Even when mature, Semi will not need a high speed, highly automated production line. The building will be small compared to Fremont or even GF3. The body will probably be composite (e.g. fiberglass or CRP) so no need high precision stamping dies that take a year to make. We could see dozens on the road before we know much about factory location.

Per information on Wikipedia, as of Q1 2018 there were 2,000 pre-orders for the Semi with Pepsico ordering 100 and UPS 125. So building "handfuls of Semis in the corner of the existing factory" won't cut it. I suspect that if parts and service meet expectations and the Semi proves to be reliable, the cost savings will result in repeat purchases from these initial buyers. Tesla needs to be ready to scale production.

Tesla Semi - Wikipedia

 

[KarenRei]

Quality trolling right here

Elon Musk on Twitter

 

[Webeevdrivers]

Yeah, I was only checking SR+.

Sitting in the leaf waiting for my wife in the supermarket. I tried a few configurations and they all come up as 3 weeks. A local Facebook group member reports his wife ordered a plus white on white and was given a 4 week window. I would think 3 to 4 weeks is probably a good number to use.

 

[kbM3]

There were some earlier diagrams (don't have a link at the moment) that showed 4x battery packs in Semi. Which if accurate could be 4x200kWh (e.g. potentially Roadster battery packs). 4 packs also corresponds with 4 drive units, e.g. one pack per drive unit (with power balancing), for added redundancy.

But there's little hard data to go on here.

800 kWh for 500 mile range would be incredible. He mentioned < 2 kWh / mile, so hitting 1.6 would be quite a milestone.

However, he’s also recently mentioned closer to 600 than 500 miles, so 800 kWh or less seems too good to be true

 

[KarenRei]

800 kWh for 500 mile range would be incredible. He mentioned < 2 kWh / mile, so hitting 1.6 would be quite a milestone.

However, he’s also recently mentioned closer to 600 than 500 miles, so 800 kWh or less seems too good to be true

If they think their production cell energy density will be higher by the time of production, then that could also be another explanation. But yes, we have little data to go on.

 

[2virgule5]

There's a lot of serious issues with this latest article. Robb Holland writes:

"For comparison, I could just about get four laps of the Nordschleife out of a tuned 400 horsepower Ford Focus RS. The P100D+’s output isn’t known, but it’s likely significantly more than that."​

Um, the power output IS known, and with precision, turn by turn, for the entire course. Tesla tweeted this graph labeled Mechanical Power [kw] for each point on the course from their telemetry: (tweeted on Sep 19, Jalopnik published on Sep 21 so no excuses)

Reading the graph, we can see that max power was 500 kw, and that was achieved at numerous points throughout the entire course. Full throttle at the exit of each corner, heavy regen going into the next corner, just like a proper EV racecar would do.

And there's further bonus points missed by Robb Holland: The graph shows no power rollback during the lap, so we know there was no overheating. But that analysis escaped Jalopnik.

Then, Holland's article piles this dreamy falsehood onto their growing heap of errors:

"Tesla claims that when it returns next month, it could get closer to seven minutes and five seconds. But the automaker’s released no video or data or anything to back this up, so for now, it’s just a claim on Twitter."​

This statement is comically false. Tesla's tweet from Sep 19 literally says: "Here’s some of our initial Model S Plaid data to keep you buzzing until we return".

Tesla also provided this telemetry augmenting the lap times: (apparently 'physicks' is different in Germany, and just plain unintelligible at Jalopnik)

This article actually raises some larger issues with Jalopnik: (author and editors both)

• does Jalopnik not understand that 500 kw is equal to 670 mechanical hp?
• are they deliberately ignoring data that makes their storyline less compelling?
• do they even comprehend the density of data Telsa provided for Plaid laps?

Clowns. At this point, the only thing missing from this comedy sketch by Jalopnik is the big red nose. At least they're already wearing their useless floppy shoes, now on public display in this latest article. Jalopnik: the Place for *BS*. That was a Hard fail.

View attachment 457807

Finally, Tesla said the telemetry data in their tweet should keep us buzzing for a month. That's how long it'd take to extract the data, tabularize it, and do proper physics analysis. You know, applying F=MA and all that "data or anything to back this up"? Any bets that when a proper data analysis comes, it won't be from Robb Holland and Jalopnik?

Cheers!

Let alone that they could not help but push again the narrative that Tesla is likely to be late delivering the Plaid (and linking to a 2014 FUD article)... Like if Tesla did not deliver model 3 ahead of time, vs all so-called competitors that have yet to deliver credible competition. It doesn't matter in the end, if Elon says that they can get to 7:05 it means that they are likely to reach it at some point, and whenever they see it fit they'll do it with a production-spec car.

 

[2virgule5]

800 kWh for 500 mile range would be incredible. He mentioned < 2 kWh / mile, so hitting 1.6 would be quite a milestone.

However, he’s also recently mentioned closer to 600 than 500 miles, so 800 kWh or less seems too good to be true

I missed the information that there would be 4 packs, but it makes sense - 250KWh each make for a 'mega' charger rate, and also increases energy from regen (200+KWh?) which will be greatly needed on the Semi.

 

[elasalle]

That's the thing. Tesla has never had "demand problems". They've had logistics problems (some quarters worse than others). Whenever they want more demand, they expand to new markets, but they have to make that decision well in advance, to set up the local logistics infrastructure and to produce relative to the local requirements. You can move vehicles around within a given market, but it takes time - and between markets only if they have the same requirements, and with significant delay. And moving vehicles around costs money. Logistics can be messed up in terms of rates of purchases in a given market not matching up to forecasts, difficulty getting vehicles to said market, or difficulty distributing them once they get to said market.

Tesla faces a much harder job with logistics that "competitors" because it's basically doing JIT with an uneven rate of consumption. Most automakers have months of inventory at any given location; they just "restock" wherever inventory gets low. Tesla has to rely on forecasts.

End Of quarter, I'm more concerned about the outcome, not the problem. cheers!!

 

[Fact Checking]

Quality trolling right here

Elon Musk on Twitter

Oh yes:

Porsche's 7:42 Taycan record was also carefully set so that their cash cow 4-door and 2-door ICE models would not be cannibalized by Taycan demand: the Panamera's best Nürburgring result is around 7:38, and most of the production 2-door 911's are in the 7:2x range.

So Porsche's problem: if they magically grow a much more capable EV platform (they won't), if they were beating Tesla with a 7:1x result they'd still be cannibalizing their own ICE offerings, which are still supposed to earn high margin profits in the next 3-4 years to make the Taycan investment break-even.

The Taycan is all about protecting Porsche's "racing car" niche and the halo effect, to keep customer jumping ship to Tesla. This image and their market leader position got seriously damaged today, and Elon's announcement that the Roadster 2 is expected to beat the 6:44 all time production car lap record on the Nürburgring will make everyone who truly wants to own the best racing car at minimum wait another year, before spending $200k-$300k on a Porsche that might be obsolete on arrival.

now he's got Porsche in sights as well, who should fear a $100k 4-door Nürburgring champion with 100,000/year production capacity far more than the originally planned $250k 2-door Roadster 2 Nürburgring champion with a few thousand sales per year.

 

[lklundin]

I still have little to say of relevance to TSLA.
However, as a Tesla Model 3 owner I have several ideas and wishes for improvements to the Tesla software.
I could find no place for Tesla owners to collect and share opinions on such ideas and wishes, so I created one.
Given that it is Sunday, I hope that it is not to far OT to pimp that thread here:
Software wishlist

 

[acoste]

I made a reply, but it's stuck in the pending queue (probably because it includes pictures)? Maybe I should have just linked to the pictures.

Have you also sent those pictures to Elon who thinks these cars have 7 seats, while it is clear from the pictures that the red car is missing at least the rear seats and the blue is missing at least the passenger seat???

 

[KarenRei]

Have you also sent those pictures to Elon who thinks these cars have 7 seats, while it is clear from the pictures that the red car is missing at least the rear seats and the blue is missing at least the passenger seat???

View attachment 457867

Okay, I've been being IMHO extremely patient with you (some other people have already started blocking you), but you're STILL making claims about the cars without providing any piece of photo evidence showing what you're claiming? After all this? This is going beyond annoying.

Either kindly show a picture that clearly shows missing rear and jump seats, or please bugger off somewhere else. Thanks. Not "here's some vague silhouette that I'm interpreting as a roll cage which I'm then going to use to interpret that there's no rear seats" - an actual picture of what you're claiming.

 

[Artful Dodger]

Oh yes:

now he's got Porsche in sights as well, who should fear a $100k 4-door Nürburgring champion with 100,000/year production capacity​

First was BMW with the Model 3. Next is Porsche (unless they abandon ICE with haste). That's the "Power of Plaid".

Cheers!

 

[Tslynk67]

I'm shaking my head reading that. They're talking about right-brain brute force optimization without any creativity.

I once made a program to fit geocoded maps of every single road in Iceland (including lots of road metadata, altitudes, etc) - gigs of raw data - into just a couple megs that could be rapidly loaded by a website and processed by Javascript. How? Rather than storing all of the points that define each road, I took advantage of the fact that the points on roads are not just random, but relative to the points that came before, and furthermore, curves on roads tend to just continue along. So I wrote a program to calculate optimal splines that could represent dozens or even hundreds of points that define a road, only switching splines whenever the deviation got too great to be represented with a single spline. Representing data as splines not only made it far smaller, but it made it easier for the client side to interpolate the roads to arbitrary resolution and with much greater speed than having to draw every single point.

In Cruise's case, it's not exactly the same thing (1d vs. 2d data), but the same sort of logic applies. Your dataset is too large to cache? Don't just try to come up with a way to crunch an awkwardly large dataset; look for a better way to represent your data.

Lossless compression, basically. Smart!

 

[Fact Checking]

The Taycan was also pre-production (the "record" was "set" before the official production start date given out by Porsche), stripped and with a roll-cage, and on non-stock, unnamed "summer tires" (you can only configure the vehicle with all-seasons) (Porsche gave out deliberately misleading wording, stating that it was on "production 21-inch wheels wrapped in summer tires" - the wheels are indeed stock (the 21" 'Mission E' wheels), but the tires are not available stock).

BTW., Porsche PR apparently insists (but not on record ...) that they only stripped the Taycan interior of rear seats and interior to counter the mass increase of their roll cage.

But a roll cage isn't overly heavy, a fiber one is I think below 50 kg, and carbon-fiber roll cages exist as well:

Carbon Fibre Roll Cage - Anti Splinter Protected | Oakley Design​

Plus the Taycan is a small car so requires a smaller roll cage.

So what curb weight did the Porsche Taycan have on their 7:42 run, exactly? I presume Porsche used the car weight scales at the Nürburgring that racing officials are using. Why isn't Porsche disclosing the exact weight measurement the machine performed?

 

[KarenRei]

New hitpiece from Dana (yes, she's trying to keep the solar thing alive, because why not?):

https://www.bloomberg.com/news/articles/2019-08-29/more-tesla-solar-panel-cases-emerge-in-wake-of-walmart-suit?utm_content=tech

What, make any attempt whatsoever to compare fire incidence rates between Tesla and other installers? Why on Earth would you ever do that? Would ruin a perfectly good hit piece.

 

[2virgule5]

Have you also sent those pictures to Elon who thinks these cars have 7 seats, while it is clear from the pictures that the red car is missing at least the rear seats and the blue is missing at least the passenger seat???

View attachment 457867

This is becoming a silly, circular argument. Neither Porsche nor Tesla have registered their run, so what the car had or had not is basically a game of who you trust (and clearly you don't trust Tesla, and I don't care).

What is clear to me is that Tesla will likely make an official run at some point, so we'll know what's inside ; they will market a Plaid version within a year and will make a point of running a 'production car' on the track as well when they have something to sell (not doing so would make this whole PR stunt counterproductive). What I suspect Porsche is doing is scratching their heads to try to beat the unofficial Tesla time, because frankly at this point they are probably more embarassed than Tesla. And if they saw that coming and have better things to show then great for all of us. Have a nice week-end.