The only way to show the most accurate estimate is to know where the driver is going, and many drivers don't always input their destination.
It can also vary dramatically based on how you drive, and in fact, if you do have a destination, it will give hints if relevant (e.g. "keep speed below 70 mph to reach destination").
You can switch the EPA-rated miles for a battery percentage indicator by just tapping it in the UI.
Let's say you live in a state with a 55mph speed limit.
The car has a lifetime average of 0.300 kWh/mi over 4 years of driving.
It's December, ambient temperature is 20F and it is snowing.
Your last 10 drives consumption rate was 0.400 kWh/mi because its cold out.
There's A LOT of smarter things the car can do than use its factory set 0.250 kWh/mi consumption to spit out the range on dashboard.
Erring on the side of caution would also always be better, whereas it is erring on the side of showing the absolute best state that I as an individual never experienced in 4 years of ownership.
It's not erring on the side of the absolute best state. The EPA-rated miles represent a standardized mix of highway and city miles.
The best state would be over 600 miles [1].
It's erring on the side of consistency, as it's also representing the battery's state of charge. It would only serve to confuse users to try and guess whether the current trip is a highway or city drive, will end in the next minute or two hours from now, which way the wind will be blowing and whether or not you're driving uphill, etc. It'd just be a useless guess, instead of an at least consistent one.
If you tell it where you're going, it already gives you a great estimate, as it doe stake into account all of those factors.
Look, I get it, some people love these cars and will stand no criticism, fine. I had one, they are lovely. We are allowed to want things we like to be better.
But all the "just use nav estimate" arguments miss that some people drive more than one place?
If I am driving out an hour to do some shopping, visit my parents, and then drive back.. it may be a total of 90mi but the first leg is 30mi. Do I need to pre-enter my round trip as multiple waypoints to get an accurate view of whether I will make it home or not?
Seems silly.
Just guesstimate a more accurate dashboard view range like you do in other place sin the UI.
I get it, too. I've seen some people that resort to ad hominems online. We are allowed to evaluate suggestions critically.
The other places in the UI are either based on either the destination you've entered or the last 5, 15, or X miles driven. They only make sense in that context: Just using one of those numbers on the dashboard would be misleading and inconsistent.
Guessing based on the Wh/m of your last ten drives might be a closer estimate, but would be wrong if your next drive isn't like your last ten. It might even be more accurate when guessing your last nine drives, so then you learn to trust it, but those were city miles, and for the next drive, it's highway miles. And now the estimate is way off.
It's just meaningless. We'd quickly learn that the number given is a bad estimate that can't be trusted and needs to be adjusted based on the way you'll actually be driving and other real world conditions... and that's where we already are. The only difference is that currently it's at least consistent, both between trips in the same car and from one car to another. This additional guesstimate would just add more confusion.
Look I don't really care about Teslas so I don't really have any idea, but the article says that Tesla does NOT use the standardized EPA-mix, but instead does their own testing and gets those tests EPA approved. The EPA spokesperson said they were following the rules, but only technically. Do you have alternate information showing that it actually is the standardized mix to produce the estimate?
The data alone should tell you that Tesla is up to something.
Almost every real world range test shows that Tesla systematically underperforms in the real world relative to spec, worse than substantially all other carmakers.
It's a shame too because Tesla are towards the top of the the real world range pack, so they don't really have to lie.. but it's like a compulsion for some people.
What is mind boggling and hilarious is that there's a Porsche rated EPA 225mi and a Tesla rated EPA 348mi and they both end up achieving 280mi.
Notably the Model 3/Y seem to have a wider EPA-to-reality delta than the Model S in the 75mph test.
This seems to align with my 4 year ownership experience. Tesla Model 3 LR rated 310mi but only hits 200mi with C&D. It's interesting how much the 3/Y range seems to drop from 70 to 75mph.
Most accurate by what standard? Range depends on a bunch of factors. The range you got going to the store yesterday is not the range you will get driving to your friend’s house tomorrow.
The car doesn’t know where you are going, whether it is city or highway driving, whether or not there is an elevation change, unless you tell it by navigating somewhere.
It could take a guess, but that guess still has plenty of chances to be wrong if your next X miles of driving is not going to be similar to your previous X miles of driving.
Why should the car use an average .250 kWh/mi hardcoded rate on a car I have owned for 4 years when it knows my lifetime average is 0.300 kWh/mi?
There's plenty of smarter things it can do by default other than "its hard, meh".
They are gonna figure out how to do coast to coast self driving this year but can't project a reasonable battery range estimate given temperature/driver history/weather? Do they need more GPUs?
That might be a useful metric but it won’t always be the most accurate.
If you mostly drive around town, that estimate will be way off when you get on a highway to go to Grandma’s house.
There’s probably a better way to do it, but Tesla seems to be optimizing for avoiding the “why does the website say 300mi but my car shows 200mi fully charged” support question, in exchange for a different set of support questions.
I think ideally the car would give a best guess estimate, along with a clear breakdown of why this is more or less than the rated range. I just don’t think that’s clearly the “most accurate” option. Most accurate requires knowing where you are going.
So make it a menu option - range estimator: best/worst/spec.
It feels like one of those Muskisms where one bending of the truth requires more and more stuff down the line.
If they weren't over optimizing for the EPA test to give almost unachievable range, then they wouldn't need to have the car range meter lie as well. But if you fib once, you need to keep fibbing and keep the fib straight.
We are HN tech crowd here so its all fun & games hobbyist stuff.
But normies do not want to get into these types of inane subtleties.
Arguably the front & center, always-there range display should be the most accurate, not the least accurate.