Percentage or Miles/Kilometers : which do you use and why?

[gecko10x]

I've always used Percent.
After reading most of this thread, I decided there might be merit to using miles, so I tried it this morning.
It lasted 30 seconds before I realized precision doesn't equal accuracy. It makes no difference if miles is more precise, because it's not accurate.
And I actually want a very imprecise number so that I'm not fooled into believing it to be accurate! So, I'm back to percent.

 

[AlanSubie4Life]

because it's not accurate

How could you tell it was not accurate (it's a measure of energy)? I'm only aware of one way to check the accuracy when driving.

Definitely use whatever you prefer, though! There's a reason it is configurable!

 

[jerry33]

I've always used Percent.
After reading most of this thread, I decided there might be merit to using miles, so I tried it this morning.
It lasted 30 seconds before I realized precision doesn't equal accuracy. It makes no difference if miles is more precise, because it's not accurate.
And I actually want a very imprecise number so that I'm not fooled into believing it to be accurate!

The trick is being consistent. Neither actually is. The percentage is based on a capacity estimate, the mi/km based on a formula. with mi/km you soon know what the difference is between how you drive and what the formula shows. But neither really matters because the only time it should be necessary is on a trip (unless your commute is really long) and then the Nav system gives you the estimated percent remaining when you reach your destination and tells you to slow down if it thinks things are dicey.

 

[gecko10x]

How could you tell it was not accurate (it's a measure of energy)? I'm only aware of one way to check the accuracy when driving.

Because by definition it will only be accurate if you always use the rated energy for each distance you drive. I find that to be highly unlikely; nigh impossible on short trips.

 

[AlanSubie4Life]

Nav system gives you the estimated percent remaining when you reach

mi/km based on a formula.

Definitely agree that for when you care about arrival %, the Trip tab on the Energy page is the way to go.

However, I suspect that both the Trip tab and the miles (and % too) will all be equally vulnerable to errors in the state of charge estimation (sort of analogous to a phone battery dropping from 30% to 0% in an instant due to errors in the tracking of energy available - though I'm sure Tesla is MUCH more careful to ensure this does not happen).

 

[gecko10x]

But neither really matters because the only time it should be necessary is on a trip (unless your commute is really long) and then the Nav system gives you the estimated percent remaining when you reach your destination and tells you to slow down if it thinks things are dicey.

Right. This whole discussion is quite pointless. Precisely what the internet was designed for

 

[AlanSubie4Life]

Because by definition it will only be accurate if you always use the rated energy for each distance you drive.

I guess it depends on your definition. I always find it is accurate, when I calculate (Rated Miles Delta) = Trip kWh / 234Wh/rmi

But I definitely get your point, in terms of utility. I am just saying that it is generally very accurate, in terms of what it measures (energy used/available (for delta/remaining, respectively)). Of course, you can just use the trip meter/since last charge for that, too (with the exception of losses when in park)!

 

[DopeGhoti]

What do you mean by “estimate” in the above? Can you describe what you think this number represents? I’ve described several times above what I mean by “estimate” - as with any measure, it is an estimate limited by accuracy (and precision) of the measure. But I am still unclear on what you mean by “estimate.” And I am not clear on why you call it an estimate of “range” (since that would imply some sort of relationship to distance).

You can have the MFD either display a percentage of stage-of-charge, or a numeral followed either by 'miles' or 'kilometers'. I don't think it's unfair to categorize what by all appearances is a statement of a certain distance which is inversely correlated with the state of charge of the car's battery to be an expression of estimated range remaining. Nor would I feel it unfair to classify an estimate of range remaining as an expression of distance.

If I say to you, "50 miles" or "50 kilometers", is your immediate association in your mind "how far is that?", or "how many kilowatt-hours does it take to project me that distance?" Indeed, the latter thought incorporates as a matter of necessity that distance.

Clearly any such estimate would be only as accurate and precise as the inputs fed into its formulae and the predictive accuracy of said formulae.

And why exactly is it “indirect”? (Other than requiring a unit conversion?) 1rmi = 234Wh (for my car).

To me, saying it is indirect is kind of like saying that kilometers are an indirect measure of distance, if we are used to measuring distance in miles. 1.609km = 1mi

That’s the core of what we need to determine if we want to agree/disagree on this.

Indeed, that seems to be the core. Miles, Kilometers, Light-Years, Furlongs, and Fathoms are all direct expressions of distance.

How far is 234 watt-hours? It's impossible to know until you know how far a watt-hour will take you. That's why it's indirect.

 

[AlanSubie4Life]

If I say to you, "50 miles" or "50 kilometers", is your immediate association in your mind "how far is that?", or "how many kilowatt-hours does it take to project me that distance?" Indeed, the latter thought incorporates as a matter of necessity that distance.

I agree that that is the core of people's confusion. That's why I always just try to refer to the units as rmi rather than mi.

to categorize what by all appearances is a statement of a certain distance

Again, certainly understand the confusion (it says "mi" right there!). But to be clear, while it is to all appearances a distance, there is no "distance" in your battery pack. It contains energy. And that rmi number is a measurement (estimate, see above) of that energy - nothing more, nothing less.

Anyway, it sounds like we're in agreement on what it represents, confusing as it may be.

or "how many kilowatt-hours does it take to project me that distance?

I definitely think of it more this way now, if I am traveling in a Tesla. A year ago, not so much.

How far is 234 watt-hours?

Wh are a measure of energy. I never said anything about using it to measure distance. How far that energy will take you depends on the mode of transport, how high you are going, etc. I could also use it to boil a few gallons of water.

 

[DopeGhoti]

Isn't it fun when it turns out that two people are cordially, if vehemently, arguing about something in which it turns out they've been in agreement the whole time?

 

[AlanSubie4Life]

Isn't it fun when it turns out that two people are cordially, if vehemently, arguing about something in which it turns out they've been in agreement the whole time?

Ha. I do think you'll find that the meter is fantastically accurate when it comes to measuring your energy consumption! It is correct over a very wide range of conditions.

 

[jerry33]

However, I suspect that both the Trip tab and the miles (and % too) will all be equally vulnerable to errors in the state of charge estimation (sort of analogous to a phone battery dropping from 30% to 0% in an instant due to errors in the tracking of energy available - though I'm sure Tesla is MUCH more careful to ensure this does not happen).

I've been using the Nav for over six years and 126K miles, and since the trip graph was enable a few years ago, it's been a no-brainer.

 

[darth_vad3r]

Here’s an example to illustrate why a unit of energy scaled by a rated consumption ratio for the specific car wins hands down against all alternatives.

You arrive at the airport rental car lot, and they say “pick any Tesla Model 3 you like”. The weather is nice and you drive like a grandma, so you don’t care if it’s performance, AWD, or RWD. You need to deliver a package in person and immediately return to catch a connecting flight. Top secret spy stuff here. You have to obey the speed limit so you don’t get noticed by police. You have no time to charge at the destination nor time for a detour to charge. Your destination is 100 mi away, no elevation change, so you want to drive 200 mi flat, there, then back again. You’ve calculated that you only have 30s to decide which car to select and be able to make the round trip and still catch your flight.

Yes this is a ridiculous contrived experiment ... but it’s a setup just to illustrate to everyone that “rated miles” is *the* ONLY choice here...

There are 6 cars. None have badges. Which do you choose based only on a quick glance at the gauge, where:

1. Display says 75%
2. Display says 64%
3. Display says “250 estimated miles” (based on last X miles of driving)
4. Display says 48.8 kWh
5. Display says 43.8 kWh
6. Display says “210 rated miles”

Options 3, 4 & 5 don’t exist, but they have been discussed as alternatives, possibly superior even.

With a quick glance at these cars, which one would you choose?
Which one can make it your 200 mi round trip?
Can more than one of them make it? If so, which has the most safety cushion?

(1) 75% sounds like more than (2) 64%, is it? What trim? What’s 100%? Is the range degraded? 75% of 240 SR+ is only 180. Not good enough.
64% of 310 is 198.
Neither of these has 200 rated miles remaining. If the LR had 65% it mignt ... unless it was degraded by 1%.

(4) 48.8 kWh is useless unless you know the trim and the constant. It’s an LR AWD at 245 Wh/mi and would only have 199 miles, that’s not enough.

(5) has even less energy at 43.8 kWh, so it’s no good, right? Well, actually, it’s an SR+ at 219 Wh/mi, it has 200 rated miles which is a smidge more than (4) even with a -5 kWh handicap!

(3) has the biggest number though, 250(!), so take that one, right? Well... what did the last X miles of driving look like? Do we even know what X is? Even if we know, it’s 30 miles ... do we know if the car was driven 30 comparable flat miles to our projected flat trip? Nope, it just descended 30 miles and only has 10 kWh left! To infinity and beyond!

So only (6) is guaranteed to have more than 200 rated miles, since, you know, the display actually tells you it has 210 rated miles.

Not some cryptic x% of some 100% value you don’t know for sure, nor some cryptic kWh.

For bonus you could have choice (7) be the best choice and it could say “100 estimated miles” but have just come up a steep climb and actually have the most range of all.

TL; DR
Rated miles IS energy for your specific car scaling kWh by appropriate factors so that it’s the same regardless of degradation or even in any trim you look at.

If you drive like an EPA grandma, you’ll get that range. If you drive even slower, you’ll get more, if you drive faster you’ll get less. Yes this is variable. All the other choices are too ... but they depend on variables out of your present control—past driving pattern, who drove your car last? Your lead foot brother or your grandma? From where? Uphill, downhill?

% requires you know what ‘100%’ is. This can change any day, even if you think you know it. Why use something that changes day to day when you can use something that’s consistent, repeatable and accurate day to day?

kWh can be equivalent, *if* you know the constant. It’s different for every car. If you own 2 different ones or are choosing a rental like above, it’s more difficult to compare. In the scenario above, who cares if the Model X has 50 kWh left, the SR+ with only 40 kWh will drive way further.

Rated miles is the great, accurate, consistent, repeatable, equalizer.

All hail the rated miles, bow before them.

Bow I said!

 

[jerry33]

Rated miles is the great, accurate, consistent, repeatable, equalizer.

Although it's pretty conservative.

 

[derotam]

You’ve calculated that you only have 30s to decide which car to select and be able to make the round trip and still catch your flight.

Wait! Hold on here...but how long did he take to calculate that 30s time period? Was that 30s from the time he walked up to the cars, or after he finished his calculation? How did he account for the time to calculate the 30s? Did he know how long it was going to take him to calculate how much time he had?

SO....MANY....QUESTIONS!!!!

 

[KenC]

Nice hypo, but how realistic is it? Maybe at car rental places they should use the range number. However, in non-rental cases, people are driving their own cars, not some mystery car. Whether my car shows % or miles, I can do an easy mathematical conversion from one to the other. I just think that range in miles gives a false precision where people are obsessing about the range shown changing a mile from what they are used to. How many threads started are actual problems with the battery and how many are just false alerts due to BMS calculation error or something?

Here’s an example to illustrate why a unit of energy scaled by a rated consumption ratio for the specific car wins hands down against all alternatives.

You arrive at the airport rental car lot, and they say “pick any Tesla Model 3 you like”. The weather is nice and you drive like a grandma, so you don’t care if it’s performance, AWD, or RWD. You need to deliver a package in person and immediately return to catch a connecting flight. Top secret spy stuff here. You have to obey the speed limit so you don’t get noticed by police. You have no time to charge at the destination nor time for a detour to charge. Your destination is 100 mi away, no elevation change, so you want to drive 200 mi flat, there, then back again. You’ve calculated that you only have 30s to decide which car to select and be able to make the round trip and still catch your flight.

Yes this is a ridiculous contrived experiment ... but it’s a setup just to illustrate to everyone that “rated miles” is *the* ONLY choice here...

There are 6 cars. None have badges. Which do you choose based only on a quick glance at the gauge, where:

1. Display says 75%
2. Display says 64%
3. Display says “250 estimated miles” (based on last X miles of driving)
4. Display says 48.8 kWh
5. Display says 43.8 kWh
6. Display says “210 rated miles”

Options 3, 4 & 5 don’t exist, but they have been discussed as alternatives, possibly superior even.

With a quick glance at these cars, which one would you choose?
Which one can make it your 200 mi round trip?
Can more than one of them make it? If so, which has the most safety cushion?

(1) 75% sounds like more than (2) 64%, is it? What trim? What’s 100%? Is the range degraded? 75% of 240 SR+ is only 180. Not good enough.
64% of 310 is 198.
Neither of these has 200 rated miles remaining. If the LR had 65% it mignt ... unless it was degraded by 1%.

(4) 48.8 kWh is useless unless you know the trim and the constant. It’s an LR AWD at 245 Wh/mi and would only have 199 miles, that’s not enough.

(5) has even less energy at 43.8 kWh, so it’s no good, right? Well, actually, it’s an SR+ at 219 Wh/mi, it has 200 rated miles which is a smidge more than (4) even with a -5 kWh handicap!

(3) has the biggest number though, 250(!), so take that one, right? Well... what did the last X miles of driving look like? Do we even know what X is? Even if we know, it’s 30 miles ... do we know if the car was driven 30 comparable flat miles to our projected flat trip? Nope, it just descended 30 miles and only has 10 kWh left! To infinity and beyond!

So only (6) is guaranteed to have more than 200 rated miles, since, you know, the display actually tells you it has 210 rated miles.

Not some cryptic x% of some 100% value you don’t know for sure, nor some cryptic kWh.

For bonus you could have choice (7) be the best choice and it could say “100 estimated miles” but have just come up a steep climb and actually have the most range of all.

TL; DR
Rated miles IS energy for your specific car scaling kWh by appropriate factors so that it’s the same regardless of degradation or even in any trim you look at.

If you drive like an EPA grandma, you’ll get that range. If you drive even slower, you’ll get more, if you drive faster you’ll get less. Yes this is variable. All the other choices are too ... but they depend on variables out of your present control—past driving pattern, who drove your car last? Your lead foot brother or your grandma? From where? Uphill, downhill?

% requires you know what ‘100%’ is. This can change any day, even if you think you know it. Why use something that changes day to day when you can use something that’s consistent, repeatable and accurate day to day?

kWh can be equivalent, *if* you know the constant. It’s different for every car. If you own 2 different ones or are choosing a rental like above, it’s more difficult to compare. In the scenario above, who cares if the Model X has 50 kWh left, the SR+ with only 40 kWh will drive way further.

Rated miles is the great, accurate, consistent, repeatable, equalizer.

All hail the rated miles, bow before them.

Bow I said!

 

[darth_vad3r]

Although it's pretty conservative.

Super conservative, but consistent. Well, or maybe we could say it’s the opposite of conservative actually, since probably most people will use more rated miles to drive real miles.

 

[darth_vad3r]

Wait! Hold on here...but how long did he take to calculate that 30s time period? Was that 30s from the time he walked up to the cars, or after he finished his calculation? How did he account for the time to calculate the 30s? Did he know how long it was going to take him to calculate how much time he had?

SO....MANY....QUESTIONS!!!!

There was a version in my mind where I actually mentioned this calculation was done whilst on the flight to avoid these questions

 

[darth_vad3r]

Nice hypo, but how realistic is it? Maybe at car rental places they should use the range number. However, in non-rental cases, people are driving their own cars, not some mystery car. Whether my car shows % or miles, I can do an easy mathematical conversion from one to the other. I just think that range in miles gives a false precision where people are obsessing about the range shown changing a mile from what they are used to. How many threads started are actual problems with the battery and how many are just false alerts due to BMS calculation error or something?

Wasn’t meant to be realistic, was meant to illustrate why it’s superior at a glance.

Obsessing over it is a separate issue entirely

Using percent is just sticking your head in the sand. You need to pay attention to notice real problems, and filter out noise.

And there are actually some people own multiple models of cars of different trims.

50% in the X vs 40% in the 3 is a lame comparison.
So is kWh.

Rated miles is obviously the only way it makes sense to compare your car to another, and thus also the best way to compare your car to ITSELF at another point in time (like months weeks or days later or even just hours later).

 

[KenC]

I have Stats. It'll tell me if my estimated range is varying.