Growing efficiency Gap: R2 w/ 21" vs. 20" tire sets

[Thebandit]

With enough samples it doesn't need to be a controlled test set...

-Rich

Only if the samples are random. These are not. The owners are making a choice of what tire to get based on their personal factors. It could easily be that the people doing the most highway (high speed) driving are opting for the 21s..... and their greater average speed is outweighing the greater efficiency of the tire in the results.

This thread is causing significant pain to my research education and experience brain lol

Sponsored

 

[mkhuffman]

With enough samples it doesn't need to be a controlled test set...

-Rich

That's not true.

What if most are the 21s are sold to people up north where it snows, and most of the 20s are sold in Florida?

 

[richpike]

Only if the samples are random. These are not. The owners are making a choice of what tire to get based on their personal factors. It could easily be that the people doing the most highway (high speed) driving are opting for the 21s..... and their greater average speed is outweighing the greater efficiency of the tire in the results.

This thread is causing significant pain to my research education and experience lol

And it could just as easily be the people getting 20s like those wheels better and will drive the tires off them (like I did with our R1S). Your argument that someone buying 21s only because they drive on the highway is arguably more off than an argument that the overall set of data is pretty darn random at this point (**the lack of ability to understand sample size is causing my math through differential equations and multiple statistics classes to feel pained ).

-Rich

 

[Thebandit]

And it could just as easily be the people getting 20s like those wheels better and will drive the tires off them (like I did with our R1S). Your argument that someone buying 21s only because they drive on the highway is arguably more off than an argument that the overall set of data is pretty darn random at this point (**the lack of ability to understand sample size is causing my math through differential equations and multiple statistics classes to feel pained ).

-Rich

Bro.

I hate to be that guy, but your understanding of research statistics is way too limited to be making the claims that you are. You're literally simply just wrong. You're making claims that are completely unsupported and nonsensical. Don't pull the multiple statistics classes thing with me. That's a fight you definitely don't want to get into.

 

[richpike]

That's not true.

What if most are the 21s are sold to people up north where it snows, and most of the 20s are sold in Florida?

Fair point. It's more that because it is getting enough samples it is getting a reasonable test set - the chance of everyone driving the same who got 20s is just as unlikely as everyone driving the same who got 21s.

-Rich

 

[kanefsky]

The only way you can truly prove is to do a same day test between two R2's with different rims/tires sizes on the same roads driving the same speeds and terrain from 100% to 0% to compare results at the end.

Side-by-side test: yes. 100% to 0%: no.

Testing 100% to 0% is more about accounting for things like nonlinear response of the battery percentage display, bottom buffer, etc. I think you could test the efficiency of two different tire/wheel combinations with a much shorter test that wouldn't require the mobile DC fast charger that OOS uses to rescue the dead vehicles on the side of the road. You also wouldn't need to be as careful about things like wind speed, altitude, elevation gain/loss, climate control settings, etc. like the OOS test as long as you did everything the same with both vehicles.

 

[richpike]

Bro.

I hate to be that guy, but your understanding of research statistics is way too limited to be making the claims that you are. You're literally simply just wrong. You're making claims that are completely unsupported and nonsensical. Don't pull the multiple statistics classes thing with me. That's a fight you definitely don't want to get into.

Not here for a dick measuring contest - my point is you're arguing everyone drives the same because they chose 20s or 21s. That is nonsensical. And I'm not the one who pulled out my "research and education" experience to start.

-Rich

 

[mkhuffman]

Fair point. It's more that because it is getting enough samples it is getting a reasonable test set - the chance of everyone driving the same who got 20s is just as unlikely as everyone driving the same who got 21s.

-Rich

You need enough samples in a controlled test to be able to draw conclusions. More and more uncontrolled test data just means you have more and more data you can't draw any conclusions from.

 

[Thebandit]

Not here for a dick measuring contest - my point is you're arguing everyone drives the same because they chose 20s or 21s. That is nonsensical. And I'm not the one who pulled out my "research and education" experience to start.

-Rich

That's not what I'm arguing. I am saying that your claims about sample size and statistics are completely wrong and betray a complete misunderstanding of the fundamentals of statistics and statistical analysis. You are repeatedly saying things that are completely nonsensical and completely unsupported by the available data. Whatever you think you learned from your multiple classes, you either missed the point or didn't actually understand.

 

[richpike]

That's not what I'm arguing. I am saying that your claims about sample size and statistics are completely wrong and betray a complete misunderstanding of the fundamentals of statistics and statistical analysis. You are repeatedly saying things that are completely nonsensical and completely unsupported by the available data. Whatever you think you learned from your multiple classes, you either missed the point or didn't actually understand.

But that's what you said:
Only if the samples are random. These are not. The owners are making a choice of what tire to get based on their personal factors. It could easily be that the people doing the most highway (high speed) driving are opting for the 21s..... and their greater average speed is outweighing the greater efficiency of the tire in the results.

So would you agree the the data will eventually converge?

-Rich

 

[Thebandit]

But that's what you said:
Only if the samples are random. These are not. The owners are making a choice of what tire to get based on their personal factors. It could easily be that the people doing the most highway (high speed) driving are opting for the 21s..... and their greater average speed is outweighing the greater efficiency of the tire in the results.

So would you agree the the data will eventually converge?

-Rich

It's one example of the systemic error that simply waiting for more data won't necessarily fix.

 

[DaveA]

If you ask me, the biggest efficiency hog isn’t wheels and tires…it’s the hvac system. Pre cooling the vehicle for 20 minutes in 95° you can almost watch the percentage go down.

 

[ksurfier]

It's one example of the systemic error that simply waiting for more data won't necessarily fix.

I respectfully disagree. The EPA city/highway split is one of the more interesting pieces of evidence in this debate because it suggests the 21” wheel/tire setup should become more advantageous as speed increases. The published EPA figures are roughly 114 city MPGe versus 109 for the 20”, about a 5% advantage, but 96 versus 89 MPGe on the highway, nearly an 8% advantage. In other words, the EPA data suggest the gap should widen at highway speeds rather than remain fixed.

That makes the real-world results harder to dismiss. If the 21” advantage comes mainly from lower rolling resistance or better aerodynamics, sustained 65–75 mph driving should be where it becomes most obvious. Yet the available owner and independent highway tests have not shown a consistent 21” advantage, while the Rivian Roamer aggregate is currently showing the opposite, with the 20” averaging roughly 8–9% better efficiency.

There is also an interesting inconsistency within the EPA data itself. The underlying certification-cycle results appear to put the two configurations only a few percent apart, yet the published range ratings are 330 miles for the 21” versus about 307 miles for the 20” — roughly a 7.5% difference. That raises the possibility that the certification and adjustment process is magnifying what began as a relatively modest measured efficiency difference.

So the overall pattern is unusual: the raw EPA data show only a small 21” advantage, the EPA highway numbers suggest that advantage should increase with speed, and the published range figures make the gap larger still. Yet the emerging real-world data do not show the expected highway advantage and, in aggregate, currently point in the opposite direction. That does not prove the 20” is inherently more efficient, but it does make a controlled same-vehicle 20” versus 21” highway test increasingly important.

 

[VegasWeezy]

If you ask me, the biggest efficiency hog isn’t wheels and tires…it’s the hvac system. Pre cooling the vehicle for 20 minutes in 95° you can almost watch the percentage go down.

I am in the camp that we do not have a) a controlled set of test data to draw conclusions from; b) the sample size is STILL way too small to draw any meaningful conclusions from; and c) almost all of these vehicles being used for this comparison are still within their break in phase.

On the HVAC point specifically, and on some of the others in general... I have an R2 on the 21s and my efficiency is atrocious (currently 1.82 mi/kWh). There are several reasons for this including 1) there are less than 500 miles on the vehicle so the tires are not broken in yet; 2) I live in Las Vegas where it has routinely been 110+ degrees and I have been running the sh*t out of my HVAC (including the ventilated seats); 3) most of my drives have been for less than 5 miles; 4) I live on the edge of the valley, so all of my drives typically have some steep elevation changes as I drive down into the valley and then back up; and 5) I have been having a lot of fun launching from a dead stop or otherwise testing the acceleration capabilities.

The point being, in my driving alone, there are too many variables, and my particular use case is not likely representative of the other 30 R2s on 21s.

 

[Crab]

Good thing I am not in Rivian Roamer with my 20s because I have been flooring the car every time I drive. Its just too damn fun

Sponsored