Acceleration times - Fiction and Truth

MacGuffin

Forum Addict
Joined
Mar 29, 2008
Messages
8,329
Location
Wilhelmshaven, Germany
Car(s)
'17 Ford Mustang GT Fastback
Leading German car magazine "Auto, Motor & Sport" asked the question, how close to reality and honest the carmakers are about the acceleration figures of their cars from 0-100 km/h (0-62 mph for our imperial friends).

Obviously this acceleration figure (as unimportant as it may be in daily traffic) is one of the key factors by which the performance of a car is valued -- especially by those who don't own it :p

But under which circumstances are those numbers being determined? Well, AMS says, that to get better numbers, the choice of a lightweight driver, empty tank, especially grippy tires and road surfaces are popular methods to pimp up the numbers. But how big is the difference between lab conditions and real life conditions?

AMS does tests close to real life conditions. They are using racing tracks, Hockenheim for example, or the runway of an airport. All cars have a full tank and two persons inside. Measurements are taken in both directions and the average is taken.

On their homepage, AMS lists 50 cars from previous reviews and compares the acceleration numbers from 0-100 km/h the manufacturers tell, with real life conditions. For the test runs all electronic helpers are off and measurements were made by GPS. For cars with a manual, the optimal number of revs for a quick start were determined at first.

Of course this is nothing new but it's nice to have a list in print for comparison.

Here's the results. They are sorted by how much (in percent) the numbers differ. Smallest difference up, biggest difference down.

Audi Q5 3.0 TDI:
Audi says 7.2 sec.
Reality says 8 sec. (That's 11.1 % difference)

Corvette ZR1:
Chevrolet says 3.6 sec.
Reality says 4 sec.

Citroen C3 VTi 95:
Citroen says 10.6 sec.
Reality says 11.8 sec.

Audi A5 Cabriolet 3.2 FSI quattro:
Audi says 6.9 sec.
Reality says 7.7 sec..

Mazda 6 1.8 MZR:
Mazda says 10.2 sec.
Reality says 11.4 sec.

Renault Clio 1.6 16V 130:
Renault says 9.3 sec.
Reality says 10.4 sec.

Opel Insignia Sports Tourer 2.0 CDTi 4x4:
Opel says 10.1 sec.
Reality says 11.3 sec.

Cadillac CTS-V:

Cadillac says 4.2 sec.
Reality says 4.7 sec.

Ford Kuga 2.0 TDCi 4x4:
Ford says 9.9 sec.
Reality says 11.1 sec.

Maserati GranTurismo S:
Maserati says 4.9 sec.
Reality says 5.5 sec.

VW Touran 1.2 TSI BlueMotion:
VW says 10.6 sec.
Reality says 11.9 sec.

VW Caddy 1.6 TDI 5-Seater:
VW says 13 sec.
Reality says 14.6 sec.

Audi A4 allroad quattro 3.0 TDI:

Audi says 6.4 sec.
Reality says 7.2 sec.

Jaguar XF 3.0 V6 Diesel S:
Jaaaag says 6.4 sec.
Reality says 7.2 sec.

Fiat Punto Evo 1.4 16V MultiAir:
Fiat says 10.8 sec..
Reality says 12.2 sec.

Mitsubishi Lancer Evolution:

Mitsubishi says 5.4 sec.
Reality says 6.1 sec.

BMW 120i:
BMW says 7.7 sec.
Reality says 8.7 sec.

Alfa Romeo Gulietta 1.8 TBi 16V:
Alfa says 6.8 sec.
Reality says 7.7 sec.

Land Rover Defender 110 TD4 Station Wagon:
Land Rover says 15.8 sec.
Reality says 17.9 sec.

Fiat 500 1.4 16V:
Fiat says 10.5 sec.
Reality says 11.9 sec.

VW Tiguan 1.4 TSi 4Motion:
VW says 9.6 sec.
Reality says 10.9 sec.

VW Golf Plus 1.2 TSI BlueMotion:
VW says 10.2 sec.
Reality says 11.6 sec.

Aston Martin DBS Volante:
Aston says 4.3 sec.
Reality says 4.9 sec.

VW Polo 1.2:
VW says 16.1 sec.
Reality says 18.4 sec.

Nissan GT-R Black Edition:
Nissan says 3.5 sec.
Reality says 4.0 sec.

Corvette Cabrio:
Chevrolet says 4.1 sec.
Reality says 4.7 sec.

Chevrolet Spark 1.2
Daewoo says 12.1 sec.
Reality says 13.9 sec.

Corvette Z06:
Chevy says 3.9 sec.
Reality says 4.5 sec.

Toyota Prius:
Toyota says 10.4 sec.
Reality says 12 sec.

Fiat Dobl? 1.6 16V Multijet:
Fiat says 13.4 sec.
Reality says 15.5 sec.

Volvo S60 2.0T DSG:
Volvo says 8.2 sec.
Reality says 9.5 sec.

Mitsubishi ASX 1.8 DI-D 4WD:
Mitsubishi says 10 sec.
Reality says 11.6 sec.

BMW 118i:
BMW says 8.7 sec.
Reality says 10.1 sec.

Lexus LFA:
Toyota says 3.7 sec.
Reality says 4.3 sec.

Fiat 500C 1.2 8V:
Fiat says 12.9 sec.
Reality says 15 sec.

Hyundai i20 1.2:
Hyundai says 12.9 sec.
Reality says 15 sec.

Mercedes E350 BlueTec:

Mercedes says 7.3 sec.
Reality says 8.5 sec.

BMW Z4 sDrive23i:
BMW says 6.6 sec.
Reality says 7.7 sec.

Mercedes E220 CDI BlueEfficiency:
Merc says 8 sec.
Reality says 9.4 sec.

Smart fortwo Cabrio mhd:
Smart says 13.7 sec.
Reality says 16.1 sec.

Fiat 500 1.2 8V Start Stop:
Fiat says 12.9 sec.
Reality says 15.3 sec.

Porsche Cayman:
Porsche says 5.8 sec.
Reality says 6.9 sec.

Citroen C3 VTi 120:
Citroen says 8.9 sec.
Reality says 10.6 sec.

Lotus Elise:
Lotus says 6.5 sec.
Reality says 7.8 sec. (that's a 20 % difference, btw.)

VW Sharan 2.0 TDI BlueMotion:
VW says 10.9 sec.
Reality says 13.1 sec.

KTM X-Bow:

KTM says 3.9 sec.
Reality says 4.7 sec.

Renault Laguna Coup? V6 dCi 235 FAP:
Renault says 7.3 sec.
Reality says 8.8 sec.

BMW 320i:
BMW says 8.2 sec.
Reality says 10 sec.

Mitsubishi Pajero 3.2 DI-D:
Mitsubishi says 11.1 sec.
Reality says 14 sec.

Daihatsu Cuore 1.0:
Daihatsu says 11.1 sec.
Reality says 14.5 sec. (That's 30.6 % difference)

Source: http://www.auto-motor-und-sport.de/news/sprint-schummler-werksangabe-vs-messwert-1921038.html
 
Last edited:
The one car that comes to mind immediately is the Corvette - many owners have gone faster in their showroom-stock 'Vettes than Chevy claims the car can do.

That said, I'm not one bit surprised that manufacturers try to cheat the numbers a little where they can. It's just marketing.
 
These numbers are pretty pointless with a passenger and a full tank of fuel. I'd like to have known how far the manufacturer's number is off what you can achieve right out of the showroom.
It's quite obvious that loading the car up will kill the acceleration figure, and it's no big surprise that the manufacturer doesn't factor in a passenger.
 
Though I think this is an interesting test, one thing that irks me about it is why two passengers? And why an absolutely topped up full gas tank? I think a more realistic test would be 1/2 or 3/4 of a tank with one passenger and maybe a computer bag.

This test (IMO) really punishes the cars like the Elise and really favours the cars with big torquey motors like the 'vette and the cadillac.

Edit: Beaten by narf, but he sums my point almost perfectly.
 
Last edited:
These numbers are pretty pointless with a passenger and a full tank of fuel. I'd like to have known how far the manufacturer's number is off what you can achieve right out of the showroom.
It's quite obvious that loading the car up will kill the acceleration figure, and it's no big surprise that the manufacturer doesn't factor in a passenger.

It only seems pointless but those are actually the "Top 50" with the biggest difference in acceleration numbers. Obviously other cars showed smaller differences. The keyword here is comparableness and you cannot compare, when one carmaker gives numbers, that are only 5 % off and the other carmaker gives numbers, that are 20 % off.

The AMS approach is good for having a real comparableness. For that the conditions have to be the same.

And about the second person, well... if it's ME in the car, there is no need for a second person *lol* And guess what? With one exception I was never able to match the data sheed numbers in reality, too.
 
Last edited:
You can not deduce incomparableness of the manufacturer test data just from the fact that this test yields different percentage differences for each car.

Let's assume the only difference between two tests were one additional occupant and a full tank. The absolute weight gained would be pretty similar, bar differences in tank sizes.
Would the percentage difference between the two tests be constant for every car? No.
Take the Q5 from the top of your list, it is quite a heavy car. Adding one occupant and some fuel slows it down, but not by much. 1955kg + 150kg is a small change, +8%.
Take the Cuore from the bottom of your list, it is quite a light car. 750kg + 120kg (not 150, the Cuore has about half the tank capacity) is a huge change, +16%.
The weight-related percentage difference in acceleration should in theory be twice as high for the Cuore. In the test above the percentage difference is almost triple, most of that can be explained with this weight difference.
 
Last edited:
Though I think this is an interesting test, one thing that irks me about it is why two passengers? And why an absolutely topped up full gas tank? I think a more realistic test would be 1/2 or 3/4 of a tank with one passenger and maybe a computer bag.

Edit: Beaten by narf, but he sums my point almost perfectly.

Fair point, well made. Even with my car there is quite a difference - 19 gallon fuel tank and one passenger is a lot different from 9 gallons and just a driver.
 
Additionally, there is no reason whatsoever to bring that extra weight. They say they want comparable results under the same conditions. Taking that extra driver or fuel does not affect that, you could just as well have comparable conditions with the standardized 75kg driver (add lead if necessary) and some level of petrol in the tank.

Why did they add that weight, you might ask? To let readers see "OMG manufacturers are ripping us off!" and make them click through a 50-page photo series instead of just giving them the table. Cha-ching.
 
You can not deduce incomparableness of the manufacturer test data just from the fact that this test yields different percentage differences for each car.

Let's assume the only difference between two tests were one additional occupant and a full tank. The absolute weight gained would be pretty similar, bar differences in tank sizes.
Would the percentage difference between the two tests be constant for every car? No.
Take the Q5 from the top of your list, it is quite a heavy car. Adding one occupant and some fuel slows it down, but not by much. 1955kg + 150kg is a small change, +8%.
Take the Cuore from the bottom of your list, it is quite a light car. 750kg + 120kg (not 150, the Cuore has about half the tank capacity) is a huge change, +16%.
The weight-related percentage difference in acceleration should in theory be twice as high for the Cuore. In the test above the percentage difference is almost triple, most of that can be explained with this weight difference.

If your assumption would be right, then the bottom of the list should be dominated by small, light and underpowered cars.

But it isn't. The runner-up is a heavy SUV.

Obviously makers of other small, underpowered cars do not have a 30 % difference between data sheet numbers and real world test numbers. Otherwise they would be there, too, to accompany the Cuore.

That's what I mean with comparableness.

Additionally, there is no reason whatsoever to bring that extra weight. They say they want comparable results under the same conditions. Taking that extra driver or fuel does not affect that, you could just as well have comparable conditions with the standardized 75kg driver (add lead if necessary) and some level of petrol in the tank.

There is no standardized driver like that (what would be, if he gained 2 kilos? Would he be fired?), as there are no regulations or laws. So everyone is free to come up with their own conditions. AMS decided to put two people in the car and make the measurements with a full tank but it doesn't really matter.

Yet I think it's only fair to measure such numbers under average conditions. Could also be one person and an empty tank or four persons, a full tank and a fully loaded trunk. So two persons with full tank is actually an average driving situation. And it is a very good source of data, because it has been done by AMS that way for decades.

But nevermind that. The important part is, that there are some carmakers, who see it the same way and measure in average situations and other carmakers measure under ideal conditions. And as long as that is the case, any kind of independent, standardized measurement is good, because it gives the potential customer the ability to compare.

Saying that AMS is only doing it to draw more readers, is a bit simple. Who else but the car press is able to make such comparisons? I think the truth is more like some carmakers pimp up their cars' acceleration figures to gain a marketing advantage.
 
Last edited:
I've never given it much thought, to be honest. I've always seen those numbers as an indication of what the cars are capable of, not what your typical average acceleration will be. It's kinda like on TopGear with their power lap board. It's an indication of what those cars are capable of, with an expert driver behind the wheel, not what you personally can do with it. If I really gave a shit about squeezing an extra few tenths of a second off my 0-62 time, I would not invite extra passengers and cargo along for the ride.
 
I'd like to see average and peak accelerative force become more common.
 
Why would someone test the Prius for its acceleration times ? I thought that would show up on the 'mileage - truth vs fiction' thread.
 
If your assumption would be right, then the bottom of the list should be dominated by small, light and underpowered cars.

But it isn't. The runner-up is a heavy SUV.

Obviously makers of other small, underpowered cars do not have a 30 % difference between data sheet numbers and real world test numbers. Otherwise they would be there, too, to accompany the Cuore.

That's what I mean with comparableness.

You're thinking of the conclusion "some cars' original measurements are not comparable in their acceleration ==> you have to put two people and a full tank into the car to make them comparable". I was arguing that the other direction, "two people and a full tank yield different results ==> original measurements must be incomparable", is not a logical conclusion - whether it's true or not is another matter.

There is no standardized driver like that (what would be, if he gained 2 kilos? Would he be fired?), as there are no regulations or laws. So everyone is free to come up with their own conditions. AMS decided to put two people in the car and make the measurements with a full tank but it doesn't really matter.

There is such a standardized driver already, he's used for determining the curb weight of cars. See DIN 70020, full tank + 75kg driver + all tools and fluids.

Yet I think it's only fair to measure such numbers under average conditions. Could also be one person and an empty tank or four persons, a full tank and a fully loaded trunk. So two persons with full tank is actually an average driving situation. And it is a very good source of data, because it has been done by AMS that way for decades.

Acceleration under average conditions are a different matter though. The manufacturer numbers are lowest possible acceleration times. Comparing average with lowest is even more pointless.

Saying that AMS is only doing it to draw more readers, is a bit simple. Who else but the car press is able to make such comparisons? I think the truth is more like some carmakers pimp up their cars' acceleration figures to gain a marketing advantage.

I'm not saying they made the tests to outrage readers. I'm saying they designed the test parameters to stir up more outrage than necessary. See above, if you compare average figures with lowest figures you are bound to get larger differences.
A bit like TomCat said, who would bring passengers on a drag run and expect the lowest possible number?
 
This test is just another one that proves 0-60mph/0-100km/h tests are totally irrelevant. Maybe these times with AWD automatic cars are useful, because the time depends only from power and transmission. But manual FWD/RWD car? Traction between current tires and current asphalt with current tire pressure makes a huge difference, as does the skill (or if he bothers to do it) of the driver to find the optimal revs to launch the car and then change just as fast as the gearbox allows. There are too much factors to make these "tests" totally irrelevant.

And all this is only after the weight thing narf very well pointed out.
 
Best way to go is to get one of those little people they put on horses during races as a driver and have no fuel in the tank.....


What?
 
Not exactly news.

The 0-60 (0-100km/h) thing only started when various factors dictated car makers could no longer brag about outright top speed in their marketing.

They massage fuel economy figures in ways to suit themselves, we can hardly expect them to do any different with acceleration.

The thing to remember is that your new car will not be as fast off the line or as economical as the brochure says, nor will it get you anywhere near as much sex as the TV commercials imply.
 
nothing new and I guess there are just too many variables to take care of, so I would give those times a decent berth anyway.

I bet manufacturers pop in enough fuel for about a mile of running, one lanky streak of piss for a driver and tell him to forget the clutch and just slam the gears in. Then theres what rubber you have, how rough the tarmac is, what the ambient conditions are like. Just too many to get right or replicate.
 
Why did they add that weight, you might ask? To let readers see "OMG manufacturers are ripping us off!" and make them click through a 50-page photo series instead of just giving them the table. Cha-ching.

Hey LeVeL, can I borrow you for a minute?

I'm not one bit surprised that manufacturers magazines try to cheat the numbers a little where they can. It's just marketing.

Thanks. :D
 
Hey LeVeL, can I borrow you for a minute?

Indeed. You could even say the magazine is as bad as the manufacturer, one making sure the numbers are very low while the other is making sure the numbers are very high. Both for the same basic reason.
 
Top