Model S First Drive Reviews

[spatterso911]

My rookie understanding from reading previous posts here is that they simulate rollout (ie. just lower the numbers in order to get better marketing $$ from BMW based on questionable extrapolations).

Well, 3.7 seconds is the reported time for a well-executed launch control launch. This is not as easy as it sounds, however, and I don't typically place much stock in launch control launches. I prefer to use the rolling launch, or an off the line launch without LC. Rolling launch is better so as to control for wheel spin/traction control influences at takeoff.

I'm very impressed with how much Tesla got right out of the gate. I suspect that with a set of Michelin Pilot super sport tires, the slalom and skid pad number would be better. Braking is tops in class, despite the weight disadvantage.

[strider]

Can't do launch control and a rollout simultaneously. The published 0-60 time with LC launch is 3.7

Why not? You initiate an LC launch and then don't start the clock until the car hits 3mph or whatever fudge factor they want to use. I prefer 5-60 times which DO eliminate LC and are more real-world (LC launches are HARD on ICE's and will roast clutches and void warranties if you do it often). I haven't seen MSP 5-60 but IIRC the new M5 was 4.4 5-60 so MSP should beat it.

[CapitalistOppressor]

Can't do launch control and a rollout simultaneously. The published 0-60 time with LC launch is 3.7

That's just not true. Technically a "rollout" happens because of how staging works at a drag race. The timing tree at a drag strip has 2 light sensors, a prestage warning sensor and the timing sensor (which activate lights on the tree).

The prestage sensor and the timing sensor are next to each other and the intent is for drivers to place their front wheels on the timing beam (to block it from the sensor) which is the starting line (The prestage warning is right before the staging sensor to alert the driver that he is close to staging). A tire is round and compressed at the bottom, and the shallowest possible stage occurs the instant when some portion of the lower front edge of the tire (the sensor is just above ground level, and the higher it is the longer the potential rollout) blocks the timing beam.

The timer does not start until your front tire rolls far enough to unblock the sensor, so an aggressive shallow stage with 12"-16" of rolling distance can cut a lot of time off your ETA. A decent car will be going 3 or 4 mph already before timing even begins (though this varies a lot in reality) so any review which includes a "rollout" is basically ignoring the first 0.3-0.5 seconds of the run where your tires might be spinning smoke and not fully hooked up yet.

In a match race it doesn't matter because you still have to beat the other car to the finish line and a shallow stage just makes your reaction time look worse while making your E.T. look better. But E.T. is important by itself for qualifying and securing sponsorships (and many other things) so how you stage, and what kind of rollout you get is important. On the other side of the coin, drivers who want to make themselves look good can deep stage (push a bit past the starting line and reduce rollout) so they can reduce their reaction times.

For 0-60 acceleration numbers, it's totally meaningless. It still represents real useful data if a magazine is comparing two vehicles using the same methodology, but if two different magazines using different methodologies print acceleration times, you can't really compare them. Fundamentally, the only "real" time is an actual measured time from the moment the car starts accelerating.

Reviewers don't even use actual timing lights anymore. They just use VBox or something like it that is a super accurate GPS and accelerometer. Vbox will list your true 0-60 time (and quarter mile time) as well as compute with a rollout adjusted E.T. for any rollout up to 16". That way they can do their test on a city street if they want and get accurate data on the acceleration characteristics of the car.

VBox includes the rollout data because it's designed for NHRA racers who want to test their vehicle without an expensive timing light rig or even a real drag strip. Just find a flat stretch of road with no cops around, turn on your VBox and punch it. It will precisely measure your acceleration and distance traveled and break out your 1/4 mile (or any other distance) without the need for a finish line timer. It will also provide adjusted E.T. with various rollout scenarios, trap speeds and a ton of other data that you can use to tune the car.

As far as launch control goes, it doesn't affect your rollout. Launch control is just a way to configure the car for maximum possible acceleration once you launch. The rollout happens after you launch regardless of whether you are using fully engaged launch control for a max speed launch or if you just slightly depress the gas pedal and drive to the finish line at 5mph (in which case it will take quite awhile to move 12" compared to a max launch).

Launch control is very useful in a drag race, but is itself quite deceptive in terms of measuring a car's real world acceleration. The 2013 M5 is reported by Car and Driver as going 0-60 in 3.7 seconds (which is with DCT and their fudged rollout) but the rolling acceleration (meaning you are already moving when you press the gas pedal) for 5-60 is 4.6 seconds. Even considering that the "true" 0-60 for DCT is closer to 4.1 seconds, it just goes to show how much launch control matters when accelerating from a stop.

In real life, like accelerating after turning onto a freeway on-ramp, launch control is meaningless and it's the rolling acceleration numbers that matter. I'm also less convinced now that MSP will significantly outperform M5 in rolling acceleration after looking at the Roadster data from 2009, but I don't understand why yet so I'm just waiting on the data. I'm mostly surprised by how close to M5 it is in 0-60 as it is, so I expect the rolling times to be competitive regardless.

[William13]

Tesla should advertise "real world" 0-60mph in 4.4 seconds. After learning how drag races and times are "calculated," I think that Tesla could make the case that their car is faster in real world driving acceleration. It also is democratic allowing an untrained driver like me to simply floor the car and shoot off. The smoothness also makes the acceleration feel safer than I feel using launch control in my brother's GTR.

[jcstp]

Tesla should advertise "real world" 0-60mph in 4.4 seconds. After learning how drag races and times are "calculated," I think that Tesla could make the case that their car is faster in real world driving acceleration. It also is democratic allowing an untrained driver like me to simply floor the car and shoot off. The smoothness also makes the acceleration feel safer than I feel using launch control in my brother's GTR.

so as a quote, "even your grandmother can accelerate to 60 in in 4.4 seconds, as long as she can reach the pedal"

[spatterso911]

That's just not true.

It's absolutely true. CO, in order to execute an LC launch on the M5, you must do the following steps...
- DSC completely off
- Shift setting at most aggressive manual setting (there's a higher shift setting only available with DSC off on the M5)
- Bring car to a stop
- Hold shifter in the forward (downshift) position until you see a checkered flag in the cluster
- While continuing to hold shifter, floor the throttle
- Adjust launch revs if desired using the cruise control stalk
- Release the shifter to go. The car will handle shifting (despite being in manual mode) as long as you keep the accelerator floored.

The car must be stopped, and you cannot be in motion when initiating the launch control, because the first clutch is not engaged. The computer is setting the engine revs to optimize the launch to avoid wheel spin, then optimize the shift points for you so that you just floor it and go. To know when you hit the rollout distance, you must make a very good guess, which is very subjective. Why bother??

You are absolutely correct in that the "rollout" is faked by the reviewers as an estimation, but it is useless, because there is no true rollout, no timing involved. If you never mentioned rollout, the time is the same. 3.7 seconds according to C&D. It's been repeated as well, but as I have said before, LC is really an inappropriate way to determine the real-world acceleration of the car, given that 99% of people will not engage LC ever in the car's lifetime. It IS the most accurate way to determine what the car is capable of, however, and I do believe the car can punch 60 mph in 3.7 seconds or less given perfect launch and optimal weather, road and tire conditions.

[ManuVince]

No that's normal. Here are the tires for 2013 BMW M5 -

Pilot Super Sport
Front - 265/35ZR-20 99Y
Rear - 295/30ZR-20 101Y

2012 Cadillac CTS-V

Pilot Sport
Front - 255/40ZR-19 96Y
Rear - 285/35ZR-19 99Y

Edit: Hmm.. writing it down next to the numbers for MSP I notice that MSP has narrower tires than M5 or CTS-V. I think thats what the first number is. Width in millimeters. Might that explain the underperformance in turns compared to M5 and CTS-V? Any tire experts out there?

Certainly would seem to me to reduce rolling resistance by having them be more narrow, which increases range. As is, MSP is decent as opposed to exceptional in handling which might make for a good tradeoff to get more range.

Edit2: And by decent, I mean compared to the best in the world in this class. Outstanding compared to normal sport sedans

First number is the tire radius in milimeter.
Second is tire wall size in milimeter.
Third number is width in inches.

So the cadillac and M5 have bigger wheels but narrower.

[Adm]

Sorry Manuvince,

First number is tire width in mm
Second number is tire wall height in % of tire width
Third number is rim diameter in inches

I believe the rest is max speed and max capacity weight

[raymond]

I believe the rest is max speed and max capacity weight

Yup, but the other way around: Wikipedia Tire code

[strider]

The car must be stopped, and you cannot be in motion when initiating the launch control, because the first clutch is not engaged. The computer is setting the engine revs to optimize the launch to avoid wheel spin, then optimize the shift points for you so that you just floor it and go. To know when you hit the rollout distance, you must make a very good guess, which is very subjective. Why bother??

Did you read CO's description of how roll-out works? You creep up to the start beam, COME TO A COMPLETE STOP, and the go from there. The clock starts when you pass out of the beam. How is that incompatible w/ LC?

You are absolutely correct in that the "rollout" is faked by the reviewers as an estimation, but it is useless, because there is no true rollout, no timing involved. If you never mentioned rollout, the time is the same. 3.7 seconds according to C&D. It's been repeated as well, but as I have said before, LC is really an inappropriate way to determine the real-world acceleration of the car, given that 99% of people will not engage LC ever in the car's lifetime. It IS the most accurate way to determine what the car is capable of, however, and I do believe the car can punch 60 mph in 3.7 seconds or less given perfect launch and optimal weather, road and tire conditions.

So show me that test - a true standing-start 0mph to 60mph. C&D uses roll-out so their's is a bogus number. From C&D's forums: Acceleration Testing Methodology? - Car and Driver Backfires

BMW says 4.2 or 4.3 depending on xmission.