A New Performance Paradigm: The ‘Speed Per Dollar’ Index

Don May 31st, 2007 4:10 PM Link

I think I just had a seizure.

chuck goolsbee May 31st, 2007 5:03 PM Link

Wow… Looks at the complete fusterckluck between the HP & weight columns!

Vehicle weight has gotten a bit out of control of late.

Good job guys, I like the metric.

–chuck

Phillip May 31st, 2007 5:11 PM Link

The SpD favors used cars. I bought a 96 Corvette for $12,500 last year that has 330hp and weighs 3350lbs. Using your calculation it has an SpD of 7880. That means its over twice as much fun as a new Mustang GT and 18 times more fun than a Saleen S7 TT.

Cognoscente May 31st, 2007 5:21 PM Link

Very interesting way of looking at things. My new WRX wagon works out well using that calculation.

I’ve linked to this page on my site:
AutoConcourse

ryan May 31st, 2007 5:22 PM Link

of COURSE it favors used cars. just because something’s fun the first time around doesn’t mean it’s not fun the second, or third time around. but comparing apples to apples, it does give a nice little comparo between all new cars, or gives you an idea as to “bang for the buck” of preowned. i’d like to see how a preowned mitsu evo compares to a preowned ‘vette anyday, as some cars can be steals in the preowned market while others are craptastic depending on market value.

this formula is what it is, and that’s a benchmark. if you can’t handle the turbo spool of a mazdaspeed3 and instead prefer a N/A application, like a mustang GT, then you can omit the turbo applications and focus on N/A, or vice versa. it will all come down to subjectivity in the long run anyway, but for finding the best bang for the buck, i think it works.

now to see what my 2003 focus with mods compares too….

ryan May 31st, 2007 5:36 PM Link

4650 is the answer. but if i turbocharge it, i’m in the 6k range! wahoo!!!

makes one think of what can be done with a preowned car, actually. like maybe a preowned SRT-4 after mopar stage 2 put in…

i’d like to see the variable of the slalom put in. that’ll give you a decent factor of the handling of the vehicle, and it’s something that can be (somewhat) closely replicated.

Jawn-son May 31st, 2007 5:44 PM Link

some how work in handling G’S and braking, then it would really show

the best of a cars capabilities, with out actually driving one

kyle May 31st, 2007 6:01 PM Link

i think maybe including a cars lateral G rating might help (just muliply everything by the lateral G rating). as it is right now, a new Honda civic SI makes 3424 points. However, multiply in its .85 lateral G and it drops to 2910.

Likewise, take the Z06 and multiply it by 1.02 lateral G and you get a couple extra points, not much, but still some.

Jeff May 31st, 2007 6:09 PM Link

My Metro works out to 3437.5. Should bump to over 4600 with my new engine.

55hp, 2000lbs, $8000 bought new in ‘95.

Mark Arnold May 31st, 2007 6:23 PM Link

Yeah, this really only works for new cars I suppose…

I did a couple of my former rides:
‘87 BMW 325is (168hp ÷ 2800lbs) x 10,000 ÷ $1200 x 100,000 = 50,000SpD

‘89 BMW 325is (168hp ÷ 2800lbs) x 10,000 ÷ $3600 x 100,000 = 16,666SpD

…am I winning?

Strada Auto Store May 31st, 2007 6:32 PM Link

Innovative way of looking at “go fast value”.

Thanks

OT:�Speed Per Dollar� Index - Benzworld.org - Mercedes Benz Discussion Forum May 31st, 2007 6:42 PM Link

[…] OT:�Speed Per Dollar� Index Any thoughts? I came up with about 2020 for the SLK32 using a rough weight of 3200lbs and an original cost of about 54K (couldn’t remember what it was). Winding Road � Archive � A New Performance Paradigm: The ‘Speed Per Dollar’ Index __________________ “Any society that would give up a little liberty to gain a little safety will deserve neither and lose both.” Benjamin Franklin […]

smokyburnout May 31st, 2007 11:21 PM Link

mark- blah, though i won…
95 jeep cherokee
190 bhp
3000 lbs
$1500
42,222 spd

Mena May 31st, 2007 11:43 PM Link

Mine comes out to 3095. It’s an 07.

Matt June 1st, 2007 12:20 AM Link

smart fortwo: (40 / 1609) x 10000 / $16500 x 100000 = 1507 spd (CDN $)

mini cooper: (115 / 2524) x 10000 / $23,500 x 100000 = 1938 spd (CDN)

Even using the Canadian $$$ which is way higher than the US, I’m doing better than the Porsche breed of cars.

Ken June 1st, 2007 7:17 AM Link

These numbers are mostly approximations:
06 MINI Cooper S: (168/2700)10000/27000 X 100000=2305SpD
The car the MINI replaced:
91 Plymouth Laser: (91/2800)10000/100 X 100000= 325000SpD
I don’t think the equation works at all for used cars. I’d much rather have my new car than my $100 Plymouth with less than 100hp.

Rick W June 1st, 2007 8:53 AM Link

This is a similar to the considerations I make when shopping for a car. I think it is even more relevant if you simply plug in the 5 or so cars in your budget range when you are ready to buy than to say “my mazdaspeed has a better SpD than a Z06″. It would be perfect if handling and braking were to be worked into the equation. There is also a lot of intangibles that of course can never be accounted for in such an equation.

Mark June 1st, 2007 9:55 AM Link

Using NEW car prices, my 2007 G35x would be:

306hp / 3703lb x 10000 / $33950 x 100000 = 2434SpD

Not bad, just above a Corvette Z06!

tweezy June 1st, 2007 11:07 AM Link

I did this for my 2004 Acura RSX Type-S which I bought used. I think you should be using the the MSRP for the car if it were new, not the price I actually paid.

Using the new price I get 2934.3SpD, just above the Z06
Using the price I paid used I get 3912.4Spd just below the Ariel Atom.

Does that mean this car is getting better with age? LOL.Seriously, the 2934.3SpD is a bit more realistic I’d think in this case.

Steve L June 1st, 2007 11:40 AM Link

Naturally, some jerk is going to bring up motorcycles. I guess I’ll do it.

2006 Honda CBR600RR
HP = 105.6, wet weight = 431 pounds, 2006 retail price $8999

The SPD index comes out to 27,227.

You might say that there’s no fair comparison because these cars can be daily drivers and keep you dry in the rain. As you know, those arguments aren’t true of the Ariel Atom. So, what does this metric tell us? Is the CBR600RR really 7-times “better” than an Atom? What does better mean? If I had the money, I’d go with the Atom, but they’re both equally impractical (although motorcyclists would argue that the 600 could be a daily ride). It’s an interesting metric, but who is going to choose the Mazdaspeed 3 over the Elise, Z06, or CCX, given the chance? The one thing this does is allow the Mazdaspeed buyer (that’s me) to brag that, for the money, I got one of the best deals going. Of course, by the time I’ve explained that to the person next to me in the Evo he’s a quarter-mile down the road (I really have to work on my reaction times).

adam June 1st, 2007 11:48 AM Link

My ‘03 Civic Si is pretty high up there.

I paid $15,900 for it back in ‘03 (no one wanted the hatch in south Georgia). It has 160 hp and weighs 2800 lbs.

3593 SpD

ryan June 1st, 2007 12:08 PM Link

lol, those are some awesome numbers people are coming up with.

but i think we have to look at the inflation rate if you start using preowned vehicles, as well as the MSRP. i know, i know, nobody pays MSRP on a preowned vehicle, but at least you can tell the bang for the buck initially. i think, though, if you’re looking between a 03 VW GTI and a preowned mustang, you’ll get a nice little number difference between them, almost validating your decision.

i’d tell those guys to hit the drawing board again. multiply the skidpad by 10000, making it a nice factor in the equation, and then see what comes up.

but by now, we’re actually getting to the crux of cars, and that’s something a bit more intangible, the FEELING.

Steve June 1st, 2007 12:36 PM Link

Interesting to run it with Torque, too.

I like it.

Nathan Russell June 1st, 2007 12:39 PM Link

I agree with Kyle: factor in lateral g’s. It’d be a clean and simple way to add a metric for those of us who prefer the twisties over 1/4 mile straights.

Sunil June 1st, 2007 12:49 PM Link

A fun, but not well-conceived idea. The issue with used cars highlights the main problem: that in this equation price has a disproportionately large effect on the final number. This is clear because the
“Price Point” and “Speed per Dollar” columns are nearly the same. The “Speed per Dollar” value does not take into account that there are diminshing returns to capital.

I think this would work if you used some sort of nonlinear, perhaps logarithmic scale for price. To underscore this point, someone on VWVortex pointed out that a Hyundai Accent GS crushes all comers with a “Speed per Dollar” value of 4281.

Johnny Rocket June 1st, 2007 5:36 PM Link

We all know the used car factor is no good with the formula. The new car factor works, though. The top picks are obvious. We all know the Ariel Atom is speed monster, the Corvette Z06 is a bargain bruiser, and the MazdaSpeed3 is the perfect car for a miniscule budget.

Aaorn June 1st, 2007 6:25 PM Link

Interesting idea, but I have an improved method that weights real world performance. Lap times are what indicates performance…HP and weight are merely ways to predict performance. For example, if you compared an AWD car with a FWD car of equal HP the AWD car would receive a penalty for weight and cost and have an equal HP score. The AWD car would be able to out accelerate the FWD car in low traction situations, making it faster around a track or down a strip. What about brakes, suspension, and tires? The list goes on and on. Of course, it’s hard to get track data so I see your dilema.

Winding Road invents “Speed per Dollar” index | Auto-Central.org June 1st, 2007 6:37 PM Link

[…] Read | Permalink | Email this | Comments […]

Automotive Articles Magazine » Just another WordPress weblog » Blog Archive » Winding Road invents “Speed per Dollar” index June 1st, 2007 6:50 PM Link

[…] Read | Permalink | Email this | Comments […]

eAuto Mall » Winding Road invents “Speed per Dollar” index June 1st, 2007 7:01 PM Link

[…] Read | Permalink | Email this | Comments […]

io June 1st, 2007 7:33 PM Link

The cheapest compact sport is arguably the 2007 Honda Civic Si. The sedan gets 3070 with navi, 3141 without. That puts it at fifth place.

The Coupe gets 3237, also at fifth.

The Scion tC gets 3234.

"Speed per Dollar" index - DTM Power June 2nd, 2007 12:57 AM Link

[…] "Speed per Dollar" index It occurred to me that enthusiasts could use a new equation to better quantify a vehicle�s capabilities in view of its cost�something like horsepower-per-pound-per-dollar, a metric that more artfully balances power, weight, and financial outlay. Thus, the �Speed per Dollar� concept was born. Being something of a buffoon at mathematics and graphics, I sought help to turn the idea into a meaningful standardized equation that would yield useful numbers. The staff at WINDING ROAD rallied around the concept and contacted the fine folks at Funnel Inc. (motto: �Information design that makes the complex clear�), and the concept became a reality. The basic calculation they came up with is as follows: (Horsepower � Weight) x 10,000 � Price Point x 100,000 = SpD We think this SpD metric is uniquely useful because it weighs a vehicle�s accelerative capabilities against its cost. There�s no doubt that the Bugatti Veyron 16/4 is a tremendous technical achievement, but as a value proposition, well, it fails miserably�a development that should come as a shock to exactly no one given its plutocratic price tag. We like the SpD because it celebrates affordable sports like the Mazdaspeed3, Chevrolet Corvette, and the Ariel Atom. It isn�t perfect (it doesn�t take into account a vehicle�s handling envelope), but nevertheless, we think it�s a better evaluation of enthusiast vehicles than existing calculations. A few sample equations: 2007 Ford Mustang GT: (300 hp � 3356 lbs.) x 10,000 � $26,370 x 100,000 = 3390 SpD 2007 Lotus Elise: (190 hp � 1984 lbs.) x 10,000 � $42,990 x 100,000 = 2228 SpD 2007 Saleen S7 TT: (750 hp � 2950 lbs.) x 10,000 � $580,000 x 100,000 = 437 SpD Link Winding Road � Archive � A New Performance Paradigm: The �Speed Per Dollar� Index Thought this would be an interesting topic to talk over. __________________ […]

Jonathan Fung June 2nd, 2007 1:14 AM Link

Um, in case nobody noticed yet, you can sort of simplify that equation. Why multiply by 10,000, then divide by 100,000 when you can simply divide by 10? Simple cancellation of zeros to make it easier to calculate. I present you with the NEW AND IMPROVED SpD INDEX EQUAITON!

(Horsepower ÷ Weight) ÷ (Price Point x 10) = SpD

Jonathan Fung June 2nd, 2007 1:15 AM Link

And just parenthetically, yes I know I accidentally misspelled “equation” above.

Jonathan Fung June 2nd, 2007 1:19 AM Link

Parenthetically again, I realize I’m an idiot. That equation doesn’t simplify that way! Go me, and my arbitrary addition of parentheses. Please, ignore my past two comments. XD

Bill June 2nd, 2007 6:33 AM Link

What all of this mathematical gobbledegook overlooks is the sex factor.

Most red-blooded guys, if they were honest, would admit that they want to buy the horniest-looking performance car they can afford (or usually not afford).

Put a sex factor into the equation and then see where commuter vehicles like your Mazda end up in the list!

ryan June 2nd, 2007 11:13 AM Link

jeez bill, then that list would almost have to be inverted….

and after taking a quick peek, with a few exceptions, it WOULD almost be inverted…

and johnathan, i did that too, realized it didn’t work, and erased a previous post before i submitted.

lap times are tough to replicate, as you’d have to have EVERY car at a certain track at the same general conditions. plus, it totally shoots the concept of “finding your own” fun factor. i can cruise the interwebs and find the details i need to calculate my car, where i couldn’t do that with “lap times” included.

how about this formulation:
((horsepower x (Quarter Mile x slalom)) / weight) 10,000 ÷ Price Point x 100,000 = SpD

now let the comments proceed!

Lance June 2nd, 2007 7:40 PM Link

I vote for quarter mile times used instead of HP and weight. Quarter mile times are a better indicator of “speed” than trying to estimate it with power and weight and also better than 0-60 times whichd don’t emphasize the transmission enough (many cars require a shift right a the 60 mph mark).

Turbo Tuning » Winding Road invents “Speed per Dollar” index June 3rd, 2007 8:09 AM Link

[…] Read | Permalink | Email this | Comments […]

jnaggs June 3rd, 2007 11:48 AM Link

i just want to point out that the most awesome car never made, the dodge razor, would easily be at the top of this list

fearless fred June 4th, 2007 12:12 AM Link

As others have noted, the formula is clearly faulty, because it makes the price completely bias the outcome.

What you want is a flexible formula where the outcome dependends on how much money means to you, i.e. if you have more money to spend you don’t mind paying disproportionaly more to get a car with higher HP (or other features) than a mazdaspeed3.

So I suggest the formula:

((weight/HP)^N)*$

(the lower the better). Where N indicates how rich you are, i.e. higher N means you care about money less.

So if you take for example N=2 as a nice starting point, you get this top “bang for your buck” cars which looks like a pretty accurate “enthusiasts” list (in order of best score):

2006 Chevrolet Corvette Z06 64890 6.2
2006 Chevrolet Corvette Coupe 43690 8
2006 Ford Mustang GT Deluxe Coupe 25140 11.5
2006 Dodge Charger SRT8 35320 9.8
2006 Nissan 350Z Coupe Base 27650 11.2
2006 Mitsubishi Lancer Evolution RS 28679 11.3
2006 Dodge Viper SRT-10 Convertible 81895 6.7
2006 Chrysler 300 C SRT-8 39920 9.8
2006 Pontiac Grand Prix GXP Sedan 27330 11.9
2006 Nissan Altima 3.5 SE 23600 12.9
2006 Chevrolet Impala SS 26330 12.3
2006 Mitsubishi Lancer Evolution IX 31399 11.3
2006 Subaru Impreza WRX STI 32995 11.2

(first number after the name is $, second is weight/HP)

if you increase to N=5, you get the list where price is almost no object, but still favouring good bang for the buck cars:

2006 Saleen S7 Twin Turbo 555000 3.7
2006 Chevrolet Corvette Z06 64890 6.2
2006 Dodge Viper SRT-10 Coupe 83145 6.7
2006 Chevrolet Corvette Coupe 43690 8
2006 Ferrari F430 Coupe 168005 6.7
2006 Lamborghini Gallardo Spyder 195000 6.7
2006 BMW M5 Sedan 81200 8
2006 BMW M6 Coupe 96100 7.8
2006 Lamborghini Murcielago Coupe 288000 6.3
2006 Morgan Aero 8 Convertible 109434 7.7
2006 Dodge Charger SRT8 35320 9.8
2006 Chrysler 300 C SRT-8 39920 9.8
2006 Mercedes-Benz E-Class E55 AMG Sedan 81800 8.7

If you decrease N to very low, like N=1.2, you get the “budget power” list (N=1 is too low to factor in performance):

2006 Chevrolet Aveo Special Value Sedan 9350 22.9
2006 Suzuki Aerio Sedan Base 13999 17.2
2006 Saturn ION 2 Sedan Manual 11925 19.9
2006 Kia Rio Base 10770 21.8
2006 Mazda MAZDA3 i 4-Door 13710 18
2006 Kia Spectra LX 12895 20
2006 Ford Mustang GT Deluxe Coupe 25140 11.5
2006 Nissan Sentra 1.8 13200 20.2
2006 Nissan Sentra SE-R Spec V 18300 15.5
2006 Nissan 350Z Coupe Base 27650 11.2
2006 Mazda MX-5 Club Spec 20435 14.5
2006 Chevrolet Cobalt SS Sedan 17400 16.6
2006 Nissan Altima 3.5 SE 23600 12.9
2006 Hyundai Accent GLS 12455 22
2006 Toyota Corolla XRS 17880 16.3

So… what N do you rate yourself at? I’m clearly an N=1.8 person
(this data is taken from an excel sheet which lists all US 2006 cars for sale with HP, weight, price… the last number is pounds per HP)

Dustin June 4th, 2007 12:46 AM Link

The zeros DO cancel out, and there are no parenthesis needed multiplication and division. Couldn’t help myself, I just got my engineering degree. I like the performance metric as a car nut and a tightwad.

George June 4th, 2007 10:06 AM Link

There’s no fun factor here, none of the many attributes that make up a great car. We all know there are bang for the buck cars but how fun is the car to drive, isn’t that what it all boils down to anyway? I think any time an equation shows a Mustang is rated higher than a Lotus that should ring some common sense bells. Here’s my take (lets give those SUVs a fighting chance):
(Horsepower ÷ Weight) x 10,000 ÷ Price Point x 100,000 x Cup Holders = SpD
BTW - I’ll take this opportunity to say I never see these bang for the buck cars at a track day. Can a new mustang GT even survive a full D/E day or weekend? I might be in the minority but I would love to know how these bargain burners hold up at a track day event.

Siminn June 4th, 2007 10:45 AM Link

I think a bored nerdy 3rd grader came up with this equation.

…Speed Per Dollar is the new benchmark? at Didn’t You Hear… June 4th, 2007 2:30 PM Link

[…] [Winding Road - Speed Per Dollar] […]

Khalen June 4th, 2007 5:34 PM Link

2006 BMW 750Li - 1034 SpD

1999 Lincoln Navigator - 1064 SpD

2002 Lexus SC430 - 1142 SpD

2001 Acura CL 3.2 Type S - 5352 SpD

All cars were brand-new when purchased except the Acura, which cost $14,000 in ‘04

Ryan Katko June 4th, 2007 9:14 PM Link

Well, my 2003 Mazdaspeed Protege ranks in pretty good. 170hp / 2843 pounds x 10,000 / $20,000 x 100,000 = 2989 SpD. Granted i bought the car last year for $14,000 so if we used that price then its: 4271 SpD =)

Keith Tanner June 5th, 2007 12:45 PM Link

My lawnmower comes in at about 1,400,000 SpD, new prices. Winner!

My home-built Lotus Seven replica also does okay, scoring about 13,500 SpD.

VehicleBuzz.net » Blog Archive » Comparing Cars–SpD August 26th, 2007 7:51 PM Link

[…] Then Winding Road got in the game, and proposed the Speed/Dollar index, which is HP/weight/price * fudge factor. […]

Josh March 16th, 2008 4:34 PM Link

So lets see how my 1985 Chevy 3/4 ton pickup does:

(I’m not 100% sure on the weight, but 5000lbs seems reasonable)

(180hp / 5,000lbs) x 10,000 / $700 x 100,000 = 51,489

and now my 2006 Mustang GT Convertible:

(300hp / 3750lbs) x 10,000 / $32,000 x 100,000 = 2,500

I guess I have to disagree with this metric. Sure, driving my old pickup was fun, since I could leave the keys in it, and I could ram into anything I wanted without much care. Still, I’m pretty sure the old truck isn’t 20.5 times as fun as my Mustang.