Recently, we tried to find out how you could get the cheapest horsepower possible — what car packs the most ponies in for the fewest bucks. I thought the results were quite interesting and surprising, and I wanted to see how the spread looked, visually. And that's pretty much what I did, and that's why we're here.
Using a two-axis graph (I actually started to make a 3-axis graph, with year on the Z-axis, but that became a complicated mess really quickly) with HP on the Y axis (BANG) and cost (BUCK, in 2014 US dollars) on the X axis, you can see a pattern emerging pretty quickly.
It's not a surprising pattern at all. We have big holes in the areas of powerful, cheap cars, and a somewhat smaller hole for expensive, weaker cars. There's a lot of clustering in the mid-low power/mid-low cost area, which isn't too surprising.
I think the biggest lesson here is that there is no volume discount on horsepower. In fact, the more you get, the more expensive each one tends to be. Since HP is a power rating of an engine and not mass-producable plastic doohickeys, this isn't too shocking, but it is striking to see how intense the effect is. Look at the extremes — the per HP cost of a Veyron compared to a Tata Nano, for example.
Also notable is that the '68 GTO was a pretty great value, HP/$-wise, and a Bristol Blenheim 3 is, not shockingly, a terrible one. I'm pretty sure everyone's monocle is safely held on their faces after seeing that.
Hopefully, this is interesting to you as well. Print it out on a shower curtain and use a dry-erase marker to add in cars while you bathe!
UPDATE: I can see I'm getting a lot of guff from the tilt of the chart. It was a design decision to make it feel a bit more dynamic, but I see your points. So, I just want to say I'm listening, and for data-heavy charts like this in the future, I'll work on having more legible approaches. Thanks for the commentary.
(see it really big here!)