What makes the AMX so fast?

[Jymis]

Yah TJN, you are right.  With the twists/elevation change/short straights. Sandy Creek was meant for the Javelin. I'm sure it won't be so dominant on say Rd America with the long straights.

[Spiny Anteater]

I think the Speed rating is the PI eater.

...

IIRC Dirt and Fndr were calculating what each 0.1 worked out to be in mph.  I know that the AMX is limited to 127.9 in my build.  Round up to 130.  Divide it by 3.0 and each 0.1 is worth 4.3 mph.  Still talking about speed rating here to be clear.

AMX = 130 mph calculated top speed with a 3.0 rating

Using that one example as a rationale the Barracuda would be capable of a much faster top speed and would be quicker around the available tracks as well as the mythical PI track.

Barracuda S has 1.7 more top speed points and that equates to 73 more mph on the top end.  (Assuming this is correct and I'm sure it is less across more examples)  That is a lot of top speed and a huge deal in the overall times on a large number of tracks.  Even on Infineon all of my cars are pushing over 110mph on the short straights.  On Road Atlanta or Sebring a car with the ability to go 20 or 30 mph would easily overtake the AMX with similar handling abilities.  At 73 mph the poor AMX is a sitting duck on the straights.

I think the top speed aspect is a bit less dramatic than this. I can't say I've tried this out, but I know that all the stats range from 3.0 to 10.0, so 3.0 covers all maximum speeds from a given value (probably 137mph or some other strange figure) and below. I'd suggest the best way is to find the top speed of a car with a speed rating of 3.1, subtract that from the top speed of a car with a rating of 9.9, and divide the value by 6.8 to give the increment per unit speed rating.

That 4.3 value is definitely too high as it suggests a car with a rating of 10 is 301mph faster than a car with a top speed rating of 3.0 To lay one limiter on that, my hyper-powered Zonda has a rating of 10 and will only hit 251mph on the Mulsanne (although I suspect the benchmark figure is probably closer to the values used for the speed rating). I suspect the increment is <2mph/0.1 speed rating - one for me to look at tonight I think...

[Fit4aking]

I meant to as Dirt if he and Fndr ever finished with their data mining to get a figure for that little experiment.

[Spiny Anteater]

Had a check last night to see what I got. I deliberately chose stock cars in my garage (needed to be in the garage to see the benchmark figures) and only went for high and low values so that I had a good range to divide by and so get closer to the exact figure. Of course, I didn't think until later that I should have also checked a middling car to get an indication of whether the progression is a straight line (arithmetic) or curved (geometric). But, on the assumption that it is arithmetic, the results were:

Ferrari 458: Top Speed 217.soemthing mph, rating = 9.7
'79 Camaro: Top Speed = 129.something mph, rating = 3.1

Dividing the difference between the top speeds by the difference in rating (6.6) gave me just under 13.5mph per speed rating point. That would make the Barracuda on the first page at least 23mph faster than the Javelin.

[DirtDriver]

It is not a straight line increase. One PI point for speed between a rating of 2 and 3 was less than one PI between 9 and 10, so we could not get a simple 1 PI equals XX mph kinda deal.

DD

[fndrbndr]

IIRC, it was some kind of log linear function, as Dirt suggests.  I think that project came from FM2, though, so I'm not sure whether 3 uses the same system (though I suspect it does).

[BarbecuePete]

I don't think there's any real correlation between speed rating and actual top speed really, for example my LeMans missile RX7 in FM3 had a speed rating of something like 8.5, but did 262mph...

I try not to look at the stats too much, I think they can be misleading. I guess one way to test would be if you could build 2 different cars with the same stats and see how they stack up against each other?

[GICheeze]

The Speed stat is probably the least reliable of the bunch.  The reason - Aero parts affect it, adjusting the Aero parts does not.  You can catch the TS for the car with default aero settings, max settings, and min aero settings.  The benchmark and testing will produce 3 different TS results, but the Speed Stat will remain the same.

$

[Spiny Anteater]

I have to agree with you on that Blooze, but I always felt that the gearbox was more often the limiting factor. As the speed rating was based on default settings, the race gearbox in particular often limited the rating. I would bet that it was the gearbox which was lowering Pete's RX-7 speed rating.

[BarbecuePete]

I would bet that it was the gearbox which was lowering Pete's RX-7 speed rating.

You'd be right Spiny, the RX7 had a sports tranny.

It was that car which started me off looking at my builds on other cars to see if it was worth dropping race transmissions for sport or even stock. And that's where I first noticed the speed rating not changing unless the race tranny was added, but it didn't have the same effects on PI as it does in FM4.