What about a PAX class?

[Loren]

Been thinking about this for a while, just haven't found time to verbalize it... or to flesh out my thoughts. Bear with me, it's entirely possible that I'll talk myself out of this in the process of gathering all of the details! I'll start by explaining PAX for those who don't know, then get into how we might maybe be able to apply it to FAST.

What is PAX? It's the "professional autocross index", also known as RTP or "racers theoretical performance" index. It's been around for decades, and is derived from annual analysis of SCCA autocross results including everything from regional to national level competition. Here's the 2017 Index

As a driver, your PAX index is determined by what class you are in (SCCA class). If you're in the fastest possible class (A Modified), your index is 1.000. Every other class has an index that is less than one. It works like this:

CL Time PAX IndexTime
AM 32.000 1.00 32.000
AS 38.000 .819 31.122
HS 39.500 .786 31.047

The faster your class is, the higher its index will be. And the resulting Index Time represents how fast each driver would be in the theoretical "A Mod" car. As you can see, it levels the playing field and potentially allows drivers to compare times across classes.

Or so the theory goes.

Flaws with the PAX system: It's really only remotely accurate when you are driving "the car" for your class, and you have it prepared to the limit of the rules. If your car is anything less than that, you're not comparing apples to apples.

Now, personally, I've always thought it was bullshit at the local level because almost none of us have a car that is fully prepped to the rules. You don't have the tire flavor-of-the-month? Well, you might be a couple tenths or more off. You didn't spring for the $2,000 Penske shocks, or the $1200 custom valved Bilsteins? There's another half-second, and it goes on and on. But, that's just, like... my opinion, man.

It seems that it's become QUITE popular in recent years, and if you go to a local SCCA or Martin event, you'll probably find a lot of people more interested in who's in contention for top PAX time than who's got FTD. And I certainly get it. At the local level, sure it takes talent... but FTD often goes to the fastest car that shows up with a reasonable driver. PAX is going to more-or-less show who the better drivers are. (as long as their car is at least reasonably prepared)

[Loren]

So, how do we apply this to FAST? Well, if you've been paying attention in any of our classing discussions over the years, you've surely seen PAX indexes come up as reference data. FAST classing is rooted in SCCA Street (formerly "Stock") classing, we just have more liberal prep rules, and we combine classes so that we don't have 56 classes for 100 drivers. We use PAX indexes to help us decide how to group classes, or verify that our intended groupings make sense.

Here are the current FAST classes, with their associated SCCA classes and PAX Indexes:

S0
SS - 0.835
AS - 0.833
S1
BS - 0.826
CS - 0.819
S2
DS - 0. 811
S3
FS - 0.814
GS - 0.806
S4
ES - 0.807
HS - 0.798
M0
SS - 0.835
AS - 0.833
M1
BS - 0.826
CS - 0.819
M2
DS - 0. 811
FS - 0.814
M3/M4
ES - 0.807
GS - 0.806
HS - 0.798

As you can see, we try to keep the PAX numbers for each class fairly close, but we're also trying to group "similar cars" together. And we feel that FS (muscle cars) and GS (heavy, but powerful... not quite muscle cars) fit together well, especially with our rules that allow GS cars to fit wider wheels and tires. And ES/HS are a good fit because they are all lower-powered light, nimble cars. But, the PAX numbers are always within a margin of .009.

If you take .009 on the typical 40-second course you get 0.36 seconds. That's our "margin of error", and I think it's pretty good. Well within the "noise level" of driver repeatability from event-to-event on the local level. I doubt that the PAX index itself is any more accurate than that given the wide disparity of cars and prep levels that we see.

I think I'm done with the "background" info...

[Loren]

No, really, how could this apply to FAST?

Well, what we'd have to do is devise our own PAX index (based on the existing PAX indexes) for each class. And, given that it doesn't have to be (and can't be) 100% accurate, maybe we can come up with something that will work well enough?

We need to look at each class individually, especially the ones with oddball rules. But, in general... one could argue that THEORETICALLY, FAST stock classes should be just a tiny bit faster than their SCCA counterparts because we allow a LOT more mods. Take the fastest car in any given class. Give it the allowance of wider wheels, modified control arms for more camber if needed, whatever diff and transmission gearing you want, lightweight flywheel, performance clutch... there are just LOTS of things that we allow that SCCA doesn't. I'd say just bumping the standard PAX indexes up by .01 would be a given. That's something like .4 seconds.

And then there's S5. It's got a tire and diff LIMITATION that none of its source classes have. Therefore, it should be SLOWER than HS. Gotta be at least half a second. Maybe a .01 deduction?

That pretty much takes care of Stock classes. Mod classes... that's a tough one. We need to establish how much faster we expect those classes to be vs. their stock counterparts. We could do that by comparing Stock ("Street") classes to their corresponding "Street Touring" classes and get a reasonable idea.

Examples:
BRZ/FRS - .801 in DS, .822 in STX = .021 diff
2010 MX5 - .810 in CS, .830 in STR = .020 diff
1995 Miata (no LSD) - .794 in ES, .818 in STS = .024 diff
Mazda 3 - .786 in HS, .800 in STF = .014 diff
Mustang V8/V6 - .804 in FS, .820 in STP = .016 diff

Average difference is about .019, which equates to 0.76 seconds on a 40-second course. For simplicity, let's just call it a .02 improvement over Stock for now.

STS & STF has a a 225 tire width, where our M4 has a 205 tire width. Plus the LSD restriction in M4 That probably makes M4 .01 slower than M3.

So, if we take the average PAX for each FAST combined class, add .02 to Stock classes, and a total of .04 for Mod classes, we end up with:

S0 - 0.844
S1 - 0.832
S2 - 0.821
S3 - 0.820
S4 - 0.812
S5 - 0.792

M0 - 0.854
M1 - 0.842
M2 - 0.832
M3 - 0.812
M4 - 0.802

And for R1... we could assume "A Modified, 1.000". But, realistically, what we're more likely to see at our events is something between CP (0.85x) and EM (0.90x) with the occasional FM (0.91x). It might be more fair to our typical R1 participant to call it 0.900.

[Loren]

Okay. This is all just me "spitballing".

If this is an idea that interests you, please discuss. (if not, then there is ZERO point in you getting involved in this thread)

If you want to devise better spitballs than I did, have at it. If you want to test my loose example "FAST PAX" indexes against some actual results, knock yerself out.

We "could" retroactively apply a PAX class to the current season if enough people wanted to and could come to a concensus on how to do it quickly enough. But, more likely, we're looking at next season.

Assuming anybody wants to do it.

[Carracer]

You have the 2016 PAX chart loaded up in your example.

I personally liked someone's idea ( I think it was Preston ) of just running a completely separate class that was a PAX class that followed SCCA rules. Simple, if you know about pax, then great. Sign up and run that! Otherwise keep it raw time. I say this driving an S4/ES that would have considerable gain if we ran a PAX'like system.

Now, following your table, S5 and S4 should have close PAX I think. A Miata even without a diff and on 195s could run very similar times to S4. We do have the risk of someone running the mythical "perfect S4 car" An MB or MR2 on 225s with 8" wheels + header and exhaust. That would SMOKE everyone in S4 with an ES legal car. So I guess an argument could be made that S4 or any other class has a lot more headroom than S5.

I'm going see what this looked like on the last event...

[puncturina]

The timing system allows for entry of a pax factor by class, right? You could just add another report by pax times for everyone, like the report by times, just to satisfy our collective curiosities. If there aren't that many people committed to a pax class, it would still be interesting info.

[Carracer]

https://docs.google.com/spreadsheets/d/ ... sp=sharing

OK, PAX round 2 of the FAST points season.

[Loren]

The timing system allows for entry of a pax factor by class, right? You could just add another report by pax times for everyone, like the report by times, just to satisfy our collective curiosities. If there aren't that many people committed to a pax class, it would still be interesting info.

Exactly. Once we decide on an appropriate "PAX Index", that is easy enough to do.

[Loren]

https://docs.google.com/spreadsheets/d/ ... sp=sharing

OK, PAX round 2 of the FAST points season.

That actually looks pretty realistic. Not bad for a shot in the dark using outdated PAX indexes. (they really don't change much year-to-year unless there's a big classing shake-up)

[Loren]

Oh, I think I noticed something that I'm surprised Philip didn't catch. When writing that 3rd post, I was monkeying around with the numbers, and may not have edited things right. If you take the "stock" PAX and add .02 to it, you end up with indexes that are faster than ANY of the Street Touring classes. I think I did a difference of .01 instead of .02. I don't think any of our FAST Mod indexes should be more than about 0.85 because we're on street tires. 0.86 is CSP... think "Elliot Harvey's Datsun on race tires". Nobody should be anywhere near that on a street tire. I guess M0 at 0.852 is reasonable. Being the "anything on street tires" class, it has potential if somebody decided to get really serious about development.

As I mentioned, I was just spitballing. We can get more "formal" about it and try to get things right if'n y'all want to.

As Eva mentioned, even if we don't do a "PAX Class", we could publish "PAX Results", which are interesting.

[JasonS]

I'd be down for a pax system. If nothing else it's another source of friendly competition, which is always a good thing imo.

[jmdoc66]

From someone one who is always getting shellacked by those pesky miata's, I actually like raw times. Keeps me on my toes and always looking for that extra tenth of a second.

[lddavis91]

PAX SUX

[Loren]

If this is an idea that interests you, please discuss. (if not, then there is ZERO point in you getting involved in this thread)

If we publish PAX results, you don't have to look at them.

If we offer a PAX class, you don't have to compete in it.

[lddavis91]

I have a long history of saying "PAX SUX" when any discussion of PAX is brought up and a reputation to uphold. One of the things I rave about when talking about FAST to those outside the group is the lack of PAX BS, that and the split day format and noon start times. However, you are correct, ultimately I can just ignore it here if it comes about just as I have elsewhere in the past. But really, if you have to break out a calculator to determine who "won the race" are you really "winning"? ;-)

The only thing sincerely that I would add is I think attempting to apply an SCCA class structured PAX to a non SCCA class structure is just an even bigger can of worms than PAX is in and of itself and would only make it suck even more than it already does. If you really want to start PAXing things just start using SCCA classes (I say that and I actually prefer FAST classes/rules to SCCA classes/rules for the most part). Using SCCA classes and rules, while it would suck in many ways, would make things easier for people who want to run FAST events and and all other autocross events and would make PAXing things easy as well.

[CaptainSquirts]

I think publishing pax results would be an ok thing. The people with fast cars can be happy with their quickest raw times, and slower cars that paxed well compared to raw times can be happy about placing high there. Make both ends of the spectrum happy.

[CaptainSquirts]

https://docs.google.com/spreadsheets/d/ ... sp=sharing

OK, PAX round 2 of the FAST points season.

Oh boy, in a world where I get to choose who is faster than me. If raw times Ken is faster, if pax times, phillip is faster. Decisions, decisions...

[Native]

If we offer a PAX class, you don't have to compete in it.

So, in essence, it would become pick your competition - raw time vs pax time? Would we set pax up as "a class by itself" with it's own championship, or have we even gotten that far yet? I guess I'm a little concerned that if someone in a class we're trying to populate, like say M4, decides to just run Pax, now is M4 short on competitors again? If I choose to run Pax, and win, can I also win my "class" at the same time? Or is it one or the other? Did this paragraph even make sense?

[Z1NONLY]

But really, if you have to break out a calculator to determine who "won the race" are you really "winning"? ;-)

The PAX system is an attempt to quantify *driver's* performance. This can't be done with raw time when so many drivers are in so many different cars at different levels of preparation.

You could argue that the PAX system fails to do this too but, in most cases, it does a far better job than raw times. If I run an A-mod car and get a faster time than Philip in his ES Miata, am I really "winning"?

I'm fine with the current FAST classing system, but I would not have a problem with publishing PAX results if they chose to do so. It could be tricky to get multipliers for FAST classes that work as well as SCCA class PAX stuff though. (And their multipliers are not perfect to begin with)

[CaptainSquirts]

Sorry I don't know a lot about PAX but does it even factor in tires? Car classing is one thing but someone running on all seasons/summer tires vs one running on extreme performance tires are going to be huge differences. If pax doesn't even incorporate tires then isn't it all just moot then?