BRI Index Error for AM

[Otto]

As you may have read, we were puzzled by the substantial advantage that AM seem to have over all other classes. Mike Gagen properly pointed out the error as follows:

quote

BRI issue.

So Roland was in HM last year but the new rules have put his HM car in with AM. Seems only fair then the BRI in AM would default to the higher of last years HM or AM indexes. AM is 1.074 and HM was 1.108. Perhaps the new AM BRI should be 1.108 to reflect the recent rules change. Last year no one had an issue with Roland's times and BRI in HM, right?

unquote

Mike is right and the BRI Index for AM used since the beginning of this year when class progression was changed should simply be corrected to reflect the 1.108 Index. The BRI Index would then make more sense as, for example, right now JI Class, which flows into AM, has an Index of 1.0830, which is higher than the current AM Index at 1.074. This is obviously incorrect.

Not to take anything away from Roland's great performance but it looks like now Jack Miller would be the new BRI leader for Fontana. Congratulations!

[Mike]

Mike Gagen properly pointed out the error as follows:

If Roland's BRI TTOD is rescinded I suggest we all buy him a beer at the next TT, maybe then he'll still be my friend.
Maybe something for all my other AM buddies would be nice too.

Otto you owe me a point by!

[Otto]

Mike:

Roland's time is still outstanding. However, in the interest of fairness, the AM BRI Index should be changed to 1.108, which more properly reflects the current potential of that class. I am sure Roland would agree because he doesn't want a hollow victory and now has a new target to shoot for. If the number is not changed, AM will win the BRI every time (as the results show so far this year) and everybody will lose (more) interest in this ranking. By the way, who makes that change? Carl? Kary? Robert?

Regarding the point-by, with your "new" car, you will be passing by me so fast that I won't even have time to give you a point-by.

[ttweed]

If the number is not changed, AM will win the BRI every time....

Well, although Roland didn't run today, the highest placed AM car was fifth on the BRI at the autox, behind cars from the GS, PP, CS, and KP classes, so this theory may not be accurate.

As I have been saying in the other threads about this, the BRI is an autox index, and should probably not be applied to time trials. Make a new index for that. Adjustments to the BRI should be based on autox performance, as it always has been. Time trials and big tracks are a different animal. There has been an anomaly in this one case due to the elimination of HM class, for sure, but I'm not sure applying the old HM index to AM is the answer. There are other problems with the BRI that need to be addressed due to the other class changes and additions in this year's rules as well, but I would like to see solutions developed systematically from AUTOX results and performance, not haphazardly based on the results of one recent time trial at one of the fastest tracks we run.

BTW, Clark's AM car was 4th TTOD (ignoring the X-class Lotus car), behind SG's 914, JGW in Harry's GT2 and Steve Dente's GT3, so it's not like it was poorly driven. I doubt that Roland's car would have placed much higher.

TT

[Otto]

Tom:

I do agree that that a separate BRI should be developed for TTs. However, in the absence of it at this time, the current BRI should be properly stated. The current AM number is a blatant error because you cannot have a class that flows into AM having a higher BRI Index than AM itself. This is the case with JI which as a BRI of 1.083 while AM, the next higher class, currently shows a BRI of 1.074.

Further, I bet Roland's car, presumably much more developed than Clark's car, would have done better in the AutoX. The principle guiding the BRI is most developed car performance driven by best driver. That is Roland at the moment for AM class. Please note how far ahead he was over all other AM drivers at the TT, which simply showed the potential of that class and that should be reflected in the BRI. Is 1.108 the right number? At least there is some basis to it because it was the number that was valid in HM when it existed. As Mike Gagen pointed out, the new hybrid class should be rather called HM and not AM and then likely there wouldn't be any discussion about the proper Index. Semantics. Bottom line, the preponderance of the evidence and results so far do show that this new catch-all class currently called AM should have the higher pre-existing Index of 1.108 to make any sense.

Who makes the correction of this error?

[ttweed]

The current AM number is a blatant error because you cannot have a class that flows into AM having a higher BRI Index than AM itself.

This is certainly clearly the case and I can't argue that with you. The AM number needs to be changed, but it needs to be done in concert with the indexes for the new classes, KM, O and P (which may also effect AR index as well), and the changes need to be based on autox data and potential performance, not TT data.

Further, I bet Roland's car, presumably much more developed than Clark's car, would have done better in the AutoX.

I can't agree with this completely. I have my doubts that Roland's car would have been much faster at the autox. A little, perhaps, but until he runs again on a short course (and he only did one autox all last year), we won't know for sure, but I don't think we can extrapolate directly from his domination at the Speedway to equal autox performance. Many cars that are set up for big tracks are not that good on a tight, slow course and vice versa.

Is 1.108 the right number? At least there is some basis to it because it was the number that was valid in HM when it existed.

What number should KM have then? You are misunderstanding the class change that took place. HM became KM, not AM. Only the lowest class of HM cars was shifted to AM, the rest of the higher potential cars went to KM. This makes it a better argument to say that KM should have the 1.108 index (or even a little higher, due to the addition of the potentially faster O and P classes which feed it), and AM should only be adjusted to be slightly higher than JI (the highest feeder class), say, 1.09x or something. If HM becomes 1.108, then KM has to be much higher, which will push AR even higher than that. You must look at the whole progression of classes, not just single out this one change and try to push it through.

Sorry, Otto, but I think you are making a knee-jerk reaction to a single event result at a TT that has little to do with the historical data for the BRI, and are basing your suggestion for the new index on some faulty reasoning about how the classes changed this year.

Who makes the correction of this error?

Carl Scragg has been the semi-involuntary "keeper" of the BRI index, and I don't think he has really had the time or the inclination (or the performance data) to look at the overall picture and make adjustments. If anyone is qualified to do it, he is the guy.

TT

[ttweed]

Kary-

I have not seen the indexes you are using now for the BRI calculations published anywhere. Certainly this page on the website is obsolete, as it does not contain the new classes.

Can you post an updated list so that we can see all the current indexes and how they relate for the sake of this discussion?

Thx,
TT

[kary]

Kary-

I have not seen the indexes you are using now for the BRI calculations published anywhere. Certainly this page on the website is obsolete, as it does not contain the new classes.

Can you post an updated list so that we can see all the current indexes and how they relate for the sake of this discussion?

Thx,
TT

Here is the database dump of the current time trail BRI index numbers:

944,1.04
QSS,1
PP,1.115
PS,1.09
PSS,1.065
OP,1.085
OS,1.06
OSS,1.035
NP,1.08
NS,1.055
NSS,1.03
MI,1.115
MP,1.08
MS,1.055
MSS,1.03
LP,1.07
LS,1.045
LSS,1.02
KM,1.12
KI,1.086
KP,1.061
KS,1.036
KSS,1.011
JI,1.083
JP,1.058
JS,1.033
JSS,1.008
IP,1.042
IS,1.017
ISS,0.992
HI,1.066
HP,1.041
HS,1.016
HSS,0.991
GP,1.04
GS,1.015
GSS,0.99
FI,1.059
FP,1.034
FS,1.009
FSS,0.984
CP,1.025
CS,1
CSS,0.975
AR,1.13
AM,1.074
AI,1.042
AP,1.011
AS,0.986
***,0.961
X,3

[kary]

Here is a scatter plot that shows things a bit better.


www.group9motorsports.com/videos/BRI_plot.doc

[jgunn]

Kary/Carl -- just so you know, even though I am running the GT2 in PP it is a 6 point car (4 for tires being 30 treadware and 2 for being 315s) and is really a PS car. I think this is important info for those trying to accurately formulate the BRI numbers, as the car runs AR times in PS trim...

Imagine what it could do in AR trim? Lose 500lbs, more rubber (GT3RSR flares, etc.), slicks, etc.

Our sample size is too small and there are just not enough really good drivers in all the classes driving maxed out cars to be able to accurately create a fair BRI.

The only really decent way to set the index would be to have a test occur one day with all the best cars in their respective classes running on sticker tires (best brand/type per points/class constraints) with the same very good driver (2 to 3 for averaging and repeatability) in a controlled test situation. At the end of the day with proper testing we could pretty much nail down a decent BRI, but that is the only way I see it happening.

[gulf911]

You are misunderstanding the class change that took place. HM became KM, not AM. Only the lowest class of HM cars was shifted to AM, the rest of the higher potential cars went to KM.

Hi Tom, I don't think that is correct. HM was right above AM and flowed from HM to AR like AM to AR , the last time I looked at the page.

Speaking from the AM perspective, this should have never happened...

[Otto]

Tom:

My motivation for this post is to focus attention on a flagrant error, which is the current BRI Index of AM relative to the classes that now flow into AM. As you well agree, you cannot have a lower class (JI) with a higher Index (1.083) than the class it flows into, AM (currently 1.074). We all recognize the limitations of the BRI but this flagrant error has to be corrected for the BRI Index to remain reasonable.

The way classes are structured currently, A/C/F/G/H/I/J flowing into AM, the old HM Index (1.108) makes the most sense for this new catch-all class currently called AM to recognize the high development potential of this class. As stated earlier, Roland's car, the current AM "top dog", was in fact an HM car last year, subject to the 1.108 Index, providing additional evidence that 1.108 is the more appropriate Index to use for the new AM class. KM, by the way, which is the only other current 41-54 point class flowing into AR, has a higher Index of 1.12, as you may have already noticed.

That said, would this new 1.108 number be fair to the lower car classes A/C/F/G, which previously were the only ones that flowed into the old AM? From all the opinions we have heard, likely not. To maintain the motivitation of these lower classes maybe the old AM class should be reinstated with A/C/F/G flowing into it with the old Index of 1.074. Classes H/I/J would flow into the old HM with the Index of 1.108. The "new technology" higher classes could also be subdivided into two 41-54 point classes, say KM made out of K/L/M (1.12 Index) and OM made out of N/O/P (? Index). This suggested arrangement would result in four 41-54 point classes flowing from six current 21-40 point classes, which is not unreasonable. AR would be reserved for 54+ point cars, or if you wanted, you could raise that point threshold, since it is the ultimate "formula libre" class.

Anyway, that is my reasoned opinion. Carl, how do we proceed from where we are?

[Jad]

Remember, there is no correct BRI. It is based on the very flawed classes we have to begin with, so perfection will not happen. It is also based on a small sample that we have in each class, so a class with several strong drivers is going to get a tougher BRI than the same class with no strong drivers. The BRI is somewhat customized to our little corner of the world. When a new element is added, like rules change or a dominate driver, the indexes must be adjusted as necessary to get what seem like reasonable results for discussion and comparison amoung a larger group of friends. There is no car that I know of that is maximized for its class, some use all the points, but not optimally in my knowledge. My car would be much faster with slicks than Toyo's, but I just adjust my car to be competitive with others in the class. The BRI should be the same, if AM always wins, then change the index so a variety of classes have a chance and a really good drive could get recognition whether it is in a GT2 or a 356. For the most part, regardless of class or car, the same people keep appearing in the BRI top ten, so it must be a least a reasonable system that just needs constant adjustments like with Rolands domination. Now, if I can just figure out how Jack's class needs a new index, then I can finally get the dolphin trophy

[Carl Scragg]

When I changed the BRI for 2005, I tried to adjust the indices in a manner that is consistent with the new Z8 class structure. It looks like I missed the fact that JI now flows into AM. Obviously the index for AM should be greater than that of JI. I can put on my Carnac turban and come up some new numbers that are not so inconsistent, but I'm open to suggestions.

[ttweed]

Hi Tom, I don't think that is correct. HM was right above AM and flowed from HM to AR like AM to AR , the last time I looked at the page.

Yes, Dan, and that is exactly what KM class does now. All that happened this year was that the three lowest classes of the old HM cars (H, I and J cars, i.e., all the torsion bar suspension 911s), are now progressing to the AM class instead of HM, where previously only cars from A thru G went there. All the 964 and later cars from K thru P class (coilover types) now progress to KM, and thence to AR, the highest class. Look at the chart on the back of the Reg/Tech form. There are only two modified classes, AM and KM, just as there used to be only AM and HM, both progressing to AR, the highest class. It was renamed KM because K is now the lowest class to progress there, but it's position in the hierarchy is the same as HM used to be.

The difference is that our G-cars used to be the highest class to progress to AM, but now the H, I and J classes have been merged with us from above-- but all of HM was not merged with AM, just the 3 lowest classes. The K, L, M, N, O and P cars that were previously also potential HM cars (or unclassified, in the case of O and P) make up the new KM class. Thus my assertion that KM is the new HM, only with greater speed potential than HM ever had because of the addition of the new O and P cars (GT2 and GT3), which when fully developed to KM level will force AR class to be raised as well.

Jad and JGW are correct, we are not going to have any index that is perfect, ever. Even the PAX in SCCA which has a huge sampling size with fully developed cars and top-level drivers to compare, has problems and is revised constantly. However, I don't agree with Otto that the old HM index should be applied to AM, as not all the HM cars were moved to AM, only the lowest tier of HM was merged. The current KM index (which I am seeing for the first time in Kary's post above) is too high, IMHO, at 1.12, if AR is 1.13. Unfortunately, we cannot change the class structure as Otto suggests to provide 4 Modifed classes until next year at a minimum. However, we can adjust the index to give a more progressive spread between JI, AM, KM and AR to reflect the speed potential of those classes more fairly. I think it ought to look something like this:

AM= 1.095
KM= 1.115
AM= 1.135

In all likelihood, if we begin to see fully developed KM cars in the future both KM and AR will have to be raised further. If you gave JGW 53 points in that GT-2 (which is already running AR times with 6 points) what do you think would happen?

YMMV,
TT