Go to previous topic
Go to next topic
Last Post 11 Aug 2014 01:58 PM by  Forcednduckshn
New Jersey Champ Tour PAX/RTP Index Results
 29 Replies
Author Messages
rtp.rick
Veteran Member
Veteran Member
Posts:2434


--
05 May 2014 06:08 PM

    Attached are PDFs of the PAX/RTP Index Results from the New Jersey Champ Tour.  As usual, the "Class Pos." column indicates Final Class Position.

    Seems like the Street Modified Category has certainly stepped up their game in 2014.

    H's & K's,
    Rick Ruth

    Attachments
    AceInHole
    Basic Member
    Basic Member
    Posts:210


    --
    05 May 2014 08:47 PM
    If "stepped up their game" is short for putting Tom O'Gorman and G.J. Dixon in the car, then sure, the game has been stepped.
    geezerracer
    New Member
    New Member
    Posts:21


    --
    06 May 2014 07:49 AM

    Great comment. Notice how DSP jumped as well, where the big car adjustment was putting Sammy in it. IMO it exposes the big flaw in pax, there are a dozen or so drivers in the country that can put a big wheel at the top of the pax list and the rest of us poor slugs have live with the results.

     

    jwbrockman
    New Member
    New Member
    Posts:91


    --
    06 May 2014 08:19 AM
    It sounds like you're describing the "big flaw" in using timing equipment, not the "big flaw" in PAX.
    z06fun
    New Member
    New Member
    Posts:47


    --
    06 May 2014 10:23 AM
    The results are also going to be skewed because of the cold temps, even colder than DC. Having a co-driver and tires that worked in the cold was everything.
    jeffh
    Basic Member
    Basic Member
    Posts:258


    --
    06 May 2014 12:07 PM

    How many others are questioning the index? I know Rick didn't sign up for it but match tours are using it to determine an overall winner. DAX seemed to be headed in the right direction of having something which was manageable throughout the year. 

    3% for STX is a pretty big margin to make up as 3.75 seconds just wasn't out there. As close to the difference of switching to R-Comps.

    OZMDD
    Basic Member
    Basic Member
    Posts:131


    --
    06 May 2014 01:07 PM
    Sour grapes. Results were skewed in a totally different direction at MW, with only slight differences in the pax for prosolo. Fundamental rule of statistics is that they don't predict an outcome, only describe trends.
    The Nebulizer
    Veteran Member
    Veteran Member
    Posts:1817


    --
    06 May 2014 02:00 PM
    Posted By geezerracer on 06 May 2014 07:49 AM

    Great comment. Notice how DSP jumped as well, where the big car adjustment was putting Sammy in it. IMO it exposes the big flaw in pax, there are a dozen or so drivers in the country that can put a big wheel at the top of the pax list and the rest of us poor slugs have live with the results.

     

    Indexing is an insanely difficult task - much harder than most give it credit. It is not simply about pooling data and taking averages. The data is sporadic and fluctuates tremendously. The class make-up varies greatly as well and each class has its own unique characteristics. Until known alien drivers like O'Gorman or Strano running well prepped car across every class on every surface under ideal conditions for easy direct comparison, indexing is a best guess endeavor. It's not magic or a true statement about reality, it is a best attempt to add inter-class competition. PAX is a tool and has its limitations (and, DAX certainly does too - I am still plugging away at its shortcomings.).

    As for O'Gorman in SM with a PAX of .870, he appears to have just driven exceptionally well (I hear he is pretty good ). Even with an outrageously high index of .890 for SM, he still would be 12th (at .879 he'd still be 1st).

    It's tempting to look at PAX results for a single event and blame the index for something that appears off. But, the index could be a perfect reflection of reality and we would still see odd things here and there. It's the nature of the imprefect and highly variable data.

    Rick has taken on a very difficult job and handled the criticism admirably and unflinchingly. Thanks, Rick!

    jwbrockman
    New Member
    New Member
    Posts:91


    --
    06 May 2014 02:37 PM
    Single driver/class anomalies are inevitable, but an entire well-attended category only placed one car in the top quarter of results (and just barely in the top quarter at that).
    OZMDD
    Basic Member
    Basic Member
    Posts:131


    --
    06 May 2014 02:40 PM
    Unless the conditions were not favorable for that type/setup of class. HS/GS cars were dominant at the MW Pro, but HS doesn't even show up at SD until the 2nd page, IIRC.
    mlane350z
    Basic Member
    Basic Member
    Posts:238


    --
    06 May 2014 03:57 PM
    Meh, I was a single driver and ran early in the day....it is what it is.
    ST///M
    New Member
    New Member
    Posts:11


    --
    06 May 2014 07:09 PM
    I'm newer at this, and maybe I'm reading into it too much, but I think the early responses were based on the quote "SM really upped it's game". The reality is there was a reason for that and it had nothing to do with the structure of the class or preparation of the cars in it compared with a week, month, or year earlier. You put alien national champions in decent cars and this happens. It's impossible to appease everyone, I would hope the fastest and slowest drivers in every class are thrown out to determine PAX, as is the case in any true statistical analysis. Not sure that happens here or not? Again, just assuming. I think the National PAX is decent (likewise, I think the Pro Solo index could use a little help) and I appreciate someone going through the effort. Personally, I like faster street legal cars on good tires, it really can be done rather affordably. Much better at being told I'm a slower driver by someone with a 6-8 second slower raw time and a calculator than I am at, well, being 6-8 seconds slower. Just personal preference, what truly matters is what happens in class.
    jeffh
    Basic Member
    Basic Member
    Posts:258


    --
    06 May 2014 07:52 PM
    Posted By OZMDD on 06 May 2014 01:07 PM
    Sour grapes. Results were skewed in a totally different direction at MW, with only slight differences in the pax for prosolo. Fundamental rule of statistics is that they don't predict an outcome, only describe trends.

    Sour grapes? Speaking up is more like it. Created a post on trying to fix the index 10 years ago when I cared about PAX. I haven't for the last 7 or so years until the SCCA started running tours with it. Similar non productive responses as yours were received.

    jeffh
    Basic Member
    Basic Member
    Posts:258


    --
    06 May 2014 08:09 PM
    Posted By ST///M on 06 May 2014 07:09 PM
    I'm newer at this, and maybe I'm reading into it too much, but I think the early responses were based on the quote "SM really upped it's game". The reality is there was a reason for that and it had nothing to do with the structure of the class or preparation of the cars in it compared with a week, month, or year earlier. You put alien national champions in decent cars and this happens. It's impossible to appease everyone, I would hope the fastest and slowest drivers in every class are thrown out to determine PAX, as is the case in any true statistical analysis. Not sure that happens here or not? Again, just assuming. I think the National PAX is decent (likewise, I think the Pro Solo index could use a little help) and I appreciate someone going through the effort. Personally, I like faster street legal cars on good tires, it really can be done rather affordably. Much better at being told I'm a slower driver by someone with a 6-8 second slower raw time and a calculator than I am at, well, being 6-8 seconds slower. Just personal preference, what truly matters is what happens in class.

    It disappoints me to see people write it off as "alien" drivers when it's not close to the case. Any of the drivers in the car would have finished top in PAX using the current SM index. It's just one glaring example of the inaccuracy of using this index.

    Match Tour Rules - Rule #7

    People spend a lot of money to compete at national events, My thoughts of being a good data point to build off of might of been far fetched. At a match tour being 3%, even 2% off the pace is not close to being competitive for the day 2 shootout. 

    Index and bump classes are run at Pro's as well. The Pro Solo format uses an index much different from Match tours. A difference of 2.1% when comparing the 2 classes discussed (unless that was changed from the rule book)


    jeffh
    Basic Member
    Basic Member
    Posts:258


    --
    06 May 2014 08:13 PM
    Posted By z06fun on 06 May 2014 10:23 AM
    The results are also going to be skewed because of the cold temps, even colder than DC. Having a co-driver and tires that worked in the cold was everything.

    Conditions were mostly ideal for Street Touring.

    jeffh
    Basic Member
    Basic Member
    Posts:258


    --
    06 May 2014 08:28 PM
    Posted By The Nebulizer on 06 May 2014 02:00 PM

    It's tempting to look at PAX results for a single event and blame the index for something that appears off. But, the index could be a perfect reflection of reality and we would still see odd things here and there. It's the nature of the imprefect and highly variable data.

     

    What can be done to fix it for an event when it counts? What to do when cars which set the bar are made faster? What to do when the index was not necessarily accurate in the first place? Is the answer really to wait until next year or maybe the year after?

    marka
    Veteran Member
    Veteran Member
    Posts:2258


    --
    06 May 2014 08:43 PM

    Howdy,

    Posted By Jeff Hurst on 06 May 2014 12:07 PM

    How many others are questioning the index? I know Rick didn't sign up for it but match tours are using it to determine an overall winner. DAX seemed to be headed in the right direction of having something which was manageable throughout the year. 

    3% for STX is a pretty big margin to make up as 3.75 seconds just wasn't out there. As close to the difference of switching to R-Comps.

    Match tours are not using the PAX index to determine an overall champion.  Match tours now use the index only to determine the top 8 from each heat that advance to the shootout.  Once there, they run on their class winners dial in, similar to ProSolo.

    Its also not accidental that heats have classes in them that "go well together" index-wise.


    Its not perfect by any stretch of the imagination, but it isn't just straight PAX indexing either.

    Mark

    (and I'm not meaning to detract from your point here... My CS RX-8 apparently would only speed up less than half a second on a 60 second course if I prepped it for STX.  That seems pretty unlikely to me.  That's assuming, of course, that you think an RX-8 is a competitive car in CS.  I'm very happy to be in a position this year where I don't care if the PAX (or DAX) index is "accurate").

    AceInHole
    Basic Member
    Basic Member
    Posts:210


    --
    06 May 2014 08:44 PM
    Hurst figured it out: the top 5 raw times were all SM category cars, and 3 of the drivers in my car would've taken the top index (including me despite coning much faster runs). G.J. would also have been there if he wasn't dirty on all 3 day 1 runs, although despite carrying a cone he still would've been 4th overall with the SM index.

    The honest truth? The courses in NJ favored cars on R-comps that were well setup for "roadcourse" elements. The street tire cars just didn't seem to accelerate or brake through the turns and transitions as well, resulting in much lower speeds once you got to the corners that you'd otherwise want to hold a speed through in steady state. Add in a couple spots that having a decent amount of power really helped (hit my 73mph limiter in 2 spots on day 2) and it's easy to see why SM/SSM would do well.

    I think for Match Tours, if the heats are split by category the PAX factor should level out (as you'd be competing with other classes within the category, or hopefully close to it). If the course is well balanced then it's even less of an issue.
    sjfehr
    Advanced Member
    Advanced Member
    Posts:634


    --
    06 May 2014 09:13 PM

    I get chastised every time I bring it up, but the only way to have an index that's worth a damn across widely disparate events is to have an index that accounts for surface and course dependencies. It doesn't need to be perfect, it just needs to be better than it is now to warrant the change.

    AceInHole
    Basic Member
    Basic Member
    Posts:210


    --
    06 May 2014 09:40 PM
    Posted By sjfehr on 06 May 2014 09:13 PM

    I get chastised every time I bring it up, but the only way to have an index that's worth a damn across widely disparate events is to have an index that accounts for surface and course dependencies. It doesn't need to be perfect, it just needs to be better than it is now to warrant the change.

    A surface modifier you potentially could do with enough events at a given site, but for course dependencies I think there are so many variables that you'd just be introducing more error.  

    conevadr
    Basic Member
    Basic Member
    Posts:217


    --
    06 May 2014 10:54 PM

    I took a look at PAX results for the five Tour events so far this year.  I counted the number of times that each category landed in the top 25. 

    S = 8

    SR = 18

    ST = 21

    SP = 34

    SM = 23

    P = 13

    M = 8

     

    George M

    AceInHole
    Basic Member
    Basic Member
    Posts:210


    --
    06 May 2014 11:53 PM
    Wouldn't those numbers be skewed by how populated each category is? You'd also have potential concentrations of talent in certain classes/ categories, though I'm not sure how you'd depict that. Perhaps number of National trophies/ championships per category....
    conevadr
    Basic Member
    Basic Member
    Posts:217


    --
    07 May 2014 08:07 AM

    I actually enjoy looking at how the realities of our world show up in the numbers.  But I'm not a statistician, and know enough that there are too many variables for a quick look at some of the numbers to allow for drawing any valid conclusions.  I'll leave that to the experts like RR and the rest.  But yes, certainly the number of participants in a category is one of the variables that could influence how many land in the top 25.

     With that said, while I truly appreciate the hard work that RR puts into PAX, I am already on record as questioning the accuracy related to the street category.  This has all been discussed already.  Even so, I do hope that the eventual adjustment to what I believe is a conservative skew will come sooner rather than later.

     

    George M 

    jeffh
    Basic Member
    Basic Member
    Posts:258


    --
    07 May 2014 09:39 AM
    Posted By marka on 06 May 2014 08:43 PM

    Howdy,

    Posted By Jeff Hurst on 06 May 2014 12:07 PM

    How many others are questioning the index? I know Rick didn't sign up for it but match tours are using it to determine an overall winner. DAX seemed to be headed in the right direction of having something which was manageable throughout the year. 

    3% for STX is a pretty big margin to make up as 3.75 seconds just wasn't out there. As close to the difference of switching to R-Comps.

    Match tours are not using the PAX index to determine an overall champion.  Match tours now use the index only to determine the top 8 from each heat that advance to the shootout.  Once there, they run on their class winners dial in, similar to ProSolo.

    Its also not accidental that heats have classes in them that "go well together" index-wise.

    Its not perfect by any stretch of the imagination, but it isn't just straight PAX indexing either.

    Mark

    (and I'm not meaning to detract from your point here... My CS RX-8 apparently would only speed up less than half a second on a 60 second course if I prepped it for STX.  That seems pretty unlikely to me.  That's assuming, of course, that you think an RX-8 is a competitive car in CS.  I'm very happy to be in a position this year where I don't care if the PAX (or DAX) index is "accurate").

    They run on dial-ins which is great. The not so great part is if you break out, you still move on. 

    And great driving the past 2 weekends. It was awesome watching you slide the RX-8 around for you last run in NJ.


    jwbrockman
    New Member
    New Member
    Posts:91


    --
    07 May 2014 11:20 AM
    Posted By conevadr on 06 May 2014 10:54 PM

    I took a look at PAX results for the five Tour events so far this year.  I counted the number of times that each category landed in the top 25. 

    S = 8

    SR = 18

    ST = 21

    SP = 34

    SM = 23

    P = 13

    M = 8

     

    George M


    I don't have the data in front of me to run this myself, and don't take this as complaining about someone else's efforts, but a more useful calculation would be the percentage of vehicles entered in each category to place in the top quarter (or half, or top 10%, a ratio rather than a fixed number) at each event. This would more clearly show how the categories are being represented by PAX. 

    OZMDD
    Basic Member
    Basic Member
    Posts:131


    --
    07 May 2014 12:22 PM
    Ultimately, the only perfect solution is to avoid using an index for competition and only run cars from the same class on raw time. Since the SCCA seems to be moving away from that "pure" approach and implementing indexes to diversify competition/events, the pax/dax argument will probably linger forever, like so many other controversial topics on these forums.

    I think you have a choice to participate or not, and build/buy to succeed within the current ruleset. If "index" is the name of the game, learn how to play it. That's my plan.
    marka
    Veteran Member
    Veteran Member
    Posts:2258


    --
    07 May 2014 01:48 PM

    Howdy,

    Posted By Jeff Hurst on 07 May 2014 09:39 AM
    And great driving the past 2 weekends. It was awesome watching you slide the RX-8 around for you last run in NJ.

    Thanks!

     

    You're not doing too badly yourself against some stiff competition.  :-)

     

    Mark

    conevadr
    Basic Member
    Basic Member
    Posts:217


    --
    08 May 2014 12:47 PM

    I took another pass at this.  This time I captured all of the 2014 Tour data from 

    http://home.comcast.net/~paxrtp/rtp2014.html  into a spreadsheet.


    I matched all of the class data with the appropriate category (S/SR/ST/SP/SM/P/M), except for dropping CAM, FSAE and Jkarts.  In total they only had 25 entries and none were in the top 50 on pax.  Ladies classes are included in whatever category was appropriate.  KM was included in M (Mod).  And I assumed that the rest of the data was correct.  Later on I noticed that there were around 50 entries at the bottom of the Dixie report all at position 225 with no time.  I left these in rather than going back and re-calculating everything.  Since they were at the bottom, they would have had the same impact across the board.  Here's what I found...


    1102 total entries counted


    #competitors  Category  %of total competitors  #intop25pax-#intop50pax  %intop25pax-%intop50pax

    83 M  (7.5%)   8-13   9.6%-15.7%

    105 P (9.5%)   13-25  12.4%-23.8%

    267 S (24.2%)   8-23  3.0%-8.6%   
    (six of 8 in top25 were at Texas)   

    111 SM (10.1%)   23-34  20.7%-30.6%

    135  SP (12.2%)   34-53  25.2%-39.3%

    71  SR  (6.4%)   18-32  25.4%-45.1%

    While not a separate category I also calculated SR for all except those in SSR.  This was a very small amount which indicates that most non-SS stock category drivers have gone elsewhere, probably to Street.
    16 SR - SSR  (1.5%)   1-4  6.2%-25.0%

    330 ST (29.9%)    21-
    69  6.4%-20.9%


    As I mentioned before, I realize the limited statistic usefulness of such a small sample.  Therefore, I hesitate to draw any real conclusions.  I assume that RR and others will be doing a much better job of this going forward.  Also, my agenda as a mostly locally focused Street category competitor in an indexed PRO class is to help push for an adjustment to be made to the Street category paxes sooner rather than later.  All of the Street category pax stuff has already been discussed, so no need to rehash.  But with a few 2014 events behind us, I thought it was interesting to share how things are shaking out.  Specifically that the numbers/percentages of Street category competitors reaching the top 25/50 is much lower than the other categories.  There could be many reasons for this, only one of which is that the pax values are in need of some adjustment. 


    George M



    rtp.rick
    Veteran Member
    Veteran Member
    Posts:2434


    --
    11 Jun 2014 07:39 PM
    Mr. Nebulizer, I have placed new PDFs above.

    H's & KI's,
    Rick Ruth
    Forcednduckshn
    New Member
    New Member
    Posts:2


    --
    11 Aug 2014 01:58 PM
    Posted By Jeff Hurst on 06 May 2014 08:09 PM
    Posted By ST///M on 06 May 2014 07:09 PM
    I'm newer at this, and maybe I'm reading into it too much, but I think the early responses were based on the quote "SM really upped it's game". The reality is there was a reason for that and it had nothing to do with the structure of the class or preparation of the cars in it compared with a week, month, or year earlier. You put alien national champions in decent cars and this happens. It's impossible to appease everyone, I would hope the fastest and slowest drivers in every class are thrown out to determine PAX, as is the case in any true statistical analysis. Not sure that happens here or not? Again, just assuming. I think the National PAX is decent (likewise, I think the Pro Solo index could use a little help) and I appreciate someone going through the effort. Personally, I like faster street legal cars on good tires, it really can be done rather affordably. Much better at being told I'm a slower driver by someone with a 6-8 second slower raw time and a calculator than I am at, well, being 6-8 seconds slower. Just personal preference, what truly matters is what happens in class.

    It disappoints me to see people write it off as "alien" drivers when it's not close to the case. Any of the drivers in the car would have finished top in PAX using the current SM index. It's just one glaring example of the inaccuracy of using this index.

    Match Tour Rules - Rule #7

    People spend a lot of money to compete at national events, My thoughts of being a good data point to build off of might of been far fetched. At a match tour being 3%, even 2% off the pace is not close to being competitive for the day 2 shootout. 

    Index and bump classes are run at Pro's as well. The Pro Solo format uses an index much different from Match tours. A difference of 2.1% when comparing the 2 classes discussed (unless that was changed from the rule book)


    Just going through some threads here and wanted to echo what Jeff says.   Although it does have some effect, course designs don't alter PAX results that much when comparing SP/SM cars to ST cars.  The ST paxes are disproportionately "hard" (i.e. they are off by many seconds compared to the top SP and SM cars) for many of the sites that we run on.  I was setting top/near top PAX in my BS car at many events last year and now have fallen off consistently between 2 to 4 seconds slower on PAX results this year compared to other drivers that stayed in the same car now that I am running STR. The idea that anyone thinks it is just "alien drivers" is why PAX is viewed with such disdain by classes with poor PAX indexes. I won my class, the biggest class in NJ, by 3 seconds, over some multitime national champions but finished something like 4 seconds off the pace on index.  If you put me or Jeff Hurst in that SM 240SX on that day you would quickly see our PAX times jump up 3-4 seconds, and recognize how much of a joke it is to use PAX as a means to isolate driver from car.  I challenge anyone who thinks otherwise to arrange for either of us to get in these cars and prove it with data.

    Thanks,



    ---