INDEX SYSTEM

[Gonz]

i took the YTD results and did a quick calculation of average score per event, for everyone who ran 5 or more events. You can see that sandbagging is a key to winning the championship. You also need to make it to as many events as possible, of course.

What's interesting that you might not notice throughout the year, is that some of the top drivers scored under 100 TOTAL points in three separate events
That's called playing the system like a fiddle. Of course it's all part of the nature of AutoX, making the most out of the rules.

Here's a look at the drivers who had the best average finish throughout the year. Sandbagging kills your average score, but does wonders for your index. Go figure.........

Total First Name Name EVENTS Avg Score
1608 Laurence Casey 9 178.67
1045 Steven Henri 6 174.17
1720 Frederic Kelley 10 172.00
1536 Wade Chamberl 9 170.67
1506 Peter G 9 167.33
1332 Brian Karwan 8 166.50
1299 Joe Seward 8 162.38
1611 Hobbs533 10 161.10
964 Fernando Puig 6 160.67
956 Jeff Duncan 6 159.33
1106 Stephen Catlin 7 158.00
947 Hil O'Herlihy 6 157.83
942 George H 6 157.00
778 Meilyng Wurmaka 5 155.60
1240 Denise Dersin 8 155.00
1235 Gonzalo Puig 8 154.38
771 Zachary Catlin 5 154.20
917 Miles DeFeyter 6 152.83
759 Mike Robbins 5 151.80
1508 Juan Guzman 10 150.80
1198 Bruce Wakefield 8 149.75
898 Dekonti Sayeh 6 149.67
1045 Matt Canada 7 149.29
1490 Rick Ross 10 149.00
742 Jerry Williams 5 148.40
1034 Slava Burmaka 7 147.71
1322 Fencer 9 146.89
1598 Ed Chan 11 145.27
1595 David Lantz 11 145.00
1301 Bill J. 9 144.56
1153 Kyle Coeyer 8 144.13
704 Al Schmidt 5 140.80
1126 Mark Anstrom 8 140.75
1124 Jerry Olson 8 140.50
702 JCV 5 140.40
982 Benjamin J 7 140.29
1820 James Moran 13 140.00
1669 Ted Joseph 12 139.08
692 Forrest Pafenberg 5 138.40
967 Barrie G 7 138.14
1365 Patti Woworth 10 136.50
680 Bigjon 5 136.00
951 Andrew Huo 7 135.86
813 John Davidson 6 135.50
677 Ted Cooper 5 135.40
1873 Michael Moran 14 133.79
799 Jacob Hurd 6 133.17
1856 Dan Rockholt 14 132.57
793 DimitriB 6 132.17
1056 Isaac O'Neill 8 132.00
1447 Carlos Guzman 11 131.55
656 Aaron C. 5 131.20
1045 Dan Kraus 8 130.63
1434 Mahmoud 11 130.36
781 Craig M. 6 130.17
1685 Bouchon 13 129.62
636 J. Nev 5 127.20
1016 Jeffrey Mortimer 8 127.00
761 Brian Tan 6 126.83
1268 Steve Mitchell 10 126.80
748 Ian Pimentel 6 124.67
747 Nicholas Fontaine 6 124.50
1119 Thomas Lytle 9 124.33
1490 Grease Truck 12 124.17
993 James Low 8 124.13
1482 Travis Finlay 12 123.50
1110 Ryan C. 9 123.33
1724 DartTurbo 14 123.14
1590 Bryan Rutkowski 13 122.31
711 Richard Stowell 6 118.50
1396 Woody oodworth 12 116.33
1603 Ed O'Donnell 14 114.50
1600 Motomoron 14 114.29

[Tedzilla]

Gonz,

Check your math. 4 events were dropped. My 1669 was for 10 events, or a 166.9 average. My total points for 12 events were 1923, or a 160.25 average...
Repeat after me; there are lies, damn lies and statistics.

Ted and his TI-83

[BugBomb]

I agree, Ted.

If you are going to average every event, you need to incude every event. My average should be 166.92 (2337/14).

Dan's average gets even better at 167.5 and Jake is at 164.7.

[BugBomb]

Sandbagging kills your average score, but does wonders for your index.

I disagree. Look at Brian. He is potentially FTD at any event. His average is 166.7 points per event. If he sandbagged at one event, he might get 90 points, but the following 3 events would probably look something like 200, 198, 190, which would balance out his average.

[Gonz]

Gonz,

Check your math. 4 events were dropped. My 1669 was for 10 events, or a 166.9 average. My total points for 12 events were 1923, or a 160.25 average...
Repeat after me; there are lies, damn lies and statistics.

Ted and his TI-83

sorry.

So "total points " isn't the sum of all events................ ok, now I really give up on this INDEX system I mean does F1 or NASCAR let you "drop" your slowest result in your championship battle ?? Consistent performance is part of the game.

So Dan then was the No1 point scorer. He just needs to drive a little slower in his "off weeks" next year

I redid the worksheet anyway.

Total Pts Name Last Name Events Avg per Evnt
Laurence Casey 9 178.67
Steven Henri 6 174.17
Frederic Kelley 10 172.00
Wade Chambern 9 170.67
Dan Rockholt 14 167.50
Peter G 9 167.33
Michael Moran 14 166.93
Brian Karwan 8 166.50
James Moran 13 164.77
Bouchon 13 162.46
Joe Seward 8 162.38
Hobbs533 10 161.10
Fernando Puig 6 160.67
Ted Joseph 12 160.25
Jeff Duncan 6 159.33
DartTurbo 14 158.00
Stephen Catlin 7 158.00
David Lantz 11 157.91
Hil O'Herlihy 6 157.83
George H 6 157.00
Meilyng WBurmaka 5 155.60
Denise Dersin 8 155.00
Ed Chan 11 154.55
Gonzalo Puig 8 154.38
0 Zachary Catlin 5 154.20
0 Miles DeFeyter 6 152.83
Mike Robbins 5 151.80
Juan Guzman 10 150.80
Bruce Wakefield 8 149.75
Dekonti Sayeh 6 149.67
Matt Canada 7 149.29
Rick Ross 10 149.00
Jerry Williams 5 148.40
Slava Burmaka 7 147.71
0 Ed O'Donnell 14 147.50
Fencer 9 146.89
Motomon 14 146.43
Bryan Rutkowski 13 145.00
Bill J. 9 144.56
Kyle Cooeyer 8 144.13
Travis Finlay 12 143.83
Grease Truck 12 142.17
Al Schmidt 5 140.80
Mark Anstrom 8 140.75
Jerry Olson 8 140.50
JCV 5 140.40
Benjamin J 7 140.29
Carlos Guzman 11 140.00
Mahmoud1 11 138.45
Forrest Pafenberg 5 138.40
0 Barrie G 7 138.14
Patti oodworth 10 136.50
Bigjon 5 136.00
Andrew Huo 7 135.86
John Davidson 6 135.50
Ted Cooper 5 135.40
Woody oodworth 12 134.50
Jacob Hurd 6 133.17
DimitriB 6 132.17
Isaac O'Neill 8 132.00
Aaron C. 5 131.20
Dan Kraus 8 130.63
Craig M. 6 130.17
J. Nev 5 127.20
Jeffrey Mortimer 8 127.00
Brian Tan 6 126.83
Steve Mitchell 10 126.80
Ian Pimentel 6 124.67
Nicholas Fontaine 6 124.50
Thomas Lytle 9 124.33
James Low 8 124.13
Ryan C. 9 123.33
Richard Stowell 6 118.50
Andrew Stowell 6 113.17
James Therry 7 113.00
Frankie Patton 8 112.50
Kyle Bowker 5 109.80
Marina Martin 10 108.20
Elron Solis 5 101.00
Jo Cooper 5 94.60

[kyle.bowker]

Gonz, the title at the top of the points results indicates the "Total Points" is the total for each driver's 10 best results that season. We arrived at this number 10 near the end of the season when we realized there would be 14 auto-x events for the year. We never know if we're going to have 11 events or 15 events so all season long we just say you can drop up to 4 lowest scoring events and the picture becomes more clear towards the end of the season as to how many events we'll actually count towards total points. If you want to find out the absolute total (as opposed to the top 10 total) all you have to do is sum across the columns. You have to be a consistent participant and consistently do well but we don't want people to kill themselves trying to attend every single event. We know people have lives beyond CDC and auto-x and we don't want the series to be too punishing.

[TubeDriver]

Welcome to the club!


Gonz said "ok, now I really give up on this INDEX system "

[FredK]

I wish I hadn't driven the Miata my first event, killed the average.
Fredk

[BugBomb]

As far as the theory of increasing the number of averaged events for the current index, I don't believe that will be effective against sandbagging. I put together a little spreadsheet that uses my indexes for this year (including only 3 sandbag events) and compares a 3-event-average to a 5-event-average. The difference in the indexes and the resulting indexed times is very minimal.

http://spreadsheets.google.com/pub?key= ... yVjaXWXKYw

[ProDarwin]

Interesting comparison of two people throughout the season with zero sandbagging:



The last column should read % of possible points (100 for the first 2 rows, 200 for the second 2). Don't know why its getting cut off - its an absurdly small image.

Notice that the index magnified Ed's lead over myself because he was less consistent throughout the season. I actually screwed up and deleted the raw data before I could get St. Deviation from it. I'm curious what those #s would look like

[kyle.bowker]

I'm not sure I understand what your comparison is trying to show. At any rate, you're talking about a minute difference between you and Ed - 0.7%, 2.8%, 1.8%, 4.0%. Do you think you're able to drive within 4% of you or your car's capability with 90% confidence and reliability? Also, while neither one of you may have intentionally sandbagged did either one of you make changes to your car throughout the season or drive someone else's car that may have resulted in you driving slower?

[ProDarwin]

What I'm trying to show is the how the index can artificially increase or decrease a persons score in comparison to their raw. After dropping 4 events, Ed is leading me by 2.8% raw. After adding index times into that (which are entirely based off raw times throughout the season) he beat me by 4%. Thats a 1.2% increase in his lead due to the index alone. Thats 0.54 seconds on a 45 second course. That's a fair amount.

[TubeDriver]

Gee, if that is not enough how about the performance of Ross and me this weekend?


I got a 45.9, Ross got a 46.0 both in S2000. Pretty close right?

Ross scored 6th in the CDC Index, I scored 41st in the index.


How can anyone continue to defend this system? It is completely screwy.

I'm not sure I understand what your comparison is trying to show. At any rate, you're talking about a minute difference between you and Ed - 0.7%, 2.8%, 1.8%, 4.0%. Do you think you're able to drive within 4% of you or your car's capability with 90% confidence and reliability? Also, while neither one of you may have intentionally sandbagged did either one of you make changes to your car throughout the season or drive someone else's car that may have resulted in you driving slower?

[ProDarwin]

Ooops, I made a math error (removed 4 events of mine even though I missed 2). New spreadsheet here:

http://spreadsheets.google.com/pub?key= ... 4z0tzYPrFg

Drop the lowest 4 scores, and before the index Ed is winning by ~4.7%, afterwards that number jumps to 6.05%. 1.25% is still a significant change.

I only threw Standard Deviation in there to point out the difference in consistency in raw times. You can see that going from Raw to Raw + Index, the standard deviations become much closer. Eds/Mine goes from 1.75 to 1.5. Interestingly enough, after 4 events are dropped, that number doesn't change as much.

If this actually works - Kyle this board is crazy right now.

[kyle.bowker]

How can anyone continue to defend this system? It is completely screwy

Simple. Your arguments are for a different system, not necessarily a better one. Each system has pros/cons and there exists compromises. Ross used to be a crap driver according to the CDC index. So at the last event either you drove really poorly or Ross improved a great deal - either by driving better or by improving his car to make it go faster. Remember, at CDC a fast driver in a slow car is nearly always faster than a slow driver in a fast car.

[Gonz]

Remember, at CDC a fast driver in a slow car is nearly always faster than a slow driver in a fast car.

they were both fast drivers in fast cars on this day. Ross drove great! Pete came through under pressure and kept the title of "fastest S2000" with one very quick final run.

But it looks like he's got some real competition now from another S2000. Something that's been overdue. It was fun to watch them no matter where the index places them.

What would have been less fun to watch, would have been if they had both driven slowly to trump the index system.....

[TubeDriver]

I agree but for one driver to get 6th overall and the other to get 41st makes no sense.

Remember, at CDC a fast driver in a slow car is nearly always faster than a slow driver in a fast car.

they were both fast drivers in fast cars on this day. Ross drove great! Pete came through under pressure and kept the title of "fastest S2000" with one very quick final run.

But it looks like he's got some real competition now from another S2000. Something that's been overdue. It was fun to watch them no matter where the index places them.

What would have been less fun to watch, would have been if they had both driven slowly to trump the index system.....

[kyle.bowker]

How does it not make sense? The CDC index is based on how much each driver's performance improves over time. Ross improved more than you did, therefore the index rewarded him more.

[TubeDriver]

Kyle,

Two drivers in virtually identical cars score within a tenth of a second.

One gets ranked in 6th place.

The other driver (who beat the 6th place driver) gets ranked in 41st place. 41st place!


How does that make any sense? I feel like I am in bizarro world, if that actually makes any sense. It certainly does not make any sense from a sports/competitive standpoint.

So the CDC system rewards inconsistency? The primary difference between professionals racers and amateur drivers is that professionals are consistent and amateur are much less consistent. The CDC system REWARDS inconsistency (or I guess you can just not report your best times at certain events like our current champion did). This makes NO sense.

Kyle, let me be blunt. There are maybe 3-4 people (including the two CDC index designers) who actually like the current "system". Everyone else wants a change.

How does it not make sense? The CDC index is based on how much each driver's performance improves over time. Ross improved more than you did, therefore the index rewarded him more.

[kyle.bowker]

There are maybe 3-4 people (including the two CDC index designers) who actually like the current "system". Everyone else wants a change.

Ed is the sole designer of the CDC index system. I don't like or dislike the current index. To me, it is what it is. "Everyone" hasn't yet voiced an opinion so I don't think you can state with any authority that "everyone else wants a change." The point you raise about consistency is valid.

[TubeDriver]

Your right, I should have said something like "the vast majority of CDC members who care enough to post their opinion about the current CDC system appear to favor a change".

If you just accept the flawed premises of the current system then yes, the current system has some operational and internal consistency. It lacks external validity (a nice way of saying it has very little to do with reality).

In a related manner to the problem with consistency, I have to say that on further reflection, not reporting your times (for a competitive advantage) seems almost a step beyond typical sandbagging. Unless everyone can drop times from events in order to manipulate their Index?

There are maybe 3-4 people (including the two CDC index designers) who actually like the current "system". Everyone else wants a change.

Ed is the sole designer of the CDC index system. I don't like or dislike the current index. To me, it is what it is. "Everyone" hasn't yet voiced an opinion so I don't think you can state with any authority that "everyone else wants a change." The point you raise about consistency is valid.

[BugBomb]

I think we are straying further from the solution - as stated by the man himself.

On the practical note, I personnally think that getting the results done is a pain the butt as it is. If we go to another format, we need to figure out who is going to do the results. If Pete wants to be responsible to get the results done, then we can talk about changing the format.

Ed doesn't seem opposed to changing the format. He is opposed to being responsible for the new format.

[TubeDriver]

Getting results with the current system (when every score is dependent on the 3 previous scores) must be a pain in the ass. The sysem I propose would be much simpler to do.

I will be organizing 2 CDC HPDE events next year so I am NOT volunteering to do the scoring.

I already volunteered to handle all the "what class is my car and what is my index" e-mails.

Mike, Ed and kyle have not agreed to switch over to a new system but lets clarify:

Ed, is it that simple? CDC gets a new index system if someone steps up to the plate? because I think we have several folks willing to do this.

I think we are straying further from the solution - as stated by the man himself.

On the practical note, I personnally think that getting the results done is a pain the butt as it is. If we go to another format, we need to figure out who is going to do the results. If Pete wants to be responsible to get the results done, then we can talk about changing the format.

Ed doesn't seem opposed to changing the format. He is opposed to being responsible for the new format.

[ButtDyno]

Yeah, IMHO it's pretty simple. The registration form already asks for your PAX score. We could either keep that or replace it with a select box that asks for your SCCA class (with a textarea for listing your mods if you're not sure).

At that point all you need are the person's times and pouf, instant results - no going back and looking at the last 3 events (or more than 3 if the person has been MIA, right?)

I'd be happy to help

john

[kyle.bowker]

It lacks external validity (a nice way of saying it has very little to do with reality).

To say the CDC index is out of touch with reality is disingenuous. If by external validity you mean SCCA, who says we need to adopt their approach? PAX is a rather arbitrary number meant to provide some level of parity amongst cars of vastly different design, construction, and performance capability. Its success in providing a "level playing field" varies from one event to another. In reality, there is no such thing as a level playing field in racing. That is unless everyone has the same amount of time, money, and skill and is running the same tamper-proof spec car that is prepared to the same standard by the same people.

In a related manner to the problem with consistency, I have to say that on further reflection, not reporting your times (for a competitive advantage) seems almost a step beyond typical sandbagging. Unless everyone can drop times from events in order to manipulate their Index?

We take members at their word. If someone says they were off course, then they were off course. One way to solve the issue of inconsistent performance resulting in a favorable index is to allow the index to move in only one direction.