Vote: Heads-UP in S/SS

[Ed Wright]

Originally Posted by Jack Matyas

At Englishtown there were probably 20 cars that could have contended for the #1 spot and most chose not to -- simply because of the AHFS . We really need to lower the indexes and get everyone racing to the finish line again .

I agree. Many Stock guys don't like the idea, I don't have a dog in that fight anyway. I sure think the SS indexes need lowered, 1/2 sec anyway. If you can't run 1/2 under in SS, you ain't tryin'.

[Lynn A McCarty]

AHFS if done the right way with a little science, sandbagging wouldnt matter. The trigger method for running fast would be immaterial. The only thing that would matter would be the curve for the whole engine. The only exception I could think of would be an obsure data point for an engine with no "like" data points.

Statistically, the slow cars and the average cars are more important in a true calculation. The really fast cars would only raise a red flag if they are way above the group of the other fast cars. These could be evaluated on a case by case basis for cheating, innovation, money spent, engineering prowess, unfair parts allocation. If you cant find anything wrong with what he is doing, then you must let him go.

The proper statistics bear out if a combo is soft on its original factoring which is the true shame of the current system. There is nothing wrong with a fast car. It is the American way. If the average of the combo is normal, and a guy is fast, unless you can catch him cheating, he should win. This isnt China.

[Ed Wright]

Originally Posted by Lynn A McCarty

AHFS if done the right way with a little science, sandbagging wouldnt matter. The trigger method for running fast would be immaterial. The only thing that would matter would be the curve for the whole engine. The only exception I could think of would be an obsure data point for an engine with no "like" data points.

Statistically, the slow cars and the average cars are more important in a true calculation. The really fast cars would only raise a red flag if they are way above the group of the other fast cars. These could be evaluated on a case by case basis for cheating, innovation, money spent, engineering prowess, unfair parts allocation. If you cant find anything wrong with what he is doing, then you must let him go.

The proper statistics bear out if a combo is soft on its original factoring which is the true shame of the current system. There is nothing wrong with a fast car. It is the American way. If the average of the combo is normal, and a guy is fast, unless you can catch him cheating, he should win. This isnt China.

Well stated.

[Michael Beard]

Originally Posted by Lynn A McCarty

AHFS if done the right way with a little science, sandbagging wouldnt matter. The trigger method for running fast would be immaterial. The only thing that would matter would be the curve for the whole engine. The only exception I could think of would be an obsure data point for an engine with no "like" data points.

Statistically, the slow cars and the average cars are more important in a true calculation. The really fast cars would only raise a red flag if they are way above the group of the other fast cars. These could be evaluated on a case by case basis for cheating, innovation, money spent, engineering prowess, unfair parts allocation. If you cant find anything wrong with what he is doing, then you must let him go.

The proper statistics bear out if a combo is soft on its original factoring which is the true shame of the current system. There is nothing wrong with a fast car. It is the American way. If the average of the combo is normal, and a guy is fast, unless you can catch him cheating, he should win. This isnt China.

Makes sense to me.

[Don Kennedy]

Lynn : I have forgotten you great version on how to do the AFHS the correct way . Could you again explain it to all of us so we can all be on the same page .

[Tom P]

The SS Hemi Shootouts they had a Mission for several years ran those cars all together at the end of the qualifying session. It was a good show, the classes have lots of appeal if it was marketed correctly. I think things like YouTube might help spread the appeal if the people watching can figure out where they can see it in person.

[John Quinn]

Originally Posted by Don Kennedy

Lynn : I have forgotten you great version on how to do the AFHS the correct way . Could you again explain it to all of us so we can all be on the same page .

I don't remember Lynn's formula but I think it uses a bell curve. If not it should, if you put all the runs of one type of car into a bell curve the index can be the mid point of the curve. If there are several cars in the class you don't have to worry about the fast ones breaking the index too hard. If there are just a few cars of your type, you will have to protect the index the way it is now so nothing is lost.

It would work really well in Comp, they could run all out without getting those stupid index adjustments. Maybe racing can get back to the concept of running hard to win and leave driving the stripe to bracket racing where it belongs.

[Tony Curcio]

Lyn McCarty has been talking about this formula for several years. This is the first time I ever saw more than one person agree. It really is the way an actuary or a true statistician would go about this.

Once the formulae are entered, it takes no more effort to use it than the current system. It would, however, be a little harder for racers to know when it is time to "save" a combination by running slower at a national event.

Lyn- do you have a working spreadsheet that can be submitted as a model to the Sportsman Council?

[Lynn A McCarty]

Everyone has the software on their computers if they have Excel. You data entry all the runs into a statistical bell curve. First you would have to clean up the data so that it is all meaningful. It doesn't have to be perfect.

1.) Correct all runs to the same correction factor for Temp/Press/Humidity
2.) Throw out all triggers (since they are meaningless for the new data)
3.) Organize "like" engines. If there is an argument of the "likeness" then compare them separately to the group.
4.) Run a statistical curve for all of Super Stock and one for all of Stock
5.) Compare each group to the overall curve.

We have a saying for Continuous Improvement. Once you get the data into a reasonable form, you Plan-Do-Check-Act. Just because it wont be perfect is not a reason not to improve what you are doing. This is why they call it "tend analysis" We are looking for trends with a scientific basis, not perfection. A logical thinking human still must make the calls. This gives him a much better tool to make intelligent judgments.

John is correct Comp needs it just as bad or worse than us. Run a heads up class that if you run too fast you automatically get factored? What if they did that in funny car or Pro Stock? It is ridiculous.

Tony is right, it would be more difficult to start, but once written it would be so much easier on NHRA, Racers, and overall fairness. NHRA has an IT department that I am sure could implement mechanics of this easily. First we would must accumulate reasonable data. You guys all know comparing runs corrected only with actual altitude or runs at 1500 feet below sea level is unworkable. So lets stop doing silly things like that.

[Lynn A McCarty]

http://www.qualityamerica.com/knowle.../CQEIVH3e.html

http://www.qualityamerica.com/knowle...C_Concepts.htm

http://www.qualityamerica.com/knowle..._Run_Tests.htm