Advancing to finals - how many do you do??

[Stan Pope]

We only have 23 scouts and that's if they all race. We are doing point this year with a preliminary and a finals. We will give away 4 trophies. How many should we advance to the finals - 8? Using the gprm software.
...
But I am still a little confused on how many to send to the finals, any help is appreciated.

For points racing, the chart used in finals should be of the PN or CPN variety. On a 3 or 4-lane track this is 3 or 4 or 7 or 13 racers.

Pick the number that give you good confidence that those deserving a trophy make the finals even if they get a worse draw during the preliminaries. Usually about twice as many finalists as trophies gets them all in. For 4 trophies, I'd run 7 finalists.

If your preliminaries are den by den instead of one big competition, then the picture is different. You have to go deeper into each den to have the same assurance.

[derbyspeed]

I was thinking along the same lines - to keep it at an even number so all 4 lanes would be racing.

Thanks, Mike

[derbyspeed]

Yes we just do it as a total pack race even though we keep track with the software as to who goes the farthest in each den. We have a 4-lane track this year, I always thought going with 8 finalist would work but you are saying to get better results we should go with maybe 7 or 13.

Thanks, Mike

[Stan Pope]

Yes we just do it as a total pack race even though we keep track with the software as to who goes the farthest in each den. We have a 4-lane track this year, I always thought going with 8 finalist would work but you are saying to get better results we should go with maybe 7 or 13.

Thanks, Mike

If your management software uses PPN or Stearns charts, then you get mostly charts that satisfy PPN or Stearns criteria. There is a subset of PPN charts that satisfy the more demanding PN and still more demanding CPN criteria. Those add more precision to racing outcome by producing exact opposition balance! Those charts only exist for a few car counts for each track width. 7 and 13 are such counts!

[derbyspeed]

Great! Thanks so much for your time Stan! This is my first year running the race and I wanted to make sure it was as accurate as could be with the points method.

Mike

[Stan Pope]

Great! Thanks so much for your time Stan! This is my first year running the race and I wanted to make sure it was as accurate as could be with the points method.

Mike

Good plan!

PPN schedules are slightly less accurate than PN and CPN schedules. The reason is that there is a limited imbalance in opponents. Consequently, I prefer to use them as screening for finals rather than as final ranking.

With screening, you just want to have high accuracy at getting the fastest cars among the finalists. If you want to ultimately find the 4 fastest cars, you want a number of finalists that will, with high probability, include the four fastest. If you projote just 4 from a PPN chart the probability is lower than if you promote 5, etc. Gradually, this probability climbs to or near 100%, typically at about twice the number of finalists as trophies.

If you want to see some examples of applying these concepts, look at http://members.aol.com/standcmr/pwmevalc.html. Both Stearns and PPN charts are used there. Latter examples apply the screening and finals concept.