Page 1 of 1

Determining Finalists for Pack Derby?

Posted: Tue Oct 14, 2014 8:38 am
by Vitamin K
So, I am looking to determine the best way to run our Pack's Pinewood Derby, and I would appreciate feedback on how to run the grand finals for the 'top in Pack' awards.

Here are the things we've determined thus far:

- We will run individual perfect-n schedules for each group ( each den is a group, plus a 'masters' class ). First, second and third place will be awarded for each group.

- Group size will be somewhere between 3-12 racers.

- All groups, save masters, will be eligible to qualify for the grand finals.

- The finals will be a perfect-n schedule for those that qualify.

- We are using GPRM for software.

So my question is: what is the best way to choose which racers go on to the final round? Should I just take the top two or three finishers in each group? Should group size determine how many racers to send to finals (top 25℅?)? I want to be efficient, but I also want to be fair.

Suggestions?

Re: Determining Finalists for Pack Derby?

Posted: Tue Oct 14, 2014 1:30 pm
by ngyoung
IIRC from when I was in a large pack that had multiple dens in each rank we took the top 3 from each den to race for rank champs. Then rank winners race for the grand pack champ. Then for districts all the pack champs race for district champ. They also hold an open scout race in round at districts to compete for a final spot to race with the pack champs.

Some of it kind of depends on how many lanes you use. All the packs in our district use the same 8 lane track so it can take a long time to get through each round when running 8 heats/car. For my son's pack which is small and we all race together and just have 1-3 winners for the pack. When we were at districts in was a pretty long event to get through and getting to the final 8 only really eliminated a few cars. I kind of wish they would have just took the top 4 or something to race in only the middle 4 lanes in the finals. I have seen some other events online with the 8 lane tracks and them only using a portion of it. Just recently I learned one of them they dedicated a lane to just a camera car to capture the cars going down the track.

Re: Determining Finalists for Pack Derby?

Posted: Tue Oct 14, 2014 1:38 pm
by Vitamin K
Top 3 from each group is the direction I'm leaning, since there are three awards to be won in the finals. That means that, theoretically, even if the three fastest cars are in the same group, they still will have an equal chance of winning the top prizes.

It does mean that cars in a particularly small group will get an automatic "pass" into the finals, but if they aren't up to snuff, they'll quickly fall out of the running. No actual benefit, but maybe a perceived benefit from somebody whose car might have been faster than those in a smaller group, but still slower than the top three in their group.
ngyoung wrote:IIRC from when I was in a large pack that had multiple dens in each rank we took the top 3 from each den to race for rank champs. Then rank winners race for the grand pack champ. Then for districts all the pack champs race for district champ. They also hold an open scout race in round at districts to compete for a final spot to race with the pack champs.

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 12:59 pm
by Stan Pope
You don't say if you are scoring by times or by points. It matters!

PPN Charts scored by points have good accuracy, but not absolute accuracy! And accuracy can vary depending on which cars get scheduled to which position in the chart and how many entrants in the group. Because of these considerations, I take more racers to the finals than just the top "nr of trophies". By simulation, I select the number to assure that 99% of the time, the truly fastest "nr of trophies" cars placed at least that well. For instance, in the simulation, if 90% of the time, the three fastest placed among the top 3, 94% of the time the three fastest placed among the top 4, and 99% of the time, the three fastest placed among the top 5, I'd advance 5 racers to the finals!

PPN Charts scored by times have good accuracy, but comparing separate group's results assumes that the track was invariant through the whole event. With Cub Scouts around, it is not unlikely that somewhere along the day, the track got jostled enough to change its characteristics by a few milliseconds. However, the results within a group should be accurate enough to cut off at exactly the top "nr of trophies."

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 1:22 pm
by birddog
Here's how we do it:

Top 2 fastest racers from each Den advance to finals. We generally have Tigers, Wolves, Bears, Webelows I and Webelows II dens, so that is a total of 10 racers advancing to the finals. Then, I also advance the "top 5 fastest cars" from the prelims. That ensures that even if the top 5 cars are all from one den, they get a fair shot at the 1-3 places we give out in the finals.

Birddog

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 1:27 pm
by Vitamin K
Ah, I was hoping to hear from you, Stan! Figured that you'd have some good thoughts on the matter.

So, sorry for my omission of data. We would be running the race by points. I actually didn't know PPN charts made any sense to use with timed racing. If you're just going by average times, it's pretty dead simple to know who to send to the finals...even if you even want to bother with a final round at all!

Is there some kind of algorithm that I can apply to determine how many cars to send to the finals? I want to make sure that the parents understand why I've chosen a particular cut-off for sending cars from any particular group to the final rounds. I also want to avoid sending too many cars, since if a large chunk go on to the finals, that's another large chunk of people feeling left out.
Stan Pope wrote:You don't say if you are scoring by times or by points. It matters!

PPN Charts scored by points have good accuracy, but not absolute accuracy! And accuracy can vary depending on which cars get scheduled to which position in the chart and how many entrants in the group. Because of these considerations, I take more racers to the finals than just the top "nr of trophies". By simulation, I select the number to assure that 99% of the time, the truly fastest "nr of trophies" cars placed at least that well. For instance, in the simulation, if 90% of the time, the three fastest placed among the top 3, 94% of the time the three fastest placed among the top 4, and 99% of the time, the three fastest placed among the top 5, I'd advance 5 racers to the finals!

PPN Charts scored by times have good accuracy, but comparing separate group's results assumes that the track was invariant through the whole event. With Cub Scouts around, it is not unlikely that somewhere along the day, the track got jostled enough to change its characteristics by a few milliseconds. However, the results within a group should be accurate enough to cut off at exactly the top "nr of trophies."

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 1:29 pm
by Vitamin K
By "fastest", should I assume that you're running your races by time, as opposed to by points? :thinking:
birddog wrote:Here's how we do it:

Top 2 fastest racers from each Den advance to finals. We generally have Tigers, Wolves, Bears, Webelows I and Webelows II dens, so that is a total of 10 racers advancing to the finals. Then, I also advance the "top 5 fastest cars" from the prelims. That ensures that even if the top 5 cars are all from one den, they get a fair shot at the 1-3 places we give out in the finals.

Birddog

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 1:31 pm
by birddog
Yes, we run our races by time, not points. I've never seen a PWD race run by points. I suspect that would be confusing to a lot of parents. Times are easy to understand.

Birddog

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 1:47 pm
by Vitamin K
Without touching off on a debate that's already been done in a few threads, I think that points is quite easy to understand, and makes the event an actual 'race', as opposed to a time comparison. It does, however, make it somewhat tricky when you want to compare racers from different groups against one another.
birddog wrote:Yes, we run our races by time, not points. I've never seen a PWD race run by points. I suspect that would be confusing to a lot of parents. Times are easy to understand.

Re: Determining Finalists for Pack Derby?

Posted: Wed Oct 15, 2014 2:43 pm
by gpraceman
Vitamin K wrote:It does, however, make it somewhat tricky when you want to compare racers from different groups against one another.
That is an apples to oranges comparison, since they do not compete directly with one another. Points are based on direct competition.

It all boils down to what you define "fastest" as. Least points? Lowest times?

Anyways, when it comes to advancing racers to a Grand Finals, it is really up to the race organizers if they want each race group represented equally or not. Most that run a Grand Finals do go for equal representation from the different race groups. That, of course, is more advantageous with smaller groups, but the trophies will still come down to which cars are more competitive.

Re: Determining Finalists for Pack Derby?

Posted: Fri Oct 17, 2014 2:05 am
by Darin McGrew
Vitamin K wrote:Top 3 from each group is the direction I'm leaning, since there are three awards to be won in the finals.
That's what I would go with.

We award only 1 prize for the fastest car overall, so we take only the 1st place car from each age group into the final "grand champion" race.

Re: Determining Finalists for Pack Derby?

Posted: Tue Dec 23, 2014 1:30 pm
by Nitro Dan
One of the Packs I work with actually takes a unique approach to this. They advance based on Points and Times. They take the top three winners based on points from each den rank and advance them to the finals. Then they go back and compare the times within all the den rank races and advance the total top 15 fastest cars to the finals. Now this may sound like you are advancing nearly 30 cars to your finals, but many of the cars that are the fastest times are also your lowest by points. In the seven years that they have been doing this method, the maximum that they have ever advance by times that did not already qualify by points was 4 cars (most years it was just an increase of 2 cars). This puts anywhere from 15 to 19 cars in your finals and helps equal out those den ranks that have a few racers in it as compared to those den ranks that might have a lot of racers in it. The finals are then ran by points.

-Nitro Dan