How do you define being the "Fastest"?

[gpraceman]

Consistency and staying power play different roles in each. So, you have to decide first what you mean by "fastest".

I thought this was worth a thread of its own. How do you define being the "fastest'? Is it the lowest cumulative time? Lowest average time? Is it the single fastest time? How about median time? or Most wins?

For our races, it has been the lowest cumulative time, which generally means that the fastest cars needed to run consistency throughout the course of the race.

So, what is your definition?

[Darin McGrew]

How do you define being the "fastest'?

Our track doesn't have a timer. It has a finish gate that gives us first, second, third, and fourth place.

We schedule cars in an "approximate round robin" schedule (called "chaotic rotation" on the "race methods" page of my personal web site). Every car races eight times: twice in each lane, and against as many different opponents as possible.

We assign 1 point for first, 2 points for second, etc., and then we use golf-style scoring: The low score (usually 8 points, for 8 first place finishes) wins, and so on. Ties are broken with pairs of run-off races, swapping lanes for the second race.

If we were to use a timer, then I'd want to drop the extreme times for each car and compare the sums of the remaining times. But I'd still use the same kind of schedule, so as many kids would get as many races as possible.

But I think a better improvement would be to use the "approximate round robin" schedule as a qualifying round. Then, if we're awarding the top n places, we'd move the top 2n cars into the finals.

[Da Graphite Kid]

In the past our Pack has used double elimination (stop cringing Stan!) due to our two lane finish line that only shows which car was first to finish, so we defined the fastest car as the one with the most wins (or fewest losses). This year we will be using a finish line that can give us run times off of all three lanes - but we will still define the fastest as the one with the most wins. As I feel that there are too many variables that the boys have no control of that could contribute to a faster or slower run time for each pwd car. Besides, they eliminate cars at the next level up (District) based on a points elimination system and so they also award by the most wins. Keeping a commonality between Pack level and District level races is an important consideration and should be done where practical/possible.

I just answered that with the assumption (uh oh! ) that what you meant by "fastest" was the one pinewood derby car that is awarded a 1st place trophy, in a speed classification, for a certain rank or level. Starting this year, since we will have a finish line capable of doing so, we will also be making an award to the pwd car that had the single fastest run down the track.

Da Graphite Kid

[Stan Pope]

I just answered that with the assumption (uh oh! ) that what you meant by "fastest" was the one pinewood derby car that is awarded a 1st place trophy, in a speed classification, for a certain rank or level. Starting this year, since we will have a finish line capable of doing so, we will also be making an award to the pwd car that had the single fastest run down the track.

Da Graphite Kid

Yes, that was the intent of the question. I think that it is really neat when your competition method awards the first place trophy to the car which best satisfies your concept of "fastest".

[gpraceman]

If we were to use a timer, then I'd want to drop the extreme times for each car and compare the sums of the remaining times. But I'd still use the same kind of schedule, so as many kids would get as many races as possible.

This was one option I didn't list, but had heard before. My question would be, why drop the fastest and slowest times, or even just the slowest time? Is it to try to remove any influence beyond the control of the racer or just to make up for a racer having bad heat?

To me, I find it hard to just throw out data, without specific justification. If you look at a race as an experiment, good experimental procedure (or at least what I was taught) would have you keep all of the outliers, unless you can specifically isolate why those points were "bad". Otherwise, you could alter the true results of the experiment (race) because telling data was left out. There is value in investigating what caused those outliers, to improve future "experiments".

[Darin McGrew]

My question would be, why drop the fastest and slowest times, or even just the slowest time? Is it to try to remove any influence beyond the control of the racer or just to make up for a racer having bad heat?

Basically, it would be an attempt to get "typical times" and to ignore anomalies.

To me, I find it hard to just throw out data, without specific justification. If you look at a race as an experiment, good experimental procedure (or at least what I was taught) would have you keep all of the outliers, unless you can specifically isolate why those points were "bad". Otherwise, you could alter the true results of the experiment (race) because telling data was left out. There is value in investigating what caused those outliers, to improve future "experiments".

In a true experiment, we'd identify the source of the outlying data points. If they were produced by the car design, then we'd keep them. If they were produced by external factors, then we'd control for them.

In a derby, there are other priorities. A perfectly fair system is useless if the event becomes boring. The nature of most derby lubricants makes excessive rerunning of heats undesirable.

[PWTom]

Our Pack has an electronic finish gate that determines the order of finish, but not the elapsed time. We use a system like Darin's with lowest scores taking home the hardware. Ties are settled with one-on-one races in altermating lanes.

[Jungle Jim]

Most Packs in our Council do not have any type of e.t. timing device. Most have the light system and as such losers get a "punch" in their card until eliminated or race conclusion. A few years back a pc-based timing device was implemented at a Council race and it was "unsuccessful".

[gpraceman]

Most Packs in our Council do not have any type of e.t. timing device. Most have the light system and as such losers get a "punch" in their card until eliminated or race conclusion. A few years back a pc-based timing device was implemented at a Council race and it was "unsuccessful".

Seems like a but of a brutal system to a little cubs esteem.

What was it about using the timing system made it "unsuccessful"? Was it not accurate? or were people more used to the "old way"?

[Jungle Jim]

Most Packs in our Council do not have any type of e.t. timing device. Most have the light system and as such losers get a "punch" in their card until eliminated or race conclusion. A few years back a pc-based timing device was implemented at a Council race and it was "unsuccessful".

Seems like a but of a brutal system to a little cubs esteem.

I know that in the Packs I assist they give out a variety of awards and as such it helps to keep the kids into all that is going on. Of course there are kids (and regrettably parents) who no matter what will be upset. How "brutal" it is depends a lot - IMO - upon the parents.

What was it about using the timing system made it "unsuccessful"? Was it not accurate? or were people more used to the "old way"?

I would say a combination of things. The system was one in which everybody raced down the track 3 times for your best time. The results were kept on a PC and no one except the guy running it could see any results until the race was done. Upon conclusion, the results would be announced and that was that. No Head-to-Head racing. Nothing reaffirmed. And a lot of suspicion. Overall I heard more gripes at that race than any other. I'm not saying there aren't better ways then the ones being used, but sometimes it's hard to get people to change.

[Darin McGrew]

What was it about using the timing system made it "unsuccessful"? Was it not accurate? or were people more used to the "old way"?

I would say a combination of things. The system was one in which everybody raced down the track 3 times for your best time. The results were kept on a PC and no one except the guy running it could see any results until the race was done. Upon conclusion, the results would be announced and that was that. No Head-to-Head racing. Nothing reaffirmed. And a lot of suspicion. Overall I heard more gripes at that race than any other. I'm not saying there aren't better ways then the ones being used, but sometimes it's hard to get people to change.

Wow... I can see why the system was unpopular.

As I said before, absolute fairness isn't the only criterion for a derby system. People want to see their cars race against other cars, and they want to see their own cars race more than 3 times. And it would sure help to have some clue how their car is doing during the whole process.

[ExtremePWD]

We just confirmed tonight at our committee meeting that we are switching from a double elimination (gag) to a combination partial perfect / perfect N format. We will be using points even though we will have timers. What was unanimous was that we need to make the scoring visible so that people could track how individuals were doing. This was obvious before when you had a big double elim chart on the wall. Without a method of tracking progress along the way the event becomes a large collection of single events with no apparent interaction and then magically at the end a winner emerges. I can imagine how this would quickly loose the crowds interest. Any sporting event I can imagine provides some type of visual measure of the relative position of the competitors throughout the competition.

[Stan Pope]

What was unanimous was that we need to make the scoring visible so that people could track how individuals were doing. This was obvious before when you had a big double elim chart on the wall. Without a method of tracking progress along the way the event becomes a large collection of single events with no apparent interaction and then magically at the end a winner emerges.

The three approaches to display that i know of are making big charts to hang on the wall, projecting transparencies (where you can mark on the transparency), or using a computer managed chart and projecting the computer image. Do you have another solution? What are you planning to do?

[gpraceman]

What was it about using the timing system made it "unsuccessful"? Was it not accurate? or were people more used to the "old way"?

I would say a combination of things. The system was one in which everybody raced down the track 3 times for your best time. The results were kept on a PC and no one except the guy running it could see any results until the race was done. Upon conclusion, the results would be announced and that was that. No Head-to-Head racing. Nothing reaffirmed. And a lot of suspicion. Overall I heard more gripes at that race than any other. I'm not saying there aren't better ways then the ones being used, but sometimes it's hard to get people to change.

Wow... I can see why the system was unpopular.

As I said before, absolute fairness isn't the only criterion for a derby system. People want to see their cars race against other cars, and they want to see their own cars race more than 3 times. And it would sure help to have some clue how their car is doing during the whole process.

I do believe that if you are using a timing system, you must either have a timer that will display the finish order or times per lane, or use a computer with a large TV or projection system to display the real-time results (my preference), or post the results after each heat (overhead projector or a large piece of poster paper). The crowd must see what the timing system, software and computer operator (if used) are doing; otherwise people will be suspicious of the the technology and/or the operator.

[ExtremePWD]

What was unanimous was that we need to make the scoring visible so that people could track how individuals were doing. This was obvious before when you had a big double elim chart on the wall. Without a method of tracking progress along the way the event becomes a large collection of single events with no apparent interaction and then magically at the end a winner emerges.

The three approaches to display that i know of are making big charts to hang on the wall, projecting transparencies (where you can mark on the transparency), or using a computer managed chart and projecting the computer image. Do you have another solution? What are you planning to do?

We will be using a computer with race management software projected on a screen on the gym wall. If we didn't have access to the equipment then I would possibly try using a 4x8 dry erase whiteboard that I have in my basement. These can be made fairly inexpensively with a sheet of white melamine panelling and permanently marked with a grid if desired.