The Case for a Good Track

A good racing method can overcome the uneven lanes on a poor track -- some of the time!

We interrupt this web page with the following late breaking news!!
As of January 14, 2003, Pack 146 now has a new state-of-the-art track! We decided to go for the aluminum plating from Beta Crafts, and a MicroWizard timer with the finish order display.
(We really liked our previous timer from New Directions. It performed flawlessly for six years. One time a car actually went airborne and crashed into the timer's top cross piece, hitting it wheel-first. The collision actually bent the axle about 15 degrees! The car was never the same after that, but the timer continued to work perfectly. Alas, we finally felt the need for a digital display.)
Anyway, our track came out as good as we'd hoped! It's almost 50 feet in total length, and the actual racing distance is about 83% longer than that of our previous track, which is the one from the Cub Scout Leader How-To-Book. Set up and take down are bigger jobs now, as is storage, but it's definitely worth it.
We created a variation of the Challenger design. The construction of the support frames was almost identical to the plans from Beta Crafts, but we opted for a longer hill, followed by a shorter runout. Thus, our hill is not as steep and our transition is more gentle as compared to the Challenger. The ratio of hill to runout is very similar to that of the How-To-Book track. The starting gate is similar in design to that of the How-To-Book, but with some improvements. And again borrowing from the How-To-Book design, we created drop-down legs which allow for easier storage.
The best authority I've found so far on track building is Larry Bosworth's site, but having now been part of a track building effort, I would offer the following general comments about track construction:

Use the best tools you can. Table saws, mitre saws, drill presses...I very much doubt that our finished product would have been as good without these "high-end" tools.
Spare no expense. Don't cut any corners on wood or hardware.
Have some woodworking experts on your team. Being good at building PWD cars doesn't necessarily mean you're an expert, unfortunately. We had several good people on our team, thank goodness, and it shows in our end product.
Be as painstaking as you can stand to be. Measure twice to cut once, of course. If two pieces of wood need to be exactly the same length, then make certain that they are but cutting them together, instead of measuring separately and cutting separately. The more effort you put into being precise, the better your track will turn out!

And now, back to our previously scheduled web page...

A Complementary Perfect-N racing chart is one that satisfies the following conditions:

Every car races the same number of times in each lane.
Every car races every other car the same number of times.
Every race between two cars has a corresponding rematch where the cars switch lanes.

This is about as strict a set of criteria that can be used to run a set of Pinewood Derby races. Anything more would result in too many races to run in a reasonable time.
Intuitively, such a format would never put the fastest car in second place. Sure, maybe if the lanes were bad enough, the fastest car might lose a race to a slower car. But that race would have a associated lane-reversal rematch which would be won, of course, by the fastest car. Thus, the best the slower car could achieve is a tie. Right???
Wrong! Consider a hypothetical District finals, with conditions as follows:
There are four cars: A, B, C, and D.

Car A is 2 inches faster than Car B
Car B is 4 inches faster than Car C
Car C is 6 inches faster than Car D

There are three lanes: 1, 2, and 3

Lane 1 is 5 inches faster than Lane 2
Lane 2 is 3 inches faster than Lane 3

The following Complementary Perfect-N chart is used:
          Lane 1   Lane 2   Lane 3

Heat 1       A        B        C
Heat 2       B        A        D
Heat 3       D        C        B
Heat 4       C        D        A
Based on the relative lane and car speeds, and assuming no random elements, here are the race results:
           1st     2nd    3rd

Heat 1:     A       B      C
Heat 2:     B       A      D
Heat 3:     B       C      D
Heat 4:     C       A      D
Even though Car A is objectively the fastest, it finishes second behind Car B. How could this have happened using a Complementary Perfect-N Chart?
A careful examination shows that Car B's victory was due to fortuitous lane assignments. Car B won both races against Car C, because they raced in fairly even lanes. However, Car A lost one its races to Car C, because they raced in the most uneven pair of lanes.
Of the available racing methods, Complementary Perfect-N is one of the most accurate at correctly identifying the fastest cars. Yet, in this case, it was not enough to overcome the obvious lane inequties. For less accurate race methods, such as double-elimination and Stearns, the ill effects of lane inequities will be even more severe.
Regardless of the race method you use, be advised that identifying the fastest cars will be error-prone if the lanes of your track are "more uneven" than the cars that are racing. And at higher levels of racing (e.g. District or Council races, and probably a lot of Packs out there) the best cars will usually be very evenly matched.
In other words, if accuracy in determining winners is one of your priorities, you might want to check out the following links:

Darin McGrew has a nice page of Four-Lane Track Plans. He also has a section on tracks on his Derby Resources page.
Ron Morrison has several pages describing the various components of Pack 240's very impressive formica-covered track.
At his San Diego 500 site, Larry Bosworth has a large (and still-growing) section on tracks which includes information about the new aluminum track plating.
Don DeYoung has a number of track oriented links on his Pinewood Derby Tracks page.
For that final race between the top two cars, which two lanes of your multi-lane track should you use? This essay by Stan Pope helps you answer that question.

Last updated on January 27, 2003, 11:00 AM
Copyright 1998-2003 © by Cory Young. All rights reserved.