Elimination Race Methods - Pinewood Derby Times Article

[gpraceman]

So what did you all think of the Elimination Methods article in the latest issue of MaxV's Pinewood Derby Times, Volume 3 Issue 7?

Here's the comments from an email that I sent to MaxV (I don't know who wrote the article though) about a couple of points in there that I had to dispute:

Shorter Event Time - With an elimination method, the total number of heats is generally less than with a rotational method. Therefore, with efficient race staging, the overall length of the event can be shorter.

I’ll have to dispute this point, and I think it is a common misconception. A typical double elimination chart will produce nearly twice the number of heats that a Perfect-N Type or Lane Rotation chart would for the same number of racers. If you have a 4 lane track and race two of the brackets at once, then you would end up with about the same number of heats. The statement may be true for some race methods like Stearns that create a lot of heats (which is probably where the misconception originated), but not others. Even with running on a two lane track, you can run twice per lane with PPN, PN, CPN or LR charts and still have about the same number of heats of the corresponding DE chart, and each racer will race 4 times instead of many going two and out.

Racers = 16
DE Heats = 31
PN, PPN, CPN or LR Heats = 16

Racers = 64
DE Heats = 125
PN, PPN, CPN or LR Heats = 64

Easy to Stage - Several methods exist for staging elimination races either manually or with a computer. The methods are generally easy to implement, thus minimizing pre-race effort.

The same could be said for many non-elimination methods. I have heard from a lot of people that get confused about all of those brackets, who should go where, and trying to draw it all out big enough so the crowd can see. With a non-elimination chart, you end up with a simple grid of lanes versus heats. Many of these can easily be generated with software and posted on a wall or projected on an overhead (try doing that with double elimination chart of any significant size). Since these charts can be generated ahead of time, racers can know which heats and lanes they will be racing in, which can help get them to the line faster.


So what do you all think about the points in the article?

[Darin McGrew]

Shorter Event Time - With an elimination method, the total number of heats is generally less than with a rotational method. Therefore, with efficient race staging, the overall length of the event can be shorter.

When we switched to using a PPN-like schedule, we doubled the number of races each car was guaranteed to have, and reduced the time required to run the event slightly. But before that, we had been using a simple lane-rotation system though, not an elimination schedule.

The simple lane-rotation system (cars 1, 2, 3, and 4 race, then cars 2, 3, 4, and 5 race, etc.) has a number of disadvantages. There is no way to stage races in advance, since 3 of the 4 cars for the next race are in the current race. (This could be addressed with a little creativity, but if you're doing that, then you might as well go with a PPN-like schedule.) And cars that happen to register immediately before/after the fastest car are penalized since they will race against these cars 3 times. Its main advantage was that it was easy to score with nothing more than an overhead projector and a couple transparency sheets.

But back to the topic of elimination methods, I think the only way to get a significant speed advantage with an elimination method would be to limit the number of losses required for elimination, which limits how many places can be determined with any accuracy. We award 1st through 4th place, which would require quadruple elimination.

Easy to Stage - Several methods exist for staging elimination races either manually or with a computer. The methods are generally easy to implement, thus minimizing pre-race effort.

Pre-race effort isn't that much of an issue. I'd gladly invest a little more time before the race if it helps the race itself run more smoothly.

With our PPN-like schedule, we know at the beginning who is racing against whom for the duration of the event. Except for a couple tie-breakers at the very end, everything is set. We don't have to figure out who won an earlier race to know which car goes in which lane for the next race.

I just don't see a significant advantage to scheduled elimination methods. Our PPN-like schedule guarantees twice as many races to each car as a quadruple-elimination schedule would, and determines the top places more accurately. And it isn't any slower than a quadruple-elimination schedule, and there is less actual scheduling work during the race.

[Darin McGrew]

I just read the article and found a third "advantage" listed:

Facilitates Human Judging - While electronic finish lines are very valuable, races can be reasonably staged with human judges since only first place (or possibly first and second place) in each heat must be determined.

There is nothing that says that non-elimination methods need results that go beyond first place. You could easily run a non-elimination schedule and count only first-place finishes. So instead of top-placing results that look like 11111111, 11112111, 11311111, 11121121, etc., you'd have top-placing results that look like 11111111, 1111-111, 11-11111, 111-11-1, etc. You'd have a few more ties to break at the end, but that's about it.