Help the New Guy - Part I

Debates and discussions on the various race scheduling methods that can be used and their fairness and accuracy in determining the winners.
ronin718
Journeyman
Journeyman
Posts: 40
Joined: Tue Sep 13, 2005 11:39 am
Location: Sterling, VA

Re: Help the New Guy - Part I

Post by ronin718 »

Thanks, Stan, for the compliment. I try to do what I can to help the boys enjoy the race.

The problem I see with the elimination method is the same that many other posters to numerous forums have mentioned --- the potential for eliminating a "high-quality" car due to a bad draw. Bad luck could end up taking a boy out after one or two races, and where's the fun in that. At least with the "one run, each lane" method, the boys get at least as many races as there are lanes. That seems more fair and more fun.

Just my $0.02........
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Help the New Guy - Part I

Post by Stan Pope »

ronin718 wrote:The problem I see with the elimination method is the same that many other posters to numerous forums have mentioned --- the potential for eliminating a "high-quality" car due to a bad draw. Bad luck could end up taking a boy out after one or two races, and where's the fun in that. At least with the "one run, each lane" method, the boys get at least as many races as there are lanes. That seems more fair and more fun.
"The elimination method" is not just single-elimination or double-elimination. With larger groups, the elimination critera can be expanded easily to 5 or so losses. And, if more than two lanes are available, the definition of win and lose can be changed, e.g. if there are 4 cars running, the first two across the finish line might be designated as "winning the heat."

In the end, it boils down to how long do the youngsters (and their parents) want to race? Then, how much racing can you pack into that amount of time. Note that method may have some influence on attention span ... Scouts "keep their heads in it" longer if they race their own cars than if they watch track staff race their cars and if their cars are "better performing."

Factors that affect the heats per hour that you can achieve include race method, who handles the cars (owner or track staff), and the skill of the track staff.

In my experience, no-chart multiple elimination with the boys racing their own cars maintains interest and achieves good heats/hour rates with only modest track staff skills. For example, running quintuple (5 loss) elimination among 50 to 70 Cubs who race their own cars is typically completed in under 3 hours at our district races. Lesser cars race at least 5 contested heats and "are in the running" for at least the first 2 hours. Attention span is an issue, but managing the racers by groups rather than individually helps avoid delays and allows Scouts to "take a break" between their heats without risk of being AWOL for their next heat.
Stan
"If it's not for the boys, it's for the birds!"
Nooby
Pine Head
Pine Head
Posts: 99
Joined: Wed Aug 31, 2005 8:55 am
Location: Southwest, MI

Re: Help the New Guy - Part I

Post by Nooby »

Sorry, I have mastered getting quotes in a reply - so I'm going manually copy your reply.


Stan wrote:

"The nochart multiple elim format does not work well for a small number of racers (around den size), since each score group very rapidly reduces to a single member. Heats matching racers from adjacent score groups messes up the accuracy of the method."

That's why I proposed only using the first two rounds from the den race to
continue in the pack competition. Any selection of first round opponents should start out relatively random, with secessive rounds becoming less and less random. If the scores of 4 rounds of den races were carried into a 5 round pack race, the final round could easily be mismatched.

However, selecting the first 2 rounds only within a den, should not produce final results that are much out of line with with 2 rounds of random matchups. The only exception I can think of would occur in a den having an extremely small number of scouts - say 3 or less with a 3 lane track, should produce second round data that is essentially equivalent to the first round data.


Stan also wrote:

"Elimination methods also have accuracy problems when significant lane differences are present. PPN, on the other hand, tends to push final scores closer together when there are significant lane differences"

Luckily, from my observations last year, lane differneces don't seem to be a concern. Also, with any scheduling, we will endeavor to make sure that each car races in all lanes.

It seems to me if the lanes are clearly unequal, then the selection a winner is always going to be a problem.
"Nooby"
User avatar
gpraceman
Site Admin
Site Admin
Posts: 4926
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: Help the New Guy - Part I

Post by gpraceman »

Nooby wrote:Sorry, I have mastered getting quotes in a reply - so I'm going manually copy your reply.
Just click on the "quote" button at the upper right hand corner of the post that you are replying to.
Randy Lisano
Romans 5:8

Awana Grand Prix and Pinewood Derby racing - Where a child, an adult and a small block of wood combine for a lot of fun and memories.
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Help the New Guy - Part I

Post by Stan Pope »

Nooby wrote: ... I proposed only using the first two rounds from the den race to continue in the pack competition. Any selection of first round opponents should start out relatively random, with secessive rounds becoming less and less random.

However, selecting the first 2 rounds only within a den, should not produce final results that are much out of line with with 2 rounds of random matchups. The only exception I can think of would occur in a den having an extremely small number of scouts - say 3 or less with a 3 lane track, should produce second round data that is essentially equivalent to the first round data.
If the pack race were an elimination method, then including separate den results into the final might not skew the results very much. However, if the pack race is to be a "final standings" ("points") competition (which is what I think you are aiming at), then the results will be skewed by the necessarily unequal competition.

When formulating points or final standings competition, it is important that the following two attributes be as nearly equal as possible:
1. equal runs in each lane, and
2. equal runs against each other possible opponent.
To the extent there is inequity, the accuracy of results increasingly suffers.

If I'm wrong about how you plan to run the pack competition, and you are going to finish out a quadruple or quintuple elimination by using the first two rounds of den races to initially populate the score groups, then you are still introducing inaccuracies into the results. They are probably less than with final standings, tho.
Stan
"If it's not for the boys, it's for the birds!"
Nooby
Pine Head
Pine Head
Posts: 99
Joined: Wed Aug 31, 2005 8:55 am
Location: Southwest, MI

Re: Help the New Guy - Part I

Post by Nooby »

First, I'm not really sure how we will be running the race. This forum has been invaluable in raising issues that I would not have thought up on my own or within our pack meetings.

I knew that double elimintation was bad - especially for determining first, second, and third place as we did last year.

My "feel" is that I like no chart elimination (4 or 5 rounds). Originally, I wanted some sort of points standing, but I've been won over by the simplicty of the no-chart method. Since this year is my first year running the derby, simplicity is my friend.

You and others pointed out that no chart doesn't lend itself to determining den champions.

I'm trying to find a way to run no-chart elimination, which still determines den champions - more trophies = more happy scouts.

So, I think you are saying that the final results (first, second, third) should not be greatly effected by restricting the first 2 rounds to den races only. But, if I want some type of 1-70 ranking, that could be impacted by this method.

I think that is going to be solved by not focusing on a top to bottom ranking (as I initially wanted). Although, such rankings could be fun for "bragging rights" it seems to complicate the race system quite a bit, without improving the accuracy of determining the top finishers.
"Nooby"
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Help the New Guy - Part I

Post by Stan Pope »

All of the formal accuracy evaluations that I have done relate to the "trophy places," typically first through third, fourth, or fifth. The elimination procedure that you laid out, seeding based on two rounds of den races, will probably do a good job on first (almost every method does that), but will become increasingly problematic through the remainder of the trophy places. I can't quantify the inaccuracy without running a simulation and the method hasn't been modeled in any of the sim software that I have access to. (BTW, Cory Young has a nice accuracy sim program for several types of racing.)

Few methods will do a very good job of ranking the entire group. Times may do that. CPN will do a fair job. But, someone (I think that it was Auntie Beans) once offered the observation that the guys in the bottom half probably don't want to know just exactly how bad their car is. (Those were her exact words, but that was the sense.) Anyway, there is value to those racers for there to be a lot of uncertainty as to which car is the slowest. I think she was right.
Stan
"If it's not for the boys, it's for the birds!"
Post Reply