Your opinion please? Based on points Racers run across lanes

[Stan Pope]

With our PPN-like method, we schedule everyone to race once before anyone gets to race a second time.

This is a distinctive characteristic of Stearns Method. (Our PPN generator does not impose that artificial constraint, but does perform heat sorting to minimize consecutive heat appearances and equalize heat count during the competition.)

This may be astsray of the thread topic, but I'll ask anyway ... Within each "round", how do you decide on pairings and lane assignments?

[Darin McGrew]

Within each "round", how do you decide on pairings and lane assignments?

It's a fairly simplistic algorithm:

• For each race
  • For each lane
    • Find the "best" car for that lane of that race

There is no consideration for future races, or for future lanes in the current race. The "best" car is based on only previously scheduled races, and previously scheduled lanes of the current race. It is calculated as:

• No car can race against itself.
• All cars get the same number of total races.
• Prefer cars that have raced less in the current lane.
• Prefer cars that have raced less against the cars already assigned to this race.
• Prefer cars that were not in the previous race.
• Prefer cars that have had fewer races so far.
• Choose cars randomly.

I'm thinking we should change this to

• No car can race against itself.
• All cars get the same number of total races.
• All cars get the same number of races in each lane.
• Prefer cars that have raced less against the cars already assigned to this race.
• Prefer cars that were not in the previous race.
• Prefer cars that have raced less in the current lane.
• Prefer cars that have had fewer races so far.
• Choose cars randomly.

Another addition that I'm considering is to create several schedules, to rate the schedules for balance, and to choose the best schedule.

[Stan Pope]

Another addition that I'm considering is to create several schedules, to rate the schedules for balance, and to choose the best schedule.

Thanks, Darin.

You might also export the schedules for accuracy analysis in Cory's DerbySim program. It has some very useful measures. While you probably wouldn't want to do that for every race generated, you could use it to "prove" the algorithm used to "rate the schedules."

[racerock]

Actually, we've found that our PPN-like method is easier to run than the old lane-rotation method we were using. Lane-rotation has the advantage of "simplicity" only when you run the 2-3-4-5 race immediately after the 1-2-3-4 race. But if you run it that way, then you can't stage the cars for a race until most of the cars from the previous race have been returned to the start. If you don't run it that way, then it's really no different from PPN-like methods, except that its accuracy is poor.

With our PPN-like method, we schedule everyone to race once before anyone gets to race a second time. We don't have to wait for the kids in race n to return their cars to the start before staging race n+1. We get in about twice as many races with our PPN-like method than we did with the old lane-rotation method.

Thank you for writing this. The organizers of the race claimed that their method was the fastest, and "we'd be here all day if we raced the other method" (Stearns actually). After leaving their conversation and thinking about it more, discussed different staging methods using the Stearns, and two "impartial" people agreed that since you can have the "next race" already staged in a box top sitting next to the starting gate, ready to hand the starter even before the starting gate is released, you could effectively run a derby in the same time, if not quicker. You don't have to sit there and wait for the collection of the cars at the finish line, walk the track, and switch out cars...

Where is that whistling smilie, o.k., I'll use:

[Stan Pope]

Thank you for writing this. The organizers of the race claimed that their method was the fastest, and "we'd be here all day if we raced the other method" (Stearns actually). After leaving their conversation and thinking about it more, discussed different staging methods using the Stearns, and two "impartial" people agreed that since you can have the "next race" already staged in a box top sitting next to the starting gate, ready to hand the starter even before the starting gate is released, you could effectively run a derby in the same time, if not quicker. You don't have to sit there and wait for the collection of the cars at the finish line, walk the track, and switch out cars...

Where is that whistling smilie, o.k., I'll use:

Except for cars 1-5, each racer could be in and out and on their way home in the time it takes to run 6 consecutive heats. But that isn't a very good measure of the quality of the method. And, how many would leave so quickly?

Whether you have track staff stage the cars (yech!!!) or have the boys stage their own cars (Yay!), avoiding back-to-back heats allows a lot of overlap of function. We start staging cars on the starting line for heat 10 as soon as the cars from heat 9 leave the gate! (In our case, it is the boys bringing their own cars up to stage them, but the idea works either way.)

I think that you have a big education job ahead of you ... sounds like some simplistic ideas are very well entrenched in the group. How to accomplish change is the big issue, because it sounds as though any change will be resisted. Lotsa possible reasons: "This is the way we've always done it!" "It ain't broke! Why try to fix it?" "We've got a good program going and we don't want to mess it up."

Several years ago, our district races changed from using charted double elimination to no-chart quintuple elimination. That is a pretty major method change. The new plan had been "in my toolkit" for a couple years at the time, but the questions were "would it improve the program" and "if so, how to bring about the change." Ended up with application in a unit's racing (cubmaster called me a few days before about how to do his races), that unit's leader telling chairman about the experience, chairman expressing interest in trying on one of the age-group tracks. The next year, it went to all five tracks because the chairman liked it (less last minute work to prep the charts) and the boys and parents liked it (more heats in same time and more accurate results)! I wish that I could say that I planned and managed that change ... I didn't. But I helped it happen by supplying objective information at the right times to the right people.

If we can help you hone your ideas on alternatives or how to effect the change, get back with us!

[Cory]

Lotsa possible reasons: "This is the way we've always done it!" "It ain't broke! Why try to fix it?" "We've got a good program going and we don't want to mess it up."

Stan, you forgot that often present (but seldom spoken) reason: "Not invented here!".

Someone who is clever enough might even be able to make the Committee believe that whole thing was their own idea.

[Stan Pope]

Lotsa possible reasons: "This is the way we've always done it!" "It ain't broke! Why try to fix it?" "We've got a good program going and we don't want to mess it up."

Stan, you forgot that often present (but seldom spoken) reason: "Not invented here!".

Must have been some other recent post / email that I cited that one! Thought I had put it in this thread already and didn't want to be redundant again.

Someone who is clever enough might even be able to make the Committee believe that whole thing was their own idea.

Ah, yes, the "coup de grâce" delivered with such artistry that racerock would be awarded both ears and the tail to hang on his family room wall (after a visit to the taxidermist.) (Concept and some words gratefully borrowed from Tom Lehrer.)