Multiple track racing

Discussions on race planning, preparations and how to run a "fair" and fun race.
User avatar
FatSebastian
Pine Head Legend
Pine Head Legend
Posts: 2646
Joined: Wed Jun 17, 2009 2:49 pm
Location: Boogerton, PA

Re: Multiple track racing

Post by FatSebastian » Thu Jun 09, 2011 11:47 am

Stan Pope wrote:All in all, I see the method as substantially reducing the percentage of time that the competitors wait for their turn to "get on the track and race."
Under what circumstances (if any) might a PPN-type approach provide as much or more racing time compared to a bubble-sort approach?



User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Thu Jun 09, 2011 3:03 pm

FatSebastian wrote:
Stan Pope wrote:All in all, I see the method as substantially reducing the percentage of time that the competitors wait for their turn to "get on the track and race."
Under what circumstances (if any) might a PPN-type approach provide as much or more racing time compared to a bubble-sort approach?
Great

question! Den-size groups, about 6 to 15 boys running a PPN style race on 3 lanes:

Boys ... %
6 ......... 50%
8 ......... 38%
10 ....... 30%
12 ....... 25%
14 ....... 21%

On 4 lane tracks the numbers are a bit better:

Boys ... %
6 ......... 67%
8 ......... 50%
10 ....... 40%
12 ....... 33%
14 ....... 28%

When wait time approaches 50% on a PPN chart, it is likely that one or more cars are racing in successive heats. This usually slows progress, and the event seems to drag.

The central problem with racing large groups is that only 3 or 4 can race on one track at a time. (More lanes, after 4 or 5 or 6, doesn't really help much because it lengthens the time to run an individual heat.) To get more boys actively racing at a time simply means that more lanes and tracks are required. 15th Burlington provides a manageable method to apply an arbitrarily large number of tracks. (There is a limit, other than the size of the gymnasium they would be set up in... For every additional track added to the line-up, the event duration needs to be lengthened by enough time for each racer to make about two more runs so that the distribution of boys on the tracks reaches "equilibrium".)

To me, the optimal ratio of boys to tracks is in the range of 7 to 10, which puts the active % in the range of 30% to 42%. Racing is not held up while racers return from the finish line to the starting line, 7 is the smallest number for which a 4-token lane draw works without delay, and the 4-token draw is necessary to produce some mixing (but not too much) between groups of three in the line of waiting racers.

If tracks were to run "in sync" and each track had a multiple of 3 racers assigned and no provision for randomizing were applied, then each racer could only race a fraction of the total number of competitors. For example, the first boys in line on each track could not race anyone who wasn't among the first three. Comparisons in speed between the groups of 3 wouldn't happen. The page includes discussion of the phenomenum and suggests ways to avoid. Letting tracks 'run at their own rate" is a good start, and 4-token draw helps the process.


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Wed Jun 15, 2011 12:55 pm

15th Burlington (Bubble Sort) Racing Method

I have updated the page, cleaning up a number of issues. Feedback appreciated.


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Wed Jul 20, 2011 9:37 pm

I have run a couple of manual simulations of 52 cars (a deck of playing cards) racing on 5 tracks. 32 heats per track (20 prelim + 12 finalist selection), with all 52 cars participating did a good job of sending more than the 5 fastest racers on to the finals. I would actually plan to run more than 20 prelim heats per track, since that would only be 6 heats per car for prelims, and some odd starting configurations could need a few more heats.

The simulation involved some upsets based on a throw of dice. A throw of 1 and 1 caused the fastest to finish 3rd and the others to shift up. A throw of 6 and 6 caused the slowest to finish first and the other two to shift down. A third odd die was used to decide which of the next 4 racers in line would not run in the heat. (That die cast until it showed 1 through 4 pips.) Also included rule that racers with the same color and rank would tie in the race. (This was too many ties to be realistic, and it slowed the sorting some, but was not fatal.)

I'm hot to actually get some boys doing this on the tracks for real! :)


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Sat Feb 25, 2012 6:07 pm

Survived our district derby using a variation of the 15th Burlington. Overall, the clock time was positive ... I was on my way home at 4:30 after helping with cleanup vs. a couple hours later in the recent past. The "event duration" from the individual Cub Scout's perspective was 1-1/2 to 2 hours shorter than in past years (from 3 hrs to 1-1/2 hours, unless he was a finalist, in which case it was 3-1/2 hrs to 2 hrs improvement.) Grade competitions were not overlapped, so parents of two Cubs in sequential years did not have to run between venues to keep up with how each racer fared.

One negative aspect showed in several grade competitions: One track raced more slowly than the others, enough so that the system's equilibrium was upset enough that the track accumulated more racers, which further affected the number of heats that racers on that track were allowed ... fewer than the intended number of heats. The solution arrived at was for that track to temporarily promote 2 instead of 1 racer toward the Gold track for as many heats as the track had excess participants. This is not ideal, but it works unless the disturbance is severe.

More later ... I'm taking my Bride to the Theatre to enjoy some Shakespear!


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Sun Feb 26, 2012 2:19 am

Comments that I received were all positive except for one who thought it was unfair that racers initially assigned (by lot) to Silver (slow) end tracks had to win more races than a racer who was initially assigned (by lot) to Gold (fast) end tracks in order to become a finalist and might not get to run enough races to become a finalist. In this case, his fear was partially justified, in that he got stuck for a few heats in the "slow track morass". I had not yet resolved the "slow track issue. In the end, he did become a finalist, one of the last 2 selected.

I think that his boy's racer was an 80th percentile racer, and the method is designed to assure (at about 99%) that those deserving of one of the 5 speed trophies would be in the finals. For groups that size, top 5 is about 90th percentile. For racers who almost, but not quite, deserve one of those 5 trophies, the probability that they will be included as a finalist drops off gradually. In order to assure that 99% probabilitiy of including the 5 most deserving I had to pull about 16 finailsts. Among those finalists, some who were not one of the fastest 16 racers would be inlcuded and some who would rank from 6th through 16th would be omitted, all depending on random factors, including who started where and who raced whom and other interventions by "outrageous fate." But remember that the goal is to include the 5 most deserving, not the 16 most deserving! (The distinction is subtle, and many will not pick up on it until it has been explained a few times.)

The remaining respondants were not put off by the "unfairness of initial random assignment" focusing instead on better adherence to schedule, more heats in less time, and much less waiting time between each racer's heats, all as compared to recent history.


Stan
"If it's not for the boys, it's for the birds!"

User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Thu Apr 11, 2013 1:43 pm

Wotamalo District has run the "15th Burlington" (Bubble Sort) method for two years now with a excellent reception by the masses.

Questions:

1. Has anyone else tried using the method for district or council races?

2. If so, what observations can you share with us?


Stan
"If it's not for the boys, it's for the birds!"

User avatar
FatSebastian
Pine Head Legend
Pine Head Legend
Posts: 2646
Joined: Wed Jun 17, 2009 2:49 pm
Location: Boogerton, PA

Re: Multiple track racing

Post by FatSebastian » Thu Apr 11, 2013 3:17 pm

:offtopic: Six posts in a row from the same person! I think that might be a DT record. :)



User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Thu Apr 11, 2013 5:46 pm

FatSebastian wrote::offtopic: Six posts in a row from the same person! I think that might be a DT record. :)
But spread over a period of more than one year! :)


Stan
"If it's not for the boys, it's for the birds!"

User avatar
FatSebastian
Pine Head Legend
Pine Head Legend
Posts: 2646
Joined: Wed Jun 17, 2009 2:49 pm
Location: Boogerton, PA

Re: Multiple track racing

Post by FatSebastian » Thu Apr 11, 2013 6:12 pm

Stan Pope wrote:But spread over a period of more than one year!
Perhaps another record!



User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6888
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Multiple track racing

Post by Stan Pope » Thu Apr 11, 2013 9:39 pm

So far, I know of the method being used in only three districts: the original Burlington (Ontario) group, my Wotamalo district, and the adjacent Heartland district thanks to the relationship between our chairman's husband and the new chairman in Heartland district.

I'd like to collect as many experiences as I can in order to refine the process.


Stan
"If it's not for the boys, it's for the birds!"

Post Reply