What type of scheduling?

Debates and discussions on the various race scheduling methods that can be used and their fairness and accuracy in determining the winners.
User avatar
pack529holycross
Master Pine Head
Master Pine Head
Posts: 556
Joined: Wed Dec 26, 2007 5:25 pm
Location: Dr. Phillips, Florida
Contact:

Re: What type of scheduling?

Post by pack529holycross » Mon Apr 06, 2009 9:58 pm

Perhaps I am spoiled by the sheer luxury of GPRM and Stans gifts to the world in the form of PWD race scheduling, but I cannot imagine what level of "nostalgia" a person would have to ascribe to in order to continue to conduct a Unit race on a two lane track. Out of complete respect to those who have done this for a very long time, all the tools are available for you to have a top notch event on a 4-6 lane track, and insure that the boys get ALOT more racing in than on a 2 lane track. If you dont have, borrow if you cant borrow - RENT. Your kids will enjoy it much better, in my opinion.

Nicholas



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: What type of scheduling?

Post by Stan Pope » Mon Apr 06, 2009 10:08 pm

davem wrote:One other thought.

I have run races using both points based scoring and time based scoring.

I prefer points and PPN because the results are visually verifiable by the crowd.
However, most people - when given a timer - will want to race by times.

When I run timed races - several people want to examine the times in detail (some looking for errors).
When I run points - haven't had anyone question the results yet.
Very interesting and very telling! Thank you for the insight!

One of the "blessings" of electronics is "intermittent failures". Most folks using timers assume that there are no errors in the electronics. Thinking, perhaps, that "It shows the time to 4 decimal places ... Wow! It must be right!" The result is that errors that aren't obvious (which is many of them) go unnoticed and are allowed to stand without question.

I'd love to see someone develop a method (with low investment and low time overhead) that will "prove" (to some reasonable degree of precision) the validity of recorded times. So far the ideas I've seen fail on one or both investment and overhead aspects. The obvious way is to incorporate redundancy in the base system, but the user community is too naive to pay the necessary premium. So, those of us who recognize the problem look for (dream of?) alternatives that can overlay the base system to routinely crosscheck and red flag the discrepancies.

Oh, oh ... I'd better get off the soap box before Randy comes after me with a stick!


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: What type of scheduling?

Post by Stan Pope » Tue Apr 07, 2009 1:23 am

pack529holycross wrote:Perhaps I am spoiled by the sheer luxury of GPRM and Stans gifts to the world in the form of PWD race scheduling, but I cannot imagine what level of "nostalgia" a person would have to ascribe to in order to continue to conduct a Unit race on a two lane track. Out of complete respect to those who have done this for a very long time, all the tools are available for you to have a top notch event on a 4-6 lane track, and insure that the boys get ALOT more racing in than on a 2 lane track. If you dont have, borrow if you cant borrow - RENT. Your kids will enjoy it much better, in my opinion.

Nicholas
You might be amazed at the amount of racing you can get on a 2-lane track if you match the method to the environment.

Going charted, I'd balk at more than 4 lanes. I think that more lanes only gets meaningfully faster if you remove the boys from the racing! I won't do that. It takes their fun away.

Oops. Soap box apology.


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

*5 J's*
Master Pine Head
Master Pine Head
Posts: 756
Joined: Fri Apr 03, 2009 8:55 am
Location: Norway, Maine

Re: What type of scheduling?

Post by *5 J's* » Tue Apr 07, 2009 10:31 am

I have been reading Darin's and Stan's personal site's re:scheduling, however, all examples that I have found are using 3+ lanes. I can follow the concepts of these, but it gets difficult when I think about implementing any of these on a two lane track - where we run each race twice, swapping lanes, and use the average speed.

First let me give you a little more info on what we use at the Pack level. We have two laptops running (both hooked to projectors). One collects the elapsed time for each car, the second computer runs the excel spreadsheet to schedule the races. To recap - we run a "prelim" round where cars are paired up by registration number. Next we run the "time trials" where the cars are paired up by elapsed time from the prelim round (1st races 2nd, 3rd races 4th, etc.). At this point we have run two rounds and select the top sixteen who compete in a single elimination round paired up (1st races 9th, 2nd races 10th, etc.), in accordance with the tree above.

Now, I suppose in a perfect world each car would race each other car in each lane. But this would create too many heats, I suspect. What I think would work best is to progressively have the faster cars running against faster cars, while slower cars run against slower cars. I believe this would give more opportunities for each car to get a win, while you slowly bubble the best too the front.

As stated above - change would be best received if it is progressive. I may be able to change the scheduling algorithm and number of rounds, however, I suspect that I would have to retain two lanes, each heat run twice with the cars swapping lanes, and use the average elapsed time.

Given these parameters - what scheduling algorithm and number of rounds would be best? Does it make sense to have an method that progressively has the faster cars running against faster cars, while slower cars run against slower cars?



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: What type of scheduling?

Post by Stan Pope » Tue Apr 07, 2009 10:54 am

Look at the section on multiple elim racing (A "No-Chart" Racing Method at http://www.stanpope.net/nelim.html) This method does what you ask, i.e. faster cars stay together and slower cars progress down the groups to race against slower cars. We run quintuple elim on 3 lanes at district and 87% of the racers go home with at least one heat win. Two lane competition increases that percentage, but would lengthen the competition to much for the number of racers we serve. For smaller groups, 2 at a time winnows the top group down to a single car more slowly. If you started with 25 racers, Three lane competition would reduce the unbeaten group to 9, 3, 1. Two lanes would reduce it to 13, 7, 4, 2, 1. (Running n-elim on a 2 lane track needs some polishing ... there is an "odd man" in many rounds so you need to do something to equalize the probabilities that "being odd man" is apportioned fairly (randomly) among the racers.) See me if you want to pursue this, as I've some suggestions on how to do it with reasonable efficiency.

PPN charts (see http://www.stanpope.net/ppngen.html) are well defined for 2 lane tracks. Preferred to use an odd number of racers and run a bye if necessary ... the charts are better. A group of 25 racers would require 24 races for each car to have everybody race against everybody. Instead, I'd run each car 6 or 8 times (3 or 4 rounds), and pick the highest scoring 5 or 7 to run a finals (this is what Darin referred to as how he would like to run.) I think 7 racers would require 21 heats (6*7/2) to have a full round-robin and get pretty accurate results. Since lanes opportunities are balanced, you can forego the two runs per "heat" and timing.

Both of these will do a more accurate job than what you have now in awarding the trophies.


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

*5 J's*
Master Pine Head
Master Pine Head
Posts: 756
Joined: Fri Apr 03, 2009 8:55 am
Location: Norway, Maine

Re: What type of scheduling?

Post by *5 J's* » Wed Apr 08, 2009 12:25 pm

I looked at your No-Chart Racing Method page. Initially I liked the idea, as I thought it would offer more chances for somebody to win at least once. I generated a spreadsheet for a two-lane, 24 car derby, with 9 rounds to determine 1st through 3rd. While setting this up I became aware of a couple of downfalls of this method (which Stan's javascript Display N-Elim Rounds illustrates). First - eventually cars are required to run "uncontested" - this doesn't seem very interesting. You could pair this car up with somebody from a different group - and just ignore the results. Another issue is that eventually your heats taper off and cars will not race, so you have idle children. Not a preferred option. I thought about keeping the cars running, having 1st race against 24th, then once identified, have 2nd race against 23rd. However, how many times would a child want to race the same car just to lose again. Or what happens if the 1st place car develops and issue and the last place car does beat it? Now what would you do?

Anyway - my initial thought to find a method to progressively have the faster cars racing the faster cars and vice versa - presents issues using the N-Elim Rounds method.

Hmm - guess I will continue to evaluate the other methods, with the assumption I will have to stay with a two lane track, using elapsed time (though - just because I display the elapsed time doesn't me the scheduling method requires the use of the actual number).



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: What type of scheduling?

Post by Stan Pope » Wed Apr 08, 2009 4:00 pm

N-Elim:

Run the simulation (link is in the N-Elim page) and see when the first racer gets eliminated ... and when you have 1/4 of them eliminated. Calculate "when" as percent of total heats which will be roughly the same as percent of total time. When you get to quintuple elim (my favorite) you are well into racing before anyone is out, and 90%+ go home with at least one contested heat win! Few other methods can claim that! About half the total heats have been run before anyone is eliminated.

Some folks insist that the cars run the uncontested heats, some don't. It is a matter of how strongly you feel that graphite wear comes into play in a series of 10 or 15 heats. For a 2-lane quintuple elim, there are about 10% uncontested heats.

Note that if you run uncontested heats (they are quicker, but assume that they take the same time) then 22 racers are out at the 92% of the heats have been run.

Charted:

By comparison, if you run a charted race with 30 cars in a prelim race with 4 heats for each car, followed by a 7 car PN final, you will run 60 prelim heats and 21 finals heats.

From my experience, roughly 60 percent of the racers will go home with a heat win.

Note that 23 racers are out at when 75% of the heats have been run.

Reflection:

How do these compare to the currently used scheme?


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

derbydad3
Journeyman
Journeyman
Posts: 41
Joined: Sat Jan 23, 2010 4:24 pm
Location: Southern, WI

Re: What type of scheduling?

Post by derbydad3 » Thu Apr 01, 2010 4:04 am

Our pack has a 4 lane track with about 40 scouts. We run 8 heats, each car races in each lane twice and it is based off of the average time of the eight races. I like this system, and it seems accurate and fair. Additionally the times are posted on an overhead projector after each heat. It displays the order 1,2,3,4 and the times for each car.



*5 J's*
Master Pine Head
Master Pine Head
Posts: 756
Joined: Fri Apr 03, 2009 8:55 am
Location: Norway, Maine

Re: What type of scheduling?

Post by *5 J's* » Thu Apr 01, 2010 6:32 am

There are pro's to running on a two lane track. I bring this topic up as our wooden Den track is getting quite old and worn and the Cub Master mentioned getting us a new Aluminum track. Giving it some thought I'm not sure I want the new aluminum track and am thinking of building a new TWO lane wooded track.

Some of the Pro's of the two lane track

Easier to score - you just need to know which car crossed the finish line first
Head-to-Head racing seems to be more exciting to the scouts
50% minimum will win a least one race - depending on the scheduling method you can get this up to 75% or so.
It is nostalgic at the Den level - "hey that is the same track/type I raced on 30 years ago"

Sure there are con's, but at the den level, I sort of like the two lane track. Our Pack has less than 50 scouts so the duration of the event has not been a concern.
Last edited by *5 J's* on Thu Apr 01, 2010 9:36 am, edited 1 time in total.



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: What type of scheduling?

Post by Stan Pope » Thu Apr 01, 2010 9:20 am

*5 J's* wrote:50% minimum will win a least one racing - depending on the scheduling method you can get this up to 75% or so.
Kudos! That is a poorly appreciated measure.

The percentage of scouts who take at least one heat win home increases as the elimination count increases but decreases as the number of lanes used increases. It is easiest to visualize by seeing how many get eliminated before a win.

Racing 2 at a time:
Single Elim: half lose first heat (the other half win) = 50% No Wins; 50% at least one
Double Elim: half of those who lost first heat lose 2nd heat = 25% No Wins; 75% at least one
Triple Elim: half of those who lost their first two heats also lose 3rd = 12.5% No Wins; 87.5% at least one win
...

In general:
N Elim on 2 lanes: 100% - (50%)^N
N Elim on 3 lanes: 100% - (2/3)^N
...
N Elim on M lanes: 100% - ((M-1)/M)^N
where "^" indicates "raised to the power of" , (1/2)^3 = 1/2 * 1/2 * 1/2 = 1/8

The percentage is really a "limit value" as the number of participants gets very large. The actual percentage for a typical number of racers tends to be a bit higher. The reason derives from the observation that you can't race 2 people three at a time!

N Elim racing has an "ending problem" with a simple and elegant solution. But those are too far afield of the thread subject to detail here.


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

User avatar
Darin McGrew
Pine Head Legend
Pine Head Legend
Posts: 1735
Joined: Fri Jul 18, 2003 1:23 pm
Location: Mountain View, CA
Contact:

Re: What type of scheduling?

Post by Darin McGrew » Thu Apr 01, 2010 10:17 am

Stan Pope wrote:The percentage of scouts who take at least one heat win home increases as the elimination count increases but decreases as the number of lanes used increases.
Yes, fewer kids will take home a win if you're using more lanes, all else being equal. But this is somewhat offset by the increased value of a 2nd place win. And just as half the racers in a heat will win on a 2-lane track, half the racers in a heat will come in 1st or 2nd on a 4-lane track.



*5 J's*
Master Pine Head
Master Pine Head
Posts: 756
Joined: Fri Apr 03, 2009 8:55 am
Location: Norway, Maine

Re: What type of scheduling?

Post by *5 J's* » Thu Apr 01, 2010 11:29 am

Darin McGrew wrote:
Stan Pope wrote:The percentage of scouts who take at least one heat win home increases as the elimination count increases but decreases as the number of lanes used increases.
Yes, fewer kids will take home a win if you're using more lanes, all else being equal. But this is somewhat offset by the increased value of a 2nd place win. And just as half the racers in a heat will win on a 2-lane track, half the racers in a heat will come in 1st or 2nd on a 4-lane track.
Good point Darin - but this depends somewhat on the age of the racer - to some 2nd place is the first loser. That is definitely NOT my point of view, but to some 6 year olds it may be.



User avatar
gpraceman
Site Admin
Site Admin
Posts: 4607
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: What type of scheduling?

Post by gpraceman » Thu Apr 01, 2010 12:02 pm

*5 J's* wrote:Good point Darin - but this depends somewhat on the age of the racer - to some 2nd place is the first loser.
I think that is more so the case with a 2 lane track. You have a winner and a loser. 2nd place on a 3 or more lane track is still better than last place in people's minds. On a track with 4 or more lanes, 2nd place is not too shabby. It is interesting to note that 4 lane tracks are the most common by far.


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.

*5 J's*
Master Pine Head
Master Pine Head
Posts: 756
Joined: Fri Apr 03, 2009 8:55 am
Location: Norway, Maine

Re: What type of scheduling?

Post by *5 J's* » Thu Apr 01, 2010 1:17 pm

It is interesting to note that 4 lane tracks are the most common by far.
I wonder how this varies by geographic version. Seems in these parts of the world the two lane is more common.
think that is more so the case with a 2 lane track. You have a winner and a loser.
Perhaps - but remember - with a two lane track and double elimination 75% of the scouts will win a race. As Stan noted triple elim raises that to 88%. I think if you can give 88% of the scouts at least one WIN - that's is pretty good.

Granted there are other cons to this scheduling method - but my post was to point out some pro's to a two lane track.



User avatar
gpraceman
Site Admin
Site Admin
Posts: 4607
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: What type of scheduling?

Post by gpraceman » Thu Apr 01, 2010 1:47 pm

*5 J's* wrote:Perhaps - but remember - with a two lane track and double elimination 75% of the scouts will win a race. As Stan noted triple elim raises that to 88%. I think if you can give 88% of the scouts at least one WIN - that's is pretty good.
If win rate is the primary concern, then running a dynamic schedule you can get well over 90%. With dynamic racing, heat match ups are based on the performance of the car's previous runs. As racing proceeds, you get slower cars with slower cars and faster cars with faster cars. If you ran it to "saturation", all but N-1 racers will win (N being the number of track lanes). Everyone gets the same amount of racing. No two or three strikes and you are out, while others continue racing. Every scheduling method has its set of cons as well, of course, and dynamic scheduling is no different.


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.

Post Reply