There is no "best method".

Debates and discussions on the various race scheduling methods that can be used and their fairness and accuracy in determining the winners.
Post Reply
User avatar
dknowles67
Master Pine Head
Master Pine Head
Posts: 163
Joined: Fri Jan 16, 2004 6:21 am
Location: Melbourne, Florida

There is no "best method".

Post by dknowles67 »

I've posted several new topics on this forum in an effort to find the "Holy Grail" of scheduling methods for my Pack. At first I thought that if I specified the number of lanes, and number of cars racing, someone would respond with the perfect scheduling method my Pack should use. What I found was the answer is always "it depends". Various methods will always have advantages/disadvanges over other methods. I'll try to post here the "wisdom" I have accumulated so far in my quest.

I am not in charge, just a dad of 2 cub scouts hoping to improve.
My pack has always used the Double Elimination method on a 3 lane track. We do not have timers. There is a noticeable difference in the 3 lanes on our track. What I observed about this method was that the fastest 3 cars did not always get the trophies, and that once the kid lost 2 races, he lost interest in the rest of the race. We race by rank each hour, and then have an overall race for the 1st place finishers in each rank. We do not know how many cars will show up, until the race starts. There are about 100 cub scouts in our Pack this year, and 65 showed up on race day. The smallest rank had 11 cars, and the largest 25. Our Pack is growing.

In order for anything to change for next year, I will have to propose a scheduling method that meets some very strict criteria. It has to be significantly "better" than what we do now, or no one will want to change. It has to stick to the one hour time limit for each rank. It has to be no extra work for the race coordinators. It (probably) cannot involve a pre-registration, as that would make more work for the race coordinators. It cannot involve use of the internet, as we won't have access to it where the race track is set up. We may possibly be getting some timers for our track for next year.

Here's what I'm currently proposing we change - but I'm looking for suggestions for improvements on this:
I've written a software program that implements the PPN race generator from the Pope & Young web-page. The racers names are entered on the computer, and the lane assignments are show for each heat. After the race, either the elapsed time, or the finish order can be enetered. After all the races are over, the 1st, 2nd, and 3rd place names are displayed. The program schedules a 2 round PPN chart for < 20 cars, and 1 round PPN for more than that. It adds an empty lane, for 6 cars, or 12 cars, forcing them into the 7, 13 car CPN charts for improved accuracy. For 8, or 9 cars, 4 of them go to a Finals race, to determine the top 3. For 10 or 11 cars, 5 go the finals. For more than 14 cars, 7 go to the finals.
(Note that < 7 cars will already have accurate results from 2 round PPN charts, so no finals are necessary. Similarly, 12,13 cars will not need finals.)
Here's what I see as significant improvements:
1) Greatly improved accuracy in determining 1st, 2nd, 3rd place trophy winners. Means that kids will end up with fair rewards, and less grumbing.
2) All cars participate up until the finals, increasing kids enjoyment of the races, up until the finals.

Here's what I see as problems:
1) For most number of cars, it may require more heats than the DE method. I think we can still finish in an hour, with < 45 cars per rank, which is more than I ever expect to show up.
2) In the "finals" part of the race, those that didn't make the finals, may not be interested anymore. Maybe we shouldn't have finals.
3) requires a computer, and operator at the races.

Some ideas:
Put the computer screen on a projector so everyone can see.
Draw the charts on a whiteboard so everyone can see.

Suggestions?
Comments?
(sorry for a long post)
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: There is no "best method".

Post by Stan Pope »

You say that you "don't have a timer", but do you have finish line electronics to register order of finish? If so, then the scheme will work. If not and you are trying to record finish order on three lanes by eye, it can be difficult to do accurately.

You need to settle some priorities: Is it more important to award trophies accurately and have some boys not racing during the last approx. 20% of the competition? or to award trophies approximately and have all of the boys racing during the last 20%?

Or would it be appropriate for those who don't "make the cut" to continue racing meaningless heats during the finals? I'm sure that it could be arranged.
Stan
"If it's not for the boys, it's for the birds!"
User avatar
dknowles67
Master Pine Head
Master Pine Head
Posts: 163
Joined: Fri Jan 16, 2004 6:21 am
Location: Melbourne, Florida

Re: There is no "best method".

Post by dknowles67 »

You say that you "don't have a timer", but do you have finish line electronics to register order of finish?
We do not have any electronics.
I have suggested that we get some, but for now we do everything by eye.
You need to settle some priorities: Is it more important to award trophies accurately and have some boys not racing during the last approx. 20% of the competition? or to award trophies approximately and have all of the boys racing during the last 20%?
My dilema exactly. What is more important to a 6-10 year old? Accuracy of the top 3 trophies, or making sure everyone races all the time. My sons are usually in the top 3 (although they don't always get a trophy :( ). To them accuracy seems important, but they would typically be racing in the finals (if we had them) anyway. I don't have a perspective from the kids with the slower cars.
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: There is no "best method".

Post by Stan Pope »

dknowles67 wrote:
You say that you "don't have a timer", but do you have finish line electronics to register order of finish?
We do not have any electronics.
I have suggested that we get some, but for now we do everything by eye.
Lessee ... three lane track, call by eye ... look at chartless elimination. Boys managed in bunches. Three judges.

We are trying a variation this year in which, for most of the racing, each judge's winner selection advances. Only when we get to the "short hair" will we ask the judges to select a single heat winner by majority rule.

Quintuple elimination runs pretty fast with modest size groups.

You can look at the simulator on my NoChart web page to see approximately how many heats in total are required and how many heats have been completed before anyone starts dropping out. Somewhat like a preliminaries / finals framework.
Stan
"If it's not for the boys, it's for the birds!"
User avatar
dknowles67
Master Pine Head
Master Pine Head
Posts: 163
Joined: Fri Jan 16, 2004 6:21 am
Location: Melbourne, Florida

Re: There is no "best method".

Post by dknowles67 »

Quintuple elimination sounds like it would take longer than double elimination. We only try to find the top 3 places, so I would argue triple elimination would be sufficient. However all the elimination methods have a significant downside, in my opinion, in that some cars will race significantly more than others. That's why I wrote the software program to use your (Young & Pope) PPN generators. I'm probably not going to be able to convince anyone to change "The-Way-We-Have-Always-Done-It", to a method which takes longer. I doubt there will be any changes unless there is a significant improvement in fun for the scouts, or has an obvious benefit to the racers, and race organizers. I haven't given up hope for installing some electronics by next year either.
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: There is no "best method".

Post by Stan Pope »

dknowles67 wrote:Quintuple elimination sounds like it would take longer than double elimination. We only try to find the top 3 places, so I would argue triple elimination would be sufficient.
I don't care if you use it or not, but reject it for the right reasons.

We used to run 2 lanes at a time in a charted double elimination at our district races. We changed to 3 lanes at a time uncharted quintuple elimination and it takes about the same amount of time. Gives minimum of 5 contested heats.

You know how many heats required to run your boys thru a double elimination. The script I mentioned will show how many heats required to run them L at a time in an E-elimination.

Quintuple elimination gives some leeway for a bad run or two during the preliminaries. A nice feature is that it is easy to incorporate late arrivals ... a nice feature when folks are making an hour-long drive to reach the races.

Charted DE usually has a lot of byes in early rounds. Uncharted has almost none. On 3 lanes, no byes until group size is reduced to 1.

The cycle time for heats is easier to maintain because the boys are managed in bunches rather than individually.

Fairness is easy to maintain and demonstrate. Accuracy is better than double elimination. Virtually no reruns because judges couldn't make a decision. (Without electronics, this is an increasingly large consideration.) Better accuracy than 1-0-0 points method in final standings charts.
Stan
"If it's not for the boys, it's for the birds!"
User avatar
dknowles67
Master Pine Head
Master Pine Head
Posts: 163
Joined: Fri Jan 16, 2004 6:21 am
Location: Melbourne, Florida

Re: There is no "best method".

Post by dknowles67 »

Okaaaay, now I'm really confused.
I started posting on this forum with little or no idea what to do, but convinced that I could find something better than the double elimination our Pack uses now.
(where "better" is always a matter of opinion).
I've approached it with an open mind, soliciting opionions, and gathering information.
I know it is difficult to recommend a "best method", when I cannot accurately specify my own Pack's race parameters.

This year, we ran 3 lanes, DE, no electronics. The DE charting was done on a whiteboard, by an overworked race coordinator, who went on the assumption that all three lanes were equal, so it didn't matter which lane you went in, and lanes were assigned first come, first serve. There were no byes. I'm not sure what you mean. The organizer made sure we had 3 cars racing up until the end. We might have raced the last 2 against each other to determine 1st place, I'm not sure.

I have hopes that next year we will have some electronics working (I've volunteered to assemble/test them).

I also hoped to come up with a scheduling method that would be an improvement with or without the electronics.

Based on everything I had read up until this thread, I concluded that 2-round PPN charts with a short finals to determine the top 3 was the way to go (for about 8-30 scouts at each rank).

I don't want to reject any method for the wrong reason.
I'll go back and re-write all the software I did for the 2-round PPN charts, if Quintuple elimination is better, but I'm not sure I understand what you mean.

Are you recommending a NoChart method?
For Quintuple elimination, from your web-page,
I get 64 heats (9-rounds) for 3 lanes 20 cars.
For Double elimination,
I get 24 heats.

I honestly don't know how many heats a typical DE chart has for N cars racing. As a rule of thumb, I seem to remember about 2N heats. We always finished in less than an hour.

I think we could run around 45 heats in an hour, but not much more than that.
This year we had 25 boys in the Webelo rank (our largest group).
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: There is no "best method".

Post by Stan Pope »

Be not confused. Let's just understand the process.

Without electronics to register finish order, points based on finish order (as needed by PPN) are more difficult to administer. Look carefully at what your finish line judges must do. If you (and they) think that they can do the job accurately, then go for it.

If you expect to get the electronics, then that is great... you've got it nailed! :)

If not, then look for more.

First, there is a variation of final standings that says heat winner gets 1 point, rest get 0 points. Using this scoring procedure, accuracy deteriorates somewhat. I don't recall details, but Cory included the option in his simulation software.

On elimination charts, most charted DE run 2 at a time and take 2*N-1 heats. Most charts for large numbers are structured for a power of 2, e.g. 2, 4, 8, 16, 32, or 64, and they would require 3, 7, 15, 31, 63 or 127 heats. For 17 to 32 cars, they use the 32 car DE chart and make 63 runs. Most user's run byes in the early rounds to equalize graphite shakeout.

20 cars in a typical 2 at a time DE chart means 63 heats.

Uncharted 2 at a time can reduce this to a truer 2N-1 heats, but byes occur later in the racing. For 20 cars this would be 39 heats.

Elimination schemes can be adjusted to take more than one winner per heat, e.g. if 4 cars are in a heat, then the first two finishers get no elim points and the 3rd and 4th get an elimination point. I don't really like the extra complication for the judges, but it does work.

I like racing 3 at a time because there are almost no bye situations.

Most folks find that they can run heats faster uncharted than charted, because they have little lag awaiting the next competitors. Overlap of racing functions is much easier to manage. If you can run 45/hour charted, you can probably do 60+/hour uncharted.

Anyway, think through the various processes and find the one you think will work best with what you have. I'll try to help you through details of any of them.
Stan
"If it's not for the boys, it's for the birds!"
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: There is no "best method".

Post by Stan Pope »

dknowles67 wrote:This year, we ran 3 lanes, DE, no electronics. The DE charting was done on a whiteboard, by an overworked race coordinator, who went on the assumption that all three lanes were equal, so it didn't matter which lane you went in, and lanes were assigned first come, first serve. There were no byes. I'm not sure what you mean. The organizer made sure we had 3 cars racing up until the end. We might have raced the last 2 against each other to determine 1st place, I'm not sure.
I missed this implication in your early posts. 3 at a time charts are understandably shorter than 2 at a time. They are also tougher to track and get the boys organized. Running only 45 heats per hour, even with the boys staging their own cars, suggests that the guy keeping the charts was way overtaxed! The result is that overlap of various race functions doesn't happen as it could. The No-chart technique eliminates almost all of the impediments to effective overlap.

But, since you have your sights clearly set on finish line electronics, then further discussion of elimination is moot! PPN will be very effective for you.
Stan
"If it's not for the boys, it's for the birds!"
User avatar
dknowles67
Master Pine Head
Master Pine Head
Posts: 163
Joined: Fri Jan 16, 2004 6:21 am
Location: Melbourne, Florida

Re: There is no "best method".

Post by dknowles67 »

But, since you have your sights clearly set on finish line electronics
I just have to convince the race organizers.
User avatar
gpraceman
Site Admin
Site Admin
Posts: 4926
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: There is no "best method".

Post by gpraceman »

dknowles67 wrote:
But, since you have your sights clearly set on finish line electronics
I just have to convince the race organizers.
Hopefully, it won't take too much convincing. You can have a system for under $75. It can be cheaper than that if you can get some of the materials, like wood for the electronics enclosure, donated. In the Do-It-Yourself section of http://grandprix-race-central.com/finish_line.htm, you will find some plans for systems that you can build inexpensively. You can even get a commercial grade timer, in kit form, for $40. You just need to solder the components together and build an enclosure for the electronics. It can't get easier than that, without buying a fully complete commercial system.

To me, with the options these days, the "it's too expensive" excuse for not having a timing system doesn't hold water.
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