Lane Calibration

DIY tracks.
User avatar
kwengerd
Apprentice
Apprentice
Posts: 14
Joined: Thu Oct 12, 2006 8:08 pm
Location: Winesburg, Ohio

Lane Calibration

Post by kwengerd »

We have not yet built our track, but how do you address lane "calibrating",
meaning, is there a good way to tune the lanes so that each one gives the
same results eliminating the "fast lane" syndrome? I had intended to make
the starting pins adjustable thereby shortening or lengthening each lane
incrementally and calibrating them with a standard car, but using a laser
starting gate, I am not sure this is possible. We have 60 cars to run and that
number will only increase, so it is virtually impossible to run each car on
each lane times. Your thoughts please...
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: Lane Calibration

Post by Stan Pope »

Since the factors involved in run times are affected by a number of lane characteristics and ther way in which the individual cars interact with those characteristics, altering one such to "equalize" would not be fruitful.

Basic charts for timed races require one heat for each registrant and give each racer a run in each lane. 60 heats is not an overwhelming load for most schedules ... an hour plus a little if you have race management hardware/software and have the boys stage their own cars. (A little longer if you have more lanes.)
Stan
"If it's not for the boys, it's for the birds!"
User avatar
Cory
Master Pine Head
Master Pine Head
Posts: 358
Joined: Fri Jul 25, 2003 7:18 am
Location: Chantilly, VA
Contact:

Re: Lane Calibration

Post by Cory »

kwengerd wrote:Your thoughts please...
How often do you plan on re-calibrating after things have started?
User avatar
Go Bubba Go
Pine Head Legend
Pine Head Legend
Posts: 1190
Joined: Sat Nov 18, 2006 12:09 pm
Location: Northern, Illinois

Re: Lane Calibration

Post by Go Bubba Go »

Trying to "calibrate", even with a particular given car, reminds me of the saying "You can't step in the same river twice...".

There are a number of variables that will simply not be exactly the same each time (staging, graphite wear down, dust/debris on track, etc....). When you measure to the 0.001 sec with a timer, the precision gives you the sense that you are super accurate within minimal variation in the track. Doesn't quite work that way.

I would have to find the old post, but there was some analysis done on WIRL times a little while back that basically demonstrated that some degree of variation was simply inherent in such an open system.

To turn your question on it's head, I would look at how to reduce the possibility of a "slow lane" as you build i.e. ensure defects in the track are kept to a minimum (smooth transitions between sections in all lanes, extra time spent selecting "super flat" (i.e. no warp) wood, uniformly smooth finish).

To compensate for the variation between lanes, I would run the cars in all 3 (or 4) lanes, and inquire of Stan regarding the best method to run the race quickly. Don't ask me, I am still a neophyte when it comes to the PPN, Picture Perfect PN, PBJ (slight tongue in cheek here) "Scheduling Methods" arena.

Simply stated, you would like to run in all lanes (more racing for the boys!!!, makes up somewhat for variation between lanes) in the time allowed, and done in a way (times vs. points - that is discussed elsewhere) that is easily understandable and seems fair to the participants.
"Who's Grandpa's neighbor?"... Phil Davis, Down and Derby
User avatar
Cory
Master Pine Head
Master Pine Head
Posts: 358
Joined: Fri Jul 25, 2003 7:18 am
Location: Chantilly, VA
Contact:

Re: Lane Calibration

Post by Cory »

Go Bubba Go wrote:When you measure to the 0.001 sec with a timer, the precision gives you the sense that you are super accurate within minimal variation in the track. Doesn't quite work that way.
I would echo this sentiment. My experiences from testing (pretty good car running on a so-so track) years ago showed me that I could get very consistent times running the same car repeatedly down the same lane. E.g. eight heats with a 6 millisecond spread was achievable.

Data from real races -- e.g. WIRL races, with better cars running on better tracks -- is not nearly as consistent. In the less controlled environment of a Cub Scout derby, where the track is more prone to debris, getting bumped, and such, it is possible that your calibration would become invalid very quickly.

Might debris and track bumps affect all lanes equally? I suppose it's possible, but unless you know that to be true I don't see the point in a one-time calibration.
User avatar
kwengerd
Apprentice
Apprentice
Posts: 14
Joined: Thu Oct 12, 2006 8:08 pm
Location: Winesburg, Ohio

Re: Lane Calibration

Post by kwengerd »

I guess I was intending to calibrate one time at each setup, averaging 3 runs per lane and adjusting so they averaged within some increment.

We have historically ran races as a tournament style, without any software or other automation. The winner was the one car remaining. Second place was determined by re-running everyone until one was left and so forth through 4 places.

Stan - so you basically recommend running each car 6 times (we will have a 6 lane track) and the four cars that average the lowest times would be the winners?
User avatar
Cory
Master Pine Head
Master Pine Head
Posts: 358
Joined: Fri Jul 25, 2003 7:18 am
Location: Chantilly, VA
Contact:

Re: Lane Calibration

Post by Cory »

kwengerd wrote:Stan - so you basically recommend running each car 6 times (we will have a 6 lane track) and the four cars that average the lowest times would be the winners?
Several questions, just to "grease the skids" for Stan and anyone else who might comment:

1. How much time do you have for the entire event?

2. Exactly how much "electronics" do you have? (e.g. finish line that displays/records places and/or times, laptop, not necessarily connected to each other)

3. Do you plan on running everyone together, or running separate groups? (e.g. Dens, Ranks)

Thx.
User avatar
kwengerd
Apprentice
Apprentice
Posts: 14
Joined: Thu Oct 12, 2006 8:08 pm
Location: Winesburg, Ohio

Re: Lane Calibration

Post by kwengerd »

We plan on using the MicroWizard K2 with the laser starting gate and Grand Prix Race Manager software.

We would like things to last no more than Four hours.
User avatar
kwengerd
Apprentice
Apprentice
Posts: 14
Joined: Thu Oct 12, 2006 8:08 pm
Location: Winesburg, Ohio

Re: Lane Calibration

Post by kwengerd »

Missed question #3...

We plan on running everything at once. We do this as a church outreach not boy scounts so there is no real hierarchy.
User avatar
gpraceman
Site Admin
Site Admin
Posts: 4926
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: Lane Calibration

Post by gpraceman »

Four hours for 60 cars? That is an eternity to run that many cars, if you are using a non-elimination method like the Perfect-N type charts. You can run all 60 cars on all 6 lanes and then do a finals round of the top racers and get done in under 2.5 hours. That assumes your race crew is reasonably organized to keep the cars flowing to the start line and retrieve them from the finish.

I'll echo the sentiment on calibrating the track. If all racers are to run on all lanes then there is no need. If you are seeing some inequity between lanes, I would try to check out the track to determine the cause and repair. I would not try to compensate by lengthening/shortening the run distance. If visually obvious, it is bound to bring complaints, even if it was done to make things equal.
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
Cory
Master Pine Head
Master Pine Head
Posts: 358
Joined: Fri Jul 25, 2003 7:18 am
Location: Chantilly, VA
Contact:

Re: Lane Calibration

Post by Cory »

gpraceman wrote:Four hours for 60 cars? That is an eternity to run that many cars, if you are using a non-elimination method like the Perfect-N type charts. You can run all 60 cars on all 6 lanes and then do a finals round of the top racers and get done in under 2.5 hours. That assumes your race crew is reasonably organized to keep the cars flowing to the start line and retrieve them from the finish.
Agree with Randy. If you prefer scoring by average/cumulative time, there is less/no need for a finals round, although you might choose to do one anyway.

If you prefer scoring by points, then it's usually best to run prelims (all cars) followed by finals (top 7 or 9 is often used). If you run a PPN chart in the prelims, each car will race 6 times, once in each lane, and will race 30 of the other 59 cars in those races. (5 opponents per race).

Then run a CPN finals. You might choose to use only four lanes for this round, it will speed things up -- not as many cars to stage, fewer cars being in consecutive heats -- and it might give you more flexibility, as 4 lane CPN charts are the most plentiful

Operationally, the key is keeping the pipeline full -- as a given heat is about to start, you want the six cars for the following heat already at the starting gate area ready to be staged.
TAL
Master Pine Head
Master Pine Head
Posts: 300
Joined: Wed May 17, 2006 8:47 pm
Location: N.C., USA

Re: Lane Calibration

Post by TAL »

kwengerd wrote: so you basically recommend running each car 6 times (we will have a 6 lane track) and the four cars that average the lowest times would be the winners?

With a 6 lane track and 60 plus cars I would run once in each lane and use the lowest average time of 5 of the 6 runs (minus the slowest run for each car with the gprm software) to determine the winners...

If you run once in each lane there will be no need for seriously calibrating the lanes perfectly ...

Average turn around time for each heat (run) should be around one minute ...

If you have 60 cars and a 6 lane track it will be 60 heats, each racer car should come around about every 10 minutes...

6 people should be at the end of the track getting there cars after their run while 6 people are staging their cars and 6 people should be getting called ready for the next heat while 6 people are putting their cars back in confinement...

We generally average 60 to 80 heats an hour...
bcatv
Pine Head
Pine Head
Posts: 56
Joined: Wed Jan 25, 2006 8:55 pm
Location: maryland

Re: Lane Calibration

Post by bcatv »

We run 2 passes each car each lane on a 6 lane track, take the quickest 2 from each den to a the championship rounds (9 dens = 18 cars), again two passes each car each lane, then quickest 6 to the finals, 1 pass each lane. This takes about 3.5 hours, gives every boy at least 12 runs for his effort, and eliminates any concern about lane variation. Also an interesting test of the "lube wearing out" theory; my oldest son's car has made it into the finals the last two years, on graphite (HEL), and not seen much variation over the 30 runs. Run lots of runs, the boys deserve it.
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: Lane Calibration

Post by Stan Pope »

Stan Pope wrote:Since the factors involved in run times are affected by a number of lane characteristics and ther way in which the individual cars interact with those characteristics, altering one such to "equalize" would not be fruitful.
I think I detect a hesitency to accept the statement. Here are some situations in which two cars might react very differently:
1. Two cars are "rail riders'. One (Car A) rides the left side of the rail; the other (Car B) rides the right side. Four lanes vary as follows:
Lane 1: both sides of rail very smooth. Cars A and B perform comparably.
Lane 2: left side of rail very smooth; right side very rough. Car A performs better than Car B.
Lane 3: left side of rail very rough; right side very smooth. Car A performs worse than Car B.
Lane 4: both sides of rail very smooth. Cars A and B perform comparably

Similar analyses might be cited for lanes with different surface roughness and cars with different "inherent springyness."

And for lanes with different curvature and cars with different CM height.

And for lanes with diffeerent lateral slope and cars riding different sides of the rail.

Add to those factors the random variance cited by Cory et.al. and you are faced with an exercise in futility in trying to "calibrate the lanes." This is true for handicapping various lanes by time or by starting position.

You are racing by time and certainly have sufficient time to complete one full round (each car runs once in each lane) if your staff and racers are alert and trained. What you don't know is, "How long will it take to complete a round?" Suppose you had a time limit of 3 hours and you found that your "one full round" (60 heats if you had 60 cars) took 50 minutes. Your racers would feel "short changed." (Your parents might feel relieved.) The time available would easily have allowed two or three full rounds! You could arrange ahead of time with the racers that if the first round completes by a certain time, then racing will continue for another round! (Check to see if the race management software can supprt this alternative. Some software probably does not.)

Preliminaries and finals are probably not needed for en masse timed racing using charts like PPN. Simple Lane Rotation is a different story.
Stan
"If it's not for the boys, it's for the birds!"
SuperDave

Re: Lane Calibration

Post by SuperDave »

Three thoughts:
1. Are you trying to find a winner OR are you trying to provide a positive experience for the kids? Elimination methods which are perfectly understandable to adults are perfectly devastating to kids. That paint job sure looked fast, now I've lost twice and go home. In my mind (and partly because I write and sell software that does it) running every car in every lane and averaging their times is fair, feels really good to the kids, can be scheduled so that slow kids race slow kids and thus get to win a heat (feel even better) and sorted to provide lots of winners (overall, den, rank, etc.), even winning teams. After determining winners this way, having 'runoffs' is generally not positive. Imagine the controversy when the first result finishers 1 thru 4 come out in the second result 4 thru 1!!

2. In several postings by others the implication is that more lanes takes more time. Not true. Given that all cars run all lanes, the number of heats equals the number of cars. There will be a few extra heats if the number of racers is not evenly divisible by the number of lanes. For example: 4 lanes, 40 racers, 40 heats. 4 lanes, 41 racers, 44 heats.

If the math sounds crazy, consider the following:
10 racers, 1 lane = 10 heats
10 racers, 10 lanes = 10 heats (but every racer runs 10 times).

More lanes can allow the race to be shorter (in averaged time scoring) if the cars are set to run say 3 or 4 of the 6 available lanes. That is a random choice of 3 or 4 out of 6, depending on the scheduler.

3. Digital timing. Ignoring inaccuracies caused by start or finish sensors, car positioning, lighting, etc., any digital measurement of a continuous function (time, weight, not people) is always at +/- 1/2 count in the last place. Combining or comparing times makes the error worse. For example, if the start signal reads and stores the time and the finish signal reads and stores the time, then the error is worse and worse still if you compare lane 1 to lane 2. The first error can be aided by starting a clock from true zero at the start of the race, but the combination of lanes remains. And, of course, I'm ignoring other potential error sources (which could fill a book). In the Pinewood Derby a general rule is that 1/8" = 0.001 seconds. That varies somewhat of course on different tracks and with different cars but it's worth remembering. Because of the well known bell curve, [moderator cut - no advertising], 4 decimal place (0.0000) ties are quite common around 10th place overall.
Post Reply