Dumb luck and track variations

[gpraceman]

Maybe I'm being pessimistic, but face it. Child built or store bought, the top positions really seem to be more an issue of dumb luck. The finish time spread is usually lower than the standard deviation of the track. Who wins is more a measure of the points system or how heats are organised if not simply pure chance.

That has not been my experience. But then again, I did a bunch of research before running my first race to determine which scheduling method to use and to decide between points and times scoring (IMO you can run a fair race with both as long as you consider the drawbacks of each).

[PWD]

I think it probably depends on the pack and what rules you use. If your rules are that you are not allowed to change much then dumb luck might be a factor.

In our pack we have pretty liberal rules and in the last ten years I have not seen a car in the top 5 that was there with dumb luck. We are in a rather large pack with over 170 kids.

You don't have a chance to be in the top 5 without deburred axles, good alignment, graphite/moly and decent weight placement. Most cars looked to also have lighten wheels. It is very rare to see a top car with a short wheel base and I have never seen a top car that used the slots. I think there could be but the good builders build long wheel base cars and drill their holes.

Probably just depends on the pack. But this is from our ten years of experience.

[SlartyBartFast]

I think it probably depends on the pack and what rules you use. If your rules are that you are not allowed to change much then dumb luck might be a factor.

If referring to my post, I think either you missed my point or I failed to make it clear.

Dumb luck isn't involved in who gets into the top of the pack. But luck, IMO, plays a large role in final order of the top cars.

[gpraceman]

Dumb luck isn't involved in who gets into the top of the pack. But luck, IMO, plays a large role in final order of the top cars.

That really depends on the race method being used. Some race coordinators go with something familiar (like double elimination) or just go with what has been done in the past without any examination of the disadvantages of the method.

For instance, when running a typical double elimination tournament, you may (or may not) be able to determine 1st and 2nd place accurately but 3rd place on is a up to chance. I can see dumb luck being a factor when using this type of race method.

[Cory]

Dumb luck isn't involved in who gets into the top of the pack. But luck, IMO, plays a large role in final order of the top cars.

Larger than many people might think, if your track is not the greatest, and you trust my simulator.

[SlartyBartFast]

For instance, when running a typical double elimination tournament, you may (or may not) be able to determine 1st and 2nd place accurately

Really? How so? Consider the times posted in this thread:
viewtopic.php?p=23638#23638

3.186
3.185
3.166
3.162
3.160
3.196
3.162
3.153

Average 3.171.

Depending on pure dumb luck cars identical to the above could alternately win races "destroying" the competition with up to a 3/100ths gap. HUGE compared to some of the time spreads considering the standard accepted fact that ties can be common up to .001.

In the thread quoted, if each different prepared car were run against the other, luck and statistics is what will have a huge influence over the outcome.

Good track or bad.

Have you posted your simulator Cory?

[gpraceman]

For instance, when running a typical double elimination tournament, you may (or may not) be able to determine 1st and 2nd place accurately

Really? How so?

Say the "true" 2nd fastest car ran against the eventual 1st place car -- strike 1. Then this car draws the "bad lane" going against the "true" 3rd fastest car and looses (only reason being lane draw) -- strike 2. A car that was the 2nd fastest car will not get the 2nd place trophy, instead the 3rd fastest car would. You can also come up with a similar scenario as to why the true fastest car doesn't get the 1st place trophy.

Is that due to "dumb luck"? I would say so (luck of the lane draw), which traces back to a poor choice of race methods.

[Stan Pope]

Have you posted your simulator Cory?

It has moved about faster than Google and my few links can keep up with it. Currently at http://members.cox.net/pinehead/software.html.

[Cory]

Have you posted your simulator Cory?

It has moved about faster than Google and my few links can keep up with it. Currently at http://members.cox.net/pinehead/software.html.

Ah, c'mon....it only moves about once every five years. And it didn't move a single time last century.

Slarty, you'll want to fiddle with the three StDev's: car speed, lane speed, and random element. I suggest using a CPN chart, like this 7-car 3-lane chart:

7 14 3
1 3 7
2 4 1
3 5 2
4 6 3
5 7 4
6 1 5
7 2 6
1 6 2
2 7 3
3 1 4
4 2 5
5 3 6
6 4 7
7 5 1

You can generate additional charts in DerbySim format using the PPN generator: http://members.aol.com/StanDCmr/ppngen.html

Stan, the generator seems to be unreachable at the moment, although I'm sure it's not your doing. (Edit: whoops, never mind, it has just returned to life.)

[SlartyBartFast]

Is that due to "dumb luck"? I would say so (luck of the lane draw), which traces back to a poor choice of race methods.

Okay. I misread you previous post as questioning my assertion that luck affects the final placement. That last post supports it.

Yet, I contend luck is and even bigger factor. Even if the track is perfect, each car has a standard deviation of its own, due both to setup each race and the car itself...

[gpraceman]

Even if the track is perfect, each car has a standard deviation of its own, due both to setup each race and the car itself...

Variation is unavoidable. It is up to the racers to minimize the variation in the cars themselves and the race coordinator and staff's responsibility is to minimize the variations due to track lane differences and from staging. Any of these can determine the eventual outcome. Hopefully, it is not from using a race method that doesn't have all racers run in each lane of the track or though incosistent staging; otherwise, things are even more up to "dumb luck".

[Stan Pope]

It has moved about faster than Google and my few links can keep up with it. Currently at http://members.cox.net/pinehead/software.html.

Ah, c'mon....it only moves about once every five years. And it didn't move a single time last century.

You are right, of course, Cory! I was exaggerating. Please accept my sincere apology.

Meanwhile, I found another link in my web site that needs to be updated to your work's newest location! Argh! Maybe if I fix that, Google will find your new location! The software is too good to be unavailable to the deserving.

[Stan Pope]

It is as silly to say that luck plays no role in race results as it is to say that car quality, track quality, and method quality play no role. In truth, they all play a role in the final standings.

Our charter as organizers should include choosing a method that minimizes the luck component and the track component by selecting a method that maximized the component due to car quality.

The choice of method is usually dictated by what your racing needs to achieve. If my primary need is to select N representatives to district races, then I am more concerned with identifying the n fastest than in accurately ordering the N fastest.

The choice of method is usually affected by the equipment at one's disposal.

Do I have access to a track? No? Then I might have to select the fastest by flippin a coin to select the winner of each heat. (That would fully qualify as winning totally based on "dumb luck".)

Does the track that I have access to have lanes that are reasonably well matched? (more closely matched than most of the cars that will race on them) No? Then I might run an elimination style race but the coin flip or draw for lane will determine many of the heat results. That comes close to winning based on "dumb luck." Or I could run a compound PPN method that reduces the role of chance by distributing the bad lane equally and produces a ranking that is close enough to make good finalist decisions from. Then run a PN with the finalists to gain more accuracy and further reduce the contribution of the luck component. (PN on a 2-lane track with lots of participants can be tedius!)

Do I have access to a track with more than 2 lanes and that has finish place electronics, No? Then I can run a multiple elimination with random lane draw. If the elimination count is large enough, the role of luck can be reduced. It is still there, but it is diminished.

And we can keep on going, but ... I am running out of gas for tonight.

[CuriousGeorge]

My son just had his first Awana race last night. It was a lot of fun, but one major flaw. For each heat they ran a set of (2) cars, switched the lanes, and ran the same cars again. If a car won both races, it was declared the winner of the heat. If one car won and then the other, they would again switch the lanes and declare whoever won the third time the winner of the heat. They had to do a 3rd race at least two times. It was pretty obvious one lane was faster than the other, but no one complained. (not yet at least)

I think that for a couple of kids "Dumb luck and track variations" played a major role in where their cars placed. I'll say something before next years race. The odd thing was, the track had 3 lanes and they only used the two on the outsides.

What would an official do if you only had two lanes and two cars kept alternating winning (no times, just place). I wouldn't be opposed to racing them backwards like I read in another thread.

[Stan Pope]

... but one major flaw. For each heat they ran a set of (2) cars, switched the lanes, and ran the same cars again. If a car won both races, it was declared the winner of the heat. If one car won and then the other, they would again switch the lanes and declare whoever won the third time the winner of the heat.

If you stop and think about it, what was gained by repeating each matchup with a second (and sometimes third) heat?

Regarding accuracy of result: Nothing was gained in this case as a result of running more than twice as many heats. In general, a "fluke result" might get washed out by the 2nd and 3rd heats.

Regarding role of officials in the proceedings: Depends on how they decided who should occupy which lanes in the first of the set of heats. If the "loader/starter decided which car to put where", then in those cases in which the cars are more equal than the lanes, the starter/loader is deciding which will win the matchup! That is not a proper role!

Regarding role of chance in the proceedings: If first heat lane occupancy were determined by chance (by lot), then chance played no smaller role in the result than if that first heat had determined the matchup results!

One specific possibly valuable result arises from this "best 2 out of 3" approaach: Participants and spectators know when luck played a role. Expensive knowledge, timewise!