New Track - Points vs Time?? Help!

[Stan Pope]

we might try using times in a special race like an "open" or "Dad's" class after the Pack has finished. This would give us an opportunity to try it with lower consequences of failure.

Mike,

That is an excellent way to find out about it ... to gain confidence in the equipment and in your application of it.

You could also make a second "dads race" where you challenge them to "break the system" or "beat the system" to help find out the weak points in your specific setup. In this race, any race-time "cheat" is fair if it goes undetected during the race. Dunno of any would fess up to finding a way, but differences in results between this and the real race might indicate the presence of such. Announce it in advance so that they can be plotting and planning!

That might be a fun activity ... but of course you, and they, would have to be vigilent ... in both races!

Who knows, they might want a second or third or fourth chance to try more ideas. Of course, be sure to rotate jobs.

If you opt for this, I'd be interested to hear what they come up with, or even, if you believe that someone came up with something but you don't know what it was!

Have the starter be one of the dads (with or without an elevated nose car). See if he figures out how to manipulate the system to his advantage. (There is a fair chance that he will.) See if you can catch him during the racing. See if the post-race analysis of times shows what he did. The techniques for manipulation are different for the two nose types.

Tell 'em that if no one breaks the system, they just aren't trying hard enough. Dunno if they have the capability but, in a group of 8 or 10 of 'em, I'd bet that one or two could find some ways.

Then, document! And share!

[Stan Pope]

If you decide to give the adults the "break it" challenge, here is something to try to do with a manual gate: Based on the first race you think that you know which is the fastest car and which is the slowest car. To the extent that there is a random element, if those cars race again, each will migrate toward the center. But, you also know which car sits at about 33% and which at 66%. See how close you can come to making them change places without affecting the other cars placements! At least, make each of them move far beyond the mean. To the extent that you moved them past the mean you have denied that the random element caused the change.

Very likely the fastest car from the initial race will still place very high (probably 1 or 2?) and the slowest car will still place very low, probably in the bottom 2 to 4. See if you can drop the first round winner out of the top 3 and move the slowest car up above the bottom 6 ... at the same time!

This should keep you busy for a few minutes!

Of course, if you have a spring-loaded starting gate, you take away some of that capability from the starter.

[glaforge]

I will agree that two racers in two different heats that are 1/1000th a second apart are hard to validate (ok impossible by sight) by the audience. But, they are usually watching (especially there own son’s car) and can very quickly tell if something is amiss because times start to vary greatly and they can see that (a very good type of validation). However, the manipulations by a “bad apple” can have just as catastrophic effect in place scoring methods. You yourself have challenged bushmg to see if he can effectively move a racer up or down in place finishing by whatever means. Say for sake of argument he is able to do that. Now you are moving a persons score up or down just as you have done with time, thereby effecting the overall results of up to the number of racers in the heat (not just the one). Secondly, if all I have when the race is over is place finish points, it becomes very difficult if not impossible to analyze the results to see if there was really a “bad apple”. At least with times I can go back and validate the results. Late yes, but it still can be done.

The thing that I really don’t like about place finish scoring is the “perceived impact” of problems affecting only the one person, when it really can effect everybody (and differently). For example, let’s say a racers’ car is performing very well until the last few heats (or even has one or two bad heats in the middle of the race). Perhaps, the boy (heaven forbid) dropped it, Whatever the reason. Now, the only people that will for sure benefit from it are the boys that race him in those bad races and get a better score because of it. The impact to those who don’t race him could either be no impact (no racer jumps in front or behind in total score, or one goes head and one goes behind), a negative impact (someone goes ahead because the benefit from getting a better score) or positive (perhaps nobody moves ahead and that one racer falls behind). It is this unequal treatment because of one cars problem that gets me.

I agree that it is a minimal impact to the one car that had a problem, but it could be catastrophic to somebody who didn’t even race in that heat. For this reason, I don’t buy into the minimal impact of one bad race theory. At least with total time, the overall score to those in the same heat are not impacted and the overall placement is either positively impacted (the bad racer dropped below you) or no impact. Nobody jumped ahead because of somebody else’s bad race.

[Stan Pope]

You yourself have challenged bushmg to see if he can effectively move a racer up or down in place finishing by whatever means. Say for sake of argument he is able to do that. Now you are moving a persons score up or down just as you have done with time, thereby effecting the overall results of up to the number of racers in the heat (not just the one).

The statement is incorrect. I can do it without ever changing a place in a heat result... and strictly by operating the manual starting gate.

[Stan Pope]

Secondly, if all I have when the race is over is place finish points, it becomes very difficult if not impossible to analyze the results to see if there was really a “bad apple”. At least with times I can go back and validate the results. Late yes, but it still can be done.

You can't file the times when you compete by place? I thought that most of the management software would do that for you if the hardware would give times.

In fact, I think that many of the competitors would like to know their times, in order. Tells about lubrication performance, perhaps?

[Stan Pope]

The thing that I really don’t like about place finish scoring is the “perceived impact” of problems affecting only the one person, when it really can effect everybody (and differently).

This gets to the real meat of the problem with points scoring! Cars have different performance profiles. They may run well on some lanes and poorly on other lanes. They may run well at the beginning and less well in later heats. Or, they may pick up speed in later heats. The variability introduces inaccuracy into the results. You can see this happening, albeit at a very low level, if you run accuracy simulations with Cory's neat program.

The larger the group relative to the number of lanes and heats, the greater the inaccuracy and the greater the effect of such variability on final finish place. That is why I like two stage racing for large group ... a preliminary that involves everyone and a final that involves about twice as many racers as there are trophies to be decided. ("Twice" is an effective rule of thumb ... it may be refined by accuracy sims.) This yields excellent accuracy of the final result.

All that said, times are a more precise way to compare cars. There are comparison issues to be worked out:
1. What time should a "did not finish" result be ascribed? (Case: His car runs really great on lanes 1, 2, and 3, but can't seem to stay on the track on lane 4.)
2. Do you tally all of each car's runs? Drop the slowest and fastest n times for each? Or simply compare each car's fastest time?
3. Does the organization value consistent or peak performance?
The organization can certainly sort these out.

Translating the precision to accuracy of results means that you must also have a good handle on the things that can go wrong. That means that the organization must resolve the issues that can impact results. The big ones are:
1. starting gate consistency
2. equipment and software integrity

Once the method issues are resolved and the integrity issues are resolved, then, indeed, racing the clock can produce more accurate results. Resolving these issues involves a lot more than hooking the gear up to the track and loading Randy's Race Management software (unashamed plug)! Based on comments on this board, I don't think that most folks who decide to race by the clock have put in the preparation to resolve those issues. And, even if they have, they must still base their success on the faith of the audience and the participants.

Personally, I'd rather have the excitement of meaningful head to head racing and no dependency on the faith of the audience.

[glaforge]

You yourself have challenged bushmg to see if he can effectively move a racer up or down in place finishing by whatever means. Say for sake of argument he is able to do that. Now you are moving a persons score up or down just as you have done with time, thereby effecting the overall results of up to the number of racers in the heat (not just the one).

The statement is incorrect. I can do it without ever changing a place in a heat result... and strictly by operating the manual starting gate.

Perhaps you can do that without changing the place finish, even so the child or his parent will begin to notice a wider variation in times and begin to question the results (a perfectly valid thing to do).

[glaforge]

Secondly, if all I have when the race is over is place finish points, it becomes very difficult if not impossible to analyze the results to see if there was really a “bad apple”. At least with times I can go back and validate the results. Late yes, but it still can be done.

You can't file the times when you compete by place? I thought that most of the management software would do that for you if the hardware would give times.

In fact, I think that many of the competitors would like to know their times, in order. Tells about lubrication performance, perhaps?

Agreed, but only if the race management software is used or times recorded along with place finishes. Some may not take the effort to plug it in!

[glaforge]

The larger the group relative to the number of lanes and heats, the greater the inaccuracy and the greater the effect of such variability on final finish place. That is why I like two stage racing for large group ... a preliminary that involves everyone and a final that involves about twice as many racers as there are trophies to be decided. ("Twice" is an effective rule of thumb ... it may be refined by accuracy sims.) This yields excellent accuracy of the final result.

Here we will get into choices for the format of the race and their requirments with respect to the method of scoring needed. I personally refer a one stage race. Everybody races the same number of times in each lane and then the race is over. Everbody is there to congratulate the winners. If you eliminate, for example, half the racers for stage two, half the people will most likely not stick around.

Guess what I am saying is that one should decide if they want the scoring method to dictate the choices for format of the race or the format of the race to dictate the methods of scoring. Neither is wrong.

[Stan Pope]

Perhaps you can do that without changing the place finish, even so the child or his parent will begin to notice a wider variation in times and begin to question the results (a perfectly valid thing to do).

Perhaps? Perhaps??? Undoubtedly!

But how much variance is "normal"? Depends on the cars, the tracks, the hardware, the staging skills, the starting gate, ...

But, finish the thought ... they begin to question the results ... bring this to its logical conclusion ...

[Stan Pope]

Some may not take the effort to plug it in!

Please don't try to run this on manually recorded times!

[Derby Wizard]

1. What time should a "did not finish" result be ascribed? (Case: His car runs really great on lanes 1, 2, and 3, but can't seem to stay on the track on lane 4.)

In our pack race, a DNF will result in being dropped significantly if not to the bottom in standings. The added time has been 10 seconds by default...but we have not had anyone suffer this penalty to date, and they couldn't based on our rules. Rerun(s) have always resulted in a finish to date.

Our rules with respect to leaving the lane more than twice, unless a major track problem is determined by the race committee, would DQ the car. Our track is old, basically bumpy, and usually results in a car leaving the track at some point during the race in our large derby...in fact I've been to few races anywhere, even on premium tracks, where a car hasn't left for some reason...like a wheel coming off or some obscure bump/jump.

The closest we've come in having to tackle this type of an issue was a car that was so heavily weighted in the back that it was suspect to a wheelie at the slightest piece of track dust. This caused several reruns to be performed and reached the point where it would have been DQ'ed on a third rerun in the same lane. The comittee has always been leanient if there is a noticable bump in the track. During the event, this 'wheelie car' was the only car leaving the track and the car design was an big factor in the problem.

In effect, this is part of the challenge on our track and we do warn that heavily weighting the rear of the car is risky...if another car showed up like the last one it could easily end up in a DQ situation if a similar problem developed.

2. Do you tally all of each car's runs? Drop the slowest and fastest n times for each? Or simply compare each car's fastest time?

Tally all. Slowest and fastest are included.

3. Does the organization value consistent or peak performance?
The organization can certainly sort these out.

Consistant.

Based on comments on this board, I don't think that most folks who decide to race by the clock have put in the preparation to resolve those issues. And, even if they have, they must still base their success on the faith of the audience and the participants.

Personally, I'd rather have the excitement of meaningful head to head racing and no dependency on the faith of the audience.

I feel preparation has been put in place to resolve many of the issues at our event which you have mentioned. I also have faith in our system. The comparsions that keep getting raised into how to fail a system and prove that one is better than the other is a good discussion but in some posts the points made don't lend well to reality.

Moving a car up or down a place is likely to occur in many methods implemented due to the enormous amount of unknown or uncontrolled variables that can come into play. I've seen and read about many perfect engineering designs, methods, and simulations that all fail when implemented in the real world.

.....

In order to provide the 'best' method I'd have tasked to use a fact that the same two cars are going to be consistanlty ~1/1000 of a second different. (This isn't going to occur in reality.) Which car is fastest using a given method? Depends on a whole list of uncontrollable variables...now we get into the fair and accuracy discussion again...

In my opinion, the best method is the one most, ideally all, in a particular organization support that is fair. I prefer times as these can provide more accurate research and analysis concerning a car and race analysis.

Derby Wizard

[Stan Pope]

Good summary, DW!

I don't think that I am disagreeing with you in the following ...

On the distinction between accurate and fair, I've yet to see a more useful distinction than the one I posted some years ago at http://members.aol.com/standcmr/pwraces.html under the heading "Purpose". It relies on the (unstated) conventional use of the term "fair" in describing probabilities and, in particular, apparatus such as dice. Dice are "fair" or not according to the absence or presence of non-random factors. My study in mathematics pushed me toward trying to retain such distinctions to help avoid muddying the analysis with emotions, perceptions and connotations.

Of course, even with very concrete language and definitions, the perceptions of the audience and participants are still important in settling the question of fair or not. Hopefully, the terms have sufficient definition that they can assess, and perhaps argue over, the same set of variables.

Regardless of the competition method selected, it is still important for the organizer to guard the fairness aspects jealously. If partiality or manipulation is allowed to encroach on the process, an event which should be fun would become a source of distrust and hard feelings. Those would hurt the organization.

I think that elimination, final standings-points, and timed races can all be subject to such influences. I think that the challenge is greater with timed races. I also think that the challenge can be met with all of these formats.

[GoneFission]

One thing is for sure...If you choose place-scoring methods, somebody will come to you and tell you that their child should have placed higher because he had a lower total time.

We have never heard that complaint. Our finish gate reports only finish order, so no one can say anything about total time.

The second district race my son qualified for was hosted by a pack using lane rotation. They told the audience they were scoring by time - we had to ask because we couldn't tell. Each car was run twice through the same schedule. We had a solid third place car - just behind second but faster than 4th.

When the results were posted, we finished third - by a single point. As it happened, we were almost always on the track with the 1st and 2nd place cars. The 4th place was not. Therefore we got 2nds and 3rds in our heats. So did the 4th place car. They finished consistently behind us, but did not face the 1st and 2nd place finishers as many times. The scores were nearly identical, but the cars were not.

Literally, one run when we won the heat and beat the 2nd place car probably kept my son on the podium.

We weren't sure why it had been so close until the next year when we started using the same software. I went back to my picture of the finish screen, and the sort order was set to points - no time was displayed.

[Stan Pope]

We weren't sure why it had been so close until the next year when we started using the same software. I went back to my picture of the finish screen, and the sort order was set to points - no time was displayed.

Arghhhh!!!!! Derby Chairmen who say "Trust me! I know what I'm doing," just should not be trusted! If they can't explain it, they probably don't know enough to run it!

I see three errors:

1. He said he was running by time but ran by points.

2. Since he ran by points, he shouldn't use just a simple lane rotation ... opposition needs to be as equitable as possible. For instance PPN charts.

3. He let you photograph the screen!