It's that throw out question (again)

[ttweed]

Charlie still gets penalized for the no-show, which is contrary to the idea of the throw-out rule.

I am not following you here, Henry. With a throw-out rule in effect, Charlie is not penalized, he actually gains ground on Kim, because he can throw out the no-show he had on 6/06. Kim must throw out one of the 5 point rain days, given that he shows up on 10/9 and scores more than 5 points. So actually, it is Kim that is slightly penalized, having to throw out a 5 point score vs. a 0 for Charlie (given that Charlie also shows up on 10/9 and scores more than 0 points.)

What am I missing in your argument?

TT

[kary]

It is my feeling that it is unfair to people to make them attend all events to be competitive. People have rich and full lives outside our club and will need to miss an event here and there. Seems a bit unfair to penalize them in a close competition over a entire season because they need to go to a wedding or take a child to a soccer game.

When I was one of the auto-x chairs we dropped events and that was either a rainout or a missed event, whichever was better for the participate. I think you should allow for that but more importantly it should have been stated at the beginning of the season. I do not know for sure what was said at the beginning of the season but if we as an organization do anything we need to be consistent and maintain for 2004 what was agreed upon at the beginning.

Just my 0.02

[Carl Scragg]

Henry has hit upon a subtlety of allowing throw-outs to be used for rained-out events. Using the results in LSS as an example. let's make a distinction between (1) no throw-outs, (2) an unrestricted throw-out policy (in which your lowest score is dropped), and (3) a restricted throw-out policy (in which your worst timed event is dropped - but not a rained-out event).

(1) Counting all events to date, Kim leads Charlie by 10 points: 95 to 85.
(2) With 1 unrestricted throw-out, Kim would lead by 5 pts: 90 to 85.
(3) With 1 restricted throw-out, Charlie would lead by 5 pts: 85 to 80.

Now which scenario best reflects the objective of a throw-out policy? If the objective of the throw-out policy is to allow a competitor to take a vacation, missing an event without hurting his competitive standing, then I think the restricted throw-out policy does this best. Note that Charlie never placed lower than second in any event. Therefore, had he not missed that one event, it's reasonable to assume that he would be leading by at least 5 points.

However, if we wish to place more emphasis on competing in as many events as possible, then there is little difference between having no throw-outs and 1 unrestricted throw-out -- Charlie drops from first to second place in the standings due to missing an event.

[David J Marguglio]

Maybe I am just dense, but it seems like we are making this a lot more complicated than it has to be. Don’t we avoid this entire debate over the semantics of rain-day vs. non-rain-day, restricted vs. unrestricted if we just take the best X scores? Using the LSS example: there have been 7 autocrosses (scheduled anyway) and we look at your best 6 scores. Therefore Kim has 90 and Charlie has 85; this in neither punitive nor confusing. What am I missing?

[Carl Scragg]

Dave's suggestion is exactly the same as having one unrestricted throw-out, and there is nothing wrong with such a system. But in our example case, it doesn't allow Charlie to take a day off without dropping from first to second place in the standings. So if your objective is to allow people to miss a day without penalty, you need a different approach.

I think that Kary had it right in stressing the need to adopt a system at the beginning of the year and then stick with it. In reality, throw-outs tend to affected only a few special cases anyway.

[Henry Walker]

Thanks Carl! You got it and explained as well as can be said. If you still don't get it, then read Carl's post again.

The bottom line is that IF we are going to have throw-outs, then they need to be restricted to non-rain events. Otherwise the throw-outs are pointless.

Now, what was the policy THIS YEAR? Did anybody document what was decided in January?

Henry car #122

[Larry Clark]

The question of restricted vs. non-restricted throw-out dates makes no sense to me. If Joe is allowed to drop one date (perhaps to attend his daughter's wedding), why should that be affected by the weather conditions? The event of his choice should be unscored, regardless of whether it was a rainout or not.

(A separate issue is why rainouts should be scored at all, but let's not go there.)

[Kim Crosser]

As the one who started the "throw-out" question a year ago, I find it funny to be the "poster child" now.

The original proposal was simple - if there were N autocrosses, everyone just gets their N-1 best scores. If you missed one event, you had a zero that got thrown out. If you attended all, but there was a rain-out, you had a 5 that got thrown out. If you attended all and there were no rain-outs, your lowest finishing score was thrown out.

This allowed participants to miss one event without being completely knocked out of a close competition. With that scoring, I would throw out a 5, leaving 90 and Charlie would throw out a 0, leaving 85.

As several people pointed out, the rain-outs mess up this simple scheme - but, since people paid good money to attend, some points for a rain-out seem fairer than "thanks for the money, no points".

One of the factors no one has mentioned is the disproportionate effects of the place scores - why 20, 15, 12, 10, 8, etc? If only two people show up, they finish with 20 and 15 points respectively for that day.

In sailboat racing, scoring is much simpler (and fairer?) by just eliminating "bonuses" for finishing places. The usual scoring method (there are variants) is as follows:
Each person scores their finishing place (1st = 1 point, 2nd = 2 points, etc.), and the winner of a series is the person with the lowest overall score. If someone attends, but is unable to complete a given race (DNF), they score the number of actual "finishers" in that race, plus one point. If someone misses a given race completely (DNR), they score the number of actual participants, plus one point. This way, someone can miss a race with four entrants and only score as though they had a fifth place finish, whereas in our scoring scheme, it is as though you finished 12th (or worse), even if only two showed up. In the case of a complete race cancellation (i.e., a rain-out), everyone present would just score the number of actual attendees, and those not present would score that count plus one. There are usually one (or even two) throw-outs as well, to allow a missed event. (To clarify the DNF/DNR situation, if 4 people show up, but 2 break down and cannot finish, you have a 1st, 2nd, and a 2-way tie for 3rd, and anyone who didn't show scores as though they got a 5th.)

This way, your points are driven by how you actually placed in competition - you don't get a huge bonus just because your strongest competitor didn't show up.

Note that there is NO way to completely, fairly, do scoring that handles missing one or more events. Even the sailing scoring would leave me with a couple of point advantage and Charlie would have to beat me in the last event by at least 3 places to win overall honors.

I would suggest we either:
A) Go with the N-1 scoring (best N-1 scores of N events are used, rain-outs are included), or
B) Have no throw-outs

[ttweed]

One of the factors no one has mentioned is the disproportionate effects of the place scores - why 20, 15, 12, 10, 8, etc?

This is an interesting point. There are lots of different ways annual points might be determined. The local SCCA region calculates points per place for the annual series championship as a percentage of the class-winning time for the event. The fastest time in the class is normalized as 100 points for the event, and each slower time is given the number of points which equals the fastest time divided by the slower time multiplied by 100. Thus if you are only beaten by a little bit, you might score 99 points, or if you are 15% slower, you would score 85. At the end of the year, they take the best two-thirds of your results and add them up for your annual total (lots of throw-outs), but they generally have many more events per year than we do.

They also have "Participation" awards for those who make every event, and a "Driver of the Year" award for the person who indexes fastest over the whole year.

Not that there is anything terribly wrong with the way we are doing it, I'm just throwing some other methods out there for possible consideration in future years. It might make for some closer competition if you weren't dinged for 25% of the winning score when you lost by a few tenths of a second.

TT

[Henry Walker]

Kim & Tom have some interesting ideas. When I first started autocrossing I thought the scoring could be improved. Why should driver X get 20pts for besting only one driver; whereas, another day driver Y gets 20 pts for winning over ten drivers? Kim's sailing example reflects this thinking.

How about only scoring a point for each driver that places slower than yourself + one point. Example: If you are the only one in your class that day, then you get one point. If you are 2nd but defeat 5 others, then you get 6 points. Of course you better show-up for a Zone-8 event because that's when attendence is best!

Tom's SCCA example is great also. It reflects the degree that you better the next guy. You don't want to have one really bad day.

The most difficult factor is what to do with the rain-outs. Whatever scoring system, it seems the rain-outs cannot be fairly dealt with. Those who paid their $$$ should get something for showing up. But the rain-outs seem to negate any type of throw-out strategy.

I still like having some kind of throw-out option, because I want to be able to vacation plan, etc. without sacrificing my season.

Finally, I think Kim & Tom are on to something. Any more ideas on scoring systems?

Henry Walker
'67 911 #122

[Gary Burch]

It appears the fly in the ointment is the rain outs. If you received no points for a rain out then all would be good in P-land. You could have one throw out based purely on competition. Of course this isn't entirely fair to the person who makes the sacrifice to attend all of the events, but should you win by participation alone?

No points for a rainout does not seem like too drastic of a penalty to help streamline the scoring procedure. Unlike adopting a new scoring strategy, the current system is simple and easy to use. It doesn't really matter how much you lose by or what coefficient you multiply it by, you still lose. And if you win the next time you make up the difference.

A rain out is a wash out as far as points and scoring so maybe no points for one is the way to go.

Thoughts?

[martinreinhardt]

I personally think that the current system is good.

AutoX is a form motorsport and if you show up = you get points incl. 5 pts for rainout (if you paid for the event).

The F1 doesn't use throwouts, why would we? I think it just makes things complicated and in most cases doesn't make a difference for the winner anyways

[Henry Walker]

Martin Reinhardt wrote: "The F1 doesn't use throwouts, why would we? "

Martin, the F1 guys usually don't miss races for vacations or because their jobs get in the way!

You are correct in that the throw-outs don't usually affect someone's season. However, I believe it does affect at least two classes this year and that is typical.

Gary was has it right--the rain-outs are the problem. Unless the rule is that you cannot use a rain-out as a throw out. It would hurt to have attended every AX and have to lose 12 or 15pts instead of a 5pt rain-out. But that happened to me 2 years ago. I was 3rd in ***. But after the two throw-outs I was 5th. I still think it was absolutely fair. Why? I scored only thirds and fourths that year. I was not the third best driver in my class and got the placing I deserved.

That's what a scoring system should do--give you the placing you deserve.

Henry Walker
'67 911 SWB, #122 FS

[Kim Crosser]

Henry Walker wrote

Unless the rule is that you cannot use a rain-out as a throw out.

In that case, we are saying that rain-outs aren't counted in points totals - you paid your money, but too bad...

As someone who paid my money at a rain out, got ZERO track time, and was privileged to stand in the rain helping set up and tear down, why shouldn't that count as opposed to someone who didn't even show up that day?
I would be quite unhappy being told I have to keep a 5-point day and throw out a 15-point day, while a competitor missed the rain day and another event and effectively got to throw out a zero?

Throw-outs are throw-outs. If you have throw-outs, they should just be the lowest score(s) of the year, whether those are rain-outs or not.

[Gary Burch]

Let's say you attend every event, rainouts and all. The policy is one throw out, and rainouts are not throw outs. Your main competitor skips the only rain out of the year. At the end of the year he gets to throw out his only missed event, the rain out. You on the other hand cannot throw out the rain out, so you have to throw out a scoring event. You lose by 10 points.

What I am saying is if you don't get points for a rainout, Things become allot easier. By the way, thanks for helping pick up the track in the wet.