Wednesday, January 30, 2013

Scoring Experiment

I see that at the ISAF World Cup racing in Miami this week, the organizers are experimenting with a slightly different scoring system for the Laser and Laser Radial Classes. The winner of each race is given a zero score - instead of one point. All other places are unchanged, i.e. 2nd place boat gets 2 points, 3rd place gets 3 points and so on.

I can only guess that the reason is to give a "bonus" to the sailor who actually wins each race?

Back in the day when I first started racing, the standard system was to give the winner of each race three-quarters of a point. This functioned as a sort of a tie-breaker because, for example, a first and a third place would be a better score (by a quarter of a point) than two second places. Although I think the current tie-breaker system is mathematically equivalent in all realistic scenarios I can imagine.

Any idea why it was decided to experiment in this way?

What do you think of it?


Baydog said...

0 points quickly separates a winner from the rest of the pack. Quite a bit different than this system.

Tillerman said...

Aaaah. High Point Percentage Scoring. We could have a whole series of posts about the merits of different scoring systems and why you would choose different ones in different situations.

Noodle said...

The scoring system should make the fastest sailor win. Both the bonus scoring system and the low point scoring system persue that goal. (Other systems - such the the one used at the Olympics - do not).

Low point was invented to have more intense racing series with smaller margins; Low point scoring makes it harder to create a gap to the pack.

It is really about this: Should 1-2-3 beat 2-2-2? Which is the faster sailor? With bonus points 1-2-3 beats 2-2-2. With low points it is a tie, but A8.1 makes 1-2-3 the winner still.

Low points give more intense racing, while still favouring a first. Some would argue it's more fair to use low points.

Tillerman said...

Good points Noodle.

Another example...

With this zero scoring experiment, in a 7 race series, boat A with 0,0,0,3,3,3,4 beats boat B with 2,2,2,2,2,2,2 as A has 13 points and B 14.

Under the normal low points system A=16, B =14 and B wins.
B beat A in 4 out of 7 races.

It's probably only going to be a factor in a few rare situations like this but it does tend to reward occasional wins more than overall consistency.

Anonymous said...

I don't think that it changes the sailor's approach to winning the regatta. The bonus is not enough to justify taking chances to try and win any single race. Though it may in some cases change who does win. So I don't see any benefit to the racing by making the change. Personally, I think consistency is a far better indicator of sailing ability over the course of a series, so I would favor the simple low point system.

Tillerman said...

I tend to agree Anonymous. It's a minor tweak that doesn't fundamentally change the game, except occasionally to penalize consistency.

So why did they do it? Does anyone know? Has it been explained anywhere?

Anonymous said...

This bonus system may attract extra interest of people to a winner of each race. And adds more competition between the first and the second.

Dallas Dude said...

It arbitrarily makes some races more valuable to one person than to others. It is, therefore, not an objective and impartial rendering of the best sailor in the series -- the 0 system has drifted into the subjective and partial.

Tillerman said...

Interesting. In the Laser qualifying series at Miami the top 3 boats had scores...

A 0 (6) 2 3 0 0
B 3 (4) 0 0 2 0
C 2 2 0 (4) 0 2

Throwouts in brackets.
Note that the sailors were split into two fleets for the qualifying series, so for each race there were 2 firsts, 2 seconds, etc.

So these 3 sailors would have been tied on points on the regular point scoring system with 8 points each. But with the zero scores, the top two sailors are tied on 5 points with the third sailor having 6 points.

Both the top two sailors have 3 wins, and both won their last race so the tie is broken in favor of the sailor who won the next-to-last race under Rule A8.2.

So the sailor who had the worst race (a 6th for A) among these three, wins the qualifying series.

And the only sailor of these three who had five finishes of 1st or 2nd (C) comes third.

On the other hand, the outcome would have been exactly the same under the regular low point scoring system applying Rules A8.1 and A8.2 to break the ties.

So why bother?

Under another unusual scoring rule for this regatta, the sailors carry forward their overall positions in the qualifying series to the gold fleet as if they were one race in that fleet. So A=0, B=2, C=3.

O Docker said...

Sounds something like the system they've tried in the Tour de France, where they awarded bonus time for finishing first in a stage to encourage more aggressive racing among the leaders.

They abandoned the practice after a few years without much of an explanation why.

More here.

Tillerman said...

I don't really know much about the Tour de France, but don't they award all the cyclists in the leading pack the same time? Was this an attempt to give at least some incentive to actually win the bloody race?

O Docker said...

Yes, this is sort of an apples and oranges example. But both are events based upon cumulative score in a series of races. A competitor with a lead going into later stages could sit back. But if there's more of an advantage to winning the stage there's more incentive.

I don't know much about the tactics of winning regattas, but this may be why they're trying new scoring systems.

Tillerman said...

It's an accepted truism that in large competitive fleets you can often win a regatta by consistently good performances, without even winning a single race. To win a race you sometimes have to take a risk by committing strongly to one side of the course for example. If it pays off you might win the race. If it doesn't pay off you might end up with 20 boats ahead of you. A string of top 5 finishes is going to beat the sailor who has two firsts and then a bunch of finishes in the 20s and 30s.

This experimental scoring system may change that dynamic a bit, but I doubt it's a big enough change to make much of a difference. As the example showed, the tie-breaking rules and throwout already slightly favor someone with several first places and one bad race over someone with a more consistent string of results.

Now if they gave a race winner minus ten points, say....

Noodle said...

Just to confuse matters: Suppose there is one throwaway, then under normal low points it becomes a tie at 12 points (1,1,1,3,3,3 versus 2,2,2,2,2,2); Now A becomes the winner through A8.1 :-D

Tillerman said...

Right. The tie-break rules already give a "bonus" to someone who wins more races, as in the real example from Miami that I posted below earlier today.

Anonymous said...

Totally off topic but that picture looks so familiar. Didn't he win a Nobel prize?


Tillerman said...

Well there's a challenge. Who can name this scientist?

Baydog said...

L. Ron Hubbard

Litoralis said...

Marshall Nirenberg

Tillerman said...

I have smart readers but apparently smarter offspring.

L. Ron Hubbard was almost a scientist (apart from those few extra letters in "scientologist") and he did won the Ig Nobel Prize for Literature in 1994 for "his crackling Good Book, 'Dianetics,' which is highly profitable to mankind or to a portion thereof." Unfortunately that is not the correct answer.

Litoralis (aka First Tiller Extension) is correct in saying that the picture is of Marshall Nirenberg who shared the Nobel Prize in Physiology or Medicine in 1968 for breaking the genetic code and describing how it operates in protein synthesis, a somewhat more significant advance than Mr. Hubbard's contributions to human knowledge.

Anonymous said...

ISAF is experimenting with different race formats for the current Olympic cycle. For instance, the windsurf competition is run quite differently in Miami. ISAF wants to get input on these new formats. See this link:


Post a Comment