not sure if you needed this, but the formula you need to input is
(20-((p-1)/(s-1))*15
the s may be incorrect, since Sailwave has quite a few choices about 'fleet size'.
you'll need to make sure you don't have a fleet size of 1 (duh!).
Art's criticisms of high scoring are of course met with Rinderle B, Chips and Sprague-Cox.
Actually, they aren't.
Let me call my observations "comments" on the built-in biases of the various systems rather than criticisms. You might very well want the bias that a particular system has and consider it "fair," in which case my comment shouldn't be considered as a criticism.
1. In my answer to Steve I was being a bit simplistic. By a "low point" system I meant a simple or straight low point system (as in the rulebook) where your score is simply the number_of_boats_that_beat_you+1; by a "high point" system I meant a simple or straight high point system where your score is simply the number_of_boats_you_beat+1.
Those types of simple systems have some obvious biases. A straight low point system strongly favors boats in smaller fleets (since there are fewer boats you have to beat). A straight high point system strongly favors boats in larger fleets (since there are more boats that can potentially be behind you). Where a series is a "regatta" (meaning a series of races on essentially consecutive days) the number of boats isn't going to vary significantly from race to race so both these types of systems can do a fine job. Where the races in a series are spread out over a longer time than a regatta (say a season-long championship series) then the number of boats may vary greatly from race to race. If so, the biases will become important and you may not think such biases are fair.
2. Rinderle B, CHIPS and Cox-Sprague all have biases in favor of larger fleets (although less than a simple or straight high point system). In other words, they award more points to boats in races that have more boats. This is a definite bias. You may or may not consider it "fair." If you were going for "no bias based on fleet size" then I would criticize your choice of either a simple low point or simple high point system or one of these three specialized high-point systems.
HOWEVER, there are quite a few folks out there who would say that there should be a bias toward your results in a big fleet (in other words, give more points when the fleet is bigger). So, the bias built into Rinderle B, CHIPS and Cox-Sprague is fair. Steve seemed to be looking for a "no bias" system, which is why I suggested the Sunset System to him (a "percentage point" type of system).
For the record, there is an even fairer way to implement "no bias based on fleet size." You do simple high point scoring for each race and then at the end of the series you divide each boat's total points by the number of points she would have gotten if she had won every race she sailed in. Each boat's score is essentially a percentage representing her "efficiency," meaning the extent to which she approached perfection. If you win every race your percentage will be 100%. If you are last in every race then your percentage will be really low (it would vary based on fleet size).
3. A bias in favor of results in bigger fleets may be "fair," but I would question it. You might get a very small turn out in rough weather. My experience in such cases is the less-skilled sailors stay ashore. If that happens, your races with smaller fleets will have more skill on average and the results should get more credit, not less. I would argue that a "no bias" system is best as it doesn't try to make value judgments about the level of competition based solely on fleet size.
4. Compensating for varying fleet size from race to race over a long series requires some thought but probably isn't particularly hard. If your average size is 10 you probably will have variation from 8-12. Some bias based on fleet size might be acceptable.
By contrast, if you are trying to compare the results of different fleets (for purposes of an overall trophy) the sizes could vary much more. Consider two different fleets (Snipes and Lasers, say) where one fleet averages 6-7 boats per race and the other averages 14-15 boats per race. If you are trying to compare those fleets to each other then a bias based on fleet size is an obvious problem. For the record, this was the problem our club was trying to grapple with when we came up with the Sunset System.
5. Rinderle B and Cox-Sprague are both "bonus" type systems. You gain more points by going from 2nd to 1st than by going from 5th to 4th. The underlying premise is that the farther up you are in a fleet during a race the harder it is to pass a boat. So, since it is harder to improve you should get more points for doing so. Such systems were common in the past but in recent decades have mostly gone out of style. [The "Bonus Point System" in the rulebook used to be called the "Olympic Scoring System" because it was used in the Olympics. Today, the Olympic Games use the Low Point System from the rulebook instead. That is probably indicative of the support in our sport for a "bonus" type system.] These types of systems are traditional in some places but if you are picking a system to start using today I would think seriously about using a system that is not a "bonus" system.
6. For the record, there were quite a few systems in the past (like Cox-Sprague) which were incredibly hard to calculate mathematically and instead were published as fixed tables. With today's computers we don't have that kind of restriction. Still, I would urge caution if you are thinking of adopting one of these complicated systems as you still need competitors to be able to comprehend the concept, if not the details, of whatever system you choose. This is a concern with any of the complex high-point type systems - such as Rinderle B, CHIPS, Cox-Sprague and the Sunset System I recommended to Steve. I wouldn't avoid these systems for that reason along but I would expect you might have educational issues with their implementation.
Art
PS - These issues keep coming up in a slightly different form every 6 months or so. There is (or at least was) a paper on CHIPS on the Sailwave Group page. Perhaps another paper discussing high-point systems in general and the four complex high-point systems mentioned above would be helpful.
On 3/10/2011 4:06 AM, malcolm clark wrote:
Art's criticisms of high scoring are of course met with Rinderle B,
Chips and Sprague-Cox.
I believe there is a minor typo in the formula above.
In his original message, I believe Steve wanted 1st to always be 20 with last place always 5 and the boats in between spaced equally in between. For three boats that would be 20-12.5-5, for four boats that would be 20-15-10-5, etc.
With a three-boat fleet the formula above would give points of 300-292.5-285. I think you want to modify the parentheses so it would be: 20 - (15 * ( (p-1)/(s-1) ) ) [spaces added for clarity]
[I think you might have to eliminate all the extra spaces in the above formula when entering it into the formula box in Sailwave. I displayed the formula that way so one can easily see the logic. Also, I changed the order slightly to put the "15" before the fraction so the logic hopefully would be clearer. Thanks to Malcolm for showing how easily Steve's original proposal could be implemented!]
I believe s=starters(including OCS) in my version (2.02 build 5) but I have this vague recollection that in a much earlier version "s" did NOT include OCS. If you are using a really old version I would double check that (I easily could be wrong).
Art
PS - In older versions I got an explanation of the formula variables when I put my cursor over the formula box. That doesn't happen with version 2.01 build 5. I had to check the help file, which doesn't explain how OCS boats are treated.
On 3/10/2011 4:06 AM, malcolm clark wrote:
not sure if you needed this, but the formula you need to input is
(20-((p-1)/(s-1))*15the s may be incorrect, since Sailwave has quite a few choices about
'fleet size'.
you'll need to make sure you don't have a fleet size of 1 (duh!).