Following from this discussion I've prepared an example workup of individual-boat performance handicapping using most of the races run so far in 2011.

Your SOL Performance Handicap (SPH) is a relative measure to compare within the fleet. It reflects only race results. Unlike most IRL sail handicapping systems, it rates the crew rather than the boat. This approach is similar to what is used in "sports" like golf and the ultimate goal is to reduce your handicap to 1.000 or "scratch". The calculations are very simple and don't need human intervention (like I've been doing for the Tallships). SPH could be automatically added to the leaderboard and boatlist in-client.

So what's the point?

• Perfomance handicapping rewards improvement. Every race that you sail "above level" counts towards your SPH.

• SPH allows you do identify your immediate competition. With the new custom boatlists, you can choose to highlight (for example the 20 or so) boats nearest to your performance.

• Corrected time results reflect effort. If you really give it your all one race you'll see it in the results. Remember that other boats are rewarded in the same way - if you make the top few dozen there's a reason!

• New sailors have immediate feedback on how they are doing. New boats also have a bit of a honeymoon period where they'll correct out fast until their handicap settles down. This is good motivation.

• Improving your SPH gets progressively difficult. The closer to scratch you get, the harder it becomes. Finishing first overall (or even top 20) may not be enough to improve your SPH.

• Unlike the SYC rank, SPH does not rely on continuous participation. If you take a break from SOL you pick up where you left off.

I've made a few assumptions in order to get reasonable results with a limited data set, and the weighting system should be more sophisticated than what I have here. Results presented for discussion (and feel free to use them to highlight your competition).

A few things to look for: if you have a recent race where you feel like you put in an above average showing - check the corrected time ranking to see if the system caught it. SPH works basically the same way as a Time-On-Time rating. Divide you elapsed time by your SPH to see how you did vs the race leader in upcoming races.

The included spreadsheet contains SPH numbers for any boat that has at least 1 top 100 race this year. The more races you've sailed the more accurate it should be. The second sheet shows top 50 boats on corrected time for the races included. Yes there are some new names in there! That's another point.

Interesting workup, still trying to absorb the details as I have not been concerned with rankings on more than a race by race basis.

One question I have relates to whether you exclude any results or bottoms which may indicate lost interest or lost internet connection, went out of town issues which may not reflect actual performance ability?

The Texel race was one of my better efforts that I can recall finishing 12th; 2 min 10 sec behind the winner. If I am reading the chart correctly I am placed mid 30's corrected. Was I expected to do better than I did or were others moved forward due to them sailing above their norm ?

thanks for letting me down easy

I've only included results where a corrected time suggests better than current SPH. There is no way to increase (worsen) your handicap, so cruises or DNFs don't factor in.

The handicapped results are quite compressed, and anything in the top half is a good result. Winston's 2nd in the Ron Miel was only good for 47th out of 100 for instance - a middle of the pack result implies that he was sailing to his rating, but not above it.

In your case, your 9th in the Ron Miel and 15th in Scan 1 actually count more towards SPH than Texel. There are few ways of looking at the results... in the Texel race, you were only beaten on corrected time by two boats (p1 and 2) with SPH lower than your own. This suggests that your position is much more other boats doing well vs their own numbers. With only 30 races and incomplete leaderboards, boats that make it into the top 100 for the first few time are promoted.

Texel is a bit of a wildcard as it was very short and relatively point-to-point. It was very difficult to _place_ well in, but many boats were able to finish closer to the leader as a % of elapsed time.

There's no differential weighting in the example, and all fleets are assumed to be of equal strength. Something to consider once more races are included.

--- Last Edited by 76Trombones at 2011-07-12 22:56:28 ---

Very nice effort 76T and I really hope something 'official' comes of it - I think something like this would add enormously to SOL...
A little server support and this can very easily be added to the boat list(s)

That said I find it interesting you have numbers > 1... Any performance handicapping system i have seen uses values < 1, which are multiplied by elapsed time to get corrected...

Interesting also in that races I (thought I) really put a lot of time into and got decent results I didn't figure very well in - clearly others were working even harder!

Thanks for your efforts 76T, I will have a good look at this.
I realy like the idea of ranked and handicapped races and leaderboards.

Yes it's backwards based on normal Time-On-Time conventions. Huh - something to do with Tallships (how come no one has pointed this out? ) and my brain being wired for Time-On-Distance... Lucky multiply/divide are easy swaps!

For anyone who might be worried, the idea is that these results are totally supplemental to and in no way replace the normal leaderboard/SYC ranking.

As mentioned in the referenced thread, over time cumulative corrected-time results converge. The interesting bits are evaluating race-by-race individual performance, identifying immediate competition, and tracking and encouraging long-term development and participation.

================================
And SYC Memberships to ideally ??


--- Last Edited by Aaron Gage at 2011-07-13 15:59:22 ---

76T... brilliant as always!
Having re-read your post several times and reviewed the spreadsheet, forgive the dullard question(s).

1) If there is no penalty for missing races, could someone race in a single race per year, WIN it (double entendre intended)and maintain scratch rating? (extreme example)

2) How come I have 2 ratings on the spreadsheet... If asked, I'll take #100, thank-you-very-much!

This is a fabulous idea and another example of how the SOL community is unlike any of the other sailing sites I have encountered.
Positively Outstanding!

SP

There needs to be some discussion about how exactly to calculate changes (SPH working group?) but as presented, your rating is always maintained. No way to adjust it slower. Winning a race isn't enough to progress as you need to win it with enough margin. The ratings should probably adjust relative to "fleet strength" but that would be a very minor change up or down even over many months.

2) How come I have 2 ratings on the spreadsheet... If asked, I'll take #100, thank-you-very-much!

Haha it looks like when I imported the results some of your data came in with a space in front of the username. No avoiding the faster one though. I get the same result with the two sets merged.

This is a fabulous idea and another example of how the SOL community is unlike any of the other sailing sites I have encountered.

Thanks! I *think* it has potential... quite a bit of added value, and it tackles some outstanding issues as well in a roundabout way.

As an extreme example of how competitive SOL has become, I figure (with a few missing updates) that UC and rafa were within 10min of a "perfect" time in the Atlantic crossing. If that's not a scratch performance nothing is! (to 4 decimal places...)

Anyone know a quick way to get detailed results? I could do this by copying and pasting from the results section, but would be faster if a better data format is available.