Any thoughts on the merits of the ORC handicap racing approach

Was there a sea change in racers perception about the accuracy and precision that handicap ratings can fairly deliver to racers? Again… back in the day… I spent a lot of time implementing USPN’s 4 windspeed ratings for beachcats. Everyone understood that the different cat classes flew a hull at different windspeeds and they wanted this non linear performance reflected in the ratings they used. They had no concept about the precision that is inherent in a sailboat race. Now the accuracy of a ratings table of 4 windspeeds without the 100 vetted data points was bogus (and for many classes, just a percentage correction to seed the table). The sailors in the US traded the simplicity of a single number system USPN DPN or SCHRS for the perception that wind speed generated a more accurate handicap race result. By contrast the Euro sailors using Texel and SCHRS calculated two wind speed ratings tables but found that most clubs used the single number SCHRS (or Texel) systems for simplicity. They argued that the inherent noise in a sailboat race made all of the complexity inherent in selecting wind speeds somewhat useless.

A second major factor in the accuracy of a handicap race result was the impact of course selection. Now multihull sailors assumed that the continual updating of the race result data would adjust the ratings for the evolution of the kind of racing actually sailed. Dixie Portsmouth Assumed equal amounts of reaching, running and beating……aka the classic Olympic course. W/L courses are not that. The reality of course was that ratings for most classes never caught up. The RYA system that the US now uses is much much better in their implementation. SCHRS is a measurement rule that assumed W/L racing and other course configurations were more or less fairly handicapped with a single number. Course selection was determinitve when racing spin and non spin boats on handicap. An extra leg with a finish to windward would screw the spin boats on handicap.

Clearly ORC racers feel the same way about their boats performance in low medium and high winds AND the impact of the course used on the accuracy aka the fairness, of the handicap results. Since an ORC rating is not cheap these owners value the perception of accuracy yielded by correcting for course and wind speed and have gotten their clubs to run ORC race circles and score accordingly. My question is… In practice, are the ORC practices of using wind speed and course selection generating more accurate race results warranted or just an inherent bias that of course wind speed is critical to fair performance evaluation and of course it is worth doing. Ie confirmation bias. The world wide growth in ORC seems to indicate that owners perceive the rule as fair.

With respect to precision of a rating… ie how many decimal places are used…. I remember Sailwave yahoo debates about the interpretation of a tie between 2 boats on handicap corrected time. Should you or should you not add another decimal place? Another example…. The argument was presented this way. Suppose that in a one design race of 30 footers with 9 boats getting to the starboard layline at A mark head to tail and one boat on port…. So all boats are sailing to their rating ie the rating is accurate! The order of 1 to 9 is random or based on their starting position. …. The boat on port will have to duck 9 30 foot boats and so will be 30 secs behind the equally well sailed boats. So this is simply noise in the rating caused by the port starboard rule and part of racing…. But this factor means that getting more decimal places in your rating is silly. The inherent noise will always exist. Now in a one design race…your order of finish is the actual result and so the concept of rating precision is meaningless. Handicap racing will never be as precise as one design racing in sorting out sailor skill.
Any thoughts on the merits of the ORC handicap racing approach.

I think the attitude of racers is that multiple wind ranges and course
configurations is definitely more accurate than any single number
system. Naval architects would confirm that. The question becomes: Where
is the “sweet” spot, where any increase in complexity (the “cost”) will
not result in any significant increase in fairness (the “benefit”)? It
is a fine balancing act, with serious racers more willing to accept more
cost for more fairness.

I’d say there seems to be convergence on 3 wind ranges - the middle
range is what we usually sail in and the low and high ranges take
account of special conditions (but for some places, like San Francisco
Bay in the summer, the high is really the norm).

I’ve heard naval architects speak multiple times and they pretty much
say that no handicap is ever going to be more accurate than +/- 3 sec/mi
and practically the best you can expect is +/- 6 sec/mi. To put that in
numbers, a typical 0-rated PHRF boat sails around a W/L course in about
600 sec/mi. So, the best practical accuracy is about +/- 1%. For a TCF
(Time Correction Factor) rating that equates to 0.010 so a 3 decimal
place TCF is about as accurate as you could ever get and makes sense.
Any more decimal places is roulette and represents increased complexity
with no increase in fairness.

With modern computers, scoring with multiple windspeed and course
options is simple - input the data a day ahead and have the PRO tell you
which ratings to use. Sailwave does wind-indexed ratings and you could
easily have 9 ratings for each boat with 3 wind ranges and 3 course
configurations (or even more!). It is just a matter of setting it up
ahead of time. I’ve never used a ratings file for lookup but presumably
you could set that up as well (locally, we have made a CSV with
competitors and ratings and imported into Sailwave for each event).

As to “NOISE” in the system - locally we often see two boats of the same
model on average 100+ sec/mi apart after a season of racing. Years ago I
did an analysis of the boats at the Farr 40 Worlds and as I recall the
average from first to last was ~60 sec/mi - that when every boat had the
maximum number of professionals and the maximum new sails and the boats
were virtually identical because of VERY tight class rules. That means
that “crew skill” is going to be a far larger factor in results than any
rating inaccuracies. Most racers don’t want to believe that because that
want to blame someone or something else for their poor results. [In a
TCF system, 100 sec/mi amounts to +/- 0.100+.]

Handicap ratings cannot be more accurate than the science behind them.
But, with crew skill far more important to results than minor
inaccuracies in handicaps what is probably most important is racer
confidence in the system being used. In that regard, the multiple
windspeed and course options systems are definitely the future for
handicapping sailboat racing.

Art

And yet OCR ratings are 4 decimal places and race results are reported to the nearest second. Any idea what OCR’s reasoning for this level of precision? Is this just a form of virtue signaling to the sailor that OCR is the most sophisticated game available and that extra decimal place is proof that they are serious?

But, with crew skill far more important to results than minor
inaccuracies in handicaps what is probably most important is racer
confidence in the system being used. In that regard, the multiple
windspeed and course options systems are definitely the future for
handicapping sailboat racing.

I agree the confidence in the system is essential. Three wind ratings and ratings for different courses goes a long way to mitigate a “horses for courses” critique. My very limited experience is very positive.