PHRF Numbers: A Classy Standard, With More Wild Boats!


Note: If you are a new sailor on the digital grid, let me apologize for the short article below.
It’s full of jargon, statistical terms, and obscure references.
Good Grief, it reads like a work memo!
Second Life Sailing’s PHRF is actually a lot of fun; it’s based on an old SL Sailing Hotlaps’ idea started by three very ‘real’ sailors: Cory Copeland, Cybrid Keats, and Kanker Greenacre (all smart skippers and all very funny people).
So…. if you are new to Hotlaps and PHRF, please don’t read this article!!  Instead, go read an Intro here.
You can then find a massive bibliography of links and posts on the history and updates to the SL-PHRF thing at the end of a recent article located here.
Actually that sounds too tedious… If any of that stuff in those articles sounds like you, please then come back here.
And if you’re a RL sailor and you don’t find what you’re looking for… no problems!
SL Sailing is a nice emulation of the real thing, and we have a global community waiting to cruise, race, or script the limits of your imagination…
If you have a question, just post a comment and ask! 🙂

 Last week I talked about  SL-PHRF numbers and posted a table of PHRF “handicaps” for a fairly large number of boats using the 2009 Madaket course. We switched to the “PHRF 2010 Course” on January 3rd, but I couldn’t report on those numbers at that time, since I thought it was important to change the “PHRF index standard.’ Based on the 2009 data, I thought the J-Class was the obvious choice as an Index Boat.

The ‘Index’ is the boat used for comparison; it’s a bit like like ‘par‘ in golf. The J-Class ‘average, good’ lap time on the 2010 test course arbitrarily gets a handicap of ‘1.00,’ and all the other boats are assigned handicap conversions relative to that number.  A boat with a handicap of “1.13” is thirteen percent faster than  the Index boat , and a boat with a handicap of  0.74′  is 26% slower.  The handicap conversion factors make it easy to level the playing field across a mixed fleet, at least with regard to solo lap times over the past few weeks. 

I commented last time that I needed 8-10 new J-Class data points to strengthen the ‘index’ for PHRF comparisons on the 2010 Course. Huge kudos to Trapez Breen and Chaos Mandelbrot for racking up thirteen new laps that fall right in a normally-distributed range, setting a basis for a new index. Norway and Texas fell dead-on the same statistical curve, volcano or not blocking their flight path! 🙂

Let me take a few sentences to comment on the choice of a new ‘index standard’ on the new test track.  The 2010 PHRF Course is roughly two sims longer than the 2009 course and in my opinion it is a better test track.
I guessed it would be perhaps 10% longer for most boats, given the extra distance and the  headings. Trapez and Chaos’ scores  came in a bit better! The new J-Class Lap time is 9.2% longer than 2009! Wow, the individual lap times and the time shift onto the new, longer course were close enough to endorse the J-Class as a new PHRF standard.  As we add more skippers and more data points in the next few months of course the baseline standard may shift appropriately; but this is a great start, with remarkably tight lap time- concurrence.

OK!!! In the first chart below, I repeat the 2009 scores through January 2, 2010. It shows the average lap times for the 2009 Madaket course, as well as the ‘corrected’ handicaps for all the boats sailed. Where available, the 2008 column shows earlier handicap data for a  large number of boats on several other test courses. There is pretty remarkable correlation over time and across multiple test scenarios.

Woots! OK the above table is closed, ended; it’s a dead-parrot.

We’re now switching to the 2010 course in an effort to demonstrate construct validity and reliability (we’ve done this as half-dozen times before). On the new course so far we have the Index boat scores, but we also have a total of 58 new test laps sailed on a variety of other boats!

Let me give a huge shout-out to Wally Warbaum, Colin Nemeth, Glorfindel Arrow, Francois Jacques, Lance Corrimal, Fearless Freenote, Trapez Breen, Aislinn Farella, Allie Tomsen, TaffyOcean Sommerstein, Slanty Uriza, Kembri Tomsen, Pensive Mission, Armano Xaris, Jane Fossett, Chaos Mandelbrot, Vin Mariani, and ahjep Kattun for running so many laps on the new 2010 course in a wide variety of boats.


At the moment I’m only reporting on a few boats; I’m waiting for the rest of the entries to reach statistical significance. However, if you are geeky enough, the numbers so far look pretty interesting! Although there are a few thousand  entries in the current database, there are only 58 valid laps on the current test course;  it will take several months to demonstrate the current results are consistent. 

Since we are changing the Index, for me the biggest question is whether the 2010 results correlate with the all the prior 2008-2009 lap info; actually, so far it looks pretty good!
For example, the  ACA33 v2.53 ranks a fast handicap score of 1.33, much faster than a J-Class but still slower than a Tako. That’s consistent with earlier ACA scores.

Even more accurate, the  WildWind RCJ-44 demonstrates an average lap time of 10:12, meriting a  handicap of 1.33; that’s a full 33% faster than a J-Class on the same course. More importantly, on the 2009 course the RCJ-44  earned a nearly identical ranking of 1.32, evidence the scores are valid and reliable year-to-year.

Given that very tight WildWind result, let me give a huge shout out to Orca Flotta and the entire sail team over at Triumphal; Woots!!!
They’ve logged a flood of new WildWind scores on the test track, adding several new boats to the list! Welcome to the SC22, SC 27, and SC35 v2.0!!!
And although there’s only one lap entry, let me send a shout out to Lance Corimmal for adding the TR-30!!!

Lots more to come soon!!!


2 responses to “PHRF Numbers: A Classy Standard, With More Wild Boats!

  1. Woots! A few more points over the past week:
    Although there are still only a small number of laps in each category and the variability is large for most vessels, the Trudeau Rozinante seems to be settling in with a PHRF handicap of 0.73-0.74, meaning it’s 26% slower than a J-Class on the new test course.
    If that handicap number holds up, it should also apply to most other race courses in SL.
    The new ACA33 v2.53 comes up with a handicap of 1.33 on the 2010 test course, and so far that number seems consistent. It falls between the scores for the ACA33v2.5 (1.30) and ACA33v2.52 (1.37).
    Thanks to Chaos Mandelbrot as well for adding JG-44 v1.03 laps on the new course; based on a small number of runs, Chaos registered a lap time of over 18 minutes, yielding a corresponding handicap of 0.74.
    This is interesting, since a few months ago Liv Leigh and Tim Warrhol did a similar small set of laps sailing the JG44v1.01 on the the 2009 test course; they ended up with a mean lap time around 13 minutes and a corresponding PHRF Handicap of 0.88!
    That’s a big difference, and it’s not clear if it’s due to skipper bias, the new boat version, or a result of the 2010 change in ‘index boat.’
    If Liv or Tim read this, I’d love to have a few laps from either of you (and anyone else!) sailing the JG44v1.03 on the current Madaket 2010 course!! That could resolve the specific boat issue, but more importantly it could help determine if the ‘new index’ we’re using is valid.
    Thanks to everyone giving the test course a try!!

  2. It’s very Interesting you sailing work over PHRF systems inworld.
    I’ve just discover than isn’t in how fast a vessel -or motorboat, why not?- can run a course. Most the beauty and feeling than a nice and well drive skipper does.

    Regarding tech problems as causes of variety lags… is it reasonable to include or not that “undesired” runs?

    If I always regards about the replica modeling most than fantasy vessels, why we don’t take as a basis the actually “real” data from PHRF associations?

    We must remaind in the paraphrase: “SL is SL” or trying to exploit and develope accurate “Real” like?

    As permanent users, most of you longlast residents and active in “vehicles” sporters… we have that critic point to improve this “inworld”…
    So, keeping this valuable testing time, isn’t too fool take as reference for designers and creators the “Real” PHRF data?
    Are you on this way?
    Are taking contrastated dates from TYC or whatever Hotlaps?

    All kind simulator designer, Linden Labs included, are regarding into an universal weather system “inbuild” in the main program instead divergents investigations by your own?

    Why not extend side by side wheater systems that can replicate “Real” weather zones in “Real” time?

    Are you thinking about any ID+i institue to focused whole efforts of designers instead to promote division between an only landscaping/weather system?

    …Honestly, i can’t sleep well after “replicate” any event under “perfect laboratory” conditions…

    Haven’t any to say?
    Thank you in advance

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s