PHRF Handicaps, Wild Winds

phrf handicaps july 19 2009

PHRF  handicaps
Here we go! The most recent PHRF Handicaps, including a few new boats, and some tighter numbers!

In the last few weeks, despite some pretty terrible grid-crossing conditions, a group of stalwart skippers added more hotlaps scores to the PHRF database. Thanks to:

Gemma Vuckovic, Francois Jacques, Carmen Foden, joro Aya, Garrick Diesel, Justbent Clarity, Heidi Stiglitz, Angus Moonwall, Dunan Wilder, Chaos Mandelbrot, Cory Copeland, LDeWell Hawker, Oliphant Ming, Julia Ceres, Liv Leigh, Triton Sands, Hal Burnstein, Blackbird Latte, Argus Farman, Vin Mariani, Rodman Mapholisto, Alain Gloster, Jane Fossett, nobuko Criss, Allie Tomsen, Angus Moonwall, Bunnie Mills, CS Price, Emme Eales, Jehan Jameson, Liv Leigh, Naeve Rossini, Nomad Zamani, Reven Fhang, Quirky Torok, Isis Rexie, Everest Piek, Arrekusu Muromachi and Masahisa Greenwood.

Here’s the new table, and it follows the same format and conventions as  previous ones. (Note: I updated the tables below on August 12 based on Lance’s suggestion to post alphabetical and performance-sorted versions. Bosth are shown below. I’ve also color-coded the boat names, based on the wind engine.)


phrf-july-19-2009-hi-low copy

Compared to the last update in early June, the numbers are more consistent and seem to be falling in line with the results from 2008 PHRF (where available). Thanks to Live Leigh, we’ve added a new boat too, the Catfish 33 catamaran. We need to get numbers from other skippers on it, but my guess is Liv’s PHRF lap probably hit it on the nose for this boat: Her lap came in with a perfect, corrected Handicap score of of 1.00 for the new fishy cat!

The JMO-60 is also new on the PHRF list, coming in with an average Madaket lap time of 8:55 and an adjusted handicap of 1.32. This makes the JMO-60 comparable racing to the RCJ-44, as predicted by the very similar polar plot results for the two boats. All three WildWind boats so far tested (RCJ-44, JMO-60, and VOJ-70) prove extremely fast, particularly compared to other boats powered by ‘apparent wind’ tested at Madaket (such as the Shelly, Catfish, and Trudeau Twenty family).

The Wildwind’s strong showing on lap testing appears to be partly due to the way it handles Apparent Wind.

Wind power.
There are three different kinds of wind most sailors think about, whether they sail in Real Life or in Second Life.

The first is called “True wind.” True wind is the force generated by the movement of air against a stationary object, such as a boat tied to a dock. It’s the same wind felt by any person standing on the dock… or by the dock itself. In the diagram below, “A” shows an example of a stationary boat, and the arrow represents a 5.0 m/s True wind. Wind is a vector with direction as well as magnitude. In this case, the wind direction is 90 degrees off the boat’s bow.

Kanker Greenacre’s Flying Tako uses a “real wind” algorithm to power the boat. Based on Kankers original scripts, several other highly popular boats are also powered by real wind, including the ACA 33, the Tetra 35, the Trucordia Yawl and the Trudeau Larinda.

Most boats move, however. That movement generates a second type of wind called a “headwind.” In example “B” below, let’s assume the wind dies and our boat’s becalmed, so there’s no True Wind. The Skipper turns on the engine and soon the boat’s traveling at 5.0 m/s. Even though there is no true wind, all the sailors onboard all feel a breeze on their face that’s generated by the boat’s forward movement through the air. That “headwind” is always coming from the bow and is always equal and opposite to the boat’s speed over ground. In “B” below, the boat sees a headwind of 5.0 m/s from 0.0 degrees.

three winds

Now let’s assume the true wind starts to blow again, and the crew puts the sails up! The boat moving through the water now sees the combination of the true wind plus the headwind, as shown in “C” above.  Adding the two wind vectors together gives you the Apparent Wind. This is important. since Apparent Wind is the wind a boat sees; it’s what actually powers the boat. Most recent boats in SL are based on an apparent wind sail engine; they represent a major advance and a closer emulation of real-life sailing.  Apparent wind was introduced in SLSailing by Owen Oyen’s ACC, and it is now commonly found in recent Trudeau boats (beginning with Trudeau Twenty), Caf Binder’s Jangars, and Mothgirl Dibou’s Fizz 3.  It’s even available in the Shelly and Leetle Cat teaching boats for new sailors.

The use of apparent wind means that as an SL sailboat accelerates upwind, the contribution of the headwind increases. The wind angle therefore shifts toward the bow, and a skipper often needs to steer away from the wind to keep the boat from stalling. For example, in order for a J-Class moving upwind to sail a course with an apparent wind angle of 40°, the boat’s angle with the real wind needs actually to be over 60° to cancel out the headwind.

This makes SL sailing a little more challenging, but far more realistic.

Three SL Wind engines
The reason I’m bringing it up in this context is my concern that the PHRF scores listed above may only be accurate within boats that use the same basic wind power algorithm. I’m quite sure that “Real Wind boats” can use the above handicaps on nearly every racecourse in SL. We’ve officially tested seven courses, and I have unofficially confirmed it on twice that many (grin, try it yourself!). I’m also “pretty sure” the handicaps work for all the “apparent wind” boats as a group. I’m far less certain it works as a comparison of “Takos ad RCJ-44s” for example, even though  the two boat classes have similar handicaps.

OK OK, the real reason I’m bringing this up has to do with WildWind boats. According to the notecards that accompany the JMO-60 and RCJ-44, WildWind boats are powered by an “apparent wind” algorithm. The boats come with a numerical HUD that reports values for the real wind and the apparent wind. However, if you get out your pencil and take the HUD values for real wind and boat speed and calculate apparent wind yourself, you’ll find out that WildWind boats  underestimate the apparent wind effects to a large degree. In the figure below,  the 5.0m/s true wind is coming from 0° and the boat is traveling on a heading of 60° to that true wind.  The boat’s speed is 4.9 m/s. The calculated apparent wind should be 8.6m/s at an angle of  30.2° [shown in yellow], but the WildWind algorithm reports it as 4 and 7.64m/s. That miscalculation makes WildWind boats much easier to handle and results in a faster lap time compared to boats that use “RL physics.”

JMO Wind Error

In the figure above I state this is an “apparent wind error.” I apologize, this is my mistake; it’s not an ‘error.’  As I stated above, Kanker Greenacre himself used Real Wind to power the Tako. Kanker’s a real sailor and knew the difference between real and apparent. He intentionally simplified the algorithm to reduce script calculations and make sailing easier for a generation of non-sailors in Second Life. I believe WildWind adjusted the apparent wind calculations to make their boats very fast and a lot of fun; I think the ‘adjustment’ in apparent wind calculation was intentional, not any ‘error.’ My issue here is just to point out that the boats are not driven by “apparent wind” but something unique, a compromise, and to comment that the PHRF scores for WildWind boats may therefore not work when applied to other boats in the list. I actually need to do more statistics to answer that question though, and I try to avoid doing statistics in August (grin).

Let me leave no doubt, however: I think WildWind boats are very fast, very pretty, and a great deal of fun. They also introduce a new algorithm, one that is a compromise between real and apparent wind. That adjustment makes the boats easy to sail, and in fact makes them pretty exciting.  Since they don’t use apparent wind, perhaps the boats shouldn’t be considered a serious emulation of real sailing however. I don’t know;  maybe that question is beyond my pay grade (grin).  Nonetheless, I admit I love sailing my JMO-60, it’s huge fun to sail and race, and because of its speed and easy handling,  I use it whenever I RD a Big-Boat Race, since I can usually make it back before most of the fleet.

For serious sailing, I’ll chose a boat that’s a closer emulation.


17 responses to “PHRF Handicaps, Wild Winds

  1. The way I read it, Wildwind is very clever not to use the full effect of the apparent wind. Because real boats do no sail as fast as they do in SL. Especially not upwind.
    If you have a speed that is say 2.0 times as fast as in RL and then compute the apparent wind, you get a very strong wind and a boat that will want to go even faster unless there is some form of damping.
    I think Wildwind simply corrects the boat speed to normal standards before computing the apparent wind.

    The old Fizz 2 used a speed multiplier for this. It could be set at the WWC setter and its default value was 1.8. By setting it to 1.0 the boatspeeds would be realistic. No matter what the speed multiplier would be, the apparent wind and all other effects would be computed according to this realistic speed. Just the actual boat movement would be 1.8 times faster. I never fully implemented it though, because I switched a few gears down in the Fizz 3.

    For a good handicap it would indeed be wise to take the actual formula used for speed computation into account. For the algorythm used will define the speed curve of the boat at all wind angles. If you would for instance have a race that would mainly be upwind and would use handicaps, you would get totally different results then by having a race that would mainly be reaching and gybing.

    But that would imply a variable handicap that is not a number but a formula. And that would hardly be workable 🙂 I’d stick to simple empirically obtained numbers.

  2. Jane explained the computation of the apparent wind very well. Now I would like to try and explain a little about the boat speeds and windangles, to throw some more light on the use and computation of handicaps.

    After being stripped from the effect of the sheet not being set to the optimal angle and the baked in limit of sailing closer to the wind than 35 degrees, this is what the Tako 2 speed function looks like:

    speed = (cos(windAngle/2)+0.5) * windSpeed;

    You can see that it wil sail faster and faster if the wind angle gets narrower. cos(180/2=90 degrees)=0, cos(0 degrees)=1.
    Just fill in the numbers and you get:
    180 to the wind: speed = 0.5 * windspeed
    90 to the wind: speed = 1.2 * windspeed
    35 to the wind: speed = 1.45 * windspeed
    This explains the steep almost straight line you get when you plot the measured boatspeed at various windangles in a curve.

    You also see that the speed is lineair with the windspeed, so a boat will simply sail 2x as fast if you double the windspeed.

    Now imagine you use apparent wind instead of real wind with just this function. You start at say 120 degrees to the wind, keep closing the sheet and you end up at 35 degrees to the apparent near tropical storm force wind. Simply because there is nothing to slow the boat down except for the line “if(windangle<35) speed = …."

    Now imagine the speed curve of such a boat: It is no longer lineair. Instead it shows a steep drop between 160 and 180 degrees (when the apparent wind will only slow down when you go faster) and whatever the windangle is, you always end up sailing 35 degrees to the apparent wind. Depending on the angle to the true wind your end speed will be faster. 120-140 will probably be the fastest and 35 degrees the slowest. (remember that 35 degrees was the fastest course on the true wind boat and 120-140 was almost the slowest)

    Needless to say that the speed of this boat is difficult to compare with a similar boat that does not use apparent wind. The outcome of the PHRF handicap will be a lottery depending on the course that was sailed.

    This Tako function is of course a basic function and most builders will probably use variations on it. The builders should be able to inform on whether or not the found handicaps are applicable in different windspeeds/windangles for their boats or not.

  3. For the record: I speak of an almost straight line to simplify things. I do know that it is actually a sine curve (cosine to be precise) with the top cut off.

  4. Just a few questions…
    How come that the Fizz 3 is “as slow” as a tradewind, and no newer fizz3 have been tested? is that because the newer fizz3 only does the newer WWC windsetters?
    And could you publish the table in two more layouts, sorted by name, and sorted by handicap?

  5. Thank you to Moth for the clear explanations of some pretty windy issues.
    I must admit I’m still uncomfortable with ‘apparent wind’ that doesn’t add up when a skipper runs the numbers in SL.
    Apparent Wind has a clear definition, and we all use the same examples to explain it to sailing students.
    For boats that use some modification of Apparent Wind, perhaps we should have a separate category to avoid confusion; maybe “Apparent Wind Lite” or “Apparent Wind (Not).”
    🙂 It’s just an idea…

  6. Thank you also to Lance for suggesting I resort the tables;
    I did so and updated them in the article.

  7. Hi Lance,

    The reason you mention is one of the 2 that the Fizz 3 is not being tested much on this course: the boat doesn’t react anymore to the style of windsetter in use at Madaket.
    The other reason gives a simple explanation as well: unlike some other boats, at 20+ knots a Fizz 3 gets severely overpowered and becomes almost unsailable. It is not made to be sailed in the conditions we test boats in for the PHRF table.

    This notion also shows a weakness of the system: it is based on the ‘old’, often-used standard of 11m/s winds in SL. The system is specific to boat types under those wind conditions. Only certain classes of boats can be more universally compared with it.

  8. And a remark to Jane:

    Thanks for mentioning me, but I do not even own a Catfish 33 catamaran. 😉
    I have seen a number of them at the Madaket line and they looked nice. But I have never sailed them.

    However I did hit in laps for Juli’s Beluga, Eclipse, JG44, Astra (new style with fizz engine) and Element. Also Surfwidow’s SWB Catboat and SWB Innovation. I see none of them back.

    Also missing are ‘good old’ boats like the Trudeau 32…

  9. Haha Liv; I’ll correct the record on the Catfish lap in the next update, and add the new scores!
    [And yes, yes… I think next month we bite the bullet and switch to an ‘index’ boat that uses apparent wind, lower wind intensity, and change Madaket to a WWC setter.

  10. With respect to whether or not the RCJ/JMO/VOJ PHRF scores are valid or not, I think they are. They are sufficient for handicapping purposes, just as they are sufficient for boats that run with real wind or apparent wind. To wit, the Wild Wind formulas simply create a hybrid between real wind and apparent wind. Their performance results are without question.

    If there *is* something to question, it’s realism. You touched on this in your article. It’s not realistic (or accurate) apparent wind. However, the fact that we have boats that ignore apparent wind entirely, and that we have varying implementations of wind shadowing, realism doesn’t really matter for empirical results.

    So, PHRF indices are valid and accurate for empirical performance of the boats in SL.

    If realism matters to you, then don’t sail non-apparent wind boats or Wild Winds. You may want to stay away from the forests of 8′ tall women, too. (I’m 5’3″. I can stare at many women’s navels standing straight up. In heels.)

  11. Naeve said: “So, PHRF indices are valid and accurate for empirical performance of the boats in SL.”
    I think thast’s one of my key questions here. I’m not sure Madaket results can accurately predict performance when comparing boats with different sail engines across the diversity of race courses.
    Let’s try it out!

  12. I don’t know what 8 ft. is like. Last time I checked my height in SL I was only 1.99 m’s tall, which is ok for a woman as long as you live in a country where Napoleon never came and that is still clueless about SI.

    Maybe I need to understand what the problem with apparent wind boats is as well… as in the situation we test them (22 knots or something) they will all default to a certain ideal true wind angles as much as ‘real wind’ boats do. As much as we can say a tako runs 8 m/s upwind at 35 degrees upwind ‘real wind’ we can say the same of an apparent wind boat if we look at it s absolute performance. Unless the formulas used for both types give different outcomes with varying wind speeds I see no issues here.

    And from my simple experience as racer: I navigate the same in ‘apparent’ wind boats as I do in ‘real’ wind boats. I can still use minimap for apparent wind boats to point to a marker and get there, ‘real’ wind values will still have a fixed optimal value in the end. The only thing apparent wind ever changed in SL is that you have to sheet more often and it requires a little more attention and training to maintain that straight line to ever the same goal.

  13. What Liv said.

    Really, for the everyday sailor and racer it doesn’t make any difference. I don’t care if my boat uses real wind or apparent wind. Once I’m at a startline my boat catches racewind … and I have to deal with it accordingly. Some boats are dumbed down, some are bitchier to sail, some are nice and easy for the casual sailor, some need your attention and skill all the time (sneeze and you lose).

    It all depends on what YOU personally prefer and want. There is no better and worse in this case.

    As long as my avie is a good 20 years younger than I and can fly and lives in her airship on a jungle platform in the sky, as long as I can just have a very expensive Admirals Cup yacht always at my disposal, a RC44 and a Open 60, and delete them after every cruise or race, as long as I can change my height, my face, even my sex (if I ever wanna do that, but why?), and hovering above some startline while discussing realistic sailing …

    … as long as I’m able to do all these things I say FUCK REALISM!!!

    I know it’s not that simple since sailing in SL is already a betrayal of SL itself, a step back if you so want. Why pay a lot of money for boats and go thru all the hassle to sail from A to B when I could just as well hit the TP button?

    Sailing like in RL would just be too boring and uncomfy for most SL sailors, so a compromise is in order. It’s now only a matter of definition where to find this compromise.

    Avatar Height v2: Orca Flotta is 187cm (6’2″)

    That is without heels. Totally unacceptable unrealistic and still perfectly average height for a normal SL woman.
    I have to admit tho that women in the sailing community are more realistically sized than throughout the rest of SL, apart from “100% Joly” maybe 🙂 At least we try to keep it down a notch to reach a slight touch of realism.

  14. I admit I’m not very much interested in video games. I like SLSailing because it’s a progressively close approximation to RL sailing.
    Building that emulation in a way that fits within the perceptual constraints of this medium is a pretty daunting task, I admit.

  15. The issue here isnt avatar height or striving for realism. It is whether or not PHRF handicaps can be used for races with multiple boats.

    I think they are fun to collect and can very well be used as long as you are careful.

    In RL the speed curves of boats are much more similar to eachother than in SL. Also you should be aware of the fact that in SL some boats sail 2x as fast in a double windspeed and some boats don’t.

    Maybe it would be wise to introduce categories and maybe reduce the windspeed a little to 8 m/s. It wouldn’t affect the handicaps much for most boats in the table. Only a few need new data.

  16. Pingback: Beating to Windward « Metaverse Sailing

  17. Pingback: BBK 137 and Nemo: Two Free Tako Trainers « Metaverse Sailing

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