Monday, March 23, 2015

Persistent Shifts Part 2



In this post I present some more thoughts about persistent shifts before I get into a few more complicated scenarios in upcoming posts, then at the bottom I have a short glossary for some terms, not all of the terms come up in this particular post, but the ones that do, I have put in bold.
 
Assumptions

As we get into a couple more persistent shift posts, I should mention that along with the invalid assumption that tactics are not a major consideration, for simplicity I am also assuming that all wind shifts happen across the entire race course evenly, so the wind direction varies only with time and do not vary according to your position on the course.  As we know in reality your neighbour can get a shift that never comes to you or that takes its sweet time.  Maybe that should be another blog post.

Examples

The real world examples of opportunities to use persistent shift strategies that I have in mind include, for example, on the prairies when a large thunder cloud passes by the lake, or on the coast it could be a slow and transition from the gradient wind to the thermal wind.

Scale

The time scale of a shift or a pattern of shifts is also very important.  When I coached in the 2015 Jensen Beach Laser Masters Regattas, there was a very slow persistent shift on the scale of multiple hours that could be seen clearly on the weather prediction websites as their website’s predicted direction arrow turned steadily around the clock through the hours and days.  Sure enough over a day of racing the course often shifted by 40 degrees, but this overall trend was generally on too large of a time scale to be used as a reliable strategy on a particular leg of the course.  The winning strategy came from playing the many and sometimes drastic smaller phase wind shifts that did not correspond to the overall macroscopic trend, they had their own causes separate from the trend that the wind prediction websites picked up.  More examples of time scale considerations are to come in the next posts.
In the diagram, I have drawn a scenario with a start line and no windward mark, so this is like a training scenario where the boat that can cross ahead of the others is in front (no worry of overshooting a mark).  The course that I have drawn (again built from arcs of a large circle as I described in the last post) has two boats A and C starting on starboard and one boat, B starting on port at the pin.  After a while of sailing the identical course and tacking at the exact same time A and B are still the same distance up a course that would have been square to the start line and square to the initial wind direction, but that reference is no longer relevant.  The red ladder rungs show the effective position of boats A and C half way up, then I have drawn ladder rungs for A, B and C at the end.  A is quite a long way ahead of C just from having started at the end of the line closer to the eventual wind direction.  B also starts off with this potential advantage, but ends up sailing with the wrong strategy.  B is probably thinking: “Great, I am on the lift!  I am on the lift!”  And if the wind was going to eventually come back right, B would eventually end up miles ahead, but since the wind just keeps on (persistently) pivoting left or backing, it is bad news for B.  By half way up B tacks to rejoin the other boats and solidifies her losses.

My main point here is that if you decide that the wind will shift one way or other, it is urgent to get to that side so that when the wind eventually comes from that direction you are the upwind boat.  There is nothing much new in this post, but I just want to lay a solid foundation before I start bringing in some weirder scenarios.
Here we are waiting for wind and sheltering from the sun at one of the 2015 Jensen Beach Masters Regattas that I mentioned

Glossary

Backing:             The wind is backing when it shifts left or clockwise.  This can also be thought of as a lift on port.    Using the terms ‘veer’ and ‘back’ is a great way to communicate clearly with your training partner and coach about the wind shifts experienced during a split tack. 

Ladder rungs:    Lines drawn perpendicular to the wind or to the rhumb line of a windward leg (the rhumb line is hopefully lined up with a good guess of where the wind will generally come from in the race).  Ladder rungs can be drawn on the water with boats passing or climbing from one evenly spaced ladder rung to the next as they zig zag up the course.  Ladder rungs can also be drawn extending from one or more boats to show which boat is further to windward or further up-course.  The distance along a ladder rung is leverage. 

Leverage:            The distance between two boats if you measure only along a ladder rung, or said another way, leverage is the he across-course component of the distance between two boats.  I have also heard leverage called lateral separation.  The opposite of leverage would be distance directly upwind or downwind.

Median wind direction:    The middle wind direction.  This is not the official definition, but it is what I mean, in Laser racing most calculations are done by intuition, even if you have a compass there is only so much mental math you can do while doing everything else.  I have not differentiated in these persistent shift posts between the statistical median and the statistical mean or for that matter the mode.  Technically if you want to calculate the median you take the two extremes of the wind direction and average them together. 

Noise:                   Quick, relatively insignificant changes in the wind direction or in some other signal.

Phase:                  In an oscillating breeze, once you have established your median wind direction you can tell that the wind is in a ‘right phase’ when the wind is to the right of the median and in a ‘left phase’ when it is to the left of the median.  You can also be in phase or out of phase with other boats if you are on the same tack as them or on different tacks respectively.  In strong fleets, staying in phase with the fleet is usually the same as staying in phase with the lifted tack, for example being on starboard in a right phase.  One fairly common scenario is to have smaller shifts coming through at a fast (higher) frequency at the same time as larger shifts that are oscillating at a slow (low) frequency.  In this scenario we have a smaller phase and a larger phase coming through at the same time.  Sometimes the smaller phase can be thought of as noise.

Pressure line:      Increase of wind velocity usually visible as a dark line of approaching rippled water

Puff:                     Localized increased wind velocity usually visible on the water as a patch of ripples

Rhumb line:             The imaginary line straight from the last rounding mark, or from the start line to the next rounding mark.

Strategy and Tactics:       These terms are often used interchangeably, but I am differentiating between them in these blogs.  Strategy is how you plan and respond to conditions like wind patterns, wind strength, waves and current before you take other boats into account.  Boats at the front of the fleet often free to execute their strategy.  Tactics are considerations relating to other boats and they should also factor into your plan.  Dirty air, leverage and rules situations are examples of tactical considerations.  Boats in the middle to back of the fleet often have trouble executing their strategy because of all of the tactical situations that they get caught up in.  In this blog entry, I am trying to focus mainly on strategy.

Veering:               The wind is veering when it shifts right or counter clockwise.  This can also be thought of as a lift on starboard.  Using the terms ‘veer’ and ‘back’ is a great way to communicate clearly with your training partner and coach about the wind shifts experienced especially during a split tack.

Thursday, March 19, 2015

Persistant Shifts Part 1


For a long time now, I have been promising to talk about strategies in persistent shifts and I have now been working on a post for a while, developing it and researching it, but as a result it has grown into a behemoth and it is quite technical.  So this blog should be more reasonable and I may release the more technical stuff that excites me in the upcoming days.

Before I get into persistent shifts, here are some common strategies that generally won’t get you in too much trouble on an upwind beat, especially in a square or well set course and oscillating breeze.

-Stay on the tack that keeps your bow pointed closest to the windward mark

-Tack on the knocks 

-Tack when you are knocked past the median wind angle

The interesting thing about persistent shifts, however, is that by blindly following oscillating shift strategies in true persistent shift situations, you lose out to someone who has recognized the persistent shift and adjusted their strategy appropriately.

In oscillating breeze we sail on the lifted tack because by definition we are expecting that at some point on the leg the wind will oscillate back to the other direction.  We zig to one side of the course on the lifted tack expecting that the wind will at some point shift at least part way back so as to let us zag back in the other direction also on the lifted tack or at least on the median so that we can make it back to the upwind mark before the layline.

In persistent shifts the wind never comes back, or it never quite oscillates back far enough for the zag in your zigzag to pay off.  The trick in a persistent shift is to get to the side that is going to be (more) upwind before the worst of the shift happens.  That way when the wind shifts to your side, or as the wind gradually moves to your side, you end up upwind of other boats without having had to sail that distance upwind as shown in the diagram below.  I made this diagram from cutting and pasting arcs of a large circle to show a perfectly even persistent shift.  The reason the starboard and port courses are not parallel is because they are sailing their tacks at different times.  By the time the boats are at the top of the course the wind is further left than at the bottom of the course which is the main point behind persistent shifts.



With persistent shifts even more than normally, you have to be careful about laylines.  To take another example shown in the second diagram below, imagine you sailed to the right layline before the shift.  When the shift that you have been calling for comes in, it puts you to windward of your competition, but it also changes the layline.  Even if it was a big shift you have made almost no gains on someone on your side of the course who stayed well below the layline and has now been lifted up to the layline.  Both of you get to sail straight at the mark.  Similarly in a gradual persistent shift, you have to ask yourself “how far to that side do I think the wind will go?”  And then try to tack onto or a bit below what you think will eventually be the layline when the wind fills in as predicted.  

Here is a more successful alternative to the above diagram: 

You have a ‘good reason’ to think that a shift no bigger than say 20 degrees could come out of the right side of the course.  Maybe there is a wind line on the horizon that matches up with a weather prediction that you heard.

You generally tack on the shifts, but whenever you are close to the median wind direction you work your way right. 

Your predicted right shift has still not come in and you are, say 20 boat lengths from the layline, so you tack onto starboard below the layline and wait.  Eventually the shift that you called for comes in and it is a satisfying 15 degrees.  That happens to be enough to lift you so that you are now pointing about three boat lengths below the mark if the wind stays in that direction for the rest of the leg.

The wind does stay because of your ‘good reason’.  You have now objectively gained about 17 boat lengths on someone who was on the same ladder rung as you before the shift (ie you were the same distance to leeward of the mark) but who had no leverage on the mark (they were more or less directly to leeward of the windward mark).


Terminology

If someone said: “take the knock to get inside the shift.”  Would you understand the jargon?

If you know that a significant shift is coming in from one side and staying, then is very important to get to that side of the course before it happens.  If you want to make gains on the fleet, you actually want to get separation on the other sailors (take a look at exactly how much by using the Shift Calculator below).  If you have a big enough shift coming it is sometimes worth taking a header (often called a knock) to get more leverage once the big one comes in rather than tacking on the small shift and staying in the middle of the course.  Or in a gradual shift, it is necessary to sail the beginning of the shift as a knock so that when you tack you will be lifted, and more importantly, you will have made it towards where the wind is coming from next (refer back to the first diagram).  Some people call this use of leverage (getting towards the wind shift before the worst of it happens) 'getting inside the shift', especially if you are between the wind shift and your competition.

I used little excel spreadsheet using the math of a simple triangle to produce a 'Shift Calculator' to approximate the gains or losses that you can make on shifts according to your leverage.  In the first box you enter the lateral separation between you and your competition in boat lengths or whatever other unit you want (this is your leverage), then in the second box you enter the size of the shift and it must be in degrees.  If you enter a positive number for the shift, it mean that the shift went to your side (you got inside the shift).  Negative numbers entered for the shift mean the shift went away from you (you were outside of the shift).  The third box gives you an output in boat lengths or whatever other unit you used: positive output is your gain and a negative output is your loss.  This approximation gets worse and worse the more windward-leeward separation that you have, but it still gives you a good idea.  The file is hosted on my Google account and it is yours to download, use, modify and share as you like by following the link below the screenshot.

  












The worst result of you reading this blog post would be if you went out sailing next time and started treating every oscillating shift as a persistent shift because with persistent shift tactics in an oscillating breeze you would actually be sailing away from the next shift and away from where the wind will eventually be coming from.  So the big question is:

          Are the wind shifts Persistent or Oscillating?

And the answer is!
          Oscillating!
(If you are not sure, assume it is oscillating, if you are sure that it is persistent then there you go)
I have had many athletes come to me after a bad race or two and say: I can’t figure out whether the wind shifts are persistent or oscillating.  Every time I tack on a shift I then get lifted so I think it is a persistent shift, but when I try to eat the header to get inside the shift I tack and then get knocked again!
In an oscillating breeze, getting lifted is not a bad thing and usually doesn't mean that it is a persistent breeze, if you are on the inside of the shift (good leverage) getting lifted is great, but even if you are on the outside of the shift  the lift is not necessarily bad because you are expecting the wind to come back to the other side as it oscillates back.  When it does come back, this time it will be your side that has good leverage and is inside the shift.
If in doubt assume the breeze is oscillating.  Have patience.  The wind will probably come back, you usually won’t go too far wrong if you sail on the lifted tack.  If you are unlucky and the wind doesn’t come back before you get to the mark, don’t freak out and sail the next leg or race as a persistent shift, look around, ask around.  There has to be some pretty compelling evidence of a persistent shift before you break out the persistent shift tactics.

When should you consider using persistent shift tactics?

 When you suspect that the wind is clocking over to a new direction AND:

-The race committee has run a change of course and set a new windward mark

-The wind forecast was for it to go in a certain direction and now there is evidence that it is starting to happen, like a change in the clouds or a wind line

-Reliable locals say something like “the wind often goes more and more left through the afternoon in these conditions” and then it does start happening

-There is a geographical feature in the race course and someone uses it to their advantage (though I am mainly considering non-geographical wind shifts here)

Don’t be the first to guess 'persistent' if you are not the local –leave that to the corner bangers, they need their moment of glory because the rest of their races aren’t so hot.


While I was training for the Miami OCR, working with Lisa Ross, she told me something that stuck with me:
I was explaining my bad race and I said something like ”I was predicting that there would be a leftie and-“

Lisa interrupts and says: “Don’t get caught predicting!”

I am mainly focused on strategy considerations in this blog post, but if you have reasonable boat speed compared to most of the fleet, just stay in phase with the other boats and be observant.  If someone takes a flyer out right make a mental note of it and notice whether their strategy of predicting a persistent shift (or some other anomaly) on this leg works out, but also notice if it doesn’t work.  Let the crazy or the desperate people take the big risks, stay near the bulk of the fleet in clean air and keep gathering information: who crosses who when they split and come back together?  Was that repeatable?  Make some theories about what is happening to the breeze, share them with a training partner between races.  Do any of the theories hold water?  Only once you have good evidence based on your experience and observation of boats, weather and race management, only then would you execute a more aggressive plan like playing the conditions as a persistent shift.  But then that is not really predicting, it is acting on evidence.  Until you have a sound, well-supported reason to break with your pack of boats, stick with them, sail fast and win the tactical battles.  When you have a good reason for your strategy that you could explain to your coach without them raising their eyebrows, then go for it, but if it is a reasonable theory some other boats will probably also try it and you should still be in touch with a pack of boats as you act on your theory.  Remember even when studying strategy that you must be sailing the fleet, not the course.