Andy's Blog.

Andy writes (mostly) about sailing, and (sometimes) about dogsledding, running, cycling & other fun stuff. Here you'll find lots of informative articles on ocean sailing (primarily), and entertaining musings on life & adventure. This blog is syndicated on SAILfeed.com.


The 'Black Swan' Method of Outfitting the Ocean Sailing Yacht

Installing the Cape Horn windvance on Arcturus.

It's 5:14 in the morning. I'm at the desk in the back of the ARC Office in Rodney Bay Marina, St. Lucia. Mia and I are on night watch here - 0200-0800 - and I'm inspired.

Earlier, I was in the process of editing the next 59 Degrees North podcast. It'll be Episode #11, and will focus on breakages and jury rigs in the 2013 ARC (the subject of my last article on here in fact). In the midst of the work, I got the idea for this column.

The ARC fleet attracts so many big and beautiful and complicated boats, that it's no wonder that stuff breaks during the course of a three-week passage. Not every boat, of course, but with a whopping 234 taking the start line, statistically, stuff is bound to go wrong, particularly with all the gear these boats carry. The most well-prepared and the best sailors definitely suffer less damage, but sometimes even bad luck can wreck your day. 

I've been obsessed lately with author Nassim Nicholas Taleb, a modern Lebanese philosopher who's dabbled in the financial markets (and made a fortune) and who spends a lot of his time thinking (and taking long walks in the wintertime without a coat). Taleb's most famous work is The Black Swan, and it is famous because it essentially predicted the financial crisis in 2008. But it's more than that.

A 'black swan' is a phrase used to describe rare events, positive or negative, that have large consequences. That is, events very unlikely to happen, but with extremely large consequences if they do. 

The phrase literally came about during a time when black swans - the waterfowl - were thought not to exist - as rare as a black swan. In fact, they do exist. And then there is the idea of proof - observing all the white swans in the world does not disprove the existence of a black swan. Furthermore, just one black swan does disprove the notion that all swans are white. So you kind of get the picture.

Anyway, a 'black swan,' or rare event with major consequences, can happen in all parts of life. Taleb uses the phrase most often to describe events in the financial markets, but it's applicable elsewhere too. Think of 9/11, perhaps the worst example - the destruction of the Twin Towers could never have been predicted, and the consequences of that destruction have forever altered American history at home and abroad. (Alternatively, on the positive side, the rise of the Internet is a black swan - nobody could or did predict it's impact on global society).

Generally, 'black swans' deal with the concept of outliers. Statisticians like to remove outliers from their samples - they wreck their models. Think of your standard high school bell curve - the highest and lowest grades in the class are removed, and the remainder are assigned along the bell curve - a few A's and F's, but mostly B's and C's. Same with a golf handicap - a 10 handicapper might be just as capable of shooting 72 as he is of shooting 90, but his highest and lowest scores are removed as 'outliers.'

Wrongly, according to Taleb, who argues vehemently against this type of thinking. He says the outliers are precisely what makes the world work. You've got to read the book - or head to his website - to go deeper into this, but the implications are profound. I'll stick to my original thesis here and apply it to outfitting an ocean sailing boat. 

In this context, then, we can think of black swans this way - if an 'outlying' event (read 'failure') involving a fitted piece of gear onboard will dramatically alter a voyage, no matter how statistically improbable the event ('failure'), there must to be a contingency in place to deal with it, or that piece of gear should not be fitted.

Take roller furling. I'm going to be in the extreme minority on this one, I realize that, but Arcturus has hank-on headsails. I even went so far as to sell the Harken gear we did have (which was nearly new when we bought the boat) in order to retrofit the old-style sails. The reasoning for me was simple, and follows this 'black swan' logic - if 99% of the time the roller furling unit works flawlessly, but 1% of the time it fails catastrophically, that's enough reason for me to live without it. (And at this stage in my life, I don't mind - in fact enjoy - going on the foredeck to change headsails. It's a price I pay for reliability).

The consequences of a given gear's failure are what you need to look at. With the furling example, the worst failure would be a partially furled headsail that can neither be furled completely, nor let out all the way in order to lower the sail on deck. This happened on several boats in the ARC fleet this year, both on jib furling systems and in-mast mainsail furlers. In both cases the sails were shredded by the time they made it to St. Lucia, and in one case the skipper feared he's lose the rig with all the vibration and shaking on the forestay that the partially furled genoa caused. Not to mention it ruined the beauty of the perfect downwind conditions they had had for the last six days of their crossing, not a minor consideration when you think of sailing as being as much a philosophical pursuit as a physical one.

In 2009 before the Caribbean 1500, the same thing happened inshore on a Hylas. The furler had stuck halfway, and the forces that the crew had applied to it to unstick it actually did cause the headstay to part (the mechanical fitting aloft had unscrewed itself). Thankfully it was in Hampton Roads before the start, and they didn't lose the rig. They installed a new furler before leaving.

Now I understand that maintenance can remove nearly all the risk in these sorts of things. Nearly. With any complicated system, you'll never remove all the risk. A hank-on headsail system is as simple as hoisting and lowering, and the consequences of it failing are far easier to live with. If the halyard were to get jammed, cutting the line will drop the sail to the deck. And so on.

Look at steering systems. On Arcturus, we retrofitted a tiller in place of the binnacle steering wheel. My dad, only half-jokingly, said 'you guys are setting back yachting 50 years!' 

To me, it was just smart, pre-emptive risk management. Losing the steering cables halfway across the Atlantic would have at the least required me to fit new ones offshore (assuming I'd had them pre-made to fit before we left), or devise a jury rig. Or worse, use the emergency tiller, which is downright impossible over anything more than 2 knots, which one ARC yacht found out this year. With a tiller directly connected to the rudder head, there's less to go wrong. And the consequences of the tiller itself breaking are easier to handle, as a jury tiller would have been much easier to fasten. 

You can - and should - apply this thinking to every system that goes on the boat, particularly if you plan to sail far from land. Here's a few more brief thought experiments regarding equipment:

  • Watermakers: the risk of a properly maintained unit failing is probably low, and likewise the consequences of it's failure is low (assuming you have enough in the tanks, or the ability to catch rainwater, which, I admit, might be a big assumption). It's more a convenience issue. Conclusion: if you've got the money, and the patience to maintain it, fit it! (I have neither).
  • Refrigeration: risk of failure is minute, consequences minute, again assuming you have enough dry stores. Mia and I have lived without a fridge for long enough - we crave yogurt and milk when coastal sailing. Conclusion: fit it! Arcturus is getting a new fridge next summer.
  • Single-line reefing: risk of failure probably low, if properly installed. Consequences of failure high if you can't reef, not to mention the myriad minor issues with lines twisting inside the boom and the enormous friction in a poorly designed system Conclusion: scrap it! Slab-reefing where you have to go forward to hook in the tack works just fine and is more robust.

And on and on.

Autopilots are another of my favorites. The likelihood of a modern, properly designed and installed electronic autopilot failing is very slim. But the consequences of that failure can ruin a voyage (maybe not safety wise, but certainly from a standpoint of enjoyment, for there is nothing I dislike more than being forced to hand steer on all of my watches. It's absolutely exhausting when doing it double-handed).

Some risks simply cannot be mitigated against. Any offshore sailors worst nightmare is hitting something in the night - a whale or container. The likelihood of such an event is minuscule - but the consequences can be life-threatening. But we sail anyway (myself with the best liferaft money can buy as a last resort - it won't mitigate the risk, but it'll save my ass if it happens).

So this then, is the 'black swan' theory in action when it comes to outfitting an offshore boat - assessing not the likelihood of gear or systems failure, but the consequences of it, and deciding if you can live with them. If you can't, that piece of gear shouldn't make it onto the boat, or you should have a simpler, more robust system as backup (a windvane backing up an autopilot, for example). It's as simple as that.

Though I wasn't conscious of it at the time, our refitting of Arcturus followed precisely this way of thinking. And I'm happy to say that it paid off - in the 23-day Atlantic crossing from St. Pierre to Ireland, we didn't have a single gear failure. In fact, the only problem at all was a small tear in the foot of the mainsail cause by it chafing on the boom during the calm spells.

Click here to see the full slideshow from the refit of Arcturus.