Real Probability of Death Passaging the N. Atlantic

[boatman61]

Originally Posted by Chotu

Brioche: exactly. The numbers really need a setting. A background. A backstory. The raw numbers by themselves don’t really tell the story.

Coming from the United States to Europe, I had always figured if the boat failed completely, as long as you had enough supplies, you would still make it. Lots of boats we lose over here end up in Ireland and whatnot.

Same applies the other way.. the current will carry you across..

[Chotu]

Originally Posted by boatman61

Same applies the other way.. the current will carry you across..

Exactly.

The boats usually make it even in dire situations

[jordanbigel]

Originally Posted by JPA Cate

My goodness! Lord love a duck!

Go sail; love it.

None of us leave earth alive.

Ann

Google, what does the saying Lord love a duck mean?
An exclamation of surprise

...Or, an American Teen Comedy from 1966!
https://en.wikipedia.org/wiki/Lord_Love_a_Duck

As for the rest (which I understood), Amen!!

[Chris31415]

Originally Posted by LakeSuperior

… I was attempting to get a order of magnitude estimate on the actual risk using reasonable estimates of the underlying numbers.

I appreciate the effort and the approach. It’s an interesting topic that I also think about.

NUMERATOR (Risk): Deaths makes sense to me since it’s probably the “best” metric we have for measuring the risk. As others have said, we don’t even have good numbers for deaths.

DENOMINATOR (Exposure): Distance, time or “per event” seem like the best options. Distance would be my choice since that seems to be most common. Using time favors the slower pace of sailing, particularly compared to aviation, but it depends what you’re trying to understand.

As always, be careful with stats, especially in the absence of good data.

“there are three types of lies: lies, damn lies and statistics"
Popularized by Mark Twain
Origin unknown

[Mike OReilly]

This discussion strikes me as the classic case addressed in the recent book by Daniel Kahneman, et al. (economist and psychologist): Noise: A Flaw in Human Judgment.

[LakeSuperior]

Originally Posted by Mike OReilly

This discussion strikes me as the classic case addressed in the recent book by Daniel Kahneman, et al. (economist and psychologist): Noise: A Flaw in Human Judgment.

Mike you need to give us a hint. Don't have the book at my fingertips. Keep in mind that physicists use order of magnitude estimation as a valuable tool.

So in the above analysis the passaging fatality rate seems to be of the same order of magnitude as operating a motor vehicle. The shakey risk number shows cruising to be 3 or 4 times more dangerous but not 10 times more dangerous.

We know cruising risk can be mitigated (but not made zero) by passaging with a 100 meter yacht or installing a boom brake or having a steel hull or having proper safety equipment and proceedures.

Brioche did not mention the case of hitting a semisubmerged object at night. This seems to be an event that is akin to being hit by another driver. It is probablistic in nature and somewhat out of ones control. Thankfullly, the probability is low but it does happen occasionally given reports here on CF.

Mitigation strategies might be installation of waterproof bulkheads, lie ahull at night, forward looking sonar, etc. But good mitigation strategies are in general not very practical and not totally guaranteed to avoid sinking. Most cruisers keep sailing and hope for the best.

I believe in "prepare for the worst and hope for the best" because in cruising and in life sh** happens and it is better to be prepared if possible.

[Drunkonwatch]

Oh no same hyperbole we often discuss about commercial fishing there are tragic accidents and there are idiots with cocaine habits and tv shows so it’s hard to nail down the actual risk to ones self and crew. Do racers affect these figures I mean to say does there attitude reflect the average cruisers death rate in the Atlantic or is it lower. A lot of the data gets clumpy like jet skiers, refugees, rowers, and dumb asses the NSTB/coast guard just report. I mean it has to be safer than the lady putting on her makeup in traffic talking on cell phone rather take my chances with flotsam connex and Indian cng tankers. But are the statistics like fishing when things happen because looking back “that boat was **** to begin with” or “he was relentless he needed money for divorce” , “drunk at watch” etc? are some of them foreseeable and non factors. I read a book by Purdeys, read a whole chapter on a disaster everyone was anchored off a sand beach stern to the tavern. The weather got snotty and all the yachts ended up on the beach. When we sailed our first time to there I thought “how stupid were these cats.” Being newer to sailing and on what I thought was clumsy Formosa no way we would have anchored like that. Is that added to the statistics in the same manner as Atlantic deaths.

[Mike OReilly]

Originally Posted by LakeSuperior

Mike you need to give us a hint. Don't have the book at my fingertips.

The point (or one of the points) Kahneman makes is that often statistical "noise" overwhelms the factor attempting to be measured. This is most especially the case when dealing with small samples and poor data integrity -- as in this situation.

In these situations, bias can more easily drive analysis, resulting in useless conclusions. This is why I keep pointing to the at least somewhat credible source: USCG data.

[LakeSuperior]

Originally Posted by Mike OReilly

The point (or one of the points) Kahneman makes is that often statistical "noise" overwhelms the factor attempting to be measured. This is most especially the case when dealing with small samples and poor data integrity -- as in this situation.

In these situations, bias can more easily drive analysis, resulting in useless conclusions. This is why I keep pointing to the at least somewhat credible source: USCG data.

I did look at the USCG data but it gave me a headache. Lots of inland boating, paddleboard, waterskiing, etc. accidents to sort. What I am trying to get at is a specific estimate on ocean passaging fatality rates.

The point Kahneman is making is well known in statistics where fluctuations in small samples cause large changes in the estimate. But not quite applicable here as we made the data up. Both the numerator and denomenators are estimates of averages over long periods of time, i.e. 40 years and we are assuming stability in the averages.

But play the what if game. If instead there are twice as many deaths per year on average for Atlantic passaging or there are 600 passages per year instead of 1000 then the risk number changes linearly with the change in the assumed values. For example, twice as many fatalities means the risk is twice as high.

[Drunkonwatch]

40 years technologies have thrown that curve way out.

[Drunkonwatch]

Sociological might be a better way of analysis if this really is a pursuit. How many people do you know who have crossed the Atlantic? How many are dead as a result?

[Jenta]

Honestly, any fears I have of passage - near or far (my longest in 10K nm sailing has only been 5 days at sea so far) - is not death. It's sailing in conditions so uncomfortable or nasty that my wife says "enough" and I'll agree with her.

I feel as though we're intelligent enough that barring the random unforeseen or extraordinary circumstance - the greatest threat to our lives will be my planning. If it was a sudden unanticipated or random event, then I think that I can accept that those can happen at sea as likely as anywhere else. Indeed, perhaps less likely.

As new sailors there are a lot of things that can catch us by surprise. Ours was riding out Dorian in Morehead City, NC on a concrete seawall after moving aboard 3 months earlier from Oregon. We "knew" about hurricanes, but we did not grasp it. We also now know that this will be a continuous learning experience and we have to be humble enough to know we will always be learning from ourselves as well as others.

As we continue this cruising lifestyle we learn the important variables and indeed have more control of them than we may give ourselves credit for. While we can't control the wind, we can plan for it. We can properly maintain our vessel and provide ourselves with backups. This to me is the kind of stuff that we should continue to share with those who are new to this and that brings us as much satisfaction as learning the lesson for ourselves.

Not saying that we can control everything - certainly we cannot. Just wanted to make that point clear.

[Adelie]

Originally Posted by LakeSuperior

FWIW, I am in the camp of going for the adventure like many CF cruisers. In fact, I have crossed the Atlantic twice on a Mason 44 as a co-owner and co-captain. For many of us it is all about getting out there and meeting the challenge. In some sense we want that glimpse into the abyss.



In a recent thread and in older threads new sailors have been very much urged to just go for it, ignore the naysayers, damn the torpedoes, full speed ahead. Any posters voicing even a hint of caution or reservation have been roundly castigated. Consequently, these threads seem to always come around to grapple with questions about the real risk of ocean passaging.



As has been said many times no one looks out for you like you. So it would be good to have an estimate of what the danger really is. Risk that is 10 or 100 times greater than being in an automobile might give pause. People have crossed the Atlantic in all manner and size of craft and most make it without incident. However, some do not. Consider the fates of sailors in Agecanonix, Escape, and Cheeki Rafiki for example.



Any hard data on sailing death rates is hard to come by or nonexistent. So, when real data can’t be had the next best approach is to make some up. We want to estimate a death rate per 24 hours of passage making in the North Atlantic. An hour searching on the internet showed me 10 deaths from 4 incidents spanning 40 years occurring during Atlantic crossings and north south passages in the US. I saw of 3 deaths in the last two seasons. I personally am aware of addition incidents that didn’t show up in the duckduckgo searches so conservatively let’s settle on an average of 1 death in all the passage making in the North Atlantic per year.



Every year for the last 30 some years there have been approximately 200 ARC yacht crossings. So counting getting there the other way gives us 400 crossing per year related to the ARC. Counting north south passages in the US and additional non-ARC crossing a 1000 passages per year might be a starting place. If we assume each passage is roughly 20 days then there are 20,000 Atlantic passage days per year. Consequently, using these estimates there is a 0.005 percent chance of a single death on a yacht per 24 hours of passaging in the North Atlantic.



Now consider that from the NTSB there are 1.34 fatalities per 100 million vehicle miles. If each vehicle is driven at an average of 40 miles per hour then it is roughly 1.34 deaths per 104,000 days of driving or 0.0013 percent chance of a death in a car per 24 hours of driving.



Consequently, if you believe this analysis, then the odds of a death passaging in the North Atlantic in a sailboat is 3.8 times higher than a death in a car accident.



While these numbers are simply estimates it is likely that passaging in the North Atlantic is somewhat more dangerous than driving around in a car contrary to the many assertions by members of CF. However, it appears the increase in risk is somewhat modest and not 10 times or 100 times riskier.

We’re those 3 death on the ARC?
If not you are using mortality rates from a larger undetermined population then applying it to the smaller population of the ARC. At least that’s what you appear to be doing.

[JPA Cate]

I did look at the USCG data but it gave me a headache. Lots of inland boating, paddleboard, waterskiing, etc. accidents to sort. What I am trying to get at is a specific estimate on ocean passaging fatality rates.

When I read the above, what it made me think of was, "maybe the reason you can't find the death stats for passages is that passage making is generally so safe that deaths on passages are actually quite rare." I'm thinking that for all the passages that Jim and I have made, in all our years of exposure, people dying along the routes was rare.

We're only an "n" of one, but it is very common among international cruisers to not have known anyone to die on passages; and among delivery skippers, many of whom are the same passage-makers, death on passage seems to be not a concern, and perhaps the reason for that is that it is actually rare. We're visiting the Australian Wooden Boat festival the past weekend. There is a contingent of tall ships here. I think in the days of exploration by engineless tall ships, there was a lot more risk than there is today:

Weather prediction is much better; communications are much better; medicine has improved hugely; current hull and rig designs allow much better windward performance--when have you recently heard of a vessel being lost due to being embayed? and so on.

Come somewhere warm and go charter a sailboat for a while, if it fits for you. Really, being out on the water is quite healthy! Seems to me like winter's getting a bunch of y'all to feeling low.

Ann

[Drunkonwatch]

Lol You folks should try to encourage sailing and learn a lesson from civilian aviation’s Demise. They failed to nurture the sport to the point it’s cost prohibitive to sport fly for the masses. People engaging keeps facilities open, parts rolling and friendly legislation etc