Misleading article.

[bilbobaggins]

I'll certainly put the suggestion of further reporting to HQ, though I'm not sure how many would read such a report (probably us anoraks who read the MAIB/MCIB stuff too!)

Numerous years ago, there was a perception, pushed by the RYA in briefings and picked up by the magazines, that 'multihulls are dangerous 'cos they keep capsizing'. This became a general public view.

The Safety Officer for the multi-crowd MOCRA around then decided to find out what was the problem - rig size, initial stability, tripping over a lee float, mishandling - so that it could be addressed.

The info came from the RNLI Rescue Records Summaries, then kept on Barclays Bank computer. The RNLI at Poole were readily persuaded to give access, and the reams of 'tractor papar' explored. There certainly were scores and scores of 'multihull capsize' incidents. In fact, there were more than the known number of cruising and racing boats!

Then the penny dropped. The Returns of Service forms had, for type of vessel, only one column for multihull. So all the scores of day-racing Hobies and Darts that fell over and were towed in/righted/attended were included in there. The huge number of 'multihull capsize' was eventually reduced to only 7 in 12 years - 5 of those French and Dutch ( chartered or RBIR racing ), and mostly in CI waters where the RNLI was involved.

However, the perception remained a long time, affecting buyers' preferences, second-hand values, and insurance premiums.

The problem came about because people used the data in ways never envisaged when the data-collection forms were designed - an abuse known to most statisticians. One cannot blame the Hon Sec at the RNLI Station. S/he just wants to fill in the form quick, after a demanding 'shout', and get the guys off home to bed.

We all need to ask just a little bit more, when 'information' derived from statistical data is presented to shape our opinion - especially when it doesn't pass the 'common sense' test.

 

[RIB_imposter]

Given also that "lives lost" rarely go into double figures each year,

really what makes you think that less than 10 people die every year within the range covered by RNLI lifeboats? I don't have any statistics to hand but if you put me on the spot I would say its more like 10 a month than 10 a year.

you can understand the sceptism about whether the 300 or so really represents the number of people who would have been dead if it weren't for the RNLI intervention.

not really that just means they must be bloody good at getting to where they are needed, when they are needed and saving lives.

I used to be equally as cynical but then one day, when I was bored, I looked over the launch and rescue stats for my two local stations for the last 5 years or so. I expected to see claims that each had saved numerous lives who in fact had only been "assisted". Actually each of those station only claimed to have saved one or two people each year at most. This despite the fact that they were each launching 50-70 times a year. Their albeit brief descriptions of those incidents left me in little doubt that had the lifeboat, a helicopter or some other urgent help not reached the person they would have died. They assisted many more people than they saved. E.g. a person who had head and spinal injuries following a cliff fall was meerly assisted. But equally with the assisted stats they still only seem to record people who were actually helped. Often they were called to windsurfers or fishermen by well-meaning members of the public who thought they were in distress. If no help was required these are recorded as Lives Saved 0, Persons Assisted 0.

Is it infeasable that from over 8000 launches they save around 300 lives ? If the true number was say 10 lives saved I think, sadly, we would need to ask if either the network was innadequate, innefective or unnecessary.

So I have reasonable confidence in their stats. Far more so than those published by the MCA/MAIB where they simply refer to the number of incidents!

I have no axe to grind, am not involved with the RNLI.

 

[Tranona]

The low numbers of deaths refer to those which are the direct result of yachting, whereas the RNLI (and coastguard) statistics cover a whole range of situations whose only connection is the sea. That is what I mean by the figures hiding the reality.

In research we have a truism "A way of seeing is a way of not seeing". So if you want to "see" the number of "lives saved" that is what you count using the definition that you determine. Nearly all of the data in the RNLI report is of the same type - simply activity counts - number of launches, number related to yachts, motor boats etc. Where "causes" are used they are at such a high level of aggregation - which treats all incidents in the chosen category as equal - as to be meaningless.

If, however you want to understand the CAUSES of incidents you need to record those. This is not done at present because RNLI and Coastguard are not judged on understanding causes, but on showing that they use resources efficiently to "save lives".

The incidents are exactly the same, just that you record them according to one measure that has little meaning in terms of prevention.

There was a time when "counting the numbers" was the dominant methodology in big ship safety. Thankfully "evidence based" methodologies as used in the aircraft industry have taken over with a resultant massive improvement in safety. This methodology is also evident in MAIB so good information, conclusions and recommendations come out of their reports, whereas all that comes out of RNLI (and coastguard) "statistics" is meaningless statements such as the one that started this thread!

 

[RIB_imposter]

The low numbers of deaths refer to those which are the direct result of yachting, whereas the RNLI (and coastguard) statistics cover a whole range of situations whose only connection is the sea. That is what I mean by the figures hiding the reality.

mmm... I think it is you not the RNLI who is choosing to be selective with your interpretation. Are you retracting your statement that "this allows them to come up with meaningless (and misleading) statistics such as "350 lives saved by the RNLI". Given that the number of deaths at sea rarely run into double figures in a typical year one has to question the reliability of such figures!" As the number of deaths at sea are much higher than you suggest. And the number of lives saved is not exclusive to "yachting". If the claim was that 350 yachtsmen were saved I agree that would be far-fetched but no one is making that claim.

Nearly all of the data in the RNLI report is of the same type - simply activity counts - number of launches, number related to yachts, motor boats etc.

ok but these are nice objective measures. as soon as you start looking at causes it is subjective. Investigating the root cause is not really the RNLIs role. And how would you decide consistently across the country to categorise an incident anyway - e.g. sailing vessel fills its tank with dodgy fuel gets towed home in a F8. How does that get classified? Contamined fuel / innadequate equipment/training to change the filters at sea / bad weather. How about a man overboard during an unexpected gybe? is that poor helm/crew training / bad luck / equipment design or failure?

Where "causes" are used they are at such a high level of aggregation - which treats all incidents in the chosen category as equal - as to be meaningless.

I have to dispute that. They don't provide you with the detailed information YOU want - but they are by no means meaningless. There are around 40 "cause categories" in their database. Any more would mean so few entries in each category as to make analysis impossible (without simply grouping them back into the sort of categories they are using!). They know it makes sense to focus efforts on pleasure vessels (>50% of callouts), and engine maintainence etc (>18%), but they know that its possibly not a priority to focus on Col regs (<0.5% of call outs to collisions) or encouraging fishermen not to snag gear (<1%) but it might be worth highlighting basic navigation and tide heights etc as >8% of cases were groundings. Their report doesn't look at these in terms of severity - but I assume that they have the data to do this - so e.g. grounding and breakdowns are common - but probably mostly not lifethreatening wheras collisions and fires potentially are.

The data and incident reports are all presumably on a central RNLI system so if you believe you can pull out informative lessons from it - then perhaps you would volunteer to work with the RNLI to do this?

This methodology is also evident in MAIB so good information, conclusions and recommendations come out of their reports,

mmm... I can point you to a number of MAIB conclusions which actually aren't supported by the evidence presented in their own report. They also regurgitate statistics in their annual reports that they have made no critical assessment of (and have no means to do so, and refuse to share the data for!).

[qupte]
whereas all that comes out of RNLI (and coastguard) "statistics" is meaningless statements such as the one that started this thread![/QUOTE] you have bothered to read their annual opperations report haven't you, before criticising their collection/publication policy or are you relying on YM to keep you informed?

 

[Tranona]

mmm... Investigating the root cause is not really the RNLIs role.

I think that is where the problem lies. Once you get past MAIB reports the only source of data is what we are discussing here, and as we see is not focused on understanding causes but in counting activities. It is therefore of little use for those who want to understand causes which means you have to look at incidents in a different way and there does not appear to be any mechanism currently to do this.

I agree about the way figures are reported. I do not know whether the YM news item is just extracts from a press release, but the message is misleading. The average over the last 7 years is 3864, lowest 3581 and highest 4075. This does not suggest any significant change this year - less than 3% higher than the average. To give such figures any meaning they have to be related to some other measure, such as level of overall activity. Otherwise all you can say is that the number varies little over the years!

As to what is objective and what is subjective, not sure that the "lives saved" measure is objective as it relies on an individual assessing what might have happened - and as I pointed out you will never know because it did not happen. The objectivity comes from the criteria that are used for the assessment, which in themselves can be subjective so we are back where we started. Unfortunately you can never prove it because you can't stop the rescue service and let the person(s) die - and anyway, nobody would be there to record what happened. So the "objectivity" is contrived.

That is why people generally prefer "counting" type "statistics" - because they give an appearance of objectivity - it must be so because numbers do not lie! Which has an element of truth only if the numbers represent faithfully what you are looking at.

 

[RIB_imposter]

Tranona,

I presume you have read the opperations report (http://www.rnli.org.uk/assets/downloads/Annual review, reports and accounts/RNLI_Statistics.pdf) I mentioned in my last post? if so how would you like to see the categories for "cause" (p12) broken down in more detail? And what do you think it would add - if only 0.2% of all incidents they respond to were "caused" by a particular sub division and then next year it was 0.3% - that would be a 50% increase, but not really significant.

The only "category" there that could probably justify sub division would be "machinery failure" which could be split to - e.g. electrical problem, fuel problem etc - but its probably not always possible to determine this and just because fuel problems are the biggest problem doesn't mean you won't have an electrical problem or fall overboard!

I agree with you that the data don't really suggest any change in number of shouts etc in the last few years. But unless there was a significant change in boater behaviour, weather, training/regulation etc then I wouldn't expect to see any sudden changes anyway.

 

[Chris_Robb]

Obviously can't comment from YM point of view, but RNLI drives the agenda. They have multiple audiences of which the two most important are government and donors. The former have to be kept at bay - and who can blame them when you see the effect of political decisions on agencies such as the coast guard. So they have to show that they meet society's needs at no cost to the public purse. Donors need to be both convinced that the organisation is indispensable and makes good use of the donated money. Read the annual report to see how effective they are in getting the message across!

Whilst they make a lot of effort to promote safe practice at sea, they do not provide useful information about incidents they deal with that do not result in a more searching MAIB enquiry. It is to their advantage to keep reporting at a high level of aggregation to hide the detail. It is like the "X deaths and serious injuries" headline where X-1 could be the number of injuries and there is no definition of what constitutes a serious injury. Good headline figure - meaningless information.

I might get into trouble here.....

When you see the prolifragate spending by the RNLI, and then compare what the French do and achieve on a fraction of the costs, it does make you wonder where they are going to end up.

I don't beleive that there will always be an endless reservoir of money to fund their excesses. It is not easy to change an instututions culture - and I predict that this will have to happen at some point. (I point out that I do make an annual contribution)

Just one example: Trevose head in Cornwall has a truely magnificent boat house set in the cliffs, built at huge costs, yet the number of launches and savings is tiny - almost to the point that the cost is just way outside reality (if you can price a life)

Time to duck......

 

[toad_oftoadhall]

I'm with Tranona on this one. Prevention is better than cure so I'd like to see better figures so we can all learn something. As for who reads the MAIB reports - a *huge* number of people.

I also find the 'lives saved' a bit dubious. How do you know if you've saved someone's life? The answer is you never do.

Thanks to whoever posted the PDF - really interesting.

Having said all of the above, it's really none of my/our business. They're independent - they collect the cash how they like and within certain limits set by the charities commission they can spend the money how they like.

I might get into trouble here.....
When you see the prolifragate spending by the RNLI, and then compare what the French do and achieve on a fraction of the costs,

I owe the RNLI a debt that can never be repaid, but I reluctantly agree. Was it Southend where they used a million pound hydraulic davit to do a job that was previously performed by a rope and tackle? I think this was around the time the Charities commission told them to start spending some of the cash they were hoarding.

Plus there's some doubt if they really need bespoke boats. IIRC self-righting Lifeboats are literally available of the shelf.

I don't believe that there will always be an endless reservoir of money to fund their excesses.

Nah, it will always be a great money spinner. Other charities look on with envy.

 

[Tranona]

Tranona,

I presume you have read the opperations report (http://www.rnli.org.uk/assets/downloads/Annual review, reports and accounts/RNLI_Statistics.pdf) I mentioned in my last post? if so how would you like to see the categories for "cause" (p12) broken down in more detail? And what do you think it would add - if only 0.2% of all incidents they respond to were "caused" by a particular sub division and then next year it was 0.3% - that would be a 50% increase, but not really significant.

The only "category" there that could probably justify sub division would be "machinery failure" which could be split to - e.g. electrical problem, fuel problem etc - but its probably not always possible to determine this and just because fuel problems are the biggest problem doesn't mean you won't have an electrical problem or fall overboard!

I agree with you that the data don't really suggest any change in number of shouts etc in the last few years. But unless there was a significant change in boater behaviour, weather, training/regulation etc then I wouldn't expect to see any sudden changes anyway.

Yes, I have read the report.

Perhaps "causes" is the wrong word to use. It is the circumstances of the incident that are relevant. Use the "Causes" in this report treats them all exactly equal, ingnoring the circumstances that turned a failure (of whatever kind) into an incident which required action. Many machinery failures (the largest category) are not life threatening and could be resolved without outside assistance - indeed many are, because many people would not dream of calling for assistance. Machinery failure that puts a vessel in danger is of a different order, and probably if you analyse all the circumtances surrounding the incident you may well find machinery failure is is only one contributing factor - but is the one that goes in the box. What we need to understand is the circumstances that result in a vessel requiring assistance.

Breaking machinery failure down into different types is almost irrelevant - that is just confusing precision with accuracy.

As I was trying to point out in my analysis of the causes of yachts foundering, by looking at the reports you can find common themes that do not necessarily reflect the categories for published "statistics". If you want to avoid having your yacht sink under you, keep clear of big ships, don't go out in extreme weather and make sure your keel or rudder is very firmly attached!

I would not have a clue what the main circumstances were in an incident reported as where "lives were saved". Were they in the middle of the channel? close to shore? were they Yachts? Motor Boats? was fire involved? were lifejackets used? and if so were they effective. All this useful information gets lost in the numbers!

 

[toad_oftoadhall]

I would not have a clue what the main circumstances were in an incident reported as where "lives were saved". Were they in the middle of the channel? close to shore? were they Yachts? Motor Boats? was fire involved? were lifejackets used? and if so were they effective. All this useful information gets lost in the numbers!

Agree. If you do something as momentous as saving a life it seems reasonable to be able to write a handful of words sketching out the basic circumstances. Christ they could sell it - I'd buy a copy every year.

 

[RIB_imposter]

Agree. If you do something as momentous as saving a life it seems reasonable to be able to write a handful of words sketching out the basic circumstances. Christ they could sell it - I'd buy a copy every year.

Many local stations have their own websites now. Many of these have a section which briefly outlines the nature of each call out. Different stations post different levels of detail, but often give you an idea how serious each shout was etc, and sometimes an idea of the circumstances. I'm not sure it will encourage people to call for help (which is what the MCA/RNLI want) to think that the detail of their f**k up is going to be splashed across the net.

Travano - I agree - just out of interest it would be interesting to see a 2D "martix" (or nicely colour coded graph) which combined "cause" and "type of craft" and "cause" and "lives saved/lost/assisted". BUT would it help the RNLI actually respond more efficiently to incidents? would it help them raise more money? would it reduce the demand on their services? if not why would the RNLI bother...

From a prevention point of view we know what the main causes of incidents are:

equipment failure, bad weather, skill/experience, lack of planning, suitability of craft etc... ...producing stats which tell you these sort of things won't help you avoid an incident (although it might make you feel better about yourself and how well prepared you are, and that would never happen to you/you would cope better). It would be unwise to focus your attention on only the most common causes of disaster - you need to cover all the bases or guess which one will bite you!

When a serious incident or an incident that has the potential to be serious occurs in circumstances that may not be expected (based on current knowledge/understanding/perceived wisdom) or where lessons could be learned the MAIB already have the systems for investigating and disseminating any important points.

 

[toad_oftoadhall]

Many local stations have their own websites now. Many of these have a section which briefly outlines the nature of each call out.

I get an e-mail every time my local lifeboat launches - but they don't seem to get their share of the 300 lives saved. Usually engine failures. Never anything interesting.

if not why would the RNLI bother...

I just don't think you're going to get very far arguing that theres no benefit from handing out information about incidents where someone nearly died.

The lessons we learned in 1979 are still fresh in all our minds and still quoted constantly.

Depending on how you measure it the RNLI have information from the equivalent of ten 1979 Fastnets each year... I can't see how we could fail to learn from that.

 

[Tranona]

.



From a prevention point of view we know what the main causes of incidents are:

equipment failure, bad weather, skill/experience, lack of planning, suitability of craft etc... ...producing stats which tell you these sort of things won't help you avoid an incident (although it might make you feel better about yourself and how well prepared you are, and that would never happen to you/you would cope better). It would be unwise to focus your attention on only the most common causes of disaster - you need to cover all the bases or guess which one will bite you!

I think that is where you are wrong! These things don't "bite" you, they are usually the end result of a string of events. Understanding the underlying causes is the basis for prevention. Why are there so few incidents involving fire? Because "everyone" knows the risks and takes action to prevent fire. The most popular and enduring articles in yachting magazines are the experience accounts because people can relate to them.

Clearly if the RNLI statistics are to be believed there is a large number of serious incidents each year from which we can learn, but about which we know little.

Using your list, if you have a well equipped boat, avoid bad weather, have the appropriate skills, an effective plan and a suitable vessel you won't get "bitten"! Precisely the point that I am making, the current data does not help understand these issues at the level of detail to help one meet that safe ideal.

 

[RIB_imposter]

I get an e-mail every time my local lifeboat launches - but they don't seem to get their share of the 300 lives saved. Usually engine failures. Never anything interesting.

So if there are no gorey details, near death experiences etc its not interesting... I am sure there are lessons to be learned from those more routine cases though. However there are about 300 boats based out of 230 stations. Since when lives are "saved" it won't always be one (could be many in a single incident) then 288 lives saved in 2008 spread around the stations probably means that some stations save, on average, less than 1 life per annum. Especially since some stations are obviously buisier than others, and some will be in areas which have lots of relatively minor incidents in well resourced areas where helicopters or other rescue resources may mean that when lives are saved its not always by the RNLI, whilst others might well be in areas where they stand a much bigger chance of being deployed / first-on-scene for a serious situation which had such serious potential consequences.

I just don't think you're going to get very far arguing that theres no benefit from handing out information about incidents where someone nearly died.

no what I am arguing is there are actually very few new lessons to be learned.

The lessons we learned in 1979 are still fresh in all our minds and still quoted constantly.

yes but there is already a mechanism for investigating and sharing that info.

Depending on how you measure it the RNLI have information from the equivalent of ten 1979 Fastnets each year... I can't see how we could fail to learn from that.

I'm not sure how you make that leap? That was one single event, with losts of very similar (in the broad spectrum of boats) vessels sailing in virtually identical conditions etc. Therefore there was lots of replicated information to add consistency. You can over analyse information and fail to apply common sense. If nobody has fallen from the top of the mast in the last 20 years and been hurt does that make it a safe opperation? If the biggest cause of sinking turns out to be seacock/stern gland failure does that change anything? does that make a collision any less dangerous.

You would also get only part of the picture. In some of the most serious incidents a helicopter would be used and the lifeboat may only tow the boat back so not a life saved by the RNLI. Many others will be rescued by Independent Lifeboats so we don't learn their lessons. In arreas like the solent with loads of boats but large numbers of lifeboat orgs this could skew the undertstanding. Conversely the number of long distance tows - as these tend to be by AWBs which are almost exclusively RNLI.

I guess my question is - what risks are out there which you think you don't know enough about? (I know thats an impossible question to answer) - but do you really think the RNLI have useful information which would be about NEW causes of accidents that aren't already known risks, and which won't be investigated/reported by the MAIB anyway (bearing in mind that I think the MAIB do at least a PE on any fatal accident).

the point is RNLI stats aren't really intended to be used for identifying risks/causes but rather to allow those who work with the organisation to understand the sort of stuff its doing and the demands on it. More detailed information that they might draw out but not publish will be used to assess the effectiveness and suitability of its fleet/training/locations etc.

 

[RIB_imposter]

I think that is where you are wrong! These things don't "bite" you, they are usually the end result of a string of events. Understanding the underlying causes is the basis for prevention. Why are there so few incidents involving fire? Because "everyone" knows the risks and takes action to prevent fire. .

or because the risks from fire are over stated? or because fire is a similar (just more dramatic) risk to that on land?

I agree that often its a chain of events. But you really think the rnli should rescue me and then analyse where it all went wrong? E.g. dismasted in bad weather rigging round prop - root cause? poor maintainence, bad planning, inexperience, not having the right gear to get the rigging cut free. Or maybe I was short manned for the voyage etc... its possibly a bit of all of them. You can do the same with every incident.

So which of these "enduring and popular" articles did you read and thing - bloody hell - that could have been me, and if only I did X would avoid it - or if only I did Y (which I wouldn't have thought of doing) I could have stopped it escalating to a rescue?

Clearly if the RNLI statistics are to be believed there is a large number of serious incidents each year from which we can learn, but about which we know little.

No you're still assuming that if you don't know the detail there must be some mystery as to what went wrong. I don't believe that is the case. Incidents involving fatalities are reported to the MAIB, incidents involving commercial craft (and leisure craft on a voluntary basis) involved in SERIOUS incidents are reported to the MAIB. They investigate and report, often in much more detail than you are seeking so we can make our own minds up.

Using your list, if you have a well equipped boat, avoid bad weather, have the appropriate skills, an effective plan and a suitable vessel you won't get "bitten"! Precisely the point that I am making, the current data does not help understand these issues at the level of detail to help one meet that safe ideal.

My list wasn't meant to be exaustive, but I do believe that an experienced skipper, using a suitable boat in appropriate weather is much less likely to get in trouble, and more likely to cope with it if they do. I obviously don't know how many people would have "ticked all the boxes" (that could reasonably been expected) and still run into trouble. But I would guess that its <20% (and maybe as low as 5%) of the lives saved - going from my own informal analysis of a few random stations in Scotland over the last 5 years. Then only a fraction of those are on the type of vessel you use (e.g. sailing), some will be specific to local conditions which might not relate to your circumstanced etc. Thats actualy a very small number to spot any meaningful trend or unexpected risks. Of those the most serious will be investigated by the MAIB anyway.

I still argue we know why people drown, why boats sink etc...

 

[toad_oftoadhall]

I guess my question is - what risks are out there which you think you don't know enough about?

Err, I don't know because they aren't published! That's the whole point.

 

[toad_oftoadhall]

I still argue we know why people drown, why boats sink etc...

Suppose we do.

If the RNLI spend in total say 1 man week documenting (say 20 words per incident) all the 300 odds lives they save, is that big loss? A volunteer could do it at zero cost - christ - Id do it for nothing.

So we learn nothing. What's the problem? We all pat ourselves on the back at being perfectly safe.

But what if every 3 years we learned one thing that might save a life... Would that not be worth it?

The more I think about this the more I think there is no reason not to and possibly a good reason to.

 

[Ken_Irby]

Wait, wait, wait, wait, wait!

This post IS, quite rightly, criticising YACHTING MONTHLY, NOT the RNLI's published statistics!
Agendas? Seems to me that some posting here are anxious to promote their own!
Chas, on Kentrina's login. So not necessarily HER point of view.

 

[toad_oftoadhall]

This post IS, quite rightly, criticising YACHTING MONTHLY, NOT the RNLI's published statistics!
Agendas? Seems to me that some posting here are anxious to promote their own!
Chas, on Kentrina's login. So not necessarily HER point of view.

I'm pretty sure nobody is criticising the RNLI's published statistics. Unless you think 'a bit more detail would be nice' is critical.

As for agendas... that's laughable.

 

[Slow_boat]

I would imagine that the RNLI are subject to the same type of statistics and 'performance indicators' that have done a good job of taking the 'service' out of public services and wrecking morale within them.

I accepted a tow from an inshore lifeboat (not RNLI) once. I was safely at anchor with an overheated engine and in no real or potential danger but was grateful for the offer of a tow in. The crew did concede, though, that it would count on their stats ans a 'rescue'

I now wonder if that 'rescue' is on a computer somewhere and could count against me in insurance checks or something.