Route and Paths

[Robert Bradley]

Last weekend I found myself giving a presentation on the evidence backing the way we search in the lowland environment. As part of this I looked at the mathematics and research behind things like leaving no stone unturned within 300m, critical separation searches and aerial, dog and neighbourhood watch searches. Obviously, as we all know, there is little or no empirical evidence into the effectiveness of dog search. We know it works, but we cannot tell a dog's sweep width and hence its probability of detection - and this, of course, has a knock-on effect in that search controllers/managers do not have any measure of the integrity or effectiveness of their search efforts (which they do have some useful form of estimation when using the other search resources mentioned.)

However, when estimating figures for POD of route and path searches for foot searchers, I was reminded of the debate about whether dogs should be used for route and path searches - and as an extension from that, whether LSDogs (or NSARDA for that matter) should be assessing dogs to do this.

One of the topics of my talk was that there was no "search based" reason why more than three foot searchers should be used on a route and path search - the extra search effort does not add much to the POD of the search and would be much better employed elsewhere where he or she would add a lot more to the POS of the search overall. Looking at dogs, although there is little empirical evidence as I have already said - the greater distance travelled by a search dog means that even with a slightly lower sweep width figure than for a human searcher (which I believe most on here would actually argue isn't true) then it is again a waste of search effort to use a dog on a route and path search.

This leads to a situation whereby a search manager has a choice - use the dog as assessed but waste its search effort that could be used elsewhere more productively, or use it for an area search, for which it hasn't been assessed, but which is paradoxically better for the misper.

In my mind, search managers are being put into a difficult position by the assessment regime of dogs, and that the assessment should be changed in order to better suit the search resource....

Thoughts anyone?

[Pete]

Just a little on the dog effectivness bit.

I personally welcome any search amnager to put foot teams over an area that I have searched with my dog, at the end of the day dogs are good but not infalable. The only way to be sure is for human eyes to search every part of the search area. Even then it only confirms that there was nobody or nothing there at the time the searchers (or dog for that matter) passed through.

With regards to route and path (hasty) again it is an emotive subject, 1 hander, dog and support searching a path equates to (for example) 2 man hours, a 3 person foot team would equate to 3 man hours however, the speed of the dog to find should also be taken into account. Dogs should be seen as a quick fix to a search not the be-all and end-all with definative results every time.

I feel that this is a subject to which there will be no end in sight, the arguments will contine for and against dependant on the person you talk to.

Pete.

[Robert Bradley]

It is not just about man-hours though; the distance travelled by a search resource has an important effect on its POD. Because a dog is likely to travel something like 5-8 times as far as its handler when searching it "covers" a massive distance compared to the route and path length and size.

The effect of this is that whilst a foot team and dog team should search the area in roughly the same time - the time it takes a handler to walk the length of the track; the dog has travelled that much further. When you consider that a three person search team has an approximately 98% POD for a route and path, the dog team can't add much to that.

The dog team put into an open area, however, could have a far greater effect on the overall search POS and it is this figure that the search manager is obviously trying to maximise.

[Pete]

Very good point Rob, I think that until there is the possibility of producing a POD for search dogs this argument could continue 'until the cows come home' so to speak. However I feel that a true POD for search dogs is an impossibility due the many variables involved. I have approached and worked with several people (most of whom have MSc's), they have informed me that it is not possible and they specialise in statistics.

Pete.

[Daryl]

Just my two penneth.

Firstly a Level 2 dog or hasty dog will search a route quicker than a foot team, this in actual fact is a requirement of our assessment of such a dog.

A Level 2 dog may stay at this level and may not progress any further, the main reason being that some breeds and also some personalities of dog will not range, therefore would not make a good area dog.

Lastly on the point of PODs - search managers are not put into a predicament because they don't know the POD of a Helicopter, Spotter Plane, Mountain Bike team etc, they just use them - just the same as a dog team.

We (some time ago) set up a number of areas to be searched, some area and some hasty, the dog unit went in first followed by the foot team. In every case the MISPER was located by both types of resource, the only difference was the one type of resource found the MISPER considerably quicker than the other. Whilst sweep width and critical separation may be difficult to put on to paper statistically, the experience of the dog handler will determine how an area is searched using their dog. What ever we (handlers) are doing it seems to be working despite the inability to offer a POD.

I am (as I am sure others will be) willing to take part in any exercises in order to prove or disprove the effectiveness of a search dog, however I do not believe that there will be a fail safe system of determining a POD for a dog.

With regard to should dogs be used in Route and Path searches - ultimately yes, just the same as a mountain bike team. Generally they move quicker than a foot team therefore searching a greater area quicker and reducing the distance that a MISPER can travel. Having said that, a dog is just another tool that can be deployed by a SAR Manager just the same as all of the other resources, and should not be used in isolation, but to compliment the other tools/resource used.

[mike]

When you consider that a three person search team has an approximately 98% POD for a route and path, the dog team can't add much to that.

98%, a high figure in my view, unless they are searching 30ft of grass each side of a track, or the search is that slow it would be a problem for the search manager.

a hasty search is exactly what it says, hasty, find that person and quickly.

ive yet to find any bit of kit, helicopters included, that can cover an area as quickly as a dog.

mike

[Robert Bradley]

Firstly, thank you for your offer of help in finding the effectiveness of a dog search - I am considering a PhD at the moment on the effectiveness of search resources (and if I do will need lots and lots of goodwill and assistance). There is actually a sound methodology, based on experiment, to give a POD figure for search resources such as foot searchers and aerial resources, like helicopters, which search managers can use to back up their rationale for their search plan should they wish, or need to. This was in essence what my presentation was about last weekend. There are a few issues with this data, however, that would need to be ironed out before it could all be used prior to putting out search resources, in order to optimally plan a search to maximise POS. This is one area in which I'm considering research.

Congratulations by the way on using the term effectiveness too. Too often searchers (and indeed some researchers) use the term POD incorrectly when describing the results of searches.

I don't know whether I have been misinterpreted here - despite the lack of evidence into the effectiveness of a dog search, I in no way doubt their ability. I just believe that the way the assessments work unfairly put dogs, handlers and search managers in a position where they are not using the strengths of the dogs' ability.

Moving back to the route and path issue - I think the results of your "experiement", Daryl, would actually back-up my rationale in this. I can guarantee that the time difference for doing the area search would be much greater proportionally than that of the R&P.

It is always important to remember that the purpose of the search manager is not to maximise POD, but to maximise POS - finding the misper quickly. In order to do this the search manager must prioritise search areas and search resources to put into them. It is here that I believe the issue lies. You have a foot team and a dog team, and a R&P and an area. Which do you put in where? From a maximising POS point of view only - the dog should be put into the open area and the foot team down the R&P. (This I believe would stay the same, despite the level of assessment of the dog & handler) This is where a dog team differs from a mountain bike team - mountain bikes cannot be used effectively to cover open areas, therefore by using them to hastily cover R&Ps you are using them to their strength.

All search resources miss, dogs, foot teams and helicopters. In order to maximise the overall POS of a search, we have to increase the risk of missing in one area, in order to increase our chances of finding in another area. This is a fact of search which many find uncomfortable but is the way it is. By introducing POD when discussing R&Ps I probably mislead readers into a false premise that POD was important in the issue - I apologise. What I was trying to imply was that foot searchers can carry out a R&P quite effectively - whereas they have a far lower POD using critical separation for an area search. The dog team will as Daryl says cover the ground much quicker to achive a similar if not higher POD. From a search management point of view, therefore, using a dog to do the R&P and the foot team to do the area lowers the POS considerable and is a bad search plan.

With regard to limiting an area using dogs, the same sort of rationale applies. Is it really a best use of resources? Mountain bikes would do this far more effeciently, for example. Here, of course, we also have to take into account the new ISRID data on mobility times for mispers. I haven't had time to study it in much detail but it seems to show that certain types of misper do not continue to move after quite short times (often before SAR teams are called out).

Finally - I know I can waffle on about this sort of thing for ages - although some believe it is, and will be, impossible to get sweep width figures for dogs (I would disagree) it actually doesn't matter. I haven't done the maths but unless the sweep width figure of a dog was less than about a quarter of that of a foot searcher - what I say must be true. Anyone on here really believe that a dog cannot sense a misper in under 4 metres (using the lowest sweep width figure from the US experiments)? When I get time I will try to estimate how low the sweep width figure of a dog must be for it to be used differently in order to maximise POS but generally most dog handlers believe that their dog can discover mispers several hundred metres away in the right conditions. And, unlike sight, with scent there are numerously more detection opportunities. And so, in order to maximise the POS for a search, in my mind (and no one has yet said anything to disprove or change my opinion on this - I'm always more than happy to be corrected) dogs are far more efficient than foot searchers at doing area searches compared to R&P searches and should be used as such.

[mike]

robert, have you spoken to John Coombes from SARDA England and Peak district MRO? if you are thinking of doing a Phd in search he will be able to give you a massive amount of advice/ paperwork on search patterns/ methods/ POD etc.
he has put together all sorts of formula etc to assist search teams/ dog teams.
he can be contacted via the sarda england site. i'm sure you will have an interesting chat with him.
mike

[mike]

Robert wrote
despite the lack of evidence into the effectiveness of a dog search,


a brave comment to put on a web site for people who work search dogs, better pass this info on to the countless people who have been found by a dog and also the 100 plus people my last police dog found during his career.
think i had better stop.

mike

[Robert Bradley]

Still doesn't amount to evidence ......

By that logic I should talk to all those people who have had lottery wins to find out how effective it is to play the lottery !

It is human nature to suffer from confirmation bias - we all tend to look for reasons to back up our own belief. What we don't look for is any reason or evidence that our beliefs do not stand up. (Doubly so, I expect in the police service where you are searching for evidence to convict - let the defence find evidence to the contrary)

One example I gave was that of false positives. A search dog indicates on an area - the handler reports this to the search manager. If it was later found the person went through that area, both the handler and search manager remember this incident and use it to back up their belief in search dogs. What they do not remember is the 10 times the dog has indicated on an area that has turned out to be nothing.

But anyway - that all moves away from the question in hand about R&Ps.

PS Thank for the contact. Should I ever get around to doing some work - I will need all the assistance I can get. (Maybe I should stop antagonising you all then )