Acer ventura

Hi I recently had the pleasure of presenting ‘Safety in Numbers’ at the International Society of Arboriculture conference in Florida last month. I’ve since worked on an enhanced transcript of the presentation to accompany the slides, and by way of thanks to Arbtalk for the platform I've been given here with some really visually heavy threads, am uploading it for those of you who might be interested. Cheers Acer ventura Safety in Numbers ISA 2015 - Handout.pdf

Hi Cassian Good to hear from you on here, and great news that you're getting on well with QTRA v5. It's a draft document at the moment. So, let me know if there's anything in it you have issue with, you think could be embellished, or any suggestions you might have about making it easier to understand, once you've taken your little red pen to it. I'd originally planned to firm it up this month, based on feedback and revisiting it after a break. However, given some of the off forum comments I've had, what's happening with the ISA LinkedIn thread about it, and that I'll be presenting QTRA v5 at the ISA conference in Florida next month, I think I'm going to delay messing about with it until the back end of August because there could be some valuable feedback, or developments, that would be worthwhile including. Cheers Acer ventura

Hi Haljam It's your risk so it's your choice. Do bear in mind though that the risk of harm from putting those trousers on is around 1/15 000, but this figure does not account for the sweaty state of them. Perhaps the thing to be more concerned about are the socks with a risk at around 1/5 000. Mind you, the data set for socks also includes tights. So it might, or might not, be a lower risk than that for you. Cheers Acer ventura

In my previous entry on this thread, I forgot to put the link to the additional post I made after Jeremy posted his ICF blog entry on the UKTC. This is a more detailed analysis of the risk from Summer Branch Drop and the factors to consider when determining what would constitute 'proportionate' risk management than the opener in the thread; which was a response to the initial Horticulture Week 'Barrell On' piece. RE: What's Barrell On? - Jeremy Barrell & Summer Branch Drop Cheers Acer ventura

Hi Haljam In the interests of balance, and to get a handle on some proportionality and perspective on what is demonstrably an incredibly low risk (it would appear to be less than 1/600 000 000 to the public in the UK based on what statistics are available), you might want to consider the contents of this thread on the UKTC. What's Barrell On? - Jeremy Barrell & Summer Branch Drop I rather suspect that no matter how hot, dry, calm, post rainfall, horizontal the branch is extending beyond the canopy, or any other causes that are claimed for a failure that is not yet well understood or clearly defined, your risk of death or injury driving to the site in the first place would be many leaps in scale of risk magnitude greater than from Summer Branch Drop whilst there. Is anyone on here aware of any climbers, or more likely groundsmen, who have been injured or killed by what appears to Summer Branch Drop. Cheers Acer ventura

I forgot to mention the Registered User also gets to play with the QTRA calculator software, which can be plugged into all manner of tree survey software and databases. Cheers Acer ventura

Hi Andy What you’re saying here makes sense to me until this last paragraph, which looks as though you’re now saying the exact opposite to the previous ones. A number alone can be difficult to understand unless it has context, which is why you’re relating the background level of tree risk as a number, with other numerical risks, to give context and establish proportionality. I do something similar in my presentations, and would be interested in your source for ‘Possession by Satan’ because it’s an amusing risk. It’s also a sufficiently high risk that you couldn’t reasonably impose such an environment on the public that could lead them to screaming “Your mother knits socks in hell!” QTRA is particularly aware of this limitation in communication of risk, which is why the risk number is coloured, with an appropriate advisory about how the level of risk relates to the Tolerability of Risk Framework, so the number has both context and meaning to the risk owner/manager. The point of expressing risk from trees as numbers here is that it’s unambiguous. It can‘t be ‘misinterpreted/misunderstood’. Conversely, if we forsake numbers as a means of defining, measuring or understanding the level of risk, then we’re simply left with the ambiguous wool of words. Without the clarity that comes with numbers, it would not be clear, for example, that a 'Low' risk of death from trees could be higher than playing Russian Roulette, or that a 'Low' risk of death can be higher than a 'Moderate' risk of death (TRAQ). Absence the numbers gained from measurement nailing down the boundaries of the words, providing unambiguous meaning, levels of risk such ‘low’, ‘acceptable’, ‘high’, ‘moderate’, or ‘unacceptable’ can easily be 'misinterpreted/misunderstood' by the risk assessor, and perhaps more importantly by tree owner/manager who is the person that has to make the risk management decision based on the information they have. Cheers Acer ventura

I see CAS have Expert Witness Day III coming up this week, with David Lonsdale amongst others, so it might be opportune follow up Jon’s point about QTRA being in court, and what that might mean, a bit further. In Harry Bowen & Others v National Trust, David Lonsdale was the Defendant’s expert and used QTRA to calculate the risk retrospectively. Consequently, the Claimant’s expert witness, Julian Forbes-Laird, decided to produce a supplementary report with the specific purpose of criticising QTRA. David Lonsdale was obliged to respond refuting the criticisms that Julian Forbes-Laird levelled. Having spoken to David about the case and this part of it, IIRC the Judge was not particularly interested in either of the supplementary reports. The supplementary reports are documents 8 & 9, and other documents relevant to the case are here: Quantified Tree Risk Assessment I highly recommend reading David's commentary (document 10) on the case. Cheers Acer ventura

Hi Just by way of keeping a record of this. The download on the first page has been updated. There's no significant change beyond me trying to steer a line between decent graphics and not making the document too large to email. The 05.06.15 document has much better images if the quality on the old one bothers you. One thing I did forget to keep a note of was the number of views it had before it was updated, which I think was something like 150+. Cheers Acer ventura

Having a quick review of this thread to see whether what I posted makes sense afterwards, I realise I've messed up here and momentarily got lost in the fog of words. This: Should read, and I'll uppercase to identify where I've gone wrong: "In it, one of the issues discussed is that a QTRA Probability of Failure (PoF) range 1 (in numbers 1/1 - <1/10) has three of four of TRAQ Likelihood of Failure ranges in it. Imminent and PROBABLE Likelihood of Failure ‘words’ are parked in the top end of PoF range 1, and the upper values of the POSSIBLE ’word’ are in there as well." Sorry about that. If one of the the moderators could fix this in the ordinal message, or draw attention to the later message that would be great. Cheers Acer ventura

Hi I was surprised to see this thread rocketing up the view charts earlier this year. The reason I was surprised is because it was written in early 2013, and a significant amount of the content has been superseded since QTRA v5 was launched in October 2013. It was then pointed out to me at a QTRA training workshop that it was being used as a learning resource by some in relation to a study assignment which was to design a tree risk assessment system. If that’s the reason you’re here then there’s still some worthwhile information in the thread, but if you haven’t yet, I would recommend having quick look at the summary of what QTRA v5 does which makes it an improvement and significant development from QTRA v4; which was the current version at the time the thread ran. The v5 thread dies quite quickly, and doesn’t provide much more information than the opening post. http://arbtalk.co.uk/forum/general-chat/72575-quantified-tree-risk-assessment-qtra-version-5-questions-answers.html Also, there’s a thread called “Tree Risk Assessment: TRAQ & QTRA - Compatibility and Common Ground”, which has a document attached to the first post, and it supersedes a similar thread that was run in 2014. There’s a lot of complicated analysis about risk matrices in there, and a lengthy discussion about tree risk assessments crossing the boundaries and becoming risk management decisions. Some of you might find this useful as well. http://arbtalk.co.uk/forum/general-chat/88697-tree-risk-assessment-traq-qtra-compatibility-common-ground.html Any questions about any of these threads, either reply on the thread, or PM me. Cheers Acer ventura

Hi Jon 50% certainty is 100% uncertainty, with the Mars Forecasts! story. And I'm 95% certain I'm presenting at the AA Conference with this: Estimating Likelihood of Failure - A VALID Approach Wednesday Speakers Are you still a QTRA Registered User? If so, you might be interested in a couple of QTRA days out that are being planned at Chatsworth for Saturday 18th July, and Windsor Great Park on Sunday 2nd August. The rough order of play is to have them like the UKTC days out with a social potter and picnic around some marvelous trees, but we’re also planning on running Probability/Likelihood of Failure calibration exercises to harvest the wisdom of the crowd, and I’ll be further road testing VALID. Having messed up a perfect photo opportunity in Sydney earlier this year, I also need some P/LoF calibration photo sequences for the ISA conference in Florida. International Society of Arboriculture Which could be interesting. Cheers Acer ventura

Hi Jules I'm much more than 'pretty sure' about the levels of risk the experts reported to the court, and this was a critical point in the case (see transcript, second joint statement, and judgment), otherwise I wouldn't have posted it. These are facts and not opinions, and I have been careful with my words; as I was when I was involved in uncovering this. Perhaps you could heed your own advice, rather than having a 'dig', and might find it useful to read document 11 (and 9 as well) in the link I provided before speculating, with selective cut and paste, whether or not you think I'm being 'libelous'. Cheers Acer ventura

Hi Chris A bit more clarification. QTRA generates a colour-coded probability that the risk owner/manager can compare to the Tolerability of Risk Framework, so I'm not comfortable calling it a 'scoring system'. Quantitative would be better. The legal context is a key element and is built into QTRA in the form of the risk assessment parameters and risk management guidance for advisory risk tolerance thresholds. TRAQ is trying to do something similar in terms of risk assessment but qualitatively (with words not numbers). However, it diverges significantly at the risk management level because the risk owner/manager has to choose one of the TRAQ outputs of Extreme, High, Moderate, or Low as their risk tolerance threshold, which effectively becomes risk management policy. The ‘Visual Tree Assessment - Estimating Probability of Failure’ day is paired with QTRA and those attending QTRA are strongly advised to attend both. However, both days stand alone. The ‘Estimating Probability of Failure’ day is entirely focused on that one element of tree risk, though there are some risk calculations in the field exercise as well. It's the most uncertain part of the risk assessment, hence dedicating a day to looking at the underpinning principles on which it is built, and calibrating PoF range opinions in the field. It's a curious thing that in the UK in particular (it's not the case in other countries), it's regularly the case that some arborists attending QTRA think they don't need the ‘Estimating Probability of Failure’ day because they know ‘VTA’. However, during the field sessions on the QTRA day when we're calibrating Probabilities of Failure, many realise the value in this and wished they're signed on for day two as well. QTRA has tried to address this with a revised description of the day here. Other Training Workshops I’ve also attached an old QTRA Newsletter where I wrote a piece that describes in much more detail what happens on the ‘Estimating Probability of Failure’ Day. As for how PTI relates to risk assessment. I’ve never been entirely clear how it deals with the risk assessment side of things. I see value in the contents, but for all the tree inspector’s knowledge of tree law, biology, decay fungi, VTA, recording data and the like, at the end of the inspection the inspector is still going to be making some form of risk assessment decision, and this has to relate to the risk management tolerance thresholds of the owner/manager. Cheers Acer ventura QTRA Newsletter Issue 14.pdf

Hi Jon Good points. The court can be hostage the experts it gets. QTRA was used by both experts in the notorious 'Poll' case (that led to the formation of the National Tree Safety Group). Unfortunately, both experts made gross errors in the application of QTRA and therefore misled the court about the level of risk attached to the Ash stem. Document 11 in the Poll section here: Quantified Tree Risk Assessment Cheers Acer ventura

Hi TTS North Thanks. You might be one of the few that's 'looking forward to reading' it. If you still have the will to live after reading it, I'd value any comments you might have. The plan is to update it as and when there's better information, but I'm planning to update from the draft version at the beginning of next month anyhow, to pick up any comments, suggestions for improvement, errors etc. Something I've spotted straight away is the pdf rendering has clobbered the quality of some of the illustrations I painstakingly took care to import via PowerPoint to image by messing about in the registry to get publishing quality of 300dpi with them. Cheers Acer ventura

Hi Andy Thanks for your post. I thought I'd deal with this first before popping back to clarify some other things about how the Visual Tree Assessment - Estimating Probability of Failure day ties in with QTRA. I had read Paul's post, hence my confusion. I think I've ended up with the same level of bafflement about it as Jules. A QTRA Registered User cannot join CAS without the requisite arboricultural accreditation (level 4), which is at a higher level than is required to attend TRAQ (level 2). So the non-arborist QTRA User cannot be an issue with Arboricultural Professional Competencies (APC) because they wouldn’t be able to join CAS. Similarly, a TRAQ accredited level 2 arborist wouldn't be able to join. It isn’t even an ‘implied’ indication of ‘better’ or ‘worse’. CAS is clearly saying TRAQ IS a BETTER tree risk assessment method than QTRA because after 10 years they are aiming to drop QTRA as an APC and replace it with TRAQ.

Hear, hear! TimberCutterDartmoor. Down with this sort of satanic science mangelwurzel thing. It be the Devil’s work I tell thee. This revelation was just visited upon me in the belly of an iron bird, returning to these god-fearing shores after an apprenticeship in a yonder land they call Salem. Whence, to inquisition a tree to tell if it be a baddun and not forsake it’s nature, I take a big bucket of sand, bury my head in it, pop a mandrake up my fundament, and wait for the disease of astonishment and the speaking of tongues. I’ve even cut off all me fingers and toes in case I become weak and get overwhelmed with an irresistible and demonic urge to try and measure. Cheers Acer ventura

Some more clarification. QTRA has had a mandatory test to be completed after training for nearly two years now. The candidates have to score 100% in the test or they do not become a Registered User. As for qualifications and experience beforehand. You don’t necessarily need to know that much about arboriculture to use QTRA because it is a proportionate Target-led risk assessment, not an exercise in measuring and describing trees and all their defects. So why exclude other professions, or tree owners/managers, who know how their land is used, and can identify gross defects? In order to generate an unacceptable level of risk in the Tolerability of Risk Framework, the Probability of Failure range has to start from Benchmark Red 1. Such a tree would have gross and easily recognisable ‘defects’, or ‘habitat features’. Where arboricultural skill is needed is in interpreting the significance of defects where the Target value is high enough, or the tree has sufficient value, to warrant the detailed risk assessment. It’s up to the risk assessor to acknowledge and work within their limitations, and know when to seek further advice about the Target or Probability of Failure, whether they be an arborist or not. Curiously, in my experience as a QTRA trainer, it is often the case that it is non-arborists who grasp the concept of tree risk assessment the easiest. With some arborists it can be the case that they have to first unburden themselves from an ingrained hazard abatement approach where they identify defects and then want to do something about them, irrespective of the risk level. They can be prone to assessing what could happen rather than what is most likely to happen. Just in case there's harm and someone afflicted with hindsight bias claims they could've foreseen it after the event. Just as a matter of curiosity, does anyone happen to know what this 'CAS Model' is, and how QTRA fails to meet it? I'd also be interested where the label 'critically acclaimed' comes from in relation to TRAQ? It implies that TRAQ's been critically reviewed and the outcome of the review was 'acclaim'. Anyone know where any of these reviews are? Cheers Acer ventura

Hi Sloth Just by way of clarification. QTRA is not about ‘hazard perception’, it’s about risk assessment. A hazard is simply something that could cause harm. Risk and hazard are importantly rather different things. Something that is hazardous can have a very low risk. With trees, trying to manage hazards without consideration of the risk attached to them, will usually result in 'risk averse' decisions which waste resources and loose benefits. I’m not sure what you mean by ‘perception’ here, but with regard to ‘precision’. QTRA makes an awful lot of effort not to be 'precise' because with tree risk assessment you can’t be. QTRA's inputs are in ranges, and its outputs are worked through Monte Carlo simulations for the very reason that you can't be that precise. I think there’s a pervasive confusion that numbers mean laser like precision and accuracy whereas words in such circumstances, even though they must have a measurable meaning and boundaries, are better because they are less precise. There’s a neat illustration that highlights this issue. A Martian, who hasn’t been slowly and surely drawing up their plans against us, lands here today and is asked to forecast the chances of it raining tomorrow. If they said there’s a 50% chance of rain the next day, they are being very precise about the fact they have no idea. They could not express how clueless they are about the chances of it raining tomorrow with any more certainty. Whilst on precision. I’m sure many of you following this thread will have seen I’ve uploaded a summary document of the TRAQ thread from last year. In it, one of the issues discussed is that a QTRA Probability of Failure (PoF) range 1 (in numbers 1/1 - <1/10) has three of four of TRAQ Likelihood of Failure ranges in it. Imminent and PROBABLE Likelihood of Failure ‘words’ are parked in the top end of PoF range 1, and the upper values of the POSSIBLE ’word’ are in there as well. Now that’s an incredible level of precision, but not obvious because the remarkable accuracy is disguised by the use of words; which can often have a significantly different meaning depending on whoever happens to say or hear them. As for the CAS thing. My only observation is that it does seem odd to me they could potentially alienate their own members who are QTRA Users. Cheers Acer ventura

Hi It's been awhile, but for those of you who have email alerts on this thread, there's now a new one. http://arbtalk.co.uk/forum/general-chat/88697-tree-risk-assessment-traq-qtra-compatibility-common-ground.html Cheers Acer ventura

Hi Andy I won’t try to second guess how long before you lose the will to live with this one . When I was first approached about doing this last year, I thought it would be a doddle. I couldn’t have been more wrong when I eventually rolled up my sleeves to set about it. Making this stuff easy on the eye is really, really difficult. So, if you have any ideas for improvement then please let me know. When/if you get to Betweenness you’ll find I’ve used a curry menu matrix to try and get the concept across. Towards the end of the putting this together I realised I could just about sum up the whole of the 50 page plus document with a simple curry menu analogy. When I tried it out on a few folk though, I got one warning it could look like I was being facetious, whereas others have said I should use it because it helps. In the end I dropped it because it didn’t fit very well with the style or context of the document. However, it still seems to me to be a neat way of encapsulating most of the important issues and complexities into a relatively easy to comprehend paragraph. Here it is: To a tree owner/manager, a TRAQ risk assessment is like placing order for a meal at an Indian restaurant for themselves, and unfamiliar others who have an unknown tolerance to heat. The owner/manager can choose anything on the menu, as long as it’s a Korma. However, the Korma could be as hot as a Madras, Vindaloo, or Phal. At an additional cost, the chef will propose ways of making the Korma as mild and bland (with reduced benefits), as the chef thinks it should be. Cheers Acer ventura

Hi I’ve had a number of requests to compile a summary of the http://arbtalk.co.uk/forum/general-chat/73829-tree-risk-assessment-qualification-traq-isa-best-management-practices.html thread into a referable and formatted document so it could be downloaded and followed without the distractions that can be the fun and games of a discussion forum. The plan is to have this as a working document that can be updated as and when improved analysis comes to the fore, or if any errors that I might have made in it are identified. It is a long, and at times necessarily complicated, document which might be best digested by reading the summary and having a nibble at the bits that take your fancy. I’m uploading it in a draft form for a month to catch any typos and oversights on my part because I’m aware I’m particularly vulnerable to word blindness with it at the moment. I’ve started a new thread because since the last time I posted, just under year ago (around 2 600 views), the view count has gone up by more than 6 000, so there still seems to be a fair bit of interest, and the content has evolved enough to warrant a clean sheet. I will maintain an email alert for any new posts on the old thread, and Steve Bullman as kindly agreed to imbed a link to the new thread in the first post of the old one. So this thread would be a good place to develop any debate about the contents further. Like last time, I’m going to post links to the thread on the UKTC, QTRA, and LinkedIn ASCA and ISA forums to let them know of the update. As with the old thread, if anything of interest comes up on the other forums about the content I’ll let you know. I would very much welcome any comments on the forum or privately about the contents. I greatly appreciate these messages and am more than happy to honour the confidentiality of anyone who requests it. So, here it is Cheers Acer ventura TRAQ - Arbtalk Document First Draft - 05.06.15.pdf

Hi Ian PM me with your email address and which ones you want and I'll see whether I can sort them out for you. Cheers Acer ventura

Hi Jules I thought I had done, and at some length. This might help Betweeness Cheers Acer ventura