Predictive Indicators Are Hogwash

lightning during nighttime

Photo by Johannes Plenio on

By Phil La Duke

I didn’t get my blog written until late last night, because a) I wasn’t feeling well, b)had started a topic that was way too long to post, and c) I was very busy, But something gnawing at me for awhile: this whole emerging infatuation with the idea of “predictive indicators” as it pertains to safety.  Let me begin by saying there are a lot more companies relying solely on lagging indicators than using a holistic approach that compares leading and lagging indicators to accurately gauge the companies’ current performance (never mind trends and likely future performance). Even within organizations with the organizations who meticulously gather lagging indicators too many do nothing to interpret the data and turn it into something that is meaningful, useful, and actionable to and for Operations.

Let me dispel the myth of safety “predictive indicators”. The term “predictive indicators” was borrowed from the world of finance and was originally applied to the ability of stock analysts to use an algorithm to predict market fluctuations, as anyone who has stock (or a retirement fund that has stocks) these predictive indicators aren’t very accurate and some believe are no better than throwing a dart at a board.  So while stock analysts still look at these predictive indicators, few of them are staking their entire fortunes on them and fewer still are betting the lives of others on them. In any case, in most cases that I have seen, when safety pundits are using term Predictive Indicators, they are either misusing the term to describe a leading indicator, or have developed a model that is impressive in its complexity but as reliable as a degenerate gambler’s system for picking the winning horse.

So what is a predictive indicator anyway? Probably the most familiar forecast using predictive indicators are weather forecasts.  It’s an apt analogy to both the stock market and performance of Safety within an organization. I tried to research exactly how reliable weather forecasts were and it was like trying to get a politician to go on the record.  As one article, put it “We know that weather forecasts are inherently uncertain. (We’re predicting the future, after all!)” What I learned is that the closer you are to the event the more accurate the prediction will be, but as you move out 5 days there is an exponential drop off in accuracy and when you get to a 14-day forecast you basically have a SWAG (silly, wild-assed, guess).  So while this lack of accuracy might prompt you to grab an umbrella on your way out the door, it’s hardly the same when it comes to dealing with financial decisions and the life and death of workers.

We need to nip this in the bud because there are non-safety executives now co opting this term and wondering why we in safety can’t use the information at hand to scry the future, and once we know what’s going to happen, can come to the rescue just in time.

I know I am once again rattling the cages of academics and con men who have spent years developing an algorithm or pretty, complex, and useless model which they want to sell to you for an incredibly low price.  The problem is we can’t predict the future outcomes of a complex system and the stock market, the weather, and safety and even a guess can be dangerous. Again, think of the weather forecast. If the forecast calls for a 70% chance of rain, that means that there is a 30% it won’t rain, and even if it DOES rain, it doesn’t mean that it will rain where you are, so most people figure, “it’s probably gonna rain” and dress appropriately.  Except me, I don’t believe in umbrellas—you still get wet and you have the inconvenience of having one hand encumbered—and forget rain hats and raincoats (what am I a cod fisherman?)

Playing the odds

I can’t tell you how many people, when I warn them of a hazard, look at me with a “gimme a break” look on their face and ask “yeah, but what are the odds that’s gonna happen?” I’m honest with them, I will tell them that the odds are low but the stakes are high.  It’s like a reverse lottery ticket. I know plenty of people who will not swim in the ocean for fear of sharks, despite the astronomical odds against being attacked. They aren’t playing the odds because the potential outcome is death and the reward (swimming) isn’t sufficient for them to chance it.  Why then do people in the workplace engage in high-risk, high-consequence activities? What’s the reward for failing to control the energy while performing tasks that require it? You save a couple of minutes? What is the probability that something could go wrong? (I’m not going to waste the time and effort to calculate the odds, but suffice to say, the odds are considerably higher of dying in a LOTO accident than it is being attacked by a shark, and swimming in the ocean is a far better reward than finishing a job 5 minutes quicker.

Duration of Exposure

While it may be impossible to predict with any sort of frequence how and when someone will get hurt we can look at leading indicators that correlate to injuries. The duration of exposure is a decent indicator of the risk associated with a given activity.  All other things being equal, a worker who spends 5 minutes in a confined space is at significantly lower risk of being harmed than a worker who spends 12 hours in a confined space. It stands to reason that, minus any other risk factors, the worker who is in a confined space is 144 times more likely to be harmed than the worker who is in a confined space than the worker who spends 5 minutes in a confined space (5 minutes x 12 =1 hour, and 1 hour x 12=144 five minute increments).  But even this is of limited value, because there is no guarantee that either worker will not be harmed in the first 5 minutes, which is why I qualified it by saying, “minus any other risk factors”. This should not be seen as an encouragement of rushing to get a job done, rather it should serve as a warning to the people who take risks because they are “only going to be in there for a minute.” Statistically, people tend to underestimate how long a task will take to complete and that means that people consistently underestimate the duration of their exposure and subsequently the risk of being killed or injured.

So let’s erase “predictive indicators” from our lexicon and focus instead on doing a better job of using appropriate lagging and leading indicators to calculate or risks.

Did you like this post? If so you will probably like my book which can be ordered here I Know My Shoes Are Untied. Mind Your Own Business or on Barnes & Note: If you are outside North America, you will want to order for the Amazon site in your country.  Did you hate this post? Did it offend you deeply? Maybe you should organize a book burning (minimum of 150 books) but be sure you are only burning my book, I don’t want you to go to a used book store and buy a bunch of cheap books and stack mine on top.

The book is a compilation of blog posts, guest blogs, magazine article (from around the world) and new material. Much of it is hard to find unless you know where to look. A second and third book has already been green-lighted by the publisher (expect fewer reprints and more new material).

Remember the holidays are coming up and this book makes the perfect gift for the person for which you feel obligated to get something for but don’t really like.

In all seriousness, I have been blogging for free for over 11 years and I think I have earned a bit of revenue so buy the damned book.