

Of all the recreational thrills one can seek, iceboating is No. 1 with me. It is an elusive thrill, since I only iceboat locally and global warming is not helping. I am a member of our local iceboat club, but I will admit that I am not tightly dialed in with the real experts in the Club. I know there are members who watch the northeast weather forecasts religiously and who will quit their job, hook up their iceboat trailers, and travel for miles for good clean ice.
I like the idea of that, but I do not have the bandwidth to pursue that dream. Instead, I hope for local ice. I am not sufficiently expert to judge the safety of our local ice, but over the years I have developed a system that I consider to be foolproof even though, as I will try to explain, it is inherently flawed on a system level.
When the temperature drops and ice appears on the river, I detour on my way to work past the iceboat club. When the small iceboats start to appear in the lot, I cancel all my appointments, break out my boat and take it to the lot. I figure that when the cognoscenti bring out their boats there is a decent chance of iceboating on the river, and I need to get ready because the iceboating window is rarely more than a day or so. (I will add the Club has an iceline too, but I like my little ritual too much to have it superseded by technology.)
Temperatures have been dropping, so I have been driving past the club in the last few days and, so far, none of the local stuff has been brought to the lot even though the river is pretty much frozen over. That means that the weather forecast is such that safe ice will not be achieved. The wait continues.


These detours made me think of an old joke.
Montana state weather forecasters were trying to determine if the coming winter would be warm, normal or cold.
They were running every model they could run. They found that the models were contradicting each other, and they were trying to figure out on how to make sense of it. Then one of the forecasters had an idea. He suggested that they would check with the local Native Americans. They had been living in these lands for thousands of years and if they looked at how much firewood they were storing, it would provide them with an additional indication of the severity of the upcoming winter.
One of the forecasters drove out to the local reservation and saw the Native Americans were doing some serious firewood stacking.
He went to the state office and told his colleagues that there was some serious firewood storage going on. Looking at the data again and with the additional knowledge of their Native American experts, they decided to issue a forecast that the winter would be “colder than average”.
All was well, but a week or so later the forecaster decided to detour to the reservation on his way to work and discovered that the Native Americans were working hard at growing their woodpiles. At the office he told his colleagues, and they decided to update their forecast. The winter would be “significantly colder than average”.
A week later the forecaster noted that the woodpiles continued to grow, and the forecast was updated to “unusually cold”.
Since now the forecasters were moving into more extreme forecast territory, they checked again a week later and the whole reservation was cutting, splitting, and stacking wood at a furious pace.
The forecasters changed their forecast to “a possibly record setting cold winter”.
This concerned the governor, and he called the weather forecasters for a meeting in his office. They grabbed their forecast printouts and went to his office. When they showed their forecast, they explained it was heavily reliant on the Native American data. The governor was a great admirer of his state’s Native American heritage and scheduled a press conference to warn the state of a possible looming state emergency.
When the forecasters went back to their office, one of the forecasters started to worry. The computer models were obviously wrong, but how did the Native Americans know it was going to be such a severe winter?
He drove out to the reservation and stopped a Native American who was furiously pushing a huge wheelbarrow of firewood.
“Excuse me Sir, can I ask you a question?”
The Native American set down his wheelbarrow and despite the autumn cold he wiped his sweating brow and tried to catch his breath.
“Sure”, the Native American said, “but make it quick I am very busy.”
“Thank you, Sir. How do you know it is going to be such a severe winter?”
“The same way we have known for years, we always check the state weather forecast.”
This is actually an engineering joke. It deals with unstable feedback loops, flawed data, bias, false experts, data testing, and the failure to take all data into account.
It is a joke, but it can give an engineer cold sweats from the memories of the times they similarly trapped themselves.
And sometimes continue to do that. Like me, in my iceboat forecasting system.
If everybody uses my system, the appearance of a single random iceboat in the lot can start an avalanche of iceboats into the lot without the chance of safe ice.
However, so far, my system holds. And the reason is that, at this stage, I am relying on true experts to lead the way. Once the true experts are removed from the system, we will simply fall into a random irrational unstable negative feedback loop.
The iceboat paradox, to a large extent, explains why we are floundering today with regard to strategic planning. Too often we have removed the experts from the system and are giving in to responding to random noise. Experts are not always right, but when there are no experts, due to irrational unstable feedback loops, we can actually have more incorrect predictions than correct predictions. In other words, without experts, we are doomed to bet less than 50%.
To properly evaluate a prediction is actually quite easy. Ask yourself who is making the prediction. Then ask: What is the subject and is the predicter expert in the subject? If not, it is just noise. Again, experts are far from always right, but even if experts are only 51% correct, it is still far better than the results of noise driven feedback loops by people who have no clue. Very few people are truly stupid, but very few people are truly expert.
Ironically today’s human society only exists because we have subject experts like doctors, engineers, mechanics, electricians, weather forecasters and farmers. Not each and everyone of them may be ranked as expert, but if they show they are capable of rational analysis, it probably pays to listen to them, rather than behave like a sheep, or a lemming, and follow a chicken afraid of drones in the New Jersey sky.

