by Judith Curry
Some reflections on my transition from academic climate research to private sector weather forecasting and regional climate change assessments.
I’ve received thousand of emails, tweets, etc. regarding my retirement from Georgia Tech. Two messages motivated the theme for this post:
Congratulations on your emancipation, and I have no doubt that you will enjoy and thrive in a skin-in-the-game world!
Good luck with your new venture, in the real world where what we pay for must work as advertised. Change from alternative reality!
Skin in the game
So, what is ‘skin in the game’?
Definition #1: If someone is said to have some “skin in the game”, it usually means that person has something to lose (whether it’s some form of ownership, money, property, or just respect) in a given situation – and that situation is usually something business related. So, the “game” in the phrase just refers a situation where there is something that has some risk involved, where people can potentially lose something.
Definition #2: To be at risk financially because you have invested in something that you want to happen. You take more ownership of something when you have some skin in the game.
Nassim Taleb has a new book in the works: Skin in the game: the thrills and logic of risk taking. Based on the snippets that I’ve seen, I don’t quite ‘get it’, but the title of one of his chapters is The skin of others in your game, which is a theme I want to exploit here.
Weather forecasts versus climate predictions
If you are a weather forecaster in the private sector, you will quickly lose your clients if your forecasts are consistently wrong. Daily forecasts are evaluated daily; seasonal forecasts are evaluated several times each year. Clearly weather forecasters have skin in the game in terms of their forecasts.
With regards to climate projections, the predictions being made now will be irrelevant in 2100, which is their target prediction date. In fact, the forecasts become obsolete every 5 years or so, as new model versions are implemented. Recent attempts to evaluate climate model projections in CMIP5 during the early 21st century have shown striking discrepancies between model projections and observations.
Defenders of the climate models and climate model projections argue that climate models shouldn’t be expected to verify on decadal time scales.
In other words, climate modelers have no skin in the game in terms of losing something if their forecasts turn out to be wrong. In fact, there is actually a perversion of skin in the game, whereby scientists are rewarded (professional recognition, grants, etc.) if they make alarming predictions (even if they are easily shown not to comport with observations).
Skin in the game: forecast uncertainty and confidence
I would like to start this section with an excerpt from my 2013 House testimony:
By engaging with decision makers in both the private and public sector on issues related to weather and seasonal climate variability through my company CFAN, my perspective on uncertainty and confidence in context of prediction, and how to convey this, has utterly and irreversibly changed. I have learned about the complexity of different decisions that depend, at least in part, on weather and climate information. I have learned the importance of careful determination and conveyance of the uncertainty associated with a forecast, and the added challenges associated with predicting extreme events. Confidence in a particular probabilistic forecast is determined by consistency of consecutive forecasts, and historical evaluation of forecast accuracy and errors under similar conditions. I have also learned how different types of decision makers make use of forecast uncertainty and confidence information.
I have found that the worst forecast outcome is a forecast issued with a high level of confidence that turns out to be wrong; a close second is missing the possibility of an extreme event.
In other words, my work with CFAN epitomizes a ‘skin in the game’ environment. Which is diametrically opposed to the ‘skin of others in your game’ approach by academics to climate change.
Working in an environment with skin in the game dramatically changed my perspective on uncertainty and confidence. This perspective (derived from my skin in the game work with CFAN) is very clear in my discussion of climate science, uncertainty, and overconfidence by the IPCC [https://judithcurry.com/category/uncertainty/]
This perspective may be unique in the academic climate community; I don’t know of any other climate scientists with skin in the game in private sector weather forecasting.
Climate prognostications: the skin of others in your game
When I saw the phrase ‘skin of others in your game,’ it immediately struck me as describing the situation of climate scientists/activists. Whereby they make dramatic prognostications of dire consequences, and advocate for emissions reductions with their attendant economic costs and detrimental effects of developing countries.
The scientists themselves have absolutely no skin in the game, other than the perversions associated with being professionally rewarded for making alarming predictions and claiming that anyone who doesn’t agree with them is a science ‘denier.’
Who among the climate activists would redirect their research funding and travel funding to efforts to reduce emissions? Who among the climate scientists/activists are making a serious effort to reduce their carbon footprint, e.g. air travel? [see Walking the climate talk]. They have no skin at all in the game. This becomes truly perverse when they insist on the skin of others in their game.
I said in my post JC in transition that I thought that the private sector is a more ‘honest’ place for a scientist than academia. In this context, in the private sector you have skin in the game with regards to weather forecasts (and shorter term climate forecasts), whereas in academia scientists have no skin in the game in terms of the climate change projections.
Making shorter term weather or climate forecasts, with some skin in the game, would be very good experience for academic climate scientists. And this experience just might end up changing their perspectives on uncertainty and forecast confidence . . . to something that sounds like my perspective.