After a week or so, I’m back.
Today, let’s discuss outcome bias and its impact on learning, especially in today’s traditional K-12 system.
Outcome bias is the assumption that good results are always the consequence of good decisions and superior performance.
Rasmus Ankersen, author, speaker, director of football at Southampton and president of the Turkish Goztepe, talks more about outcome bias in his TedTalk from 2018:
…
“…Newcastle United has all the building clocks to become a really successful football club. They got a billionaire businessman owner, they got a fan base most other clubs would die for. But despite all that, Newcastle is also known as the team that never really fulfilled its potential. But then a little bit more than five years ago, in 2012, the fortune seems to change for Newcastle because they sensationally finished fifth in the English Premier League, the strongest league in the world. That is an amazing achievement. All the experts predicted Newcastle to be battling to avoid relegation. Now they were fifth. Inside Newcastle, the excitement was bubbling too. The management team was convinced this is the start of a new era in the football club’s history. So they decided to reward the head coach and his whole coaching staff a historic-long, unbreakable eight-year contract. And they really tried to do everything they could to keep the squad together for the following season, because that’s how successful organizations are supposed to think, right?”
“The following season this happened. Newcastle almost relegated. They finished 16th, and no one had a clue what was going on. Because they had the same players, they had the same coaching staff, they even played in the same stadium. I mean, nothing’s changed. How could a largely unchanged team go from fifth to 16th in less than 12 months?”
“I’m going to try and answer that question. But first, I’m going to give you a little bit of background. Because in football, there’s a good old expression that’s frequently used by coaches, players, commentators, experts. It goes like this: the league table never lies. It kind of assumes that once the last game of a season is played, justice prevails. The team that finished top of the table is the best-performing team, and the teams that relegated are the worst-performing teams. At the end of the day, everyone gets what they deserve.”
“It’s not that simple, actually. And let me explain why. Please meet Matthew Benham. He’s a gambler. He convinced me the brightest guys in football don’t work for the football clubs. They work in the gambling industry. And I’m not talking about the amateur gambler, the guy that is placing a bet every Saturday morning to trigger his weekly rush of adrenaline. I’m talking about the pros, the guys that are using sophisticated mathematical models to place the bets. That’s the brightest guys, and Matthew is one of them. He’s an Oxford graduate in physics that decided to set up his own better syndicate in North London, basically an office full of PhD statisticians calculating the probabilities for the outcome of football games….”
…
“So here’s the thing. If you want to understand how a gambler thinks, there’s one concept you must get, it’s this: the league table always lies.”
“Why is a gambler saying this? Because he understands football is a very random game. It’s a much more random game than basketball, for example. Why is that? Because football is a low-scoring sport. There’s not a lot of goals in football. No, the average number of goals in a football game is 2.8. The average number of goals in a basketball game is above 200. Here’s the principle. The fewer goals there is in a sport, the more impact random events have. Random events like the ball getting deflected and spins into the net, or the referee making a wrong call in the last minute of the game. Ultimately, this means that the best team wins less often in football, a low scoring-sport, than in a high-scoring sport like basketball. And that’s why the league table lies. It lies more after ten games than after 20 games. But even 38 games, that is a full season, 38 games is a very small sample size. You know, not enough to strip out the randomness from the equation.”
“So a gambler is in the business of making predictions and backing those predictions with his bets. And when he decides to back a team and place a bet, he doesn’t look at where that team is in the league table. He looks at underlying performance indicators that have more predictive value, tells him more about where that team is likely to go in the future than its current league table position.”
…
“I believe that Newcastle was blinded by one of the most common management delusions, what psychologists call outcome bias. The assumption that good results are always the consequence of good decisions and superior performance. Or, let’s put it in another way: success turns luck into genius. And this is where the message becomes about much more than just football analytics, right? Because when an organization fails, it becomes very natural for us to dig for reasons, to ask the tough questions. Why did we fail? How can we do things better? But when we are successful, we don’t ask those questions. We tend to just cruise on blinded by outcome bias. We expect results to continue automatically…”
…
“…[A] good gambler has trained his brain to resist the human tendency to assume that good outcomes always mean that someone acted brilliantly. A good gambler treats success with the same skepticism as he treats failure. And this is what successful [learning] organizations should do too, if they want to stay relevant: ask themselves the questions of a gambler. Why were we actually successful? How do we separate luck from skill? Is it possible that our league table sometimes lied too?”
In the traditional K-12 system, decision-makers pay too much attention to their versions of the league table – test performance, attendance statistics, class rankings that list top grade point averages down to the lowest. All this data feeds into the very questionable practice of outcome bias. What is needed inside learning organizations worth their mettle is the data usually ignored by the K-12 decision-makers, data like how well a young learner can define, plan, execute, and evaluate their own learning.
Learning is not a linear process. Embrace the random! That way, young learners will own their own learning in a personal and authentic manner.
Til tomorrow. SVB
A Better Path to Learning 
Leave a comment