The official English Premier League fantasy football game, FPL, is currently played by over 7.2 million people, globally. Among the millions of mini-leagues that players organise themselves into, there is one that stands apart. Like many others it’s a work league of sorts, but unlike those others its members weren’t necessarily football fans before they joined; nor did they even choose to join. The league is organised by the director of the graduate training programme at one of the City’s largest banks. Each year the director corrals together her current cohort of aspiring bankers and traders into a mini-league and leaves them to compete it out. It’s a required part of the programme, and it’s fiercely competitive.
So why does FPL make it into the programme, alongside the courses in bond maths, company valuation and strategic analysis that the graduates have to take? The answer is that it sheds light on the way that financial markets operate. In financial markets participants have to make decisions under uncertainty. Many games share that feature, like poker and bridge. Warren Buffett, arguably the best investor in the world, is a keen bridge player, and David Einhorn, one of the most famous hedge fund managers, is a World Series poker player.
But what the director of the graduate training programme at the City bank recognises is that FPL is up there too. In fact, FPL has some closer parallels with financial markets: it is played over a long period of time (ten months) and it is predicated on the construction of a portfolio of assets.
That’s not to say there are not differences. In FPL players have to make around 500 decisions over the course of a season, assuming no hits are taken*. In financial markets the number of decisions required to be taken can be a multiple of that, although longer term investors will try to reduce the number. In FPL the actions of other players affect prices/value but they don’t affect points which is what you are trying to maximise; in financial markets the actions of other participants affect prices/value which is exactly what you are trying to maximise: no boom/bust cycles in FPL.
I didn’t do the graduate training programme at that City bank, but I have been involved in financial markets for over 20 years and a player of FPL for ten. There’s a bit of crossover I’ve picked up along the way:
The importance of edge
FPL is a relative game, the idea being to get more points than other players rather than the most points possible. To do that, players need an edge. In days gone by that edge may have been informational – knowing when the double gameweeks and the blanks are, knowing which players are out injured and which are non-starters. Then it became analytical. There is a rich source of data currently available and understanding how to use data around shots in the box, say, or expected goals can provide players with a useful analytic edge over rivals. Finally, there’s behavioural edge: how players make decisions under uncertainty and react to their environment. As in financial markets the ubiquity of information and increasingly of analytic overlay makes this behavioural edge all the more important.
Many of the same behavioural biases that affect the way participants act in financial markets affect the way players perform in FPL. The classic is herding. In millenia gone by it was useful mental shortcut simply to follow the crowd in flight on the assumption that they were likely running away from something pretty terrifying. In financial markets herding can cause bubbles and crashes; in FPL it can cost points. Nearly every player I know has a story of a transfer they did simply because everybody else was doing it.
Other behavioural biases common in both FPL and financial markets include the endowment effect – falling in love with what you own, against your better instincts, and recency bias – being overly influenced by what just happened (which in FPL parlance gives rise to the knee-jerk transfer).
We wouldn’t be human if we eliminated these biases completely but being aware of them is very helpful in both FPL and financial market domains.
Loser’s game
There are two types of game – the winner’s game and the loser’s game. In the winner’s game, the challenge is to win points; in the loser’s game it’s not to lose points. Professional tennis is a winner’s game. The ultimate outcome is determined by the actions of the winner. Victory comes from winning more points than the opponent wins. Amateur tennis is a loser’s game. Both players make mistakes and the victor is the one that makes fewer. Other games are similarly nuanced. Poker is a winner’s game among professionals (who fold more hands), a loser’s game among amateurs. Boxing starts out as winner’s game when the fighters try for a knockout, but morphs into a loser’s game of endurance in later rounds as boxers compete not to fall down.
Charles Ellis, an American investment consultant, wrote about this in the context of financial markets many years ago. His view was that financial markets used to be a winner’s game but increasing professionalisation turned them into a loser’s game. At some point in the past professionals stopped competing with amateurs and began to compete with each other. The upshot is that today, simply not losing money can be enough to outperform.
There is something of the loser’s game to FPL, too. Easy availability of information and analytics has created many more “professionals”. The upshot is the same: covering the well owned players not to fall behind and taking fewer calculated risks can be the most successful strategy.
Skill versus luck
There’s another common behavioural bias that affects us as humans. When things go right, it’s down to skill; when things go wrong, it’s bad luck. Psychologists refer to it as an attribution error. In FPL and in financial markets, luck undoubtedly plays a role. It’s not all luck as some naysayers would believe. If it were you couldn’t lose on purpose. Nevertheless, the role of luck is growing. This is an emergent feature of FPL that has played out already in financial markets, and it’s a consequence of rising professionalism. If the outcome is a blend of some skill and some luck, and the overall level of skill is rising, then luck will inevitably play a bigger role in the ultimate result. I posted some charts on Twitter that show that this is indeed the case in FPL.
Now there’s not much a player can do about this, except recognise it for what it is. One strategy that works in financial markets is to focus on process rather than outcome of any specific week. A focus on process also defends against some of the behavioural tics that can sidetrack a player. As the Fantasy Football Scout ‘Meet the Manager’ series showcases, the best managers have a tangible process and a style of play that fits with their personality and that they consistently play the game by.
I don’t know whether any research has been done yet on how those City bank graduates go on to perform after they are unleashed on trading floors and investment management divisions of the bank. But unless they are spending all their waking hours finessing their FPL teams at the expense of their day jobs, the crossover will bring them many benefits.
* Broadly 12 per gameweek, plus four complete squad builds.
Originally published in 90MAAT Fantasy Football Monthly Issue 02
This is an interesting post and I assume it was written in layman’s terms rather than financial because the audience is likely to be an FPL one but I see some similarities with the Akerlof Lemons model. FPL started off as a market with asymmetric information with “casual” participants simply picking players they knew or from their own teams but who were in competition with more serious operators who diligently research players on pay sites such as Fantasy Football Scout etc. The market, from an Akerloff sense, was “inefficient” and it is always easy – if you are the one with information – to make money or “win”. My observation is that in recent years, particularly this one, there is now a proliferation of free information given out to the casual player via Twitter, free podcasts and YouTube streams. The casual player no longer needs to even do any work, the players most probable to return in a GW and over future GWs are highlighted and made available to all. So in relation to the Akerlof model, the market has not failed (maybe failure in the FPL sense would be the casual players losing interest and their teams becoming “ghost ones”) and is now “efficient”. From the financial world perspective, of course, strong form efficiency would mean all public information being discounted which would mean that the only way in which it is possible to outperform would be luck, or, taking bigger risks on players who have a high variance of returns.