I sadly cannot attend the inaugural OptaPro Analytics Forum this year, but in order to cut off the Negative Nancy vibes about the state of the industry, I’ve written a handy keynote address that I would have given had I been the keynote speaker.
Hello. Though Opta have told me they want this conference to focus on the actual work of soccer analytics and not the meta crapola of which this author has become something of an expert, I hope they will indulge me a little here.
First, I’d like to thank the organizers for such a lovely spread and great catering in general. Rumours of English cuisine have been greatly exaggerated. *silence*. Right. Next, I’d also like to thank all of you for coming out in such terrible rain today. As ever, the statisticians can tell us the likelihood of bad weather occurring, but they can’t prevent it, eh? *silence* Haha. Yes.
Ladies and gentleman, I ask you, where should the analytics industry best focus their attention in order to follow the footsteps of baseball, basketball, and even (in some cases) hockey in bringing analytics into both the boardroom and the dressing room?
Should it be fans? Perhaps, as football supporters really are the engine of the game *laughter*. (Haha, of course not, it’s their money). The problem however with convincing fans that football analytics is a worthwhile pursuit and not simply the slow addition of angels dancing on the head of a pin is that you end up either preaching to the converted, overwhelming the confused, or wasting your time on ideologues posing as skeptics.
Now, to be fair, a lot of soccer fans do earnestly love statistics. But while some certainly do get what we’re trying to do here, many don’t know the difference between a Per90 and Per Game metric, and they don’t care. In my darker moments I think data companies really should just sell the raw data to them and call it a day. My advice is that when it comes to fans, be nice to your allies, engage with the earnest skeptics, and ignore the hard-hearted.
Should it be the media? Well, the problem is that a good part of our work challenges the very heart of how television and newspapers—and human beings as a whole really—have always viewed the sport of football. To them (and to most of us when we don’t have our thinking caps on), football matches are stories. They come in three acts, with goals as convenient plot points. Star players are not beneficiaries of accumulated advantage, but rather agents of destiny. Losses aren’t just a production cost, but a sign of evil. This is perfectly natural and very difficult to change.
While it’s fun to play iconoclast from time to time, I’ve come to believe that these stories, shared among fans and journalists alike, are generally harmless. It’s also not fair to say that these stories are always false or misleading; the best fiction, as someone once said, is true in the real sense of the word (hence I encourage you to disregard Taleb and keep reading those newspapers). But, as it stands, those stories should absolutely not be the driving force of how decisions are made by football clubs. Not if they want to be successful.
To that end, should we focus on managers? No. Because even if were to convince this particularly conservative, strictly football-minded lot, they’d be gone long before their ideas took root. How about technical directors? Despite appearances, DoFs are at their best when picking out bottles of wine in the Duty Free on their way to meet player agents. I exaggerate here, but not by much.
No, the big fish here are the owners. And not any kind of owner, but that rare beast that not only wants to spend his or her money sensibly and win trophies at the same time, but, crucially, really believes they can do it. Because I will tell you right now: they cannot do it without us.
To give you an idea of why, here’s a question: what goes into a player transfer? The simple answer is a club wants a player, comes up with a reasonable bid based on the player’s accomplishments and age, and then makes an offer which the other team either accepts or refuses.
That’s one way of looking at it, and not particularly an effective one. Because a transfer fee isn’t the cost of the player in question, but rather the cost of the risk associated with taking that player on. The better the player, the more likely they will continue to perform at a high level, the less risk in buying that player, the higher the transfer fee.
But most clubs, at least on the surface, calculate that risk in a very crude way. How many goals did they score? How many assists did they make? How many first team appearances for a big club playing big matches against big teams? Are they, as PSG owner Nasser Al-Khelaifi reportedly said of Newcastle’s Yohan Cabaye, “a nice guy”? This is all lovely, but doesn’t exactly justify the increasingly enormous sums clubs pay for elite players.
Within that considerable knowledge gap lies a huge opportunity for all of you in this room.
Because a smart club would want to walk into a transfer negotiation armed with knowledge, knowledge gleaned from some fairly reliable data. If I’m buying that player, I want to know what is a reasonable fee based on the likelihood of a player of his age, nationality, position, career stats, and key performance indicators increasing in value over time. Using the same data I also want a general idea of the probability that player will hit a proscribed number of targets based on any number of performance indicators and historical samples, and I want that probability given in a dollar (or pound) amount of transfer money I’m risking here. And I would also want to see if there are any like for like alternatives at a lower price, with KPIs weighted by respective league and team influence to help mitigate risk. I would also want to know whether my first team coaching staff is using the latest science to ensure that players improve as much as possible and training in a way that improves them without putting them at risk of injury.
I also want to know about the specific market I’m facing when trying to buy that player. Do some player agents command higher transfer fees for similar players than others? Do certain clubs have a record of inflated fees? Does buying at a particular time in the market inflate fees? Does the length of negotiations? Little things, big things. Things that will lessen both risk and price at the same time. What is the potential payoff in buying now and not later? Etc.
The person who will understand this most is the person who chairs the board, helps hire front office positions, and sometimes bankrolls the team. It’s the people who worked with analysts and quants, who know that success in business isn’t destiny but careful risk management, which is never immune to failure even in the best of circumstances. The owners, my friends. You need to work on the owners, or at least be patient as they come around. So don’t worry about the state of analytics. It’s just fine. Do the work, wait for the opportunities, don’t throw your pearls in front of pigs, and practice your elevator pitch should you see John Henry in the Hilton one day. Thank you.