Taken from a long article about social media but worth a read.... In 1998 Jones reviewed Money Ball the second book by then-up-and-coming nonfiction writer Michael Lewis. He gave it a rave and like the rest of the world became fascinated by the revolutionary power of analytics in sport. But as Lewis’s predicted analytics revolution consumed not just baseball, but the world as he knew it, Jones, like the rest of us, began to grow disillusioned. He began to notice the collateral damage everywhere. “I started to notice the stuff we were losing. Analytics were killing guys like Jim Fregosi,” he says of the man who inspired the so-called “Eye Test” – the old school subjective method Fregosi and other managers relied upon to spot talent before Moneyball changed everything. During his years as a sportswriter, Jones began to notice the way analytics were being misapplied, often irrationally, to the detriment of clubs throughout the global multi- billion-dollar industry of professional sport. Nowhere, he says, was this more obvious than in the world of European football. “Analytics are great for baseball, because it’s a very confined system. There’s not much movement, it’s pretty mathematical and measurable. But with football, how do you quantify the value of a defensive midfielder? A lot of it can’t be quantified, there’s too much movement, too much luck is involved. Similarly, ‘possession’ has become a huge determining statistic in football, but in real terms it doesn’t mean that much. Statistically you can easily dominate a game of football, but still lose because you let in one goal. It happens all the time.”
I saw the last half of the film a couple of months ago. The most telling thing for me was in the end credits,where it states that Billy Beane has never got a team to the last game of the season (final). I don't think he is involved in Barnsley FC anymore, but analytics can only take you so far. Is my analysis.
I watched it a couple of weeks ago, and didn't get a lot from it really, though I think that may have been to the way the film was made. I knew he'd never ultimately succeeded with his philosophy ahead of seeing the film, simply outperformed for a period. The issue I've had is that baseball is a very static game and the KPI's can be very easily tracked. You throw and you swing ad you run or jog, and you catch or throw in the field. You can start to segment those stats in allsorts of ways, but thats the nutshell. Football is extremely dynamic with so many moving parts. The elements play a part, officials can sway a game, a crowd can influence a game and one lapse of concentration can mean the difference in loss or defeat, even if all your other stats are perfect. I've no doubt that capturing large amounts of data on players can assist to get a picture of a players ability. But it has to be the right data. And it has to be in the right context, and be transferable to what you're asking them to do at your club. The most successful transfers were under Crynes approach when confined largely to league 1. Robust data collected over a good period, at competitive levels, with players who had a lot of games under their belt, but still in their early to mid 20's. The other thing, as the Moneyball film showed right at the end. And it's the same in business. If you do something different, and you aren't the wealthiest in your industry, if it works, others will copy it or buy it within 2 years and be more successful than you were. It's a small window of opportunity. I think ours may have lapsed.
I saw the film last year, and the other thing that struck me about it, aside from the fact that Beane never actually won the comp, was a quote from the analyst early on in the film about how the model wasn't about hitting home runs, it was about winning games (or something along those lines). I.e. building a team, as opposed to just signing players that have potential. In fact, I'm not sure resale value has any part of moneyball, which is more about building a competitive team on a budget. Then there's the fact that data needs context. So if a player has decent stats, what level was it at. What have other players with similar stats at that level gone on to achieve? For the successful ones, what environment / system were they successful in. I suspect we ignore the latter, as those metrics would be giving the data to say that players need some experience alongside them. Then there's the qualitative / anecdotal data about a player's character and attitude. That's why there have been so many calls on here for a Director of Football. I think the summer has shown that there is a place for using the data, as we've seen what squad has been assembled when we don't. But at the moment, the club does not know who the head coach will be, or what style of play they will be looking for players to fit. Those decisions need to be made very quickly, because until they are, the data will be useless because the recruitment team will not know what they are looking for.
I agree with most of that. But I thought I’d read somewhere that along with the performance data, we did also include info re personality in the decision making process. And I believe the point re head coach is that’s supposed to be included in the model too. ie the point is to hire head coaches who fit a similar style of football. Choosing to tweak that, with no input from experts has been the downfall. Unfortunately this season it all came crashing down due to the fitness coach leaving, no CEO to oversee replacement of key back room staff, coach or players. Leaving us with what might as well have been some random picks throughout the organisation. Schopp was clearly out of his depth, with no support network at all, and poor Poya isn’t much of an improvement.
I can't recall the figures off the top of my head but the gist seemed to be that he got his team to overperfom on a budget that was way lower than the big guns like the White Sox, Dodgers etc
