Process and Outcomes

Former Los Angeles Dodgers General Manager Paul DePodesta has a very interesting post on his blog where he talks about the importance of process in the running of his current team, the San Diego Padres. This he does with specific reference to the draft, noting that it takes so long to actually assess the outcome of the baseball draft that one must follow a sound process, and also, that luck can and does play a huge part in these decisions. One can follow a sound process and be the victim of bad luck, and vice versa. He uses the following model:

What makes this model so appealing is that it can, quite literally, be applied across the board. DePodesta acknowledges that one cannot win all the time, but that by following the correct process, one can at least increase their chances of being successful. After all, in baseball, a successful season is pegged at a 60% winning percentage. The question that occurred to me was this: how such a model be applied to football?

First of all, it needs to be noted that there can be more than one 'correct' process. This depends upon the resources available to clubs, and their relative aspirations (these factors are generally linked). In baseball, it is much easier to identify a tangible 'organizational philosophy' than in football, as has been mention on this blog before. I would argue that this is because baseball is at its essence much more quantifiable than football, or many sports. Football can be quantified, however, in certain ways. After all, that is what league tables seek to do.

The Premier League has had its top four spots yield a Champions League berth for six years now. Lets define finishing in the top four in the period between 2002/3 - 2007/8 as success (yes, all success is relative). In that period, the top four teams have had an average win percentage of 61.1% (never going higher than 65.1% or lower than 57.8%). So, the most successful teams are still only winning 60% of the time, in general (and yes, there are differences between those teams finishing first and fourth). It seems to leave a lot of room for failure (in the form of a tie or a loss), and for bad luck.

You can not win all the time, but you can follow the correct process to facilitate winning more consistently. It seems that the centrality of luck and randomness are still underestimated in football, however. A cluster of three or four defeats in a row and teams ditch their manager. This can be attributed to the much smaller sample size in Premiership football (a 38 game season versus a 162 game regular season in baseball). The stakes become higher in each game.

Ideally, teams would like to reside in the top left corner of the model, receiving deserved success. Within the top four, there are two examples of vastly differing processes, both of which have received 'deserved success' (relatively speaking). Chelsea put a structure of experienced and savvy football executives in place to make sure that there was a powerful link between Roman Abramovich's money and the direction of the first team, and that money was not merely being splurged for the sake of it (yes, the latter happened, but they still won two Premiership titles, so I, for one, will not argue). Essentially, the team was constructed with expensive signings, however. This is the process that Man City should be aiming for, in my opinion.

A second example of a good process would be Arsenal, who have built some successful teams (no Premierships, alas) on the basis of player development. Acquiring the best young talent from across Europe had two advantages, as far as I can see: it was cheaper, and it allowed the players to develop a certain style of play with Arsenal. The problem with this process is that, as DePodesta noted with respect to the draft, Arsenal will only discover the outcome a number of years down the road, and this is not always positive. However, Arsene Wenger has done a good job of managing it thus far.

So, in a basic sense, one can talk of processes in terms of the superficial policy of a club in player acquisition. Processes on the pitch are a much murkier question. Randomness and luck again rear their heads. Contradictions emerge. For example, in terms of football tactics, the area in which most number crunching was done led to the development of 'pressing' and the long-ball game (a notable exception would be Lobanovskyi's Dynamo Kiev, but they were exceptional). The idea was that you maximize the resources you have available to you by using somewhat counter-intuitive logic. Unsurprisingly, one of the modern advocates of this approach, Aidy Boothroyd, claimed to be a disciple of Billy Beane.

Another proponent, Jack Charlton, used to tell two stories about his management style. Firstly, he often quoted Alf Ramsey, who always picked the best team, not necessarily the best eleven individuals. Secondly, he pointed out that the opposition could not score when the ball is in their half (ie, advocating the long ball out of defence). There is an inherent contradiction in this approach, as the long ball is premised upon what are oftentimes random outcomes. This can work in positive and negative ways, but it is no guaranteed way of retaining possession. And the opposition also cannot score when they do not have the ball. Perhaps it's a sound process to stop the opposition, but it does not facilitate winning.

On the other hand, there are many apocryphal stories about some of the most successful managers simply telling their players to 'go out and enjoy/express themselves.' This probably needs to be taken with a huge grain of salt, although in the situation where one has eleven immensely talented individuals on the pitch, there is every chance they can play more instinctively than the eleven journeymen, for example. The dalliance with long ball in the 1980s and 1990s has mostly come to an end, and the arrival of more technically gifted players into the Premiership has ensured that keep-ball tactics are the norm (although not without the odd exception, usually a promoted team).

Processes are a little messy then when we try to apply them to how teams line up on the pitch. Tactics are like the latest fashions: they come and go. 4-4-2, 3-5-2, 4-5-1, 4-2-3-1, and variations thereon have all dominated for periods in the recent past. The only exception to this is with away games or in knockout play, where teams may alter their lineup in the hopes of getting a desired result. Given that, in many ways, teams will follow very similar processes when their teams take to the pitch (similar formations, similar passing games, similar philosophy - note the newfangled fetish for the 'Makelele role'), perhaps it is best to think of processes and the model best in terms of how teams acquire and evaluate their talent. Luck will always conspire against you at some point, but smart squad construction will generally get you a long way. Not exactly a revelation, but something that an increasing number of managers have failed to master.