“I don’t care what the stats say, he plays in a Farmer’s League” is a phrase I hear quite a bit and for me whenever anyone says it, all I hear is “I don’t have an actual argument to make about this player because I’ve never seen him play.” But despite its ubiquitous use I don’t quite understand what is a Farmer’s League? How do we decide that a league is a “Farmer’s League?” And finally, does it really matter that a player comes from another league? Does playing in Spain invalidate a player’s data points?
I think I know what people mean by Farmer’s League: they mean a league where the level of play is amateurish, so much so that some of the players have second jobs or are even actual farmers. There aren’t many leagues where this is the case anymore and none of the so-called “top five” leagues employ amateurs. Occasionally, an FA Cup tie or a Europa League clash will toss up a club where one or two of the players were recently working a second job but for the most part, almost every player linked to a club like Arsenal is a professional player who plays for a professional team.
So, we all do sort of know what Farmer’s League means and what’s annoying is when someone applies the epithet to a League like Ligue Un. No one in their right mind believes that French football is a Farmer’s League. But many people these days tend to reach for “top shelf” words when they actually mean something is just “not quite as good” or “maybe the competition isn’t as high as the league we are currently playing in”.
There’s no question for me that the Premier League is the most difficult league in the world. It has the most money and as a result it attracts and aggregates the best players. And judging by many of the metrics out there, it is the league with the two best teams and overall has the most high quality teams. I went to Nate Silver’s 538 web site and downloaded his SPI database. 538 uses a metric called “SPI” to rank every team’s strength. The metric is tested and constantly refined but it’s based on a team’s offensive and defensive strengths relative to how they are predicted to play against an average opponent. They also use history and market value to create an SPI at the start of the season and then tweak it as the season progresses. For example, Liverpool has an SPI of 93.9 right now and their average offensive output is 2.9 goals while their defensive liability against an average team is an astonishing 0.2. At the start of the season, Liverpool’s SPI was 88.7.
I was curious how the leagues ranked around the world and whether there was a “Farmer’s League” in Europe so I took the 538 club rankings and then averaged them for each league and got this chart.
As you can see, the average SPI for the Premier League is the best in the world at 73, but not too far behind that are la Liga and the Bundesliga. In a sort of 2nd tier you’ve got Serie A and Ligue Un and then from there on down there’s a pretty big drop off in terms of quality. One little thing I found interesting is that MLS is about the same level as Scottish football (fitba) which is pretty far down the table in terms of quality. Scottish football’s top two teams have an SPI rating of 69 and 66 respectively but the drop after that (to Hearts) is 44. In terms of Premier League, Celtic and Rangers would be bottom third of the table (Celtic above Wolves and Rangers above Southampton) while Hearts would be a mid-table English League championship side (just above Preston North End).
I tend to watch football from all over the world – I’ve been watching Rangers a bit since Aaron Ramsey joined, I watch Marseille, and I’ve followed Atalanta (Serie A) for a few years now. I watch Barcelona occasionally. I have also seen quite a few Borussia Dortmund matches this season and have over the years watched them as well. My impression from watching these various leagues is that the SPI rankings are close but not 100% accurate. I don’t think Bundesliga is close to Premier League and the football in Spain hasn’t been that great this year. Serie A is not a great league anymore and I also think the French league is getting short thrift in these rankings. But I can accept the model and don’t really think it’s that wild. There are, in essence, 5 top leagues and then everyone else.
But I think it’s important to note that almost all of the great talent aggregating in the Premier League comes from these other leagues! Arsenal purchased Kieran Tierney from Celtic and if there were a league out there which I would call “somewhat farmerish” it’s Fitba: maybe not because they have actual farmers but rather because so much of the football is agricultural.
Which leads me to my main point in all of this: you need to judge a player by watching him play. I learned this lesson the hard way. Nicolas Pepe, for example. I saw his data points (fine toothed comb at that) and thought that his dribbles, crosses, corners, and ability to score all made him a complete player. But if I had watched him play I would have quickly picked up his painful one-footedness. And of course, I could have judged how he played, whether he needed a lot of space, and how he meshed with teammates or if he was reliant on one or more players for his data. None of that invalidates his data.
Of course experiential research can be flawed and individuals can have agendas or misperceptions (such as, for example, calling French football’s flagship league a “Farmer’s League”) but matching data with experience gives us, usually, a good measure of a player.