Leicester, Burnley and QPR are rejoining the Premier League after shorter or longer breaks from the biggest stage. But what can we expect from these teams? Can we say anything about their chances of success, based on how they performed in the Championship or what they have done in the transfer window?
Does a high points haul in the Championship foretell a successful season in the Premier League? Plotting each teams’ points from its promotion season in the Championship against its Premier League points in the following season shows there is no clear trend in the dataset.
The vast majority of teams land within a belt around 40 PL points (often referred to as the magic number which guarantees you another season in the big league) give or take 10 points, regardless of how they did in the Championship. In other words, there is no correlation between Championship points gained and PL points in the following season. The three best Premier League seasons by a promoted side in the period we are looking at have come from West Ham (05/06), Ipswich (00/01) and Reading (06/07). They gained 73, 91 and 106 points respectively in their promotion season ranging from one the worst to the very best performance in the Championship. Looking at the plot there may be an indication that teams who gain exceptional points returns (95+) in the Championship also perform well when they get promoted. That may be true, and would bode well for Leicester (102 points last season), but the truth is we do not have enough data points to say anything meaningful about that subgroup.
Many people claim that teams who score many goals in the Championship will also score relatively more goals in the Premier League. Using simple linear regression I found a significant (95% confidence) positive correlation between these two factors which means there is some credibility to the claim. The R² for the trend line is only 0.15 though, meaning goals scored in the Championship can only account for about 15% of the variance in goals scored in the Premier League. The coefficient suggests that for every ten goals more a team scored in the Championship it is likely to score roughly three more goals in the Premier League the following season. In other words, we would expect Leicester (83 goals last season) to score approximately seven goals more than QPR (60 goals) this season – hardly a big difference. For a fantasy manager it is certainly not enough to choose a midfielder or attacker from Leicester over one from QPR who has better individual stats. Likewise there is a statistically significant connection between the amount of goals scored in the Championship and the amount of points amassed in the Premier League the following season. But here the explanatory power is even lower (12%).
I am sure we will hear it again at the beginning of this season, in the pre-match build-up to Burnley’s first game, or in Match of the Day in the evening. It is a widely held belief among TV pundits that the promoted side who has shown the most solid defensive base in the Championship will fare well in the Premier League. The dataset does not support this claim. There is not even a remotely significant correlation between the amount of goals conceded in the Championship and either the points haul or the amount of goals conceded in the Premier League. There could be two reasons for that. Perhaps the jump in quality of opposition is just so large that it doesn’t matter what you did at the lower level. The other explanation could be that the amount of goals conceded does not necessarily reflect your defensive strength. As I have previously suggested (http://englishfootballpatient.wordpress.com/2013/01/14/what-happened-to-newcastle/) the amount of big chances a team allows is a better indicator of its inherent ability than how many of those chances happened to be converted into goals. As a fantasy manager you should look at the price and quality of defenders from the promoted sides, their propensity to get goals or assists, their team’s opening run of fixtures, but not their team’s defensive record from the Championship. There is no reason to believe we can learn anything from that about how they will do defensively in the Premier League.
So the Championship record is not a great indicator of what to expect from the promoted sides. Those that score more goals in the Championship are more likely to perform well in the Premier League, but their goalscoring record can only explain 15% of the difference in performance between teams. That’s because it’s all about the money, right? The real difference maker is not how the Championship side performed, but what players are brought in to the team in the transfer window. To examine this claim I collected data about the net investments for each team from the brilliant website Transfer League. What I found was bad news for the teams looking to maintain their Premier League spot through the transfer market.
There is no clear indication that larger net investments result in more points for the promoted sides. In fact the only team that made a lot of money from transfers in its first season back, Newcastle (10/11), is among the better performing teams in the Premier League. Admittedly, the bulk of that money came in halfway through the season when they managed to secure an incredible £35m for Duncan-Ferguson-in-disguise. But the big-spending teams do not stand out as high-performers. Cardiff (13/14) was the second-highest spending team in the dataset, including a whopping fee for my non-performing compatriot Andreas Cornelius. They won the Championship with 87 points and made big investments in the squad, but still ended dead-last with only 30 points. Perhaps teams don’t spend their money wisely in the market, but maybe there is also a word of warning for fantasy managers that it takes time for new arrivals to adapt to the rigours of the Premier League.
Does that mean that money doesn’t matter then? The answer appears to be a resounding “no”. As previously shown by Kuper & Szymanski (Soccernomics) and Anderson & Sally (The Numbers Game) there is a clear correlation between total salary paid and points won in the Premier League. I collected data on the salary paid per team per season in the Premier League (courtesy of The Telegraph) and divided the salary for each team with the average salary in the league that season. What I found was a statistically significant relationship.
The market for player salaries appears to be more efficient than the market for player transfers. The more you pay your players the more points you get. There is nothing particularly ground-breaking about that finding, but it is interesting that the salary explains 22% of the variance in Premier League points. The amount of money a promoted team spends on salaries is a better indicator of its likelihood of success than how many points it got, goals it scored and goals it conceded last season and how much money it invested in transfers after winning promotion. The only problem is that salary figures are usually released retrospectively so we are unable to use this knowledge for the upcoming fantasy season. A cautious guess would be that QPR have the biggest salary budget of the three promoted sides, Leicester the second biggest and Burnley the smallest. Who will do well? It is hard to say, but my money would be on QPR and Leicester for the reasons highlighted above. Fantasy managers should be wary of the latter’s tough start to the season though (their first five games include Everton, Chelsea, Arsenal and Man Utd). Janus Nielsen writes statistical analysis about the Barclay’s Premier League on his blog the English Football Patient. He will be contributing occasionally here at Fantasybet, and can be found at Twitter here.