The league table never lies
At the end of the season, when every team has played every other team home and away, that may be true. At all other times, the league table lies. Why? Because every team’s fixture list is different. Maybe team 1 has played more matches against the top six than team 2? Maybe more of team 1′s games were at home? There are too many variables for the league table to mean anything until the end of the season. This is why it’s pointless to think about how the season will play out until Easter at the earliest, and any talk of potential champions, top four finishes or relegation scraps should wait until then. Everything prior to that is just football.
So what would make the league table more accurate?
Well, one of two things. Firstly, you could compare teams based on opponents in common. For example, Manchester United and Manchester City have both played at home to Arsenal, Sunderland, QPR and Tottenham, and away at Chelsea, Liverpool and Newcastle so far this season. United have picked up 18 points from those 7 games; City only 15. This works, but once you try to extend it to three or more teams, the number of common opponents falls so quickly that it becomes useless.
Secondly, you could predict how each team is doing based on how they did against the same opponents last season, home and away, and see if they’re doing better or worse, and what that would mean if they continue the trend. So let’s do that:
What am I looking at?
Last Season Pts is the number of points won by each team in 2011/12. This Season Pts is the points total so far for 2012/13, while Pld is the number of matches played. Points Change shows how many more (or fewer) points each team has, compared to the exact same fixtures last season. For example, Tottenham are 1 point worse off than last season, in terms of like-for-like fixtures played so far in 2012/13. Note that this is not simply comparing their points tally after 17 games, but comparing their results in the actual 17 games they have played, home or away against particular opponents.
So, what about Projected Pts?
If a team is up by a half a dozen points on like-for-like fixtures at the halfway point of the season, then it’s a reasonable assumption that they genuinely are a better team than last season (or some of the other teams are a little bit worse, which is the same thing, relatively speaking). So I project what their final points tally would be by taking their remaining fixtures, looking at their results in those fixtures last season, and assuming that they’ll gain a similar number of points in the second half of the season. The maths is slightly more complicated than that, but that’s the concept.
What about the newly-promoted sides?
Obviously, Reading, Southampton and West Ham didn’t play in the Premier League last season. To allow some sort of comparison to be made, I compare the promoted sides with last season’s relegated teams. This means Reading, Southampton and West Ham are matched with the records of Bolton, Blackburn and Wolves from last season. It isn’t perfect, but it’s the only real option. Promoted sides should be looking for positive Points Change numbers, as that suggests they are doing better than the side they replaced in the league. This is bad news for Reading, as they are eight points behind Bolton’s record in like-for-like fixtures.
So what does it tell me that the normal league table doesn’t?
Chelsea are five points better off than then they were in like-for-like fixtures last season. However, that was ten points before Benitez took over, four Premier League games ago. It certainly puts Chelsea fans’ support for Di Matteo in context, as Chelsea’s league performance was much improved prior to his sacking. Chelsea would have comfortably finished 3rd last season with those additional 10 points, and perhaps Di Matteo may have gained even more, given the chance.
Newcastle’s fall from grace this season is highlighted by them being the biggest losers – 13 points down on last season so far on a like-for-like basis. Swansea have pushed on from last season, already six points better off under Laudrup than last season under Rodgers. Stoke and Aston Villa have also improved markedly.
Arsenal fans should be concerned; they may appear to be only two points off third place in the “normal” league table, but on a like-for-like basis they are already 10 points worse off than last season. The concern is, last season they scored 37 points from these 17 matches, and only 33 points from the 21 matches they have yet to play, suggesting that their remaining fixtures are tougher. Even if they match last season’s results from here on in, they’ll finish on 60 points (only good enough for 7th last season), and after losing 10 points already, who’s to say they won’t lose more? That’s what the like-for-like table suggests – unless Arsenal can improve dramatically, they could be on for a bottom half finish.
Site Twitter: @footballistix | Author Twitter: @footballistnick
Chalkboards provided by the brilliant Stats Zone iPhone app.