With the season over, transfer targets drawn up and negotiations under way, I thought I’d set myself a brief to build a team from players in the EPL that could reasonably be expected to challenge for a place in the top 6, using stats from my player database (TPOEM).
No budget constraints? That’s easy! Bale, Van Persie, Hazard… Yes, that’s miles too easy, so I’ve set some constraints as follows:
- Cannot select any player outside of the EPL
- Cannot select any player from the teams in the top 7 from last season (Man Utd, Man City, Chelsea, Arsenal, Spurs, Everton, Liverpool)
- Cannot select more than 1 player from the same team
- Cannot select any player who played fewer than 1500mins
That makes it more interesting. And within the realms of possibility if money suddenly became no object for a team currently outside the top 7. The objective of the team that follows would be to try and oust Everton and Liverpool from the top 6 next season and challenge for 4th/5th.
Goalkeeper – Artur Boruc (Southampton)
I don’t have a great depth of up-to-date stats for goalkeepers – and although I may do some work at a later date to infer their stats from opposition player data, I’ll go with the simple ratings I currently have which placed Boruc just ahead of Krul, Begovic and Mignolet. There is very little really to tell these players apart – and the back line became a bit of a juggling game – my selection effectively hinged on Boruc and Huth vs Begovic and Yoshida towards the end.
Full backs – Danny Rose (Sunderland) and Billy Jones (WBA)
Rangel, Santon and Lowton were just ahead in the pecking order but were dominated by superior players from their respective teams in other positions. The clubs beyond the top 7 aren’t really characterised by strong attacking full backs in general, which tend to catch the eye of TPOEM, but with Rose we have a strong all-rounder who performed well despite his team’s position in the league.
Centre backs – Chico (Swansea) and Robert Huth (Stoke)
José Manuel Flores Moreno (no wonder they call him Chico) was an easy choice as central defender despite a long list of Swansea players contesting for almost every position. His stats, as I have alluded to in recent posts, really were very good last season. Huth, a bit like Jones in the previous selection, is a bit of a ‘plug’ that I made with little else to choose from. Players that didn’t make the cut included Hangeland, McAuley and Williamson (that’s right – the Newcastle player, whose stats were pretty decent when he played).
Defensive Midfield – Mark Noble (West Ham United)
Noble beat Ki, Mulumbu and Delph to take a key place in midfield. In my previous post I published that he was rated 5th best DM in the league last season so it was another straightforward choice.
Midfield – Robert Snodgrass (Norwich) and Yohan Cabaye (Newcastle)
2 more players that were top of the class of midfielders and therefore beat their teammates for selection. De Guzman, Schneiderlin and Sidwell just missed the cut. Both Snodgrass and Kebab -sorry Cabaye- were in the top 10 of my season end player awards due to performances of a very high standard, again despite their team’s inconsistent results.
Attacking Midfield – Adel Taarabt (QPR)
I know that Taarabt divides opinion, having spent 2 seasons in the top flight without reaching the requisite levels of consistency to be mentioned alongside the best attacking midfielders in the league. But he is rated highly by TPOEM and wins his place ahead of Maloney. The reason he is liked by TPOEM could also be a reason to critique the results of the model, because it places a bit more importance on action frequency than the net result of those actions. Still, last season he created 76 total chances (including goal assists) and around a third of his many shots were on target (admittedly he only scored 5). I still think that’s a good result despite playing in a poor QPR team.
Forwards – Christian Benteke (Aston Villa) and Dimitar Berbatov (Fulham)
These 2 pretty much pick themselves, although there was decent competition for places as Koné, Lukaku and Lambert were overlooked. And Michu was even behind them, believe it or not. Benteke and Berbatov (B&B) scored 19 and 15 goals respectively and can each be identified as the major reason that Villa and Fulham didn’t each finish a few places further down the table.
The team in full, plus subs bench:
[click the image below if it looks dodgy…]
I’ll use Everton as the benchmark, considering they actually achieved 6th place.
Player for player, it’s very close ratings-wise. I sorted the starting line-ups side by side (above – click the image if it looks sketchy) and highlighted in green the player who won each head-to-head. 6-5 to the hypothetical team, but only if I match Snodgrass vs Osman and Cabaye vs Fellaini, otherwise it’s 6-5 to Everton. From the picture, it would appear as though Everton have a clear advantage in defence, whilst the hypothetical top 6 team has a stronger attack.
The TPOEM measures back this up overall: suggesting that Everton would be stronger at winning the ball and retaining possession whilst the goalscoring potential of the ‘top 6’ team would be superior. This hypothetical team has a total score only very slightly higher than Everton – so close in fact that if they played each other at a neutral venue the expected result according to the model would be something like a 37% chance of either team winning and 26% chance of a draw; with a 72% chance of there being more than 2.5 goals in the game but a most likely score of 2-2 (12.7% probability)!
So could the team challenge for the top 6? Yes, I believe it just about could – assuming the players stayed fit all season and the team defended in numbers when appropriate! A rough valuation of the team would be somewhere between £75-100m in transfer fees.
Thoughts and comments welcome.