Does the league table still lie?

Around this time last season I wrote one of my favourite pieces for Trading Expert, looking at whether the English Premier League table “lied” in terms of where it placed teams.

The idea was to compare the Actual League Table with our Analyst Fair Score Table, in addition to our proprietary Model Fair Outcome Table, to see which teams had been over-achieving or under-achieving based on their general performances and key metrics.

The long and short of it is that the table did indeed “lie”.

However, to go a little deeper the big story was Leicester City leading the way, while then-champions Chelsea were floundering down in 14th. Our Analyst Fair Score table actually had Leicester in third, close behind Manchester City and Arsenal, while Chelsea were apparently fully deserving of their 14th rank. The Model Fair Outcome Table told a similar story, with Leicester again third behind Arsenal and Manchester City, while Chelsea were even worse off in 15th.

Funnily enough by the end of the campaign Leicester still led the way in one of the most remarkable Premier League seasons ever, while Chelsea’s change of manager helped them recover a little bit to finish tenth.

Anybody thinking that Liverpool being stuck in eighth after 15 games of the 2015/16 campaign was “wrong” were shown that they actually deserved to be 12th or ninth at that time, depending on whether you believed the analysts or the model, and in the end their below-par performance also held strong as they ended up still in eighth come mid-May. Southampton’s final position of sixth in the standings was also much more in line with their “fair” ranking from December, with the analysts having them seventh and the model sixth, versus their actual position at the time of 12th.

Before we dive into 2016/17, I want to finish on West Ham. The analytics community highlighted Slaven Bilic’s men last year for their over-performance in terms of goals versus shots, which our Model Fair Outcome Table mirrored by dropping them to 13th from the seventh place they actually held. That was the biggest single difference seen in the data, although the Analyst Fair Score Table pitched them at a more reasonable ninth. In the end West Ham actually did end up seventh, again highlighting that “mean reversion” can take more than a season of games to come to pass (something I’m always keen to remind people adamant that certain teams or players “can’t keep doing this every week”.)

My belief that the truest representation of the league comes from a combination of the Actual Table, Analyst Table and Model Table still holds true and that’s again how I want to approach this piece. However, this season we have worked hard to merge the Analyst Fair Score and the Model Fair Outcome together to give the most accurate Fair Scores possible. To this end I will only be using the Actual League Table and the Analyst Fair Score Table, which from here on out will simply be known as the Fair Score Table.

So without further ado, let’s begin with the Premier League table as it stands on Wednesday 11th January 2017, after each team has played 20 matches.

The Actual League Table

Figure1_ActualTable-01.png

This season the top end is definitely lacking a “Leicester”, as the traditional top six has been fully restored. There won’t have been many predicting that any other teams would have been infiltrating the title race come the halfway point of 2016/17, and so it has proven. Everton in seventh is no great surprise, though their performances have been far from convincing despite the change of manager from Martinez to Koeman.

West Brom are the “average team” in the league this time around, which is translating to a very healthy position of eighth under Tony Pulis. He looked set to come under significant pressure this year, with worrying off the field stories compounding his general unpopularity with WBA supporters due to his desperately negative style. However, some shrewd transfer business in the summer (the signing of Matt Phillips especially, who always seems to produce) has allowed him to open up a little bit and while his team keep less clean sheets they also carry more threat.

Indeed, West Brom being where they are is perhaps the first mini surprise, while Bournemouth backing them up ahead of south coast rivals Southampton also raises a few eyebrows. I have to admit that I didn’t expect to find Burnley anywhere near 12th after 20 games (and still believe they will tumble), while West Ham, Leicester, Crystal Palace and Swansea are surely this campaign’s biggest underachievers to date with rankings of 13th, 15th, 17th and 19th respectively.

West Ham have been returning to more “normal” levels recently as they have recovered injured players, while Leicester’s general performances and metrics are on the up (although this is not yet being reflected in their rolling points haul). Palace and Swansea have moved to change managers in order to arrest their slide, Palace dispensing of the thoroughly detestable Alan Pardew far too late (did you see his record in 2016?) and Swansea making the bold play of hiring Bob Bradley and then getting cold feet very swiftly when an instant upturn was not forthcoming.

The general under-performance of this quartet has allowed Burnley to climb so high, while Watford’s position of 14th also looks “false” to me based on what I’ve seen of them this season. In addition, Middlesbrough have all the hallmarks of a team who had a little boost in the early part of the season after promotion but should probably be more involved in the relegation fight than they are.

Still, my opinions aside, just how does the Actual League Table match up with the Fair Score Table? Who has been benefitting from that mystical element called “luck” more than most and who is in dire need of a rabbit’s foot? Let’s find out…

The Fair Score Table

Figure2_FairScoreTable-01-1.png

Unsurprisingly the top six holds steady, though Liverpool have gained an enormous three wins and seven points to displace Chelsea at the summit. Antonio Conte’s team have performed exactly as expected, which is possibly no surprise given the overall consistency of their professionalism and metrics following the switch to 3-4-3 (or 3-4-2-1, depending on how fastidious you are).

Manchester United are the biggest overall winners when the “fair” view is taken into account, gaining four wins and eight points to jump from sixth to third and being right in the race for the title. I dare say Jose Mourinho would enjoy this piece, and perhaps this particular revelation makes his continuing grumpiness a little easier to understand. He would be especially heartened to see his old rival Arsene Wenger being punished to the tune of losing two wins, five points and one place too, I’m sure.

Finally, to round off the top six we see Tottenham Hotspur drop a couple of places and points into fifth, while Man City gain a win and three points but hold firm in fourth. On the whole we see the range between the six teams expanding from ten points to fifteen thanks to Liverpool’s rise and Arsenal’s fall, which looks a bit too wide at this point in time.

However, the bigger issue I can see is that the dominance of the top six has been watered down somewhat, which does not really reflect what the season is increasingly telling us. As usual the big teams were the winners over the hectic festive period, asserting their dominance and pulling away further while nobody below sixth managed to take more than two wins from the four games played over Christmas and New Year.

Figure3_DifferenceTable-01-2.png

Winner

In terms of trading the outright it seems that Chelsea are credible heavy favourites (1.89 at Betfair), because they are consistently getting what they deserve with their wins, while Liverpool (6.4) are displaying much more volatility and generally failing to earn the points that their performances and metrics deserve.

Arsenal (16.00) probably shouldn’t be considered (though is that really news to anybody?), while Man Utd (17.00) have given themselves too much to do but could be a firm shout for top four if they can improve their efficiency (which seems to be happening). Man City (8.40) are a disappointing fourth and at least six points off the top whichever way you slice it, which sounds about right given their performances, though of anybody they surely have the most room for improvement. Finally, Tottenham (12.50) have managed to get into third with a small amount of fortune on their side and with performances improving and star players coming into view again they look a serious shout for the Champions League spots, if not more.

Top Four

My personal view at this stage is that we’ll see Chelsea (1.05), Liverpool (1.37), Tottenham (1.71) and Man Utd (2.04) in the top four slots come the end (most likely in that order), with Man City (1.41) and Arsenal (1.91) missing out. Feel free to remind me of this if it all goes horribly wrong, of course, though I will throw in an injuries/transfers pending disclaimer here for good measure…

Top Ten

Moving onto the middle of the pack and the battle for top ten positions, it seems that Everton (1.08) are unsurprisingly the favourites to hold firm there in seventh. Our Fair Score Table has Southampton (1.36) in eighth, which I suppose is where most people would rate them if asked to rank all Premier League teams from 1-20, and despite the on-going troubles with Fonte and the chance of Van Dijk leaving to compound things they really should make that position their own. West Brom (1.69) drop one position to ninth because of Southampton’s gain of five points, but like Chelsea have been consistently getting what they deserve (except for the fact that they should have been involved in more “unders” games, which would surely please old Tony) and look a good shout to maintain this through until May.

Tenth position itself is increasingly tricky to gauge, with Bournemouth (2.00) surprise contenders, Stoke (2.92) always in the mix and potential comebacks from West Ham (3.75), Leicester (2.96) and maybe even Crystal Palace (18) to consider. The Fair Score Table does West Ham’s chances no favours by relegating them an enormous five places to 18th, docking them of four wins and eight points, but Stoke gain a fraction to climb to tenth at the expense of Bournemouth, who have apparently taken two more wins and three more points than they should have so far in 2016. Leicester have been the great enigma this season (I’m still not buying the mean reversion excuse alone, as it’s far too simplistic) and though they lose three points in the Fair Score Table they actually gain a place to land in 14th, mainly due to the apparent over-performance of others.

My personal view here is that Everton and Southampton should be certainties, unless Southampton lose both centre-backs and fail to acquire even semi-adequate replacements, while West Brom show no signs of slipping and should hold on to ninth or tenth. After that it probably comes down to Stoke, who have been consistently average enough despite dealing with a number of key absences at different times. I can’t see a team who start Simon Francis every week (as captain, no less) maintaining a top ten place, while I won’t even count Burnley, Watford or Middlesbrough in this despite the fact Watford and Middlesbrough both come out higher in the Fair Score table.

Middlesbrough gain four places despite taking just two extra points, but should actually have a worse goal difference by virtue of deserving to score three less than their paltry actual total of 17. This is the main reason I want to consider Middlesbrough as part of the clutch of teams who will be fighting relegation, along with Watford, Burnley, Sunderland, Swansea and Hull.

Relegation

Jumping straight to Burnley, I mentioned earlier that I would never have expected to see them in 12th after 20 games and I’m not surprised to see them dropping five places in the Fair Score Table, by virtue of giving away three wins and six points (a reduction second only to West Ham). They have taken exactly what they have deserved when away (a solitary point) but are clearly over-performing at home. If their home points tally dries up even slightly then they will tumble into trouble quickly. Watford, like Middlesbrough, actually appear higher in the Fair Score Table despite losing three points, while Sunderland are also punished by the same amount but fall to 20th, behind Hull who only lose a solitary point and climb to 19th.

Swansea should apparently be three points and three places better off, which may well have kept Bradley in a job had it come to pass, though this does point to the fact that Paul Clement should have plenty to work with as he bids to do far, far better than his predecessor. Ultimately there is little hope to be found for Hull (1.20) or Sunderland (1.54), who both come out worse off in terms of points and still look to be heavy favourites for the drop in 19th and 20th, though like with the last spot available in the top ten it’s an arduous task to pinpoint who joins them in 18th.

West Ham (16.50) occupy that position in the Fair Score Table but I cannot believe Bilic’s men will be anywhere near come May, which leaves the other three teams I mentioned earlier. Swansea (1.59) should have the quality to climb, though sensible January recruitment looks important after their shameful summer in the market, leaving Burnley (4.50), Middlesbrough (6.60) and Watford (7.20) in the firing line.

Watford are clearly where the value is, but despite my exasperation with their points haul they probably have enough to steer clear, putting it down to Burnley’s home form versus Middlesbrough’s away form. As this is a Fair Score blog, then Middlesbrough’s performance here must swing it so I’ll add Burnley as the third team to go down, although as with most seasons it’s still a nightmare to call at this stage! Some smart January business could be the difference between life and death down here, though it seems the die have been cast on the managerial front.

Conclusion

Finally, let me sign off with a table to show the average of the Actual League Table and the Fair Score Table, which is probably where the truth lies:

Figure4_AverageTable-01.png

This supports my feelings that we’ll end up with Chelsea as Champions and Man Utd in the top four, while the entire top ten is in line with my view as a whole. The main differences are that Man City have usurped Tottenham in the Champions League spots, while Swansea have joined Sunderland and Hull in being cast into the abyss instead of Burnley. However, armed with the knowledge that Swansea and Crystal Palace have changed manager, plus the belief that West Ham and Leicester will improve enough in the second part of the season to move comfortably into mid-table, Burnley are handily pulled right back into the mire…

Many thanks for reading, and while you may not agree with my outlook I at least hope the use of our Fair Score Table has given you some indicators as to where the market might be overrating or underrating teams based on actual historical performance instead of fair historical performance, both in terms of Asian Handicaps and Total Goals Lines. As always when looking at realised data, not everything is as it seems.

Richard Huggan (@AnalysisRich)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s