JPL Shooting Stats & More after Gameweek 5

As we head into the 6th weekend of the Jupiler Pro League, I take a look into some basic statistics and infer how well the teams have been playing.

Shot Matrices

Dominance

Firstly, I look at the JPL team's "shot-dominance" – all credit due to Ben @experimental361 for this concept – by comparing the shots a team takes and faces, per game. 

• Standard has overall "shot supremacy" in the JPL so far: far outshooting their opponents and facing the lowest number of shots per game.   
• OHL’s defence has been the busiest, Standard’s the least busy. 
• Genk & Zulte both outshot by their opponents. Genk has the more over-worked defence of the 2, but Zulte is hovering close to the overworked territory. Both teams have top-6 aspirations – both need to limit their opponent’s more, it would appear, in their bid to do so.

Efficiency

"Shooting efficiency" (again, hat-tip Ben) is defined by the number of shots taken to score one goal. The rule-of-thumb being: 'the fewer the number of shots a team must take to score 1 goal, the higher their level of shooting efficiency, or, the better they are in front-of goal'.

• Ostende – the league leaders – have been clinical in front of goal needing roughly 6 shots to score 1 goal, which when they take over 13 shots per game, on average, implies a high return. 
• But it is Waasland Beveren, the second-highest goal scorers in the JPL, that are the most efficient at turning their shots into goals, despite taking a below-average number of shots per game.
• At the other end of the spectrum, Gent and Charleroi are so far struggling to convert theirshots into goals as well as teams further up the table are.

Defence

To assess the effectiveness of a team’s defence, we approximate by comparing the number of shots a team faces per game to the number of shots it has faced per goal conceded. Good defences limit the number of shots taken by the opposition as well as limit the number of goals they concede.

• What immediately catches your attention in the chart above is the extreme outlier: Gent. The Buffaloes have by far the most effective defence at dealing with shots in the JPL. They limit their opponents to about 10 shots per game (only Lokeren and Standard face fewer), and have conceded a goal only every 25 shots from their opponents’. 
• Standard face the lowest number of shots per game but have the third leakiest defence in the JPL. 
• Anderlecht’s defence also strays into ‘leaky’ territory despite facing the fourth lowest number of shots per game in the JPL. 
• At the opposite end of the x-axis OHL’s defensive frailties are fully exposed by this chart: they are overworked and leak goals.

Possession Game

I compare a team’s average possession statistic to their shots on target ratio (SoTR), which measures one team’s share of the sum of their & their opponent’s shots on target.

• The chart generally shows that the teams higher up the league have higher possession statistics as well as higher SoTR values. The exceptions are STVV and Zulte, which both have SoTRs than 0.5, but tend to have less possession of the ball in their games.
• Standard has the highest SoTR in the League at 0.66, translating to: 66% of the total shots on target in their games, eg if there are 15 shots on target in a game, Standard had 10, their opponent had 5. 

What's Luck Got To Do (got to do) With it?

Who need's luck when skill has spoken? Asides from just typing the worst pun on Tina Turner's classic in history, I try to assess the 'luck' JPL teams have had thus far. To do so, I turn to the PDO (credit due to James Grayson) vs. SoTR, or the 'good/lucky', matrix. PDO is the sum of shots percentage – which is the percentage of shots on targets that are converted to goals – and the save percentage – which is the percentage of an opponents’ shots on targets that are not converted to goals. PDO measures luck, averaging to 1000.

• STVV and Ostende are riding high on luck so far, with PDO values in excess of 1100, which is down to both teams scoring a high number of goals from a relatively low number of shots (than compared to, say, Anderlecht, Club Brugge & Standard, and to a lesser extent, Gent & Charleroi).
• Gent’s above average PDO is more down to their high save percentage & the stellar performances of Matz Sels. 
• Standard look to be due for a change in their fortunes some point soon. Whether or not that will come Sunday afternoon when they visit Club Brugge in what looks to be the most exciting fixture of the JPL so far this season, will be fascinating to see. 

Thanks for reading, I appreciate any feedback on how to improve. Cheers, Charlie.

Lowzer's Low Country Revue: Who is the Belgian JPL's form team?

My last couple of posts have primarily focussed on shot-level data to try and tell a story about how the team's in the Belgian Jupiler Pro League (JPL) have performed. Another way of assessing team's performances is to compare bookmaker's predictions to actual performance. Using this approach, I look here at which team is Belgium's most "in-form" team, and which team is performing above expectations...

Simon Gleave recently wrote a post investigating whether or not Liverpool is the most "in-form" team in the English Premier League. In his post he linked to an explanation of a simple methodology to use to compare expected points (from bookmakers' odds) to actual points (won by teams). I use his model here with Belgian JPL data from soccerway.com and bookmakers' odds from odds portal.

To begin, let's look at the form table of the JPL over the last 10 matches - which is the last 10 full gameweeks at the time of writing, so gameweek 18 to gameweek 27.

Belgian JPL form last 10 matches GW 18-27 (@SatThereThinkin)
Team	Actual Points
Standard Liège	23
Club Brugge	20
Anderlecht	20
Kortrijk	19
Gent	19
Charleroi	18
Genk	15
Lokeren	13
Mechelen	12
Waregem	12
Cercle Brugge	11
Oostende	10
Westerlo	9
Waasland Beveren	8
Lierse	8
Mouscron-Péruwelz	3

As you can see from the table, Standard Liège is the most "in-form" team in the Belgian JPL having taken 23 points from a possible 30 over the past 10 matches. Standard's form over gameweeks 18-27 is a whole 3 points better than their rivals Club Brugge and Anderlecht. It is perhaps unsurprising to most of you that Standard, Club Brugge and Anderlecht are the top of the form table in the Belgian JPL as they are, at least in recent years, the top 3 clubs.

Looking at the form-table another way, however, tells us another story. What I want to do is "control" for the strength of the opposition the teams face over the past 10 matches. What I mean by this is, let's say Standard played 10 minnows over the past 10 matches (ie avoided Club Brugge & Anderlecht), their 23 points from a possible 30 seems less impressive [for the record, Standard played Anderlecht in Gameweek 23 and won 2-0].

So, to control for strength of the opposition over the past 10 matches, I use bookmakers' odds as a proxy. The methodology is explained at the end but for now, I used bookmakers' odds to provide an "expected points" table. I then compared the expected points over the last 10 matches to the actual (or, observed) points over the last 10 matches to see who is "outperforming" bookmakers' expectations, which is roughly to say 'overperforming' (or, the teams in green) :

Belgian Jupiler Pro League GW 18-27 – expected vs actual points last 10 matches (@SatThereThinkin)
Team	Actual Points	Expected Points	Difference
Genk	15	8.24	6.76
Standard Liège	23	16.48	6.52
Charleroi	18	14.03	3.97
Kortrijk	19	15.07	3.93
Gent	19	17.06	1.94
Cercle Brugge	11	9.43	1.57
Club Brugge	20	19.99	0.01
Anderlecht	20	20.33	-0.33
Oostende	10	10.86	-0.86
Westerlo	9	9.99	-0.99
Mechelen	12	13.52	-1.52
Waasland Beveren	8	9.73	-1.73
Lierse	8	10.15	-2.15
Waregem	12	14.20	-2.20
Lokeren	13	15.24	-2.24
Mouscron-Péruwelz	3	9.68	-6.68

The table shows Genk - managed by Alex McLeish - as the team performing most above expectations. Genk is currently battling to remain in the top-6 competing closely with 3rd most overperforming team Charleroi. This will be an interesting battle to watch. Interestingly, the 2nd most overperforming team is Standard, adding weight to the initial claim they are performing better than expected. Standard seemed to be underperforming for quite a while and were towards the lower end of the top-6 for much of the season. They have, however, picked-up some form recently and look to be in hot pursuit of top-2; Anderlecht and Club Brugge. Anderlecht have underperformed over the past 10 games, but not by much. Club Brugge have pretty much played to expectation.

I will try to return to this as much as possible to try and tell a bigger story. For now, it seems Standard are deserving of their ascendance in the table. It also feels like Genk - on recent form - deserve to propel themselves into the top 6.

--------------------------------------------

Methodology explained:

1) Taking the example of the Genk Anderlecht game in GW 27 (which ended 1-0 Anderlecht), I took the odds of 3.3 home win, 3.31 draw and 2.18 away win. I put these all as the denominator under 1 to get a %age chance of the odds.

2) The odds then became 30% Genk win, 30% draw and 46% Anderlecht win - the 6% being the bookmakers' profit. So, I had to divide all these odds by 1.06 (or, 106%) to ensure the odds added up to 1.

3) In so doing the odds became 28.5% Genk win, 28.4% draw and 43.1% Anderlecht win.

4) For expected points: Genk (3*0.285) + 0.284 = 1.1 points; and, Anderlecht (3*0.431) + 0.284 = 1.6 points, ie Anderlecht favourites to win.