Wednesday 3 October 2012

Team Analysis and Form Indicators

I've taken a day long hiatus from my series of 6 week Team Analysis articles because I started to ask myself what the hell do the numbers actually mean?  Shot Accuracy up from 26% to 31% for example.  What does that indicate?  How good exactly is this?

So what I've done is I've created indicators for attacking and defensive performance then looked at last season's data for each indicator and graded each team A, B, C or D.  Allow me to explain...


This is how many chances a team create during a 90 minute game of football and is calculated by Total Shots / Total Minutes.  A: > 20%, B 17-19% C: 14-16% D: <14%
Teams who scored an A Grade last year for Creativity were Liverpool, Tottenham, Arsenal and Chelsea.
Teams scoring a D were Blackburn, Aston Villa, Stoke and Sunderland.


This is how often teams get shooting opportunities in the box.  The calculation is Shots in Box / Total Shots.  Further work would involve incorporating Penalty Box touches, Final 3rd Touches, etc.  
2011/12 A-Graders (>62%) were Arsenal, United, City and Stoke.
QPR, Wigan, Chelsea and Newcastle all got a D (<52%)


This is simply shooting accuracy, Shots on Target / Total Shots.  This does not vary as much from team to team as the other indicators but is still important.
A-grades awarded too (>33%) City, United, Arsenal and Fulham,
D's dished out to  Stoke, QPR, Swansea, Liverpool and Blackburn.


This indicator goes against my grain of only using underlying stats as it is a calculation of how many shots on target a team convert into goals - Goals / Shots on Target - but some teams do this very well, others do not.
A (>36%):  City, United, Stoke and Newcastle
D: (<28%):  Wolves, Liverpool, Wigan, Fulham

These are the exact same measurements as Attacking Form but with more defensive, and perhaps a tad daft, sounding names with the grading scale reversed.

This is the reverse of Creativity and is a measure of how many shots at team concede in 90 minutes of footy (Shots Against / Total Mins)
A (<14%):  City, Arsenal, Chelsea, Liverpool
D:(>20%):  Wolves, Bolton, Blackburn

This is the reverse of Penetration and measures how good a team are at keeping their opposition from shooting inside their 18 yard box. (Shots In Box Against / Shots Against).  It's a key metric.
A: (<52%)  City, United, Liverpool, Stoke
D: (>62%)  QPR, Bolton, Norwich

Couldn't really think of a good name for this but is the opposite to Shot Accuracy, my reasoning being that if players are put under pressure on the ball around the 18 yd box they'll find it harder to hit the target).  It's really an indication of how deep teams sit or how close their midfield are to their defence.  (Shots on Target Against / Shots Against).  Just like Accuracy, it does not vary a great deal.

A (<30%):  Newcastle, Everton, Sunderland, West Brom, Aston Villa
D (>33%):  Arsenal, Tottenham, Norwich, Wolves

The opposite of Conversion, this measure how often a goal is scored when their opposition hit the target with a shot.   Thinking outside the stats for a moment, this is a similar metric to Pressure in that if a team are gifting the opposition lots of space, letting them in behind, or are caught out on the counter, then shots that are hitting the target are more likely to catch the keeper out of position or in a one on one situation.
(Goals Against / Shots on Target Against)
A: (<28%)  United, City, Tottenham, Sunderland
D: (>36%)  Arsenal, Blackburn, Bolton

I think these indicators provide a bit more relevance to the numbers than simply saying Newcastle's shooting accuracy has improved from X% to Y% and puts a teams performance into context.  

I hope you’ll agree that subjectively the teams getting an A or D grade last seasons in an indicator  are the teams that had or did not have the particular strength.  Arsenal’s defence is a great example – they have a lot of the ball so don’t give away many chances (A for Frugality) but when they do they are often caught out with a high conversion rate against (D for Resilience).  The teams with  high Pressure grading are those that are well organised and like to defend on the edge of their box – likes of Newcastle, Everton and Sunderland.

The indicators might also prove useful to match teams against each other and help predict the result.  If Stoke for example have a high Penetration I might expect them to get more goal scoring chances against QPR who have a low Solidity than against a team like Fulham who concede a similar number of chances to QPR overall but have an decent Solidity, conceding less of chances in the box. 

With these indicators I can go one step further than simply predicting goals for or against, or shots for or against, and put together Team Profiles which I hope will provide indicators of teams relative strengths and weaknesses.

The good news from a statistical perspective is that these indicators individually correlate quite well with actual goals scored (~0.70 on average with Creativity/Frugality being the best at 0.77) but combined give a near perfect correlation (0.98).

I'll now get back to some Team Analysis (Newcastle and Swansea coming next) with these new indicators and grading included, but due to time spent on all this  I'll leave the rest of the Teams 'til next week and turn my attention towards GW7.

1 comment:

  1. Wow! This can only mean more good news for us readers! Looking forward to more of your in-depth and ever-engaging analysis.