Ted Knutson, the prolific writer, EPL trader and key cog of the football analytics twitterati posted a very interesting article on Squawka.com which looked at the rate in which European club teams concede shots on target. I highly recommend you check it out. This is a link to the EPL Shots on target conceded table.
If you clicked the link you'll already know what I am about to tell you: Aston Villa were the third best team in the Premier league at preventing their opponents from getting their shots on target. This was Ted's take:
As you can see, Villa gave up a lot of shots per game – 16.63 of them to be precise. However, they were comparatively good at preventing those shots from turning in to shots on target.
All true. The purpose of this fanpost is to elaborate and further explore on why Aston Villa restricted opponents ability to get shots on target. to do this I am going to look at the Game States at which Aston Villa conceded shots on target, missed shots and blocked opponents shots. I am also going to take that data and expand it for all of Aston Villa's 38 games in order to discover whether Villa improved at restricting their opponents shots.
The chart below shows the percentage of opponents shots that were on target, blocked or missed. As an extra kicker I have included the game state.
- The most frequent shot outcome in the chart above is missed shots. Every team in the league, I assume, forces their opponents to miss with a big percentage of their shots.
- Villa's ability to block shots is pretty consistent throughout the game states. Some teams display an ability to block an increased percentage of opponents shots at +1 due to the leading team trying to protect their game position.
- Villa's ability to prevent the opposition from registering shots on target is pretty strong. At Tied and Plus 1 Game State, we see ~30% of oppositions shots end up on target. This is a pretty darn good number, further questions will need to be asked as to why Villa were able o put up such good numbers.
- The real talking point where the chart above is concerned is Villa's defensive performance at Minus 1 (trailing by a goal): Over 50% of oppositions shots were off target and just 20% were on target. These are really strong numbers, especially so when we consider Villa, at a goal down, were probably chasing the game to some extent. I have no concrete explanations as to why Villa's opponents, when leading by a goal, only registered ~20% of their total shots on target (league average is 30%). Whatever defensive scheme Villa were employing at Minus 1 it was either very good or very lucky.
This chart shows Villa's ability to block opponents shots as the season progressed.
- Villa's ability to block shots when losing (ie Minus 1) took a beating in mid-season when the hammerings were frequently painful. But blocks when losing improved.
- As for Villa's shot blocking ability at Tied, winning and Close (-1, Tied and +1) we can see that as the season progressed Villa became more skilled at blocking their opponents shots, in the case of Tied and Winning Villa improved their shot blocking ability by 20% from Game 10 to the seasons end.
- Close Game State is absolutely crucial. 80% of all game events take place at this bunched state. Close goal difference has an r2=0.97 to points won, so everything that happens at Close is vital. As you can see, Villa improved all season in shot blocking ability at Close.
Shots On Target Conceded
- I have listed total, Tied and Close game state information here and it shows that as the season wore on Villa were unable to prevent the opposition from recording improvement in getting shots on target. The improvement in slight, but evident, in the first half season.
- From the halfway point on we see that improvement plateau and lurk relatively consistently around 27-29%, numbers which are still excellent.
- An admission, I can tell you that these things happened but for the life of me I cannot tell you why. Was there a tactical shift by the manager, was it a personnel issue which drove these excellent Villa numbers? Maybe it was about experience for a very young side and if it is, it should bode well for the upcoming season.
Attack And Defense
I have shown you a ton of information on Villa's ability to prevent opponents shots from getting on target. Now lets take the other side of that coin, Villa's attacking performance and look at a few ratio metrics.
- TSR Total shots for / totals shots
- Fenwick Shots on target for + missed shots for /t otal shots on target + total missed shots (basically, we remove blocked shots.
- Mathias Shots on target for + Blocked Shots for / Total Shots on target + Total Blocked shots.
- SoTR Shots on target for / Total Shots On Target
*Fenwick comes all the way from Alberta, Mathias is a stat I made up last night and originates from office at home in Barcelona. Name belongs to the girlfriend, it is probably going to make SImon Gleave really mad!
All Game States
We are looking for a metric that 'fits' closely to Villa's moving goal difference that Is included on the secondary axis.
- Villa's goal difference was plodding along, poor but not terrible, and then the barren run began. thrashings at the hands of Chelsea, Tottenham and Wigan destroyed Villa's goal difference and wiped out the happy memory of the Liverpool away win. But that goal difference recovered at seasons end and improved right at the death. Credit to this young squad at that pressure point in the run-in.
- As for the metric which fits closely to Villa's goal difference, well there is quite the spread. Fenwick comes out bottom with an r2=0.22, TSR is 2nd worst, then SoTR but Mathias is best with an r2=0.76. A lot of these correlations improve significantly if we strip out the early fixtures and there is so much more work for me to do on the other Premier League teams before we can start talking of which metric fits best.
- All of the ratios/metrics improved at seasons end. A little of that improvement was down to Villa's shot blocking, Villa's attacking numbers and more time spent at Close Game State where Villa were not terrible (especially at Tied GS).
Remember how I talked of Close Game State and how crucial it is?
Close Game State
That is some improvement in the run-in for Aston Villa. As stated earlier, I can show you what happened but not exactly why it happened.
To be able to fully understand Villa's improvement late on, and especially at Close Game State, we may need to know more about the personnel deployed or the system tweaks by Lambert. In short, it needs more but I feel comfortable in saying Villa were not a bad defensive team in terms of restricting the quality of their opponents shots. And boy, did Villa go on a nice run in the last 6 or 7 games of the season in terms of their shots output.