Twitter

Friday 13 March 2015

Macro look at combined Attack/Defense in the Premier League

As we come down the stretch, I wanted to take a quick look at the shots and goals numbers for & against. Lots of times we only see the Shot Ratios (i.e. TSR, SoTR, GF%) posted but they do not give us the actual quantity of shots each team has taken and given up.

I know scatter plots can be tough to read/decipher but, for me, this is the best way to show where teams stand.

*This is the 3rd post of 3. The 1st post covered Attack and the 2nd looked at the Defense


Attack & Defense Combined

Shots on Target Differential vs Total Shots Differential


We can see that their is a clear top 5 teams with Man United lagging behind. The race for the 3rd and 4th Champions League spots looks like they could come down to the last game of the season.  Extremely tight.

Meanwhile, Sunderland are just abysmal. How they are not sitting at the bottom table is fascinating.

Everton, stand out as outliers, along with Chelsea, in over performance of shots on target. Chelsea are thriving off this, while Everton haven't been able to capitalize.



Goal Differential vs Shots on Target Differential


*If you are not familiar with PDO please check out the Glossary along the top of the page.


Analytics tell us to expect Man United's PDO to regress toward the mean (100). Many, mistakenly, think that if a team carries a very high PDO for a lengthy period of time that it will, in turn, have a game or multiple games where the team has a terribly low PDO. This is gamblers fallacy which many confuse with regression to the mean. 38 games is a small sample and good/bad fortune can last over the course of the season (or even more), even if it is unlikely. For example, Man City finished last season with ~112.0 PDO.

**Just for clarification, I am not saying Man United's PDO will stay high (It could fall off a cliff, for all I know). I am just trying to clarify that over the last 10 games (for all teams) anything can happen. The previous 28 games do not determine the PDO of game 29. Much like flipping a coin and getting heads 10 times in a row doesn't mean the odds of getting tails is higher on the 11th flip**


That is the end of this short 'series' of blogs. Hopefully, you found it somewhat informative and/or made you think about things a little different.

Cheers,

Clarke


 

Wednesday 11 March 2015

Macro look at the 'Defenses' in the Premier League

As we come down the stretch, I wanted to take a quick look at the shots and goals numbers for & against. Lots of times we only see the Shot Ratios (i.e. TSR, SoTR, GF%) posted but they do not give us the actual quantity of shots each team has taken and given up.

I know scatter plots can be tough to read/decipher but, for me, this is the best way to show where teams stand.

*This is the 2nd post of 3 (time permitting). The 1st post covered Attack, this one will focus on Defense and the last one will combine Attack/Defense by looking at Shot/Goal Difference.

Defense

Shots on Target Against vs Total Shots Against


Southampton! Least amount of 'shots against' and 'shots on target against'. The defensive side of the ball is why they have such high TSR and SoTR's. What they have done, defensively, is both sustainable and amazing. Soton's 'money spent vs. performance' compared to the two oil-financed clubs is very 'Oakland Athletics' like.

Manchester United and Spurs are giving up far to many shots on target. United do a good job of suppressing shots but when they do get through they seem to be hitting the target (and as we will see below, hitting De Gea). Spurs on the other hand have been quite easy to play against giving up a lot of shots coupled with a high rate on target.  

In the attacking piece I didn't mention Everton because their was nothing 'out of the ordinary' with their numbers. Now, defensively, they are having an interesting season. Martinez has been heavily scrutinized for how he sets up his team and that is fair given that they sit 12th in total shots against.
They have a lot fewer 'shots on target against' than would be expected and normally this would be a good thing, even if it is a bit fortunate, but this has not led to success. We look below as to why?


Goals Against vs Shots on Target Against



Goalkeepers are voodoo. Ask Everton. Pretty much, all of Everton's under performance in the league is down to their inability to stop the ball.  They sit dead last in save % (table below), while last season they were 2nd. Everton have been both, slightly lucky (6th lowest shots on target) and really unlucky (league low save %) this season. Both these metrics should regress but it is too late for it to matter, this season. 

On the opposite end of the voodoo spectrum, we have Man United, who are getting away with it, this season. 8th in shots on target against and De Gea bailing them out. This United team are riding the variance of their goalkeeper. Will it last? Crazier things have happened.

Another notable outlier is Sunderland. Pantillimon, since coming in to the side, has been the saviour. His performance is the reason they are not buried in the relegation places.


Here is the save% of each team (sortable).
Team Save %
Arsenal 70.0
Aston Villa 69.1
Burnley 66.9
Chelsea 72.8
Crystal Palace 64.2
Everton 59.0
Hull 68.1
Leicester 65.4
Liverpool 68.4
Man City 69.0
Man United 75.9
Newcastle 61.6
QPR 69.2
Southampton 73.7
Stoke 64.6
Sunderland 73.8
Swansea 72.6
Tottenham 70.5
West Brom 69.5
West Ham 74.2
Average 69.0
* compiled using footstats.co.uk

Clarke 


 

Tuesday 10 March 2015

Macro look at the 'Attacks' in the Premier League

As we come down the stretch, I wanted to take a quick look at the shots and goals numbers for & against. Lots of times we only see the Shot Ratios (i.e. TSR, SoTR, GF%) posted but they do not give us the actual quantity of shots each team has taken and given up.

I know scatter plots can be tough to read/decipher but, for me, this is the best way to show where teams stand.

*This is the 1st post of 3 (time permitting). This one will focus on Attack. The next will be Defense, followed by combining Attack/Defense and looking at Shot/Goal Difference.


Attack

Shots on Target For vs Total Shots For

Who had QPR leading Man United in 'total shots for' before the season started? Problem for QPR is that they just do not hit the target. Some of this will be down to poor shot selection and some down to being unlucky but I think we can make a guess that QPR (and Stoke) start hitting the target at a more normal rate over the last 10 matches.

Chelsea continue to lead in 'shots on target' but Man City has made gains in recent weeks. I have been expecting Chelsea to regress to a normal rate of putting shots on net but this has not happened. Their ability to get into dangerous shooting areas has been very good. As per Objective Football, Chelsea leads the league in 'minutes played while leading' by a lot of minutes and this is probably driving the disparity between shots and shots on target.

Man United and Southampton really do not have 'Top 4' shot numbers. Meanwhile Spurs seem to be moving on up in the shots for department.


Goals For vs Shots on Target For


Liverpool really should have more goals, at this point, based on their shots on target. Kind of surprising given their solid run of form over last couple months. I cannot really see any reason why the goals will dry up, other than injuries.

Aston Villa cannot score goals this season (~18% G/SoT). Can 'Tactics Tim' simply say 'Benteke, score me some goals'? Worked for Mr. Burns on the Simpsons Episode 'Homer at the Bat' (Could not, for the life of me, find a video clip):

Burns: You, Strawberry, hit a home run.
Strawberry: Okay, skip.
(hits a home run)
Burns: Ha-ha! I told him to do that.
Smithers: Brilliant strategy sir.

If Villa start scoring it will be more to do with regression rather than Sherwood waving his magic wand. I am sure the media will see it this way, as well.

I mentioned, above, that we could make a guess that Stoke will start hitting the target more often. We can also make an assumption that they will score at a slower rate. Now this does not mean they will not score. It just means that they will score at more normal rate. Essentially, the increase in shots on target will offset the decline in scoring %. Confused? Me too.

Man, stats can be Hard to Expain...



Clarke