The PL 13/14: Man United & Quality Of Opposition

Thank ___ the Premier League football is almost back. I’m not a huge fan of international football. Yep, not even the greatest World Cup since the last World Cup particularly fired up my imagination. Probably just personal taste and Northern apathy for a national team nary seen past Watford gap.

Anyway, back to club football from here on.

I just wanted to post a quick graph and some numbers on Man United’s 13/14 results against certain quality of opposition, specifically, <50% TSR teams and >50% TSR teams.

man utd 13 14 qoo 

*All opposition TSR numbers taken at the time of fixture.

Man United went 5-4-10 against >50% TSR teams and 14-2-3 against <50% TSR teams (some opposition were very close to the 50% mark and likely crossed from “good” team to “bad” team numerous times throughout the season). Now, Man United crushed the subpar TSR teams taking 44 of a possible 57 points. Against above average TSR teams United took just 19 of 57 points.

Was this poor points return due to overly cautious Moyes tactics or simply due to the fact that United were overmatched talent wise? Not sure we’ll ever truly know, but it happened and it was likely due to both talent issues and managerial issues.

Van Gaal has plenty of room to improve this Man United side and he’ll need to start by figuring out how to play against Premier League teams that are actually good in terms of tilting the field (shots count) in their favour. 


World Cup Score Effects: Final Third Possession %

I have two types of final third possession to show you today: attempted final third possession and completed final third possession.

Attempted Possession% = attempts for/(attempts against+attempts for)

Completed Possession% = completed passes for/(completed against+completed for)

This is the data:

wc pass% chart

This is the data in graph form:

wc pass%

OK, a few things: It is no surprise to see that trailing teams post higher possession %’s (attempted) than leading teams. These pass % numbers are the foundation on which trailing teams are able to outshoot a team with the lead. Oupassing leads to outshooting the leading team.

But why are trailing teams able to outpass leading teams? Leading teams (especially at +1, and some at +2) sit back and allow the opposition time to play passes into their final third. A defensive shell operates, say, 20 to 25 to 30 yards from goal, which still leaves plenty of room for the trailing team to pass the ball into and around in that final third.

Of interest is the information at minus 1/plus 1 regarding possession% of completed passes. The -1 team posts a higher Completed pass % than Attempted pass %. How is this possible?

When a trailing team is passing the ball in the oppositions final third it likely encounters a defensive shell and numerous defensive bodies that is has to pass around, in other words defensive pressure. So how is it possible that a trailing team who faces this defensive pressure posts a higher Completed pass% than the leading team who doesn’t face this ttype of defensive pressure when attacking? I think it has to do with pass length and difficulty of pass.

Trailing teams who attack against this defensive pressure likely have more ball support (attackers and passing options) and play shorter, lower risk passes as they probe around the opponents defensive shell. This likely leads to a lot of completed final third passes. The leading (+1) team, sat deep, trying to protect its lead likely plays a higher number of long passes or counter attack passes when attacking. Longer passes into or inside the opponents final third involving fewer attackers and ball support than a trailing team would have leads to a lower number of completed passes.

Well, this is my rough theory. Feel free to have a good ‘ol kick at it if you want. It is also worth noting that once a team moves past +1 things change somewhat.

WC pass prevention%

The leading team (+1) has problems on both sides of the ball in terms of pass completion and pass prevention: It breaks up a lower percentage of the oppositions final third passes (passive defensive shell) and posts a lower pass completion percentage likely due to the longer pass with fewer attacking bodies to aim for.

We should never take team level final third pass numbers at face value for those numbers need context (game state). It is also worth treading lightly when quoting an individual player and his final thrid pass numbers and completion numbers. Those individual numbers need context, and the number of minutes spent at certain game states would be just one part of that context.



World Cup Score Effects: TSR, SoTR & Fenwick

This is the data split by game state:

WC fenwick chart

And as a visual:

WC Fenwick graph So, by TSR and Fenwick% trailing teams outshoot leading teams. If we focus on the -1/+1 split we still see that the trailing team (-1) outshoots the leading team (+1) by TSR and Fenwick but that trend doesn’t continue when we look at SoTR.

So what happened?

In the previous piece I talked about how the lack of space for the trailing team to attack into led to difficulties in getting shots on target, and how the increase in space led to the leading team having an easier time getting its shots on target, well, that shows up in SoTR.





















World Cup Score Effects: Shots On Target Rating

This post will look at the percentage of shots a team gets on target, the percentage of shots on target that are prevented and a shots on target rating number which is the for% and the prevention% added together.

WC SoT Rating chart

Remember that the game is barely played outside of the boundaries of +1/-1 so sample sizes will be an issue with any data outside of that.

See how big the advantage on both sides of the ball the team who is at plus 1 will have? The plus 1 team gets a higher percentage of its shots on target and prevents a higher percentage of the oppositions (trailing team at minus 1) shots from hitting the target.

Those numbers lead to a SoT Rating number that is 113.8 for the plus 1 team and just 86.2 for the minus 1 team. It is a huge swing in shots on target efficiency.

WC SoT rating Graph

The plus 1 team enjoyed such a tremendous boost on both sides of the ball.



WC sot rating team

Some awful SoT Prevention percentages here. Hands up Spain and Ivory Coast who should have been a lot better defensively.

World Cup Score Effects: Shot Outcomes

I’m going to focus on shot outcomes (missed, on target, blocked) at Close (Within a goal). Only 13.3% of World Cup minutes were played outside of this game state. The other stuff likely doesn’t matter too much.

Shots Numbers

WC shot outcome chart

These are the shots numbers broken down into three categories. We know more shots are taken by the trailing team (incentives) which leads to more missed and more blocked shots but less shots on target than the leading team.

So let’s break this down into some percentages.

WC shot outcome % chart

When we look at the shot types in terms of a percentage we see some different things:

  • Missed shots, as a percentage, occur, more or less, as frequently despite the different defensive pressures and defensive tactics teams employ across these three very different game states.
  • Teams at minus 1 had the shots that they took blocked far more often than when they were tying or leading.
  • Teams at minus 1 managed to get just 27.7% of their shots on target compared to teams at plus 1 who got 41.5% of their shots on target.

WC shot outcome graph This is what the above table looks like in graph form.

As we move from left to right (trailing to tied to leading) we see that the percentage of shots that are blocked decreases and the percentage of shots that are on target increases.

This is likely caused by lack of space in which teams attack into. Teams that are trailing by a goal come up against the defensive shell, or the parked bus, and find that a higher percentage of the shots they take are blocked by one of the many opposition players. That same defensive pressure makes it harder for the trailing team to get its shots on target.

The leading team finds that it can attack into space as their trailing opponents commit more bodies to attack in an attempt at trying to equalize. More space, and fewer defenders, leads to a lower percentage of shots blocked and a far higher percentage fo shots on target.

It’s all about space and numbers of bodies. More space and less bodies will lead to quicker attacks and fewer final third passes. Less space and more bodies will lead to a slower build up with more passes and the mighty difficult task of getting shots on target.

World Cup Score Effects: Shot Rate

If one were to take the number of shots taken and the number of minutes played at each game state we can create something I called shot rate per minute. This is what the 2014 World Cup looked like:

WC shot rate graph Remember how this post on Time showed that the game is barely played outside of +2/-2 so ignore the information outside of those game states due to tiniy samples.

The really cool piece of info here is the shot rate at minus 1. When a team was trailing by a single goal in the 2014 World Cup they markedly increased their shot rate per minute. This isn’t surprising really, it’s all about incentives. The leading team (+1) drops deeper with a sharper focus on protecting their important one goal lead which allows the opposition (-1) closer to thier goal which allows them to play more final third passes and take more shots (increased shot rate).

WC shot rate chart

At minus 1 a team shoots 16.47 times p/90.

At tied a team shoots 11.79 times p/90.

At plus 1 a team shoots 11.43 times p/90.

Again, this is all about incentives. It is also worth noting that not all of the extra shots at minus 1 are good shots.

World Cup Score Effects: Time

This article will be part 1 of a series of very short articles which will focus on the 2014 World Cup and score effects. Topics that will be covered include, time, goals, shots, shot outcomes, shot blocking, scoring%, save%, final third passing and likely a few other topics. I was initially going to include all these topics in one post and have it featured someplace other than here but any such article would be way too long. I also have serious concerns about how many people want to read work that is over 1000 words, not to mention the fact that I no longer wish to write anything that lengthy.



WC % Time 

Close is when the game is within a single goal (+1, tied, -1) and Extended Close is when the game is within 2 goals (+2, +1, tied, -1, -2)

These numbers fit pretty nicely with what I found when looking at the EPL: the game is barely played outside of Close. If we use Extended Close we capture ~95% of all game time.

WC Time

 This is the distribution of the chart featured above.

Team Minutes

WC Team Time Engerlaaand, Engerlaaand, Engerlaaand……

I would post this as a table for people to scrape but wordpress is a piece of shit.


Football Statistics


Get every new post delivered to your Inbox.