"...this is the most articulate and accurate piece written about the club for years!" - Tales from the Front, http://www.otib.co.uk/

Wednesday, 18 July 2012

Lies, damned lies and statistics: Bristol City

Is this the truth behind City's season?


As any of you who follow me on Twitter will know, I follow many people with a different take on football and take great pleasure in retweeting the best and most interesting.  Most of these probably pass by relatively unnoticed, as so much on Twitter probably does due to the way the timeline works and the number of people everyone follows, but one in particular caught the imagination last year.

City were unknowingly coming to the end of their worst run of the season – and that’s saying something considering the start we made! – but when I retweeted this article by Ben Mayhew - it had more reaction than to any other I can recall.

If a dose of reality was required – although to be honest right at that time it really wasn’t – then these charts brought them home.  Poor in front of goal, yes.  The worst in the entire league…blimey!

The interest shown in the charts leapt back into my consciousness when I saw the site’s proprietor had done a review of Nottingham Forest’s season recently.  When I approached him to do the same with City and he accepted, I have been eagerly anticipating the results ever since.

So here it is, a fantastic piece of analysis and commentary, with many thanks to Ben:

Over at Experimental 361, I’ve developed a graph to visualise and compare teams’ attacking and defensive performance, which I’ll be using here to briefly analyse Bristol City’s season.

In a convoluted nutshell, the horizontal axis measures the average number of goal attempts a team either creates or faces in a match, while the vertical shows how many shots it takes on average for a goal to be either scored or conceded. These are quite simple metrics, but used together they can be quite revealing (as this post will hopefully demonstrate).

I’ve used it below to plot out City’s attacking (green) and defensive (red) performance, split by (H) Home, (A) Away and (O) Overall. The axes are centred on the average values for the Football League, which allows us to split the graph into quadrants (which I’ve labelled).
























Attack
What's immediately evident is that while creating chances isn't a problem for the Robins, converting them certainly is. You'll no doubt be aware that this was a particularly low-scoring season for City, and here's the reason why: overall they took 3 additional shots to score each goal than the average Football League team. This would be fine if they were taking 3 more shots than most other teams, but the picture shows that the number of chances they create is only slightly over average for the division. As you'd expect, the team are better at both generating and converting shots on goal at home and this disparity is well within the normal range. The focus for next season simply has to be on improving finishing and / or creating more clear-cut chances, although there is some cause for optimism which I'll touch on later.

Defence
While the defence overall looks to be creeping into the worst quadrant, particularly in terms of the number of shots faced, it's interesting to see what happens when the home and away records are split out. At home, City are slightly better than average at restricting the number of shots the opposition have, but the ones that do get through are more likely to go in. In away games, they allowed the opposition to come at them a lot more, facing more than 3 extra shots per game than the average side, but repelled a larger proportion of these. Given that the overall resilience is hardly different from the League average, I'd say the coaching staff should focus on closing down the opposition a bit better to reduce the number of shots at the City goal.

Pre- and post-McInnes
It was suggested that I look at the impact of Derek McInnes' appointment, given the dire state of affairs when he joined the club, so here is the same graph split to show performance pre and post his arrival:



The attacking woes McInnes inherited are starkly visible here in the darker green - the eagle-eyed among you will notice that I even had to extend the vertical axis to accommodate the extreme home profligacy! Needing almost 23 shots to score each home goal massively skews this distribution and shows that the Robins figuratively couldn't buy a goal at Ashton Gate in the first few months of the season, despite fashioning a very respectable number of shots on goal. Whether this was down to the quality of the chances or some very nervous finishing I don't know, but the improvement under McInnes was revelatory. At home the efficiency of the strike force more than doubled which, despite a slight deterioration in away games, is a remarkable achievement. If you ignore the pre-McInnes stats, the attacking picture looks a lot less bleak and suggests that if the initial improvements made during his reign continue then the Robins can expect a much more cheerful campaign next season. He improved the team's resilience at the back too, with the team soaking up an average of 3 extra shots for each goal they conceded. They'll have to watch the number of shots they face away from home though, as the Championship's more clinical attacks won't need 14 shots to find a way through most defences.


Summing up
I’m ashamed to say that I’ve never watched Bristol City play, which in addition to leaving a gaping void in my soul also prevents me from interpreting these graphs using detailed knowledge of the team’s qualities and foibles this season. Please feel free to weigh in with your infinitely more informed interpretations in the comments section below.

So there you have it, if you weren't clear on where City's failings were and what had improved since McInnes took over then hopefully that's clarified a little for you?  I think the information is fascinating and I suppose the fact that we're at such an extreme end of the attacking chart in particular it probably won't surprise too many.

The improvemens made since McInnes took over are shown quite clearly, with the attack in particular becoming more of a threat and very much moving in the right direction from the extreme 'worst-in-the-league' status it did hold, whilst the defence has also marginally improved and looks to be becoming more solid.

However, there remains plenty of work to do and Ben promises much more content on his site next year, so it will be interesting to see how things move on next season.

If you've enjoyed this you can check out Ben's site here and follow him on Twitter by clicking here and please leave a comment so Ben or myself can respond.

Follow me on Twitter: @TheExiledRobin

http://exiledrobin.blogspot.com



4 comments:

  1. Very good analysis by Mr Mayhew - to some extent it doesn't matter if he'd not seen City play, although if he had done he'd have probably been as frustrated as the rest of us with the 'pass the ball around the opposition penalty box before losing possession or failing to shoot' style of attacking that was prevalent last season.

    Defensively the problem could also be illustrated by the decline in clean sheets from 16 in 2010/11 (only Millwall, Swansea and QPR kept more) to just nine last season...with three of those coming against teams that were eventually promoted.

    I definitely think the worst is over though. Here's to an uneventful season of mid table mediocrity in 2012/13 :-)

    ReplyDelete
  2. An interesting review. Some of the statistics could be better interpreted with a knowledge of formations, however you are not a City fan. City played one up top for a great deal of the season, they were shooting from distance. Playing 4-5-1 [4-4-1-1 sort of at times] invited the opposition onto the team to shoot as it was not an attacking formation, City sat deep.

    A change in formation and inclusion of Carey, Skuse and Stead brought improvement. There were interesting statistics that show over a season Carey in the back four keeps goals down and points up, that appeared to be true again as Carey played most of his football after feburary.

    ReplyDelete
  3. Any chance of an update for 2012/13?

    ReplyDelete
  4. Hi Ted,

    Could well belooking for a follow up either at midway or the end of the season, but in the meantime why not check out the man's blog - there's general league updates on there so you can pick out City's relation to others

    http://experimental361.com/

    ReplyDelete

Most Popular Posts