Howard Webb: Manchester United’s 12th Man?

Howard Webb: favours Manchester United?

Given his achievements in the game, Howard Webb is easily England’s most accomplished referee. It’s a fact that grinds with fans of many Premier League clubs, who don’t feel he warrants the praise he receives. It’s safe to assume Webb divides opinion; he commands the respect of his peers, but often frustrates fans with some of his decisions.

One of the biggest accusations levelled against Howard Webb in recent years is a perceived bias towards ‘bigger clubs’, chiefly Manchester United.

This post aims to look at this belief, which although is mostly held by fans, is also perhaps held by players too, as Ryan Babel’s infamous tweet showed.

First, however, a few items to point out:

• This post does not, by any means, aim to question the integrity of Howard Webb. It merely looks at decisions he’s made in matches involving Manchester United, and compares them to decisions made by other referees in similar games. Any suggestion that Webb gives Manchester United more favourable decisions than their opponents will be attributed to external factors, as opposed to any inherent bias.
• We can only look at decisions made during a match, namely yellow cards, red cards and penalties awarded (either scored or missed). There is no way to measure decisions that ‘should’ have been made; hopefully this is reflected in decisions not made (e.g. teams against Man United have disproportionately fewer penalties).
• It is also impossible to judge if the decisions made were correct, but again we look for a disproportionately high number of decisions favouring Manchester United.
• This analysis only looks at Premier League games. There are two main reasons for this; league data is more easily accessible than cup data, and cup games have a different dynamic to league matches – I was keen to avoid any bias that may arise from this.
• If there are any suggestions to improve this analysis, please comment in the space below this post. Any suggestions are hugely appreciated.

The Big Picture

Howard Webb refereed his first Premier League match on the 18^th of October 2003, a 0-0 draw between Fulham and Wolves. Remarkably, his first five Premier League outings failed to produce a goal. Webb’s first Manchester United game came against Southampton in the FA Cup in March 2005, and just under a month later, on 9^th April, he officiated United’s 2-0 loss at Norwich.

Including the match at Carrow Road, Howard Webb has taken charge of 25 Manchester United league games over seven seasons. During this period, United have taken part in 197 other league matches (up to and including United’s 2-1 loss to Wolves on Saturday). Here’s a  summary of decisions made during this period:

Two facts are immediately striking; Howard Webb has never given an opposition player a red card in a Manchester United match, and that he also appears to have a sterner approach, dishing out nearly 2 yellows per match for United, and over 2 yellows for their opponents. This is higher than the 1.45 and 1.81 figures for other referees respectively. This could be in part due to Webb’s appointments in ‘bigger’ matches, where inevitably tempers flare and stronger discipline is required; more on this later.

Broken down into home and away matches – that is matches at Old Trafford and otherwise – we see the following averages:

Webb’s no-nonsense attitude comes through again here, with yellow cards handed out at a significantly higher rate, particularly at Old Trafford. He’s only refereed 12 league matches at Old Trafford, but 5 penalties in United’s favour at a rate of 0.42 per match compared to 0.14 with other referees does raise questions. Again, more on this later.

Webb’s big match record

Before we jump to any conclusions, we must consider the skew created by the fact that 11 of Webb’s 25 Manchester United matches have been ‘big matches’; against Arsenal, Chelsea, Liverpool or Manchester City. A further two have been against Spurs – analysis on this later.

Since 9^th April 2005, United have played 33 ‘big matches’ officiated by a referee other than Howard Webb, and a further ten against Spurs. By comparing Webb’s record against other referees in only big matches, we eliminate the skew generated in the overall statistics. After all, it’s primarily in big matches that fans get frustrated at Webb’s decisions.

Webb’s distribution of yellow and red cards is very similar to other referees, and there’s no reason to suggest he favours United on this front. If anything, there’s a case for the contrary. The three penalties Webb’s awarded to Man United, at a rate of 0.27 per game, is significantly higher than the 0.12 per game normally awarded to United, but a closer look at the penalties (Arsenal at home 2007/08, Liverpool at home 2009/10, Arsenal at home 2010/11) suggest there was nothing hugely untoward. Again, he’s possibly refereed too few big matches to draw meaningful conclusions.

Spurs come into the equation because their match with United in 2008/09 sparked huge controversy. Trailing 2-0 at half time, United were awarded a contentious penalty for a Gomes foul on Carrick. Ronaldo scored, and United went on to win 5-2. Post-match, even Sir Alex Ferguson admitted the penalty was “fortunate,” and Webb was duly relegated to the Championship for a weekend. It would be 11 months before the Premier League handed Webb another big match, United’s 2-1 victory over Liverpool in March 2010, which incidentally saw another (slightly less controversial, but still debateable) penalty decision for United.

So, including Webb’s two United v Spurs matches, and the 10 officiated by his colleagues since his arrival in the Premier League, we get the following averages:

Curiously, in five of the six comparable decisions Webb favours Man United’s opponents over the Red Devils compared to other officials. He gives more yellows and reds to United than normal, fewer yellows and reds to their opponents than normal, and more penalties to their opponents than normal. Admittedly, many of these are marginal, but more importantly they show no overriding bias towards United.

The only area – an important area – where Webb favours Man United more than normal is his awarding of penalties, with 4 in 13 games at a rate of 0.31 compared to 0.16. The question is whether this is a significant increase. Unfortunately I have neither the knowledge nor resources to confirm this, but am more than happy to listen to suggestions to test it statistically.

Howard Webb: impartial and unfairly criticised

The sometimes-held belief that every decision goes United’s way when Howard Webb is officiating, particularly in more important matches, is hugely exaggerated. Although he’s refereed relatively few Manchester United matches, there’s no evidence to suggest Howard Webb gives a significantly disproportionate amount of decisions in the Red Devils’ favour. He isn’t, for example, handing out penalties to Man United every other match, nor sending off opponents willy-nilly. In fact his record of never sending off an opposing player in 25 league matches is fairly remarkable, perhaps reflective of his ability to control matches from early on.

Some of the criticism directed towards Howard Webb is hugely unfair. Although this analysis looks at decisions Webb makes, it is certainly not the best basis by which to judge referees. Officials will always get decisions wrong, and it’s unjust to isolate examples. Instead, referees would be better judged in the media and by the public by their ability to keep control of matches, and on this front Webb generally excels.

Footnote

I wrote this article a year ago, and it has proven to be immensely popular with people trying to find evidence of Howard Webb’s bias towards United. Most people cite Webb’s propensity to give United penalties as evidence of bias. The reality is that this is likely to be statistically insignificant, and what’s more, there’s nothing to say these penalties were incorrectly given. Anyone questioning the integrity of referees should recognise this.

Note: Research for this was conducted independently, and whilst great care was taken to ensure no errors in the data, it is not always possible to by 100% accurate. That said, the analysis should not be significantly affected if there are errors in the data.