Pages

Saturday 5 November 2011

Shot Conversion Rates.

San Marino almost certainly wrote themselves a permanent footnote in the history of the World Cup by scoring a goal after just 8.4 seconds of a World Cup qualifier,they also demonstrated how unusual and unlikely events can and do happen.Europe's third smallest nation with a population that could fit comfortably into the Britannia Stadium didn't accomplish their feat against a fellow minnow,but against England and on neutral soil.Five passes and an assist from Stuart Pearce,what could be simpler.Sadly for the footballing,motor racing enthusiasts the remainder of the game took a more predictable route as England ran out comfortable winners,7-1.Roughly the margin of victory  expected by the pre game ratings.

Overall the 6 goal margin of victory would not have surprised too many people,but few would have predicted a San Marino goal,let alone in the game's opening seconds especially as it was the first time the had ever led in a recognised FIFA international.England,the first 9 seconds excepted had performed in a fairly predictable manner,but Gualtieri's goal had demonstrated that the timing and order of goals can be much more erratic.As Cloughie said "it only takes a second to score a goal".


If that's the case we should therefore expect the timings of goals scored to require a much larger sample size before they begin to approach our pre game expectations.If we hold onto this thought it may help us to explain some of the more unusual patterns seen when we look at a team's ability to turn shots into goals.


In the 2010/1 season a number of teams appeared to be much more efficient a turning away shots at goal into goals compared to their efficiency on home turf.Among them,Arsenal who scored a goal every 9.6 shots at home,but needed just 7 shots on the road and Blackburn whose respective figures were 10.6 and 6.3 shots.In essence,these teams appear to be away specialists.However,football is notoriously littered with specialists of many kinds and in virtually every case the label is an illusion caused predominately by random chance and the road warriors more often than not see their stats gravitate to the league's norm.


Arsenal particularly had been in very advantageous positions during their travels in 2010/11,so I added up all the minutes they had led in a game during their away games,reasoning that whilst they were ahead,not only had they probably been very efficient in taking their chances to gain the lead,but once ahead they could look to play a slightly more conservative game,resulting in less speculative efforts on goal and increasingly as the game progresses clearer cut chances as their opponents push on in search of reducing the deficit.In short,for a large part of their away campaign they had found themselves in a position where their already high shooting efficiency was likely to remain high.And if Gualtieri's record breaker against England was a marker,they could have found themselves in those advantageous positions over a short 19 game stretch at least partly through pure chance.


When I looked at Arsenal's "time in the lead" record at home I found it was equally extreme,but in opposition to the away split.They had led at home for just 33% of the available minutes compared to 45% of the available minutes on the road.If their attacking had been cool,clinical and easier on their travels because of their advantageous away positions and increasingly speculative,desperate and more difficult at home because of the unusually large percentage of the home time they spent either trailing or drawing,then that could partly account for the difference in observed efficiencies.


Time spent in the lead,split by venue for teams who were more efficient at converting chances on the road. EPL Season 2010/11.

TEAM. % of Time
Spent Leading at Home.
*Shot
Conversion Rate at Home.
% of Time
Spent Leading  Away.
Shot
Conversion Rate Away.
Wigan 11.7 0.086 17.5 0.093
Arsenal. 33.0 0.104 45.0 0.141
Spurs. 22.4 0.093 28.3 0.106
Blackpool. 21.4 0.119 25.9 0.130
Blackburn. 29.3 0.094 20.4 0.159

*Shot conversion rates are available from a variety of sources.Some count blocked shots in the total,some don't.Therefore there may be slight variations depending upon the source used.

However,one team does not provide reliable evidence,we need a larger sample size.So I recorded the percentage of minutes a team led both at home and away for the last EPL season and matched these numbers up with their shot conversion rate for each type of venue.As we can see from the table above,five teams appeared to be more efficient on their travels and four of those five teams spent a higher proportion of their away games measured by time leading compared their games at home.


The five leading home shot conversion specialists are listed below and all five spent a larger percentage of their home time leading compared to their away games. 

Time spent in the lead,split by venue for teams who were more efficient at converting chances at home.EPL Season 2010/11.

TEAM. % of Time
Spent Leading at Home.
Shot
Conversion Rate at Home.
% of Time
Spent Leading  Away.
Shot
Conversion Rate Away.
Newcastle. 34.0 0.162 17.3 0.083
Wolves. 33.1 0.144 17.2 0.098
Liverpool. 46.7 0.145 18.3 0.101
Bolton. 27.0 0.122 9.6 0.082
Man Utd. 59.7 0.157 23.4 0.126

If we now plot the shot efficiencies against percentage of time spent leading for all the EPL teams from the 2010/11 season for both home and away games we find that there is a reasonably strong relationship between the two,r squared values are in the region of 0.4 for each graph.


























The same graph as above only this time plotted for away games.




























So there appears little doubt that time spent leading a game is correlated to shot efficiency,but the question remains as to which variable drives the relationship.Do teams lead for long periods away from home because they are particularly good at converting away chances.Or are teams able to maintain a high efficiency on the road partly because random chance has allowed their goals to fall in an advantageous order,enabling the "away specialist" to ply their highly efficient start against teams increasingly more concerned with attack than defence.

I would suggest that random chance is the defining factor.If teams have a particular ability to better convert chances on their travels you would expect that ability to persist across seasons,but it doesn't.Arsenal weren't away conversion specialists in 2009/10,neither were Blackburn or Tottenham,but Wolves,who had been home specialists in 2010/11 were apparently away conversion kings in the previous season.This immediately leads you to suspect the fickle hand of fortune isn't too far away.

You would also expect occasional seasons where away specialists predominate,but year on year larger sample sized whole league figures show the on average superior home sides having a higher efficiency than the visitors.Conditions conspire to make home teams about 0.4 of a goal superior to their visitors and better sides on the day lead over large sample sizes for longer and that appears to confirm the relationship between leading and efficiency.Over large sample sizes the away specialist disappears.

I'm not claiming that random goal scoring timings account for all the variation seen in a team's shot efficiency,but it is certainly a factor.Team ability is another,Blackburn for example may be happy to sit on a draw at say Wigan,but Arsenal would be more likely to press forward possibly leading to reduced shooting efficiency if they were in the same position.

In short,the more desperate your situation is the less efficient you become,just look at the passing stats of NFL teams who are trailing.

2 comments:

  1. Very cool. There definitely seems to be a relationship between % of time leading and conversion rates. I think you're right to wonder if clubs that lead more often have better conversion rates simply because they are a better club.

    Just a suggestion, and I don't know how hard it would be to do, but to get rid of that problem you could compare club's conversion rates when they are trailing to when they are leading. If there is a significant difference between the two, then you would be pretty sure that club's have better conversion rates because they are leading.

    Just some thoughts, and keep up the good work.

    ReplyDelete
  2. excellent suggestions,Ford.
    I think the shot conversion topics will run and run........
    Thanks for the comments :-)

    ReplyDelete