A chance is a chance is a chance?

Posted on 13 July 2011


Before we dive deep into soccermetrics territory to aim for a method to assign different value to different chances created during a football match, let’s take a big step back in time. Back to our early childhood, the school yard…

I’m pretty sure that the majority of us have enjoyed playing marbles at that time. Not just the fun of playing the simple one-on-one game, but also the trickery trade that came with it. There were those guys (or girls) you looked up against, who always had the biggest and most shiny marbles of all, making your bunch look pale and insignificant.

This simple game of marbles taught us valuable principles that we apply in the rest of our lives, particularly on the relative value of different objects. Later on, we’ve come to apply the same principle to coins, cars, etcetera. The bottom line being that not all objects are of equal value.  But we generally forget this principle when evaluating the chances created during a football match.


From the school yard to the pitch

Traditional match reports tend to provide statistics, if any, on possession average and chances created. And after a match where your team created twelve versus conceding eight chances to the opponent, you might feel hard done with a 1-1 score line. But we don’t take into account the nature of the different chances during the game.

Imagine the twelve chances team to be the huge favorite, playing an ultra defensive opponent. Their twelve chances are likely to be a range of hopeful long shots with a couple of odd through balls or shots from corners thrown in. But if an ultra defensive side replies with eight chances of its own, their chances are likely to contain a number of fast breaks among them.

So here we touch upon the issue to be sorted out before we can judge the relative value of the twelve versus eight chances during our fictional match. What’s the value of each different chance created?

In order to answer this difficult question we have to assess the chance that each chance results in a goal being scored. In order to avoid the word ‘chance’ referring to two different things (the fraction of goals scored and the goal scoring opportunity itself), I’ll refer to goal scoring opportunities as attempts from now on. So, what’s the value of having a penalty? How many attempts created from open play equal one fast break attempt? What about direct free kick attempts?


Soccer by the Numbers

A huge inspiration to dive into this matter was provided by a series of posts on the excellent ‘Soccer by the Numbers’ blog. Several of the questions raised above are answered straight away in different posts on that blog, using OPTA match data from the past EPL season. For example, the average amount of goals scored per penalty we’ve learnt to be 0.77 (lower than you’d imagined?) and the average amount of goals scored on fast break attempts to be a fraction 3 or 4 higher than attempts from direct free kicks.


Different chances

Generally six different ways to generate attempt on goal are discriminated (in decreasing order of frequency): attempts from open play, corner situations, set piece situations, direct free kicks, fast breaks and penalties. But in order to assess the value of individual goal scoring attempts, other factors have to be weighed in.

Obviously, the distance from goal affects the chance that an attempt results in a goal. Another factor to weigh in would be whether the attempt is a shot or a header. Fortunately, this type of information can be deducted from the above mentioned data.


Breaking down the data

This all being said, we can now break down all 11078 goal scoring attempts of the past EPL season by three different criteria: match situation, goal distance and shots/headers. The next figure contains an example of a pitch lay out, with different grids containing clusters of attempts and their success rates in terms of goal scored. This image only concerns shots from open play, but we could construct the exact same figure for each different match situation for both shots and headers set apart.

Goals per shot from open play, broken down by different areas of the pitch

Due to the low amount of shots fired in from the 6-6, 8-2, 8-9 areas, the odds here stick out due to a single attempt flying into goal. But the overall picture is quite clear and it confirms what we would expect: closer to goal and closer to the axis of the pitch benefit the odds of scoring from an open play shot attempt.

This kind of figure could be made for all combinations of shots and headers from open play, corner situations, set piece situations and fast break, as well as for shots from direct free kicks. For shots from penalties, the previously mentioned estimate of 0.77 goals per attempt will suffice.


In the end

Based on these measurements we now have a system in place in order to estimate the overall value of the chances created by either team during the match. Knowing how many goals a team is expected to score from its chances is of much more value than just knowing how many attempts to score a goal were made. Other applications of this method of evaluation would be to distinguish a lack of quality attempts created from a finishing problem or to evaluate defensive and goalkeeping performances. And a third option would be to plot the balance of play during the match in terms of the quality of chances created in order to graphically represent how the balance of play evolved during the match.

Evaluation the quality, rather than the absolute number or chances created seems like a worthwhile effort. And with more detailed Eredivisie data on goal scoring attempts available on, hopefully, short notice, this kind of tool might prove a valuable addition to this season’s match reports on 11tegen11.

Posted in: Uncategorized