Does the Past Research into Football Statistical Analysis influence its Future ?

Posted on: September 13th, 2016 by Jonny Grossmark


If you look at the graph above ( seen in a number of academic papers ) I think we can all agree that more goals are scored as time decays in a game.

The question is, does the time of the first goal effect the expectation of further goal(s) and does it also effect the ability of the team that scored the opening goal to hang onto the lead ?

Dixon and Robinson ;

“Clear evidence that the rate of scoring goals changes over the course of a match. This rate tends to increase over the game but is also influenced by the current score. We develop a model for a soccer match that incorporates parameters for both the attacking and the defensive strength of a team, home advantage, the current score and the time left to play.”

 Dixon and Robinson concluded that  there was a continuously increasing rate of goals for both the home team and the away team, and they thought that maybe this was caused by an increase in defensive mistakes due to fatigue.

They also found no evidence to support the idea of immediate ‘fightback’, that is, the home team equalising soon after conceding.

Survival  in terms of football, is the ability or inability of a team to prevent fightback.

Nevo and Ritov concluded that;

“a first goal occurence could either expedite or impede the next goal scoring, depending on the time it was scored. Moreover, once a goal is scored, another goal becomes more and more likely whether the goal was scored or conceded.”

Nevo and Ritov  also concluded that there were limitations to survival analysis, such as goalless matches, which cannot be incorporated into their research. In addition, their model could not account for the “fast goals”, as it was inaccurate for the first 15 minutes of a game because these events may be more random than others

Oliver Rubner and Andreas Heuer  ;

“Thus, a goal just before half time is no more influential on the further course of the match than an earlier goal. This observation suggests that it is not important to keep into account the absolute values of the scoring times. Thus, we omit the information about scoring times in order to strongly simplify the statistical description. Of course, strictly speaking a mild dependence cannot be excluded for other scores. Since testing all scores and all scoring times is impossible for statistical reasons we did not proceed any further with this question.”

“Our results also allow us to find an answer to our initial question about the origin of the large number of draws. It is the persistence of a draw, i.e. (2), rather than the ability of a team, trailing by one goal, to score an additional goal as expressed by (3). Actually, (3) would rather decrease the number of draws because the probability that a 0∶1 transforms in a 0∶2 during the last minutes is significantly larger than expected.”

“(2) In case of a draw the total goal rate becomes smaller. Thus the goal rates have to be adjusted in dependence of time and score. This is a strong deviation from the Poisson expectation. (3) In case of a lead of the away team dramatic deviations from the Poisson expectation are observed during the last 5–10 minutes of the match.”


Full Time Summary 

The mainstream are not looking at the effect of the time of the opening goal for 2 reasons ;

1. the academics do not believe that is possible to do for statistical reasons > I am not sure why as I have been  testing all scores and all scoring times in a number of leagues for the last few years and yes it was labour intensive but surely the academics can  ease that labour .

2. The mainstream to an extent believe that the effect of the early goal is based on hot air without really having looked at this area in any depth.


Expectation of a goal is dependent on the current score ( GAME STATE )  added to the time of the opening goal  added to any other variables that may occur during the game.

The reason why we see a large number of draws is because of games that have a low goal production and indeed a game with an early goal could also produce a low goal production and produce expectation of a draw .

Indeed it is correct to suggest that the away team  winning late on are more likely to > 0-2 pathway but if they opened the scoring late in the game .

The numbers game book jumped on the bandwagon to advise that teams are not vulnerable when they have scored , which is indeed not correct as the research they offered to back up that statement has a serious flaw because the sample looked at only had games that ended 1-1 at full time and as a result they did not discover that an early goal in a number of leagues and there will be expectation of a quick response , and very few early goal games will end 1-1 because of the expectation of acceleration of goal production .

Another area which the mainstream have ignored is the effect of the early goal and the goal(s) 80 + in terms of higher expectation of  a late goal then in a game with no early goal metric


My challenge to the mainstream / Academics > have a look at the effect of the time of the opening goal in individual leagues added to the half time game states and you should end up with data that gives a warm glow . Thumbs Up