Investigating goals and corners in knockout competitions

In my most recent blog I failed to find any evidence of seasonality in corners, despite previously finding a link between seasonality and goals. I also looked at the direct relationship between goals and corners and found a correlation to exist in some instances.

In this piece I will look at goal and corner rate production in first legs versus second legs of knockout competitions, more specifically in the UEFA Champions League.

When previously finding that a relationship between goals and corners existed in some leagues, I drew the conclusion that as goals and corners are results of attacks, then offensive teams were likely to produce high numbers of both goals and corners (and defensive teams the opposite).

In each league around the world, we would expect the best teams to be the most offensive ones: PSG in France, Bayern Munich/Borussia Dortmund in Germany and Barcelona/Real Madrid in Spain. It is a fair assumption to make that teams who dominate their domestic league will score the most goals and produce the most corners. This is due to them building more attacks than the other teams.

Thinking more generally, though, what else forces teams to attack during games and can make them more likely to score goals or win corner kicks?

One answer could be when a team is chasing a game having gone behind.

If a team goes 1-0 down they are more likely to push men forward in the hope of scoring an equaliser. This should be true for all teams, so even a defensive side who set out to initially play for a draw will be spurred into action to try and get back into the match when falling behind. By attacking more themselves, they will leave more space for the opposition to play into and this should generate more attacking play overall, leading to goals and/or corners.

This idea of attacking more when trailing in a match is strongly related to the knockout stages of cup competitions. If a team is behind from the first leg in a tie, then it’s the equivalent of being behind in a normal match, only with 90 minutes still to play. We can expect the trailing team to be more offensive minded than normal in the second leg and attack from an earlier point in the contest.

In the recent Europa League Round of 32 fixture involving Manchester United and Midtjylland, United were trailing 2-1 going into the second leg at Old Trafford. Despite being badly affected by injury problems, the home side set out to attack from the first whistle. They produced four corners before Midtjylland took a surprise lead in the 27th minute. This then forced United into an even more attacking stance and they managed to produce a further 10 corners – along with 5 goals – to advance to the next round with an aggregate score of 6-3.

A commonly held view in betting circles is that goals increase the chances of further goals. This is to say that the first goal in a game is a catalyst for more attacking play and ultimately more goals. We could substitute “corners” for “goals” in the last sentence too, though only if we agree that corners and goals are just outcomes of a single attack.

To see if this view was correct I looked at the first and second leg goal and corner rates from the Round of 16, Quarter-Finals and Semi-Finals of the UEFA Champions League for the past 10 seasons, excluding this season so far:

Knockout Goals avgKnockout corners

From the graphs above, we can see that goals and corners generally do increase in the second legs of knockout competitions. In fact, in the instances where there were more goals or corners in the first legs it was typically not by much. In the other instances there is overwhelming evidence of more action occurring in the second legs.

Each game, season and tie must be looked at on an individual basis, however, despite the evidence suggesting an overall trend. This season in the Round of 16, not many of the tie favourites were behind going into the second legs, and it would be no surprise if we did not see this behaviour in such circumstances.

In the first group of second leg matches played recently in the UCL, only Zenit were pre-match favourites to qualify and were trailing going into the second leg. They managed to produce seven corners before scoring a goal in the 69th minute and levelling the tie. Unfortunately for them they couldn’t build on this equalising goal and conceded twice late on to lose 3-1 on aggregate.

Zenit weren’t particularly strong favourites in any case, being given only an implied 55.56% chance of progressing by the market (4/5), so maybe shouldn’t have been expected to turn around a 1-0 first leg defeat, or to be fully dominant in the second leg.

Just like in a normal match, when favourites go behind in a game they are then expected to attack more frequently in order to retrieve the situation. I believe this will be the case in second legs of ties when the pre-match favourites are trailing, just like in the Manchester United versus Midtjylland example.

Although there weren’t any examples of this in the second group of second legs in the Round of 16 played this week, it is well worth monitoring the draw for the Quarter-Final, where an opportunity to back goals and corners could present itself.

Mark McAfee (@Avonbets)