WFC games data tracking with possession-based approach! – 06.11.2018

There are several ways on how to track and collect data. We are naturally interested to know how to win, how to score and prevent goals. When studying goals it seems logical to evaluate events connected to them - shots or shot attempts. The believe is if we analyze what types of shot attempts, from where on the field and how are generated we will be able to better understand of how to win.

Uhm, not really...


There is a way to go deeper as we can try to rate these shot attempts. We come closer to evaluating scoring chances. We can track other isolated events such as turnovers, passing plays or time possessions. As data analysts we usually try to link these events time- and/or sequence- wise and see if we find a pattern (e.g. turnover + one-timer from in close = good chance of a goal) for successful teams.


Are we getting there?


Generally, the game itself is not as simple as that to analyze. This way we still miss too many contextual things that are connected to three main components into the game: position – time – space.


Tell me more...


It matters in what position is a player with a ball, it matters how much time she/he has and what are spatial options for her/him (is any teammate open, in what formation opposition´s defense is, etc.) and these are changing in time.


For sure better teams will yield more scoring chances, possibly more shot attempts and can be successful in turnovers or possession-wise. These depend a lot on playing styles though. There are no proofs yet that more shot attempts or better possession numbers mean winning. It rather looks like (see the article) being able to master breakaway situations or yield clear path situations could be the key.


So are we still unsure?!


Public data are still not statistically significant to prove this yet. Nevertheless we move our focus from event- or shot-based approach to possession-based one expecting more. This has one more big advantage. There are many situations during games with an opportunity for one team but that play is never finished with shot attempt or miss pass happens.


Possession-based approach evaluates each possession separately searching for spatial opportunities created for attacking team. If your team yields odd man rush but does not execute the play, this is a dangerous possession created. With such data we evaluate both ability to create quality plays and also execution of these.


We will be also able to answer following questions:

  • How many organized possessions you need to create a clear path (good scoring opportunity)?

  • How much more dangerous are counter attacks than slow possessions?

  • What is the value of turnover possession vs. possession when forecheck is escaped?


It seems like shooting more or having more time possessing is not the right path to track a success. But ability to yield quality possessions (clear paths and counter attacks) could be a long-term goal for successful teams.


And this approach will be taken for WFC games this December!


Now let´s get into better understanding of what data will be tracked and publicly available from the world championships.


Firstly how do we evaluate possessions? At first we categorize the type of possession (slow, quick, counter, turnover, free hit, no attack). Then we look at events happened during the possession such as cross-field pass on offensive half, odd man rush or clear path opportunities. But what exactly are these?


Cross-field pass on offensive half is a pass that crosses imaginary axis connecting both goals on the field. We track both opportunities (openings) for cross field pass and successful executions. It seems like executing cross-field pass raises the probability of goal scoring.


Odd man rush is a situation in which attacking team has a man advantage on defense in a quick attack (or counter attack).


Clear path is a situation in which offensive player has a clear path to the net and is located in dangerous area close to the opposing goal with a real chance of having or receiving the ball. This clear path can be executed (by shot attempt) or not with a possibility to execute.


To better understand this concept let´s connect our theory with practical examples. Video analysis can show us not only the practice but also ways how to yield clear path situations.


Clear paths are very often connected to defensive mistakes. On two following examples these mistakes fall into category of both tactical and individual mistakes. In the first example we can see that being a winger require quality defensive awareness.


KOTILAINEN gif – min 20 vs SUI (CP created)


Finish winger no. 61 Peteri Kotilainen did not do his job well as he let his player (Swiss no. 16) move into dangerous area. Clear path from organized attack was created and executed by Switzerland.


In another example we can see quick attack from Finland and bad coverage by Czech players leading to another clear path opportunity.


ONDRUSEK gif – min 37 vs FIN (CP created)


Finland executed simple give and go play that can be very efficient. In this case the Czech team had a lot of time to eliminate clear path that was created. In last moment Tom Ondrusek (no. 25) could have catched his opponent to prevent this quick run of Finish defender Savonen (no. 56). This ended in a good scoring opportunity from a clear path.


Clear paths from organized or slow attacks are dangerous and are also usually preventable (as we saw above). There are cases in which attacking team just outplays with its creativity and quality the opposition. Take a look on the example.


GRAF – min 3 vs FIN (CP and goal created)


Swiss defender Luca Graf (no. 9) decided to go for a confident move as he faked the opponent (Finish no. 36) and created little space along the boards on offensive half. Confusion in Finish defense arised. In a split of a second no. 17 from Switzerland remained open in the dangerous area, received a pass from Graf and finish this clear path opportunity with a goal. One may still argue how much this was due quality of offense and how much due defensive mistake.


In other two examples we focus on counter attacks (odd man rushes). These can be created from turnovers where one team transitions quickly up the field or just from plain individual mistake. That is the case of the next example.


KOTILAINEN – min 52 vs SUI (3 on 2 situation)


It is again Finish offensive weapon Peteri Kotilainen (no. 61) who got carried away with a play and let no. 19 Mattias Hofbauer from Switzerland to sneak into the middle creating a solid 3 on 2 situation. This ended up with a miss pass but yielding opportunities like this is a way to go.


In the last example, we have Finish team overcommiting to the offense on their organized attack.


TOO AGRESSIVE – min 25 vs SUI (2 on 1 situation)


Finish team possessing the ball on the left side of offensive half. Finish defender (Talvitie no. 5) decides to continue moving forward after passing the ball to his teammate in the corner. At the same time no. 73 on the other side is also rushing straight to the goal. This puts four finish players to forward movement close to the goal line of opposition in a situation where Swiss team creates a good pressure on the ball. The ball is turned over and the most of the field is defended by single Finish player. Both slow reaction of Swiss players and quick reaction from Fins rushing back to defend prevented this odd man rush from being dangerous at its end.


Using video analysis with data tracked helps us much better understand how these dangerous situations look like and how are they created.


During the WFC games, teams and players will be analyzed using possession-based approach in data analysis. Be ready for it!

By Petr Malina