Analyzing players based on numbers, stats and metrics is a difficult task. Especially when most data sources and stats look something like a table shown below.
While there lies a lot of value in using data to identify and analyse player and team performances, it’s just an excruciating task to find out what’s good and what’s bad. The naked eye can(and probably will) get overwhelmed by the sheer quantity of numbers. To put it in perspective, there were exactly 2656 players who played in ISL season 5.
To counter this, we build metrics and visualizations which can make more sense to us in a more macro sense. One such visualization are the Radar Plots which are probably the best in analyzing and profiling players on the basis of their numbers in comparison to the whole of the league.( and bonus points, they look really artsy). An example:
Above is the radar for our beloved Sunil Chhetri, for the 5th edition of ISL. Immediately, a few things pop out: He scores a lot of goals for an attacking winger, Shoots more than most players in the ISL and has a better shot accuracy than most who do. The defensive side, is well… not that great but perhaps that can be expected of a player who’s been given a lot of freedom in attacking in his team.
Bartholomew Ogbeche left NorthEast United to join Kerala Blasters after a superb season. As the radar shows, he’s an elite goal scorer, a high volume shooter and commits to a lot of challenges (both aerial and on ground), winning a decent number of them. ShotTouch% is the ratio between the number of touches and the number of shots he takes. As can be seen, he shoots on sight. Not taking a lot of touches to take his chance.
What’s better, we can make comparisons between any two players:
Here, we can see probably two of the best strikers of ISL 2018-19 in Miku and Ogbeche and the differences between these two are highly visible. Miku is more of a dribbler, more involved in build-up play with elite numbers in assisting while Ogbeche being good on the defensive side of things and shot-taking as well.
There’s lots to gain in terms of player knowledge from these radars. The most important things being: 1. profiling players in terms of what player is good at what and how he fares in comparison to every other player in the league. 2. Identifying similar players that might be wanted in case a vacancy is created.
We at SportsKPI look to provide as much knowledge and IQ as possible from such numbers and visualisations. We are constantly coming up with new metrics such as xG, xA etc. which represent player performances in a better and more consistent manner. All of these can be integrated into the radars to make them more informative and useful for the future. While Radars aren’t the ultimate thing in analytics they’re a great start in terms of profiling and evaluating teams and players. Expect a lot of them in our future work.
Article by our Analyst – Mandeep Verma