• Amrit

Shot Selection - The Fault in Our Stars


Can Chhetri and Lanza embark on a game-changing journey that brings them closer.....to goal

Sunil Chhetri is all set to make his 100th appearance for India against Kenya and is already our highest ever goalscorer with 59 goals, so obviously, I chose today to write about the biggest weakness in his game. There was also some gossip doing the rounds last week that Manuel Lanzarote was heading to ATK for $500K, hopefully, I will prove why that is not a great investment.


Anyway, let's look at all the players that took more than 25 shots in the ISL last season. ATK are the only team without a representative - Robbie Keane took only 24 shots and I didn't include him because 24 is not 25.

Lanza is marginally under-performing and Sunil is marginally over-performing and Miku is doing exactly as he is expected to do.

Only Miku, Coro and Kalu Uche outscored Sunil and Lanza. The latter duo played in wider/deeper roles than the others which stands as a testament to how great their seasons were. They were scoring just as many goals and at a seemingly sustainable rate - 0.09 and 0.06 goals above expected per game respectively. This means that they haven't missed easy chances too often and neither have they been excessively lucky.



Digging deeper, if we look at their conversion rates and xG-per-shot (the probability that they score the average shot that they took), we can clearly see Chhetri and Lanza dropping off the top and pure strikers like Alfaro and Jeje taking their place. This indicates that Chhetri and Lanza are taking tougher shots and, logically, scoring at a lower rate.


NP Conversion is the conversion rate of the player after excluding penalties. xG per shot quantifies how easy/hard the average shot taken by the player was

But, if you think about it, they play for two of the best teams in the league. The passing and movement of these teams should be creating easier chances for them, right? If we investigate each shot taken by these players, a trend becomes clearer. Nobody else in the league shoots from outside the box more often than Chhetri, Lanzarote and Marcelinho. In case you were not aware, shooting from inside the box is waaaaaaaay smarter than shooting from outside.



Lanza has taken 39 shots from outside the box and scored 1, Sunil has 22 outside the box and also scored just 1. Miku, Coro, Jeje, Alfaro and Kalu Uche all together have 25 shots from outside the box. What this tells us is that, Lanza and Sunil are jumping the gun and shooting more often than necessary from outside the box. If we compare their conversion rates, Lanza is 2% outside the box and 63% from within it. Sunil is 4% from outside the box and 34% within it. Do you need more proof?


If we simulate their performances using the following formula,


Goals Scored From Outside Box = Shots per Game * Percentage of Shots from outside Box * Avg. xG from Outside Box
Goals Scored From Inside Box = Shots per Game * Percentage of Shots from Inside Box * Avg. xG from Inside Box
Total Goals = [ Goals Scored From Outside Box + Goals Scored From Inside Box ] * Number of Games (18)

The first two formulae are basically shots per game from each zone multiplied by the quality of that chance and tell us the average number of goals each player will score from inside or outside the box per game. The third extrapolates that over the season. It then becomes a "lazy" way for us to compare the goals scored by each player over the next season. Ideally, I'd simulate each shot many times and see how many goals they would have scored on average but....time constraints.


Top goalscorers next season,


All players on that list, except Jeje and Alfaro, scored at over 0.5 non-penalty goals per 90 minutes last year. Obviously, very few players can maintain those kind of numbers/form and it is statistically not repeatable year-on-year, except if your name is Cristiano Ronaldo or Lionel Messi. Understandable then, that my model expects everybody to score fewer goals but right at the bottom of that list are Chhetri and Lanza.


To conclude, Sunil Chhetri and Manuel Lanzarote are coming off excellent seasons and are amongst the best in the league in Goals scored (and Expected Goals), however, they needed almost twice as many shots as the players around them to get those numbers. Their shot locations are predominantly in low value zones and as such this "selfishness" could be a detriment to both their team-mates' output and the results of their respective teams. Chhetri is the best Indian goalscorer but he could get even better if he improved at choosing when to shoot. The stats say that Jeje, for example, is more of a shooting threat but probably doesn't get/take as many chances,


Furthermore, Lanzarote to ATK at $500k could backfire, especially when they have already spent big on Kalu Uche. This could tie them up tactically, and force them into playing two players that might not necessarily perform. Kalu Uche had 4 penalties in his 13 goals and will just about get you 10 goals a season. I have already demonstrated that Lanza's goals might not be sustainable. He is primarily a creative player but there are concerns on that front too - he had 6 assists last year, 5 of which were rolling it into the box for Coro, a player he is no longer going to play with.


All stats from my personal database of the 2017-18 Indian Super League. Please send your thoughts/feedback to @sgtsaltnpeppa

  • White Twitter Icon