N’Golo Kante has always been regarded as a highly-effective midfielder who is more skilled in the defensive aspects of the game. Many have wondered how the Frenchman may fit into this new system with Thomas Tuchel, because it appears that Tuchel favors midfielders who are more effective at using effective, accurate, and simple passes to keep the ball and probe the opposition defense into creating spaces to exploit. With this not being the primary role for Kante in the past, it is a fair question to ask how he may or may not succeed in this new system that the German manager has enacted upon Chelsea Football Club.
This series of articles used a statistical analysis of the data from the Lampard era during the 2020/2021 season to determine how well each midfielder will fit into the new system, and where each individual will need to improve and/or continue to excel. The first article in this series examined Mason Mount, then the focus shifted to the 2019/2020 Player of the Year, Mateo Kovacic. Here, we will be analysing the stats behind the enigma of N’Golo Kante and answer the question of how the Frenchman will fit in with Tuchel’s system.
N’Golo Kante: Individual results and analysis
Basis Descriptive Statistics
N’Golo Kante played a total of 1624 minutes and received six yellow cards. During this time on the field, he registered zero goals and two assists. Kante had 372 pressures, 62 tackles, 42 interceptions, and 47 blocks under Lampard’s management this season. Kante has been characterized – with much opposition at time – as a center defensive midfielder who is best at protecting the backline and neutralizing opposition passes and counter attacks. Others have characterized Kante as a roaming center midfielder who is best utilized in a “destroyer” role alongside another midfield partner, and this is evidenced by the fact both of Kante’s titles in the Premier League came when playing alongside Danny Drinkwater and Nemanja Matic for Leicester City and Chelsea, respectively.
Kante’s defensive abilities seem to have been on full display yet again this season under Lampard, as shown by the above numbers Kante has registered in the defensive aspects of midfield. Kante has not ever been regarded as a goal-creating midfielder, and this is something that has not changed so far this season (zero goals and two assists). Despite a low output, Kante has also not been expected to have much output in goals and assists. Kante’s expected goals are only 0.3, and his expected assists are 0.7. Exceeding his expected assists by 1.3 assists is quite impressive given his role as a sitting DM under Lampard.
Kante completed on average 87.40% of his 1033 completed passes under Lampard, and 74 of those passes were progressive passes, which are defined as a pass that travels 10+ yards towards the opponent’s goal. Kante averaged 3.22 progressive passes per game. Kante recorded 4 shots with 1 shot being on target, and 0 goals scored. This equates to a 25% shot on target percentage, with a 0% conversion rate. Kante carried the ball 900 times under Lampard this season and 80 of those were progressive carries. Finally, Kante completed 24 dribbles prior to Lampard’s dismissal.
Plots and Correlation Analyses
Kante has not been regarded as a player who consistently plays progressive passes that break the lines to create shot opportunities for teammates. However, this has never been the role of Kante, and the only time that argument could even hold validity is if used in the context of the role Kante played under Sarri in 2018-2019. The descriptive statistics support the idea that Kante is not a player who is playing progressive, big chance passes with frequency, but how does that hold up when plotted and analyzed for statistically significant correlations?
The above plot is a scatterplot of the number of progressive passes per game by Kante versus the number of shot-creating actions created per match. This plot was analyzed for statistical significance and for the strength and direction of the correlation. Kante has shown a medium-strength, positive, statistically insignificant correlation between SCAs & Progressive Passes (t = 1.39, df = 21, p = 0.180, r = 0.290). This means that Kante is more likely to create SCAs the more he plays progressive passes to his teammates, however, the trend is not statistically significant, and therefore random chance is a better explanation for the above correlation than any statistically significant interaction. It also is intuitive common sense that the more progressive passes played, the more likely shots against the opponent will occur. However, the key here is that Kante is not creating SCAs relative to his progressive passes at a rate that random chance alone does not explain.
Interestingly, Kante has more progressive carries than Kovacic does so far this season. This is an odd result given the reputations Kante and Kovacic each have as a midfielder. However, is the relationship between Kante’s touches and progressive carries significant? In other words, does Kante have more progressive carries per touch than expected by random chance? This was answered with the below plot and analysis.
The above plot shows that Kante has a medium-strength, positive, marginally significant correlation (t = 1.992, df = 21, p = 0.0596, r = 0.399). This result is so close to being statistically significant that I would argue it essentially is (statistical significance is determined by the p value and is significant when it is less than or equal to 0.05). This means that Kante is creating progressive carries more often than would be expected by random chance: in other words, Kante creates significantly more progressive carries than would be expected each time he touches the ball. This could be a massive positive factor for Kante under Tuchel, if he will be expected to drive the ball forward more so than he was under Lampard.
Another aspect of Kante’s game that should be analysed is his progressive passes versus goal-creating actions (GCAs). For readers that are unaware of a GCA, it is defined as a footballing action that leads to a subsequent teammate creating an assist or goal. So, it is possible for a single player to register more than one (one) GCA per goal scored. This is because a player could dribble past and then pass to a player who registers an assist from a pass, and this scenario would give the player in question two GCAs for the one goal scored. This correlation will elucidate more on the type of progressive passes that Kante plays.
Although this data is not normally distributed, it is still informative. First, it can be seen that Kante’s GCAs have not come from progressive passes this season. This would quickly suggest that Kante was not involved in GCAs under Lampard, and that Kante’s progressive passes are used in other ways than to set up the final pass for a goal. We still analysed any potential correlation and found that there is a medium-strength, negative, statistically insignificant correlation (t = -1.430, df = 21, p = 0.167, r = -0.298).
Finally, we wanted to ask and answer how accurate of a passer is N’Golo Kante prior to Tuchel’s arrival? Below is a boxplot of the pass completion percentage for Kante.
The above plot shows several key pieces of data: the median, 25th and 75th interquartile ranges (IQRs), the lower and upper bounds of the probability distribution function (PDF), and outliers. The IQRs are displayed by the upper and lower (75th and 25th, respectively) bounds of the grey box, and these values correspond to one standard deviation above and below the median value on the PDF, respectively (this is but one way to display and evaluate the variance around a central tendency, which in this case is the median). The “whiskers” of the plot are the upper and lower bounds of the PDF, meaning that values outside of the “whiskers” are outliers, and those are displayed by the unshaded circles.
The boxplot above shows several things. First, Kante’s median pass completion percentage is nearly 87.2%, which is a high percentage. Although this pass percentage is not as high as Kovacic and Jorghino, it is still quite a good pass percentage for a player who was used as lone DM under Lampard and mainly tasked with more defensive duties. For a lone defensive midfielder, this is quite a good pass percentage median. When Kante performs one standard deviation above the median, his pass percentage is around 92.5%, and when he performs one standard deviation below the median, his pass percentage is around 82.5%. This is the most variability of midfielders analyzed so far (which were Kovacic and Mount). Due to this greater variability in pass percentage per match, Kante has not registered any outlier performances.
N’Golo Kante may seem a bit of an odd fit for Tuchel’s system which is based around retaining the ball and using slow build-up play to break down the opponent, as opposed to be more of an anchor at lone DM in the prior management’s system. Kante may struggle in this system to some degree if his pass percentage variability cannot be reduced. This is because the 2 holding midfielders need to find the simple, easy, effective pass with incredible accuracy and consistency. Kante has nearly a 10% variability around his median pass percentage, and that variability could become a bit of a liability within a system that requires such precision and consistent passing from the 2 holding mids.
However, Kante has little to no worry on the defensive aspects of the game in the new system, as he is the best midfielder in the squad regarding defensive output. Kante is averaging 3.22 progressive passes per game, which likely will not need to increase for Tuchel. Overall, Kante will work in this system, but he will have to improve on his ball retention and execution in passing to become a more consistent midfielder who continually plays the effective ball.
Written by Travis Flock @Crossroads_CFC
Edited by Tom Coley @tomcoley49
Follow us on: