As we predicted last week, Manchester City lost their home game against Arsenal. We predicted the game to be 0:1 but in fact, the margin is bigger than we expected, which is 0:2. Most of people, even Arsenal fans, has to admit that Manchester City seem to have a better roster than Arsenal does, also Manchester City home game record this season is impressive, 7 wins, 2 draws, and 1 lost before this game. Let us take a look at the game statistically and find the reason why Manchester City lost this key game.
First of all, we want to find the winning pattern of Manchester City game; therefore, we run the EM Algorithm to see how many typical ways to win a game. The result shows that Manchester City is quite a consistent team, and they approximately and only have one way to win, since we only received one cluster in the result based on given data. The result is quite long, so we only drafted some key indicators, such as Shots, Passing Percentage, Tackle Percentage, and Interceptions. Here are the average number of these indicators when they tends to win by one goal.
std. dev. 5.629
std. dev. 0.0369
std. dev. 0.0704
std. dev. 0.0848
Then we will compare these averages to the actual performance in the game against Arsenal.
Here are the numbers for Manchester City against Arsenal,
As we can see, except Interception, all of the rest are lower than the regular performance, especially Shots and tackles. With limited number of shots, normally we can imagine how difficult the team can score, however, let us take a look at the percentage of how the team converts shots to goal and number of shots in the last four games,
We can see that in the last two away games, Manchester City maintains pretty good converting percentage when they have less shots than normal. In fact, based on our other analysis, we conclude one phenomenon about Manchester City, which Manchester City tends to have a better result in the end when they shoot a little, than when they shoot a lot. Therefore, we can exclude that the little shots make significant impact on the team. The little shots, however, limits the number of Chance Created, even though we know that Manchester City is a creative team. Here is the the table of Shots and Chance Created of last four games.
In this table, when the team shoot a lot, they tend to have less efficiency on creating a chance to score, but the total number of chances can be guaranteed to be big. When the team shoot little, however, even though most of the time they can have a better ratio converting Shots to Chances, they cannot guarantee that they can always maintain this high ratio.
Also, the drop of tackles did make an negative impact on the team. As we know, the defense of Manchester City is above average, they were scored two goals six times before this game, and twice for home games. These six games are,
9th home game vs. Burnley
10th home game vs. Sunderland
2nd away game vs. Arsenal
3rd away game vs. Hull City
5th away game vs. WestHam
6th away game vs. QPR
Here are the tackle percentages for these six games,
The average is 0.36. Compared to the Winning average 0.3741, we can come out with this assumption, tackle percentage contributes positively to the team defense, a bad tackle percentage may imply a bad team defense at that time.
In fact, according to our research and analysis, successful Aerial Duel ratio is the other two key indicators to determine how well the team can perform, besides Tackles. We use the same method as we did on tackle. We come out with the similar result.
In conclusion, Manchester City was defeated because of their under-performance on tackles, aerial duels, and their inconsistent ability to convert shots to chances. These weakness is difficult to overcome, so we believe that Manchester City will have a hard time in the coming weeks.