Chelsea's short victory road

To the astonishment of most Chelsea's fans, the current 10th position on the league table is hard to digest. As an aspiring football data-mining scientists and a Chelsea fan, I decided to take some time off in my final semester to not just code the football statistics but to compile a mini-report here to analyze on Chelsea's victories so far in the 2015/2016 as of 21st April 2016.

A far cry from last season's title wining form, victories for Chelsea have been rare. Still, I believe it is worthwhile to relate several associations to the Blue's wins this season; in short, factors like 'Days apart between matches', 'goal conversion rates from total attempts' and 'timing of goals' will be discussed to shed light on possible improvements in the near future of squad assembly and understand the trends involved in Chelsea's victories so far. Based on statistics and match report from the online website of Chelsea Football Club, I am able to gather my own observations to better understand these trends; in a nutshell, Chelsea have won 10 home games and 9 away matches across all competitions.

To clarify some caveats, the matches studied in this post entails independent home and away matches.

Days apart between matches

To begin, days apart between matches have been widely discussed as a vital factor in contributing to winning match since players do have more time to rest prior to the game and have more training sessions. On the contrary, some critics have claimed that there is a trade-off in losing their positive match form.

It was discovered that the range of days apart between the last match and a victory is 3 to 14 days; to elaborate further, the range of days apart between a last match (win, lose or draw) and a victory in a home or away setting is very similar. In a home setting, it ranges from 3 to 14 days whilst in an away environment, it ranges from 3 to 13 days. As such, this implies that there seems to be no differences if the players require more time to rest or need to play successive games to maintain their form.

Goal conversion rates from total attempts

As opposed to relating goal conversion rates from shots on target, this post seeks to translate goal conversion from a wider pool of total attempts in a match.

The goal conversion rate = Goals scored by Chelsea / Total Attempts x 100

Indeed, while I personally find that this equation is rather simplistic, it has many uses. In order to sharpen the goal conversion rate, Chelsea either has to score a lot goals with the least number of attempts. At times, the notion of the 'one shot one goal' may apply in certain games, especially in games where Chelsea is dominated by a stronger technical team like Barcelona. More importantly, this equation reflects how clinical the scorers are. Even so, chances or attempts must be created to score a goal based on the premise of this goal conversion rate.

From what I have gathered, the total attempts in Chelsea's winning matches varies across 8 to 22 chances, this is also reflective in victories at Stamford Bridge. This is as opposed to their away victories which is more aggressive; the total attempts are 11 to 24. As such, this may imply that there might be lesser pressure away from home matches enabling the players to take more shots.

With regards to the goal conversion rate across both home and away matches, it ranges from 5% to 36%. To elaborate in a succinct manner, 36% suggests that for every 100 attempts in a match will translate to 36 goals. This spread from a mere 5% to 36% suggest that there might be cases where the Blues are wasteful at goal or even clinical at times. In this regard, it is hard to predict their goal conversion rate prior to the game as chances are largely isolated within the commencement of the game itself.

Interestingly, there is an unique case where Chelsea managed to win a game despite having a lower goal conversion rate than their opponent; in a game at St' Mary, Southampton had a slight advantage of its goal conversion rate (14.3%) over Chelsea (13%). In this match, Chelsea had to come from behind from a Shane Long's goal to win through goals from Fabregas and Ivanovic.

Moving on, it is noteworthy to analyse the moving average of Chelsea's goals in their winning matches across their home games. Since the moving average line enables us to detect trends, to follow progress and report reversal, the below two graphs seeks to highlight the divergence of goals scored in independent matches with the moving average. To put it simply, in theory, as long as any points touches the moving average line, it would tend to suggest a start of a new trend (decreasing or increasing goal tendencies).

For the home game victories, the moving average line implies that Chelsea is on a roll scoring above the moving average lately. However, it is noteworthy to realize that their last win at home was roughly 2 months ago in a 5-1 victory over Manchester City.

Pertaining to the away form, the moving average line differs; even so, the upward curve towards the later games suggest that Chelsea may have started to get a knack of scoring more goals in the later part of the season; this trend can be said for the home game victories as well.

Timing of goals

Among the factors, I personally felt that timing of the goals which forces the lead is essential is securing a victory.

Based on the two graphs, it is apparent that scoring within the 1- 15mins period is significant in victories. Comparing both graphs, the fact that almost half of the leading goals scored in away matches lies within the 16-30 min imply that there is a tendency for the Blues to absorb the pressure of playing away or even maybe altering their tactics in the opening minutes. Based on the 75 - 90 mins period on both graphs shows that Chelsea do score last minutes goals in an attempt to secure a win; this is rather optimistic in the sense that the players have the 'winning' mentality.

Despite this, Chelsea should focus on winning games up till the 75 mins - be it home or away - because majority of the goals stem from before that particular minute. One implications in terms of squad rotation could be that they should start with big impact players from the start to stand a better chance in winning; it is clear that they tend to score lesser goals towards the end. And if you split the timing of goals into 1st and 2nd half breaks, it is evident that most goals (home and away) are nicked in the 1st half.

In short, these are some of the factors which contributed to Chelsea's victories this season so far. Undoubtedly, there are key factors which are not covered as well. Still, I will slowly uncover trends through data mining of football.