HOT TOPIC: A lack of rotation could hurt Arsenal’s and Aston Villa’s chances
Sun 30 Mar 2008, 13:58
When investigating the pros and cons of rotation in The Football Review 2007 we discovered, among of things, that there was no evidence of a relationship between first half rotations and second half success in the league. We have since revisited this topic to investigate whether any benefit is gained from rotating players regularly throughout the season and the answer is in fact yes. Using the rotation data we collected last season we discovered that there was a significant relationship between the total game-to-game player changes club’s made in the first 28 games of the season and the number of points they won in the last 10 games of the season (Table 1.1).
Table 1.1 Total changes after 28 games and points won in last 10 games during the 2006/07 season
| Team |
Total player Changes after 28 games |
Points won in last 10 games |
| Man United |
81 |
20 |
| Arsenal |
75 |
16 |
| West Ham |
75 |
21 |
| Newcastle |
74 |
7 |
| Liverpool |
72 |
15 |
| Chelsea |
65 |
20 |
| Charlton |
64 |
11 |
| Sheff United |
61 |
8 |
| Tottenham |
61 |
21 |
| Man City |
60 |
12 |
| Middlesbrough |
60 |
11 |
| Portsmouth |
60 |
13 |
| Blackburn |
54 |
15 |
| Fulham |
54 |
7 |
| Everton |
53 |
16 |
| Watford |
52 |
9 |
| Wigan |
51 |
9 |
| Aston Villa |
48 |
17 |
| Bolton |
39 |
9 |
| Reading |
30 |
12 |
The correlation coefficients between these two variables was 0.38 indicating a fairly strong positive relationship meaning the more a team rotated in the first 28 games of the season the higher their points total was seen to be in the last 10 games of the year. Even when allowing for the different strength of fixtures faced by each club in the last 10 games of the season the relationship between the two variables was still observed to be significant.
So what can that tell us about this season? Well, Table 1.2, below, gives the total game-by-game changes by each team after 28 games this season and their accompanying points totals so far in the last ten games of this season.
Table 1.2 Total changes after 28 games and points won in the last three games this season
| Team |
Total Changes in first 28 games |
Points won in last 3 games |
| Liverpool |
90 |
6 |
| Chelsea |
83 |
7 |
| Derby |
80 |
0 |
| Tottenham |
76 |
4 |
| Man United |
73 |
9 |
| Sunderland |
66 |
1 |
| Fulham |
66 |
4 |
| Newcastle |
59 |
4 |
| Everton |
58 |
4 |
| Middlesbrough |
52 |
5 |
| Man City |
50 |
4 |
| Blackburn |
50 |
4 |
| Portsmouth |
49 |
6 |
| Bolton |
49 |
1 |
| Birmingham |
49 |
1 |
| West Ham |
47 |
4 |
| Arsenal |
46 |
2 |
| Wigan |
43 |
4 |
| Reading |
40 |
6 |
| Aston Villa |
28 |
1 |
The results so far are very similar to last season with the two variables having a correlation coefficient of 0.30. Even though the points won sample is only from three games it does give an indication that the same pattern might be occurring this season. The most visually alarming finding is seeing Arsenal and Aston Villa in the bottom four in total player changes after the first 28 games of the season with the knowledge that both are currently displaying struggling form. For both clubs their decision to not rest players occasionally throughout the season could end up being very costly with the title and UEFA cup contention, respectively, fast disappearing from view.