How much do the big matches matter?
“Win when you’re not playing well: that’s a sign of champions.”
“It’s grinding out results at difficult places like Stoke, Bolton or Blackburn that wins you titles.”
“These are the games that matter: if they lose to their title rival, they can wave goodbye to being champions.”
Evidently, a combination of results wins teams titles. However, it’s worth looking at results between title rivals to see the true importance of these games. Primarily, does the team that wins the ‘league’ between title contenders go on to win the league title, and what conclusions can we draw from ignoring these results in the final standings?
The period in question is the Abramovich era, or the 2003-04 season onwards. The period since 2003 has arguably seen some of the more competitive football in the Premier League in terms of the number of teams involved in title races, aided by the development of the ‘Big Four’ and breaking up for the Arsenal-Manchester United duopoly.
Wishing to compare the results of title competitors, the difficulty comes from deciding which teams were actually involved in a title challenge each season; in this respect the league table can lie. For instance, in 2004-05 Man United finished 18 points off Chelsea. In February, United were second and 8 points off the league leaders, and given their reputation as strong finishers to the league season, it is safe to assume that they were title contenders for most of the season despite their distant final finish. Meanwhile last season Spurs finished 16 points short of being champions, but realistically they were never within touching distance of the league, and so are not considered title challengers for the 2009-10 season.
When comparing the final league table to the league table between title contenders (drawn from results against one another), it’s very easy to look at an 11 point differential and claim “if team x had beaten team y instead of the other way around the title would be completely different.” I don’t intend to do this. Instead, I will subtract all points from the title contender mini-league table from the final league table; a process which I will call ‘normalising’ from here. From this position, we can observe the final league table without ‘big’ matches, or alternatively matches against only weaker opposition, and see whether the games mattered from there. If it’s a little confusing, it should become clear with the help of tables and results to follow.
2003-04
Final League Table
GD | Pts | ||
1 | Arsenal | +47 | 90 |
2 | Chelsea | +37 | 79 |
3 | Manchester United | +29 | 75 |
Chelsea’s first season with Abramovich money faced up to an invincible Arsenal, and they won the league comfortably by 11 points. All top three teams led the league at one stage, and so are all considered title contenders.
Head-to-Head Table & Results
W | D | L | GD | Pts | Ars | Che | MU | ||||
1 | Arsenal | 2 | 2 | 0 | +2 | 8 | Arsenal | 2-1 | 1-1 | ||
2 | Chelsea | 1 | 1 | 2 | -1 | 4 | Chelsea | 1-2 | 1-0 | ||
3 | Manchester United | 0 | 3 | 1 | -1 | 3 | Man Utd | 0-0 | 1-1 |
United were uncharacteristically poor in the big matches, failing to win a single game.
Final League Table Normalised
GD | Pts | ||
1 | Arsenal | +45 | 82 |
2 | Chelsea | +38 | 75 |
3 | Manchester United | +30 | 72 |
After subtracting points and goals from the big games, it’s clear to see that in 2003-04, the big games weren’t all that important. Arsenal still held a seven point lead, and Chelsea would have needed a few wins and results to go their way in order to win the league. Arsenal were so superior against weaker opposition, results against their other title contenders didn’t really come into it.
2004-05
Final League Table
GD | Pts | ||
1 | Chelsea | +57 | 95 |
2 | Arsenal | +51 | 83 |
3 | Manchester United | +32 | 77 |
Mourinho’s first season at Chelsea led to an incredible 95 points, the highest in our range of years. Again, the ‘big three’ were in the title race; United may have finished 18 points off Chelsea, but were as close at 8 points in February with a reputation for doing well after Christmas.
Head-to-Head Table & Results
W | D | L | GD | Pts | Che | MUN | Arsn | ||||
1 | Chelsea | 2 | 2 | 0 | +3 | 8 | Chelsea | 1-0 | 0-0 | ||
2 | Manchester United | 2 | 0 | 2 | +1 | 6 | Man Utd | 1-3 | 2-0 | ||
3 | Arsenal | 0 | 2 | 2 | -4 | 2 | Arsenal | 2-2 | 2-4 |
Chelsea come top of both leagues, going undefeated in the big matches.
Final League Table Normalised
GD | Pts | ||
1 | Chelsea | +54 | 87 |
2 | Arsenal | +50 | 81 |
3 | Manchester United | +36 | 71 |
Again we see a situation where the big games aren’t hugely important. Chelsea hold a six point cushion in the normalised table, meaning Arsenal would have required two wins against their rivals and results to go their way. Chelsea were ruthless against weaker opposition, demonstrated by conceding only 6 goals at home all season.
2005-06
Final League Table
GD | Pts | ||
1 | Chelsea | +50 | 91 |
2 | Manchester United | +38 | 83 |
3 | Liverpool | +32 | 82 |
Arsenal fell out of the title race for the first time in the Abramovich era, with Liverpool taking their place. It’s debatable whether Liverpool were within realistic touching distance of Chelsea, but both the Reds and United improved vastly after Christmas to come close enough.
Head-to-Head Table & Results
W | D | L | GD | Pts | Che | MUN | Livp | ||||
1 | Chelsea | 3 | 0 | 1 | +7 | 9 | Chelsea | 3-0 | 2-0 | ||
2 | Manchester United | 2 | 1 | 1 | -1 | 7 | Man Utd | 1-0 | 1-0 | ||
3 | Liverpool | 0 | 1 | 3 | -6 | 1 | L’pool | 1-4 | 0-0 |
A criticism of Benitez’s early years at Liverpool was a failure to win the big matches, shown here. Only one goal against their title contenders worked against them.
Final League Table Normalised
GD | Pts | ||
1 | Chelsea | +43 | 82 |
2 | Liverpool | +38 | 81 |
3 | Manchester United | +39 | 77 |
Here we see a normalised table that is far closer. Liverpool’s poor performance in the big matches clearly worked against them; home victories would have been a start at least. Even United, five points off Chelsea, could have hoped for better away form, particularly against Chelsea in their 3-0 defeat in late April. It’s fair to say here that the big games, given the closeness of the normalised table, mattered this season.
2006-07
Final League Table
GD | Pts | ||
1 | Manchester United | +56 | 89 |
2 | Chelsea | +40 | 83 |
The first and only genuine two-horse race of the Premier League in our range of years.
Head-to-Head Table & Results
W | D | L | GD | Pts | ||||
=1 | Chelsea | 0 | 2 | 0 | 0 | 2 | Manchester United 1-1 Chelsea | |
=1 | Manchester United | 0 | 2 | 0 | 0 | 2 | Chelsea 0-0 Manchester United |
Two draws, although the second was after the league title had been decided. Both teams fielded weaker teams with the FA Cup final to come.
Final League Table Normalised
GD | Pts | ||
1 | Manchester United | +56 | 87 |
2 | Chelsea | +40 | 81 |
Two wins for Chelsea realistically wouldn’t have been enough; for the first time we can conclusively say the big matches this season made no difference to the destination of the league title, though this conclusion is soured somewhat by the presence of only two teams.
2007-08
Final League Table
GD | Pts | ||
1 | Manchester United | +58 | 87 |
2 | Chelsea | +39 | 85 |
3 | Arsenal | +43 | 83 |
We returned to a three-horse race in 2007-08, and a genuine one at that. Arsenal looked favourites till they conceded late at Birmingham City, a game famous for Eduardo’s leg-break and club captain William Gallas’ full-time antics. Man United and Chelsea were consistently in the hunt.
Head-to-Head Table & Results
W | D | L | GD | Pts | MU | Che | Ars | ||||
1 | Manchester United | 2 | 1 | 1 | +2 | 7 | Man Utd | 2-0 | 2-1 | ||
2 | Chelsea | 2 | 0 | 2 | -1 | 6 | Chelsea | 2-1 | 2-1 | ||
3 | Arsenal | 2 | 1 | 2 | -1 | 4 | Arsenal | 2-2 | 1-0 |
United’s away point at Arsenal – the only away point between title contenders this season – may just have proved crucial.
Final League Table Normalised
GD | Pts | ||
1 | Manchester United | +56 | 80 |
2 | Arsenal | +44 | 79 |
3 | Chelsea | +40 | 79 |
Easily the tightest normalised league table. Results in the big matches were hugely important, and deserved all the billing they could get. We see that United’s ability to secure a 2-2 draw at Arsenal was very important.
2008-09
Final League Table
GD | Pts | ||
1 | Manchester United | +44 | 90 |
2 | Liverpool | +50 | 86 |
3 | Chelsea | +44 | 83 |
Liverpool appeared out of the race until a 4-1 win at Old Trafford, whilst Guus Hiddink revitalised Chelsea towards the end of the season to push United all the way.
Head-to-Head Table & Results
W | D | L | GD | Pts | Liv | MU | Che | ||||
1 | Liverpool | 4 | 0 | 0 | +7 | 12 | L’pool | 2-1 | 2-0 | ||
2 | Manchester United | 1 | 1 | 2 | -1 | 4 | Man Utd | 1-4 | 3-0 | ||
3 | Chelsea | 0 | 1 | 3 | -6 | 1 | Chelsea | 0-1 | 1-1 |
For the first time, the team that won the head-to-head table didn’t win the league. Rafael Benitez had transformed Liverpool’s record in big matches, but ultimately it was the 11 draws, 7 at home, that proved damaging.
Final League Table Normalised
GD | Pts | ||
1 | Manchester United | +45 | 86 |
2 | Chelsea | +50 | 82 |
3 | Liverpool | +43 | 74 |
The normalised table reveals what we probably already knew about Liverpool; it was the ‘small’ games that were important, not the big matches. For Chelsea, it was results in the big games that proved their undoing; three of their five league losses came in these games.
2009-10
Final League Table
GD | Pts | ||
1 | Chelsea | +71 | 86 |
2 | Manchester United | +58 | 85 |
3 | Arsenal | +42 | 75 |
Last season Chelsea won the league, scoring 103 goals, capped off with an 8-0 win on the final day. Arsenal were 2 points off the pace at one stage in March, and so can be considered genuine title contenders.
Head-to-Head Table & Results
W | D | L | GD | Pts | Che | MU | Ars | ||||
1 | Chelsea | 4 | 0 | 0 | +7 | 12 | Chelsea | 1-0 | 2-0 | ||
2 | Manchester United | 2 | 0 | 2 | +1 | 6 | Man Utd | 1-2 | 2-1 | ||
3 | Arsenal | 0 | 0 | 4 | -8 | 0 | Arsenal | 0-3 | 1-3 |
Chelsea became the second team in the Abramovich era to record four wins out of four, an impressive feat.
Final League Table Normalised
GD | Pts | ||
1 | Manchester United | +59 | 79 |
2 | Arsenal | +48 | 75 |
3 | Chelsea | +64 | 74 |
The importance of the big matches in 2009-10 is clear; Chelsea drop to third in the normalised league table. We can pin down the most important result of the season: Chelsea’s 2-1 win at Old Trafford is perhaps the most obvious example in our range of years as being a true title decider.
Summary
In five out of the six three-horse race seasons, the team that won the league title also topped the mini-league between title contenders, but the question is to determine whether the big matches were significantly important. There are mathematical possibilities from the big matches to suggest the title could have changed hands. Realistically, however, this isn’t always true. Thus, 6 points becomes the magic number. In a three horse race (as all but one of the seasons are), if the second-placed team is six or more points off the title winner in the normalised table, it can be considered that the big matches were not important in that season. The reasoning behind this is that being six points off the title winner would have required the second place team to win all four big matches, and rely on other results in the big matches to win the league (i.e. the normalised league winner failing to beat the third-placed team twice). The points swing in this scenario is too large to be considered realistic. Thus:
2003-04: Big matches not significantly important
2004-05: Big matches not significantly important
2005-06: Big matches significantly important
2006-07: Big matches not significantly important
2007-08: Big matches significantly important
2008-09: Big matches significantly important for Chelsea, big matches not significantly important for Liverpool
2009-10: Big matches significantly important
We have a curious split. Clearly, judging the importance of big matches is not an exact science, but if there is a trend to be seen, more recent league seasons have placed larger importance on matches between the title contenders. This perhaps suggests the gap between even the big teams has shrunk in recent years, maybe indicative of lower levels of spending as time from Abramovich’s arrival has passed. Teams are no longer dominant in the same manner as Arsenal’s invincibles or Mourinho’s Chelsea. On the whole, it is fair to say matches between title contenders hold importance, but perhaps not as much as television companies have made us out to believe.
Note: If there are any suggestions, particularly towards the number of title contenders each season, please feel free to comment, and I’ll be happy to revise the tables. Also, many thanks to statto.com for incredibly helpful league tables.
The seasons where the big games didn’t make a difference, although the points made no difference there must be some sort of mental aspect. If they win the big games that has to give them a lot of confidence and belief that they can win the league and surely that must affect the games after?
Whilst obviously this is deliberately a very number/fact based analysis.
There are certainly other important factors to consider other than just the points given out at the end of the game.
From my own experience of playing football, the self belief that comes from beating a title challenger also makes those games extremely important and gives a sort of legitimised confidence which is hard to achieve elsewhere.
I think this is certainly worth considering if you ever attempt a more detailed analysis.
Both legitimate points, but the problem is quantifying self belief and confidence. Perhaps I’ll look at when the matches were played – wins in Feb/March/April followed by runs of form.