So with all three teams dropping points with some regularity, is there any way to tell which will prevail in the battle to win the Premier League? Hopefully there is thanks to a nifty little graph we've created which shows the form of all three clubs on a month-to-month basis since the start of the season.
What we've done is calculate how many of the available points Arsenal, Chelsea and Man United collected for each month since the season started last August and shown the figures as percentages on our graph. So for instance, if during September Arsenal won three games and lost one, that would mean they picked up 9 points from a possible 12 which in turn means Arsenal picked up 75% of the available points for that month… clear? Good.
So here's the graph and the main thing to note is that the red line represents Arsenal, the white line represents Man United and the blue is for Chelsea (naturally).
(click image for bigger version)
And while you're looking at that, let's track how the three teams did throughout each month.
August 2007
Man United's start to the season was a pretty poor one by their standards. Only one win in their first four games, along with a defeat away to near neighbours Man City, left them with a success rate of only 41.7%. Chelsea, meanwhile, won three out of four of their games with only Liverpool stopping them from winning in a 1-1 draw at Anfield. Arsenal played three games, two of which were wins - the other being a draw away to Blackburn.
September 2007
A much better month for Man United. Following their only August win at home to Tottenham, they won all four of their matches in September including a 2-0 victory of Chelsea at Old Trafford. Arsenal also won four out of four, of which two were against London neighbours Spurs and West Ham. Chelsea, however, had an absolute shocker of a month. Only two points from a possible twelve left them with a success rating of only 16.7%, causing Jose Mourinho to leave the Stamford Bridge club.
October 2007
Man United's amazing run of wins in September extended right through to the end of October as well. They scored four goals in all three of their games against Wigan, Aston Villa and Middlesbrough leaving them level with Arsenal on points and goal difference at the end of the month. Arsenal's lead at the top of the table was wiped out when they dropped their only points in October in a 1-1 draw against Liverpool. Chelsea, however, were doing much better under new manager Avram Grant. They won three out of three including a 6-0 win over Manchester City.
November 2007
Chelsea and Arsenal kept their campaigns ticking along as both teams won two games and drew one, giving them a success rating of 77.8%. For Man United, however, November saw a return to the poor form of August. After a hard-fought 2-2 draw against Arsenal, they earned a 2-0 home win over Blackburn before a shock defeat to Bolton left them three points behind Arsenal in the title race again.
December 2007
A busy month for all teams, particularly Arsenal who had a game in hand to fit in along with the six matches already scheduled for the festive season. It was in fact Manchester United who came through it best, though, winning five consecutive games before ending the year with a surprise 2-1 defeat to West Ham. Arsenal won four and drew two of their seven ties, losing only to Middlesbrough, but Chelsea lost further ground in the chase for the title despite losing only one of their six games in December against fellow challengers Arsenal.
January 2008
The new year saw excellent form on the part of all three teams, especially Man United and Chelsea who had a 100% record after their four games. Arsenal managed to win three and draw one during January, Birmingham being the only team to take points from The Gunners all month. Once again, Arsenal's slight slip allowed Manchester United to draw level with them on points, but this time United topped the table due to a superior goal difference.
February 2008
Having clawed their way back to within four points of the leaders, Chelsea saw all their hard work undone during February. Admittedly they played only twice last month, but both games ended in draws at the hands of Liverpool and Portsmouth. Arsenal pulled away from Man United at the top of the table once again thanks to two wins and yet another draw against Birmingham, but Sir Alex Ferguson's men had to endure their third bad patch of the season. A second defeat to Man City and a draw at Tottenham was only offset by another convincing win against Newcastle.
…and that brings us almost up to date. There has, of course, been one match so far this month and that resulted in Arsenal's three point lead over United being reduced to one, but look at the graph. Notice how Arsenal's line is much straighter than Man United's which has a distinctly alpine look. Does that suggest The Gunners are much more consistent and therefore more likely to win the title come May? Can Chelsea make their game in hand count and mount a decent challenge of their own?
Your thoughts, as ever, will be interesting to hear…
As a gooner, I have to say that graph worries me slightly. While the Arsenal line is around 80% the entire time, and the United line jumps between very low and very high, United are currently in a trough. If the trend continues until the end of the season, they will have a 2 month peak now, followed by a low May. And since there are far more games in March/April than May, their predicted low month would come at exactly the right time for them.
ReplyDeleteSo, going from a purely statistical standpoint:
An average United 'high' month gets 95% of points and a 'low' month gets 45%, which would result in-
March: 13 points
April: 10 points
May: 3 points
Ending the season on 90 points
An average month for Arsenal gets 80% of points, which would equal 24 points total, ending the season on 89 points.
*note, I changed some of United's numbers a bit to make them fit possible scores, i.e. they can't get two points from a game, so I lowered March from 14 to 13. Of course, it's all nonsense anyway, so I don't feel too bad about doing it :)
Chelsea have always looked like a 3rd place side. Arsenal have had far too many injuries this season and the Eduardo blow seems to have broken the camel's back. ManU on the other hand are getting healthier and playing as well now as they have all season. It's really a story of two sides moving in opposite directions. Unfortunately, ManU has the momentum and will most likely claim the title. (*string of expletives*)
ReplyDeleteOops, I just realized the mistake I made. I included March's results already, but I should not have.
ReplyDeleteSo, based on Results from the end of February (11 games remaining, Arsenal on 64 points, Man Utd on 61 points).
The update (and I believe correct) version would result in Arsenal getting 80% of their points from their remaining 11 games, which is 26 points out of 33, ending the season on 90 points. (This could come from 8 wins, 2 draws, 1 loss)
Man Utd would get 25 points from their games in March/April (9 games) and 3 points from their games in May (2 games), to end the season on 89 points. (This could come from 9 wins 1 draw 1 loss).
So in this scenario, the season would end with them in the same exact positions they are now (with the first of those 11 games finished).
I'm not sure about the graph, seems plausible enough though.
ReplyDeleteI've been tipping Arsenal to win all season, and will stick with that. There is no statistical foundation for this mind. Just an educated hunch.
I also have a hunchthat Everton will lose to Fiorentina in teh UEFA Cup but finish 4th in the prem.
Please note- my hunches are often wrong.
What a fantastic answer you gave, Adam! I love the fact that you took my statistical analysis into a new direction by trying to predict how Arsenal and Man United would finish. Just imagine if Arsenal ended the season on 90 points with Man U on 89!! I think I'd be straight down to the Patent Office with my methods of calculation, to say nothing of my graph...
ReplyDeleteChris B - you say Man U are getting healthier now and playing better... but people probably said that at the end of October and January. Do you think they're capable of one last dip in form?
Chris c Paul - The graphs just a bit of fun. I like to do this from time to time - analyse some statistics and see what slant you can put on it. All good fun... except the time I analysed some statistics and suggested that Wigan were the worst supported team in the Premier League. BOY did I have some apologising to do that day...
Barring an injury to Vidic or Ronaldo, I don't see ManU slowing. They'll continue to score goals and defend well. They play most of their difficult fixures at Old Trafford, and the toughest road test will be at Stamford Birdge. Arsenal may have to take all three points at Old Trafford or pray that Chelsea can manage a draw.
ReplyDeleteAlso, the Wigan fallout was very funny.
The perspective of the monthly point percentages was a unique and interesting one, so I ran with it. I'd be interesting to add some conditional probability information to that (such as what teams they played in low months or who was injured) but I don't really have the inclination to add that extra layer of complexity. :)
ReplyDeleteI personally think the title will be decided by the match between the two at Old Trafford. Win or draw, and I think Arsenal will pull it off. Lose, and I think the title will go to Man Utd. The wild cards are the two games against Chelsea, but I expect two draws from those matches.
If the season DOES end on 90/89 in favor of Arsenal, that would be mighty impressive.
The important thing to remember is that Arsenal are a diving bunch of cheating twats. COME ON UNITED!
ReplyDelete