The last 5 games of the League are going to decide the season, here we attempt to make sense of the stats from the past 4 years to understand the run-ins for Chelsea, Man Utd and Arsenal.

My previous statistical analysis of Arsenal’s run-in predicted that we would take something like 16 points from our last 8 games, or 65.7% of the points available.

So far we have played 3 of those games, taking 7 points out of a potential 9 – that is 77.7%. Continue that sort of form over the last 5 games and we will end on 83 points.

This means Chelsea must gain no more than 8 points from 15 for us to lead with a 1 point gap, and Man Utd no more than 10 from 15. From 5 games that is a big ask.

Arsenal really cannot afford to drop any points – that much is obvious to anyone. What this post aims to do is follow up the analysis with a breakdown of Man Utd and Chelsea’s record for the final 5 games.

Scores for each fixture from a Chelsea and Man Utd perspective:

		08/09	07/08	06/07	05/06
Chelsea
13 April	Bolton H	4-3	1-1	2-2	5-1
17 April	Tottenham A	0-1	4-4	1-2	2-0
25 April	Stoke H	2-1
01 May	Liverpool A	0-2	1-1	0-2	4-1
09 May	Wigan H	2-1	1-1	4-0	1-0
	Points	9/15	4/12	4/12	12/12
	% Taken	60%	33.3%	33.3%	100%	Av 40.2%
Man Utd
11 April	Blackburn A	2-0	1-1	1-0	3-4
17 April	Man City A	1-0	0-1	1-0	1-3
24 April	Tottenham H	5-2	1-0	1-0	1-1
01 May	Sunderland A	2-1	4-0		3-1
09 May	Stoke H	5-0
	Points	15/15	5/12	9/9	4/12
	% Taken	100%	41.6%	100%	33.3%	Av 68.7%

This table shows a fascinating variety of results in these ties over the past 4 seasons. In total Chelsea averaged taking just 40.2% of the available points, whilst Man Utd averaged 68.7% in from these 5 ties. Just for comparison, Arsenal average 75% against their last 5 opponents over the last 4 years.

Interesting reading, and something that might give hope to any Arsenal fans, of course these are just statistics, but they show the chance of slip ups are high in the last 5 games, for example Chelsea’s results show just 1 win in 4 at White Hart Lane, and the same at Anfield, with none in the last 3 years.

If Chelsea really did just take 40.2% of their points they’d have something like 6 points out of 15 – leaving them on 79, whilst Man Utd taking 68.7% would give them about 10 – leaving them on 82. Arsenal with 75% would take about 11 – putting us on 82 as well – crazy stuff.

In the last post I also looked at the actual points taken over just the last games of the season. So now lets compare the points totals for the run in each season for the 3 clubs.

Last 5 Games Points Total	08/09	07/08	06/07	05/06
Chelsea	15/15	11/15	5/15	9/15
	100%	73.3%	33.3%	60%	Av 66.7%
Man Utd	13/15	10/15	8/15	8/15
	86.6%	66.6%	53.3%	53.3%	Av 65%
Arsenal	10/15	12/15	9/15	13/15
	66.6%	80%	60%	86.6%	Av 73.3%

This shows considerable improvement on the part of Chelsea, whilst Man Utd remain similar. The Arsenal data is based on my earlier post. Again Arsenal are stronger, with 73.3% – a correlation with the 75% from the first bit of analysis. These would give Chelsea 84 pts, Man Utd 82 pts and Arsenal 82 pts (points to nearest whole number). Either of the two averages for Arsenal and Man Utd put them level.

What we can deduce from this is that Chelsea’s results from the games against their remaining 5 teams are quite random, mistakes are very possible, but the data from the final 5 games of every season still shows that they can be consistent. 10 to 12 points is a realistic target for the run in – the team that goes above this will take the title.

These results and the previous analysis also show that Arsenal are performing above the level the data predicted for our last 3 games – if we can continue that then we have a real chance.

This data obviously doesn’t predict any certainties, but I think it gives us an indication of what can happen. The continuity of matches will be important, for example Arsenal having played Barca away now face Spurs and Wigan away and it is very tough to get results over 3 hard away games.

In reality I wouldn’t be surprised to see Chelsea finish very strongly, with Man Utd it will depend on their reaction to the Bayern defeat, and whether Rooney is fit. Arsenal are the underdogs, and like United our reaction to defeat in Europe will be crucial.

The past statistics put us and Man Utd neck and neck on points, with Chelsea either 3rd or 1st – goal difference would keep Arsenal down in either scenario. It could end like that, and it would be very exciting, but then again it could be completely random.

On this basis I think the title will be won with around 85 points –  Arsenal have to win every game. Liverpool and Man City are the best chances for the others to drop points.

This has proved an interesting exercise, the remaining games will show whether these teams buck the trend. This season has had many twists and turns; there is every likelihood that there will be a few more yet.

Update: In this season’s reverse fixtures Chelsea’s run was  WWWWL – 80% and Man Utd’s was WWWDW – 86.6%. Expect something close to this.

Read more articles from our excellent Football Data section.