clock menu more-arrow no yes mobile

Filed under:

The Manchester Derby: Form by Numbers

The 163rd Manchester Derby takes place today. There are many intriguing battles between players and tensions, quite laughably, over penalties between the two clubs managers. But which club is in form, and do the underlying numbers of that form give us any indication as to who might be favourites for the game?

Anyone for 2-0 to City?
Anyone for 2-0 to City?

So, we are just hours away from this crucial derby now and we know that Man united lead man City by 3 points, this is the all important number. But how have the respective teams gained their points totals so far; have one side been better, the other luckier? Let's find out

The Fixture List

Both teams have played 15 games, Man United stand on 36 points and City on 33 points, but what has the fixture list been like?

Lets use a little something I came up with at the end of last season, it's called quality of opposition, it's not perfect but it gives an idea of how difficult a teams fixtures have been. The basic premise is to take the oppositions league position at the time of every fixture, add them all up and we get a number, we then weight this number, so a game against the number 1 ranked team in the league is given a value of 20, the most difficult game.

So, the higher the Quality of Opposition average total is, the more difficult the schedule. The higher the number the harder the schedule.

Man City Man Utd
10.83 9.43

This is pretty basic, it gives us a basic idea of the difficulty of a teams schedule so far, it doesn't factor in who the opposition is or our ideas of their traditional strength or weakness, it just takes their league position and how well the opposition is doing in terms of form.

United have had an easier schedule so far when we look at it by this method. United have had easier away games against a struggling Liverpool, Southampton, Reading, Norwich and Villa.

So it's City with the slightly harder schedule and sit just 3 points behind in the table.


Man City Man Utd
59.20 59.60
3rd 2nd

I make no great claims about possession, for without penetration or skill in the final 3rd possession becomes, by reduction, just another method in which to restrict the oppositions scoring chances.

United just shade it, though. Both numbers are excellent and a shade behind Arsenal.


Man City Man Utd
shots 266 238
against 142 176
total 124 62
total per game 8.26 4.13

So, here we see total shots.

Shots for are pretty similar, with City just edging it yet United's propensity to give up shots to the opposition at a higher than Man City do is what hurts them. If we look at each teams shot +/- per game, there really is only one winner and that is Man City. City are twice the shots +/- team that Man United are, luckily for Man United I don't put too much stock into evaluating total shots numbers.

Shots On Target ratio

Here is a quick look at the respective shots on target ratio's of the Manchester clubs. This is each teams moving average as the season has progressed. The shots on target ratio is basically a team's share of the shots in each game.


Only one winner.

Shots On Target

Below we see City's shots on target vs shots on target against as the season has progressed. We see a wide, and increasing gap between for and against as far as City are concerned, and after the 15th game Man City have a +58 shots on target number, or +3.86 per game. This is a very very good number, the league's best in fact.


But what of United's number?

Man United

Just from a simple visualization we can tell that Man United aren't the same excellent Sshot on target side that Man City are. There exists a gap between for and against, and that gap is widening but it is much narrower than their cross town neighbours numbers are. Man United are +31 in shots on target, which, in terms of a per game number is +2.06 per game. A number which is still very good, but nowhere near Man City's number.


This begs the question: If Man United are a poorer shot team, how have they managed to record so many points?

Maybe scoring and save %'s can give us a clue.



Scoring% (goals/shots on target for)

Now then. We know United have recorded less shots on target than Man City, but have managed to score 37 goals to City's 28. A very very high scoring% is the reason Man United have been able to score so many goals and rescue so many points.

League average scoring% is around 31%, so Man City are slightly below league average(which won't continue) and Man United are 10 percentage points above the league average. Man United's number is around 35% better than Man City's, do we believe that Man United are 25% better in terms of scoring goals than Man City or do we expect Man United's number to regress and thus fall from 41%? I'd say there is an element of luck at play here.

If Man City had had United's scoring % they would sit at 39 goals scored. If Man United had City's scoring% they would sit at 26 goals scored and PL points totals would look very different.



Here, in the above chart, we can see regression in action. City started from a poor early season save % and have regressed positively toward the mean, whilst United's save % has been regressing toward the mean since game 1 and has now dropped below the league average of just short of 70%. Again, elements of luck, and the quality of chances play a part in this save% number.

In 2011/12 the Manchester clubs were the two best teams in the league. Will they be the top two teams this term?




Pdo is a teams save% and scoring% added together. And it is here that we finally see a little parity between the two Clubs. City have taken a long long time to to get to the mean of 100, whilst United have been regressing toward that mean for 10 games or so now.

The First goal

Man City Man Utd
y n
n n
y n
n y
y n
n n
y y
y n
y y
n/a y
n n
y n
n/a n
y y
n n

Man United have scored the first goal of the game on just 5 occasions compared to City's 8 occasions. Why is this important? The expected chance of a win when scoring the first goal is 59.85%, all good so far, and we expect the Manchester teams to above that figure, which they are.

So what about the occasions when the Manchester teams did not score the first goal? City fell behind 5 times and United 10 times. United's number is worrying at first sight, for teams who fall behind have a win % chance of just 14.08% in the Premier League (a figure which drops to 9.84 without United). Man United, having fallen behind on ten occasions, have gone on to win 7 of those games, seven!

This is an incredible number, an unsustainably high number and one which is completely at odds with United's 2011/12 season, in which they never won having conceded the first goal. Comeback wins doesn't seem to be a repeatable skill, it varies from year to year despite relatively small changes to team personnel and strength.

Actual wins expected
7 1

Put simply, the above graph illustrates how many games United have won versus what the league average team would have won. This United team are either very good (they can score, alright) or very lucky.


As we have seen above, the two Manchester clubs are both very strong teams, but both have achieved their points totals in differing ways.

Man United have an excellent scoring % and Man City do not.

City's save % is slightly better, and City can also boast a team and tactical shape that restricts opposition shots on target at a far better rate than Man United, who are really leaking shots on target and which their save % is not helping them to prevent goals.

Both teams are excellent shots on target teams with the 1st and 3rd highest numbers in the league, but it is United and their ability to score at a very high rate, in terms of goals to shots on target, which has enabled them to score so many goals. And if Man United didn't boast such an excellent scoring % would they have been able to score their way to 7 victories out of 10 when falling a goal behind? Very unlikely.

For Man City the task seems straight forward: Prevent United from attacking in wide areas and getting shots on target, from which they score at a an alarmingly high rate, and secondly attack Man United's weak shape and take advantage of United's charitable defense and poor save%.

There should be goals, but Derbies tend to throw surprising results at us.

I feel quietly confident.

We shall find out which team can expose the others weaknesses soon enough.