clock menu more-arrow no yes

Filed under:

The Stat Man: How many goals are enough?

New, comments

Should Sunderland find the secret ingredient (psssst, it’s pace and early intelligent forward passes) to becoming a more potent attacking force, they will finally become entertaining and unstoppable...

Charlton Athletic v Sunderland - Sky Bet League One Photo by Ian Horrocks/Getty Images

There can’t be many defences in League One better than Sunderland’s. Saturday’s third clean sheet in four games certainly supports that. Especially when you consider that the only goal conceded was from a penalty. This is to be celebrated and bodes well for the promotion push ahead.

It didn’t escape anyone’s notice, however, that we failed to score. And if I bring into consideration that the single goal last week against Peterborough was from a penalty, rather than open play, I might be a little concerned that Sunderland had scored just three open play goals in four games.

But we’ve only conceded one goal in those four games, so that’s alright, right?

I’ll return to that question later.

For now, let’s take a look at how the weekend’s games alter the bigger picture.


Weekend Performances

Below are the performance wheels for the whole league that show just Saturday’s games. Note that Oxford and Crewe’s wheels are from the previous weekend as their game was postponed.

Some sides who seemed to have good weekends (lots of black, green and red) are Bristol Rovers, Milton Keynes Dons, Peterborough United and Portsmouth.

Saturday was much less good for Burton Albion and Ipswich Town, however.

Next we’ll have a closer look at Sunderland’s game against Charlton.

Charlton offered virtually nothing in attack (as anyone who watched the stream will confirm). That they offered so little is due partly to their off the field problems, but credit must also go to Sunderland who, once again, looked impenetrable.

Sunderland’s wheel shows a clean sweep for defensive performance (all red). But the right half gives cause for concern. With some better finishing - and maybe a little luck - Sunderland could have been two or three nil to the good by half time. But the issue here is again a lack of shots, then consequently, shots on target and, ultimately, goals.

Sunderland had ten shots in the game, three of which were on target. Yes, real chances were fluffed by misfiring forwards, but sides that win automatic promotion from League One average more than 14 shots per game, more than five shots on target per game, and close to two goals per game.

A better attacking side would have provided more opportunities for the forwards, misfiring or otherwise.


The Season So Far

Below are the performance wheels for the whole season to date.

Looking in good shape are Blackpool on the top row, Hull on the second row, MK Dons and Peterborough on the third row, and Portsmouth and Sunderland on the bottom row.

Bristol Rovers and Rochdale both did well at the weekend but they’ll need a few more such performances before they can improve their overall outlook.

Most pleasingly, Sunderland (highlighted in red on the bottom row) have almost a clean sweep in defensive attributes over the season. A little behind for defence are Portsmouth (shown at the bottom left).

But if we take a closer look at Sunderland, we can see an area of concern that seems likely to continue to develop. Below shows the progression of the overall average for the season after the Oxford, Peterborough and Charlton games, going from left to right respectively.

The defence looks solid overall but the section at the top right raises concerns that were also present throughout last season - after looking promising post Oxford, the goal threat has dwindled away.

Note also that accurate smart passes (SPA on the green side) are consistently absent. To remind you, a smart pass is “a creative and penetrative pass that attempts to break the opposition’s defensive lines to gain a significant advantage in attack“. It’s something many successful League One sides do to gain promotion.

Lastly, before I return to the original question, here is the performance table for the season as a whole. It takes into account the difficulty of the opposition each side has faced.

The good news is that, for the third consecutive week, Sunderland remain top. However, our old friends Portsmouth are now breathing down our necks having risen to second, only 0.001 behind.

Most surprisingly positioned are MK Dons and Blackpool. Both appear to be performing relatively well yet occupy very low positions in the actual league table. It will be interesting to observe what happens over the coming weeks with those two sides. Both are scoring fewer goals and conceding more goals than their performances suggest they should.


How Many Goals Are Enough?

I expressed concern in the introduction that Sunderland don’t score many goals. But as we’re operating under Phil Parkinson’s guidance, it seems like we’re going to have to get used to that. While a solid defence isn’t going to get fans off their seats (perhaps aside from the occasional perfectly timed tackle), it is nonetheless important and, in the right circumstances, a side can defend their way to promotion.

With this in mind, I took a statistical journey back through the last ten complete League One seasons and created the plot below.

The y-axis is points per game (needs no further explanation I hope) and the x-axis is goal quotient. Goal quotient sounds fancy but it is really as simple as ‘goals for divided by goals against’.

A logarithmic scale is used for the goal quotient axis because it forms a curve otherwise. Curves look nice but they make things more difficult than they really need to be. If there is a quick and easy way to make a straight line, it’s better to use it!

Observing the plot, one can see that the smaller the goal quotient, the smaller the points per game. And conversely, the larger the goal quotient, the larger the points per game.

The red circle I have added shows where automatic promotions happen. It roughly coincides with two points per game, which is the often used rule of thumb measure of how promotion-worthy a side is. In reality, two points per game is very often more than is necessary, but that is a discussion for another day.

Some simple examples of how goal quotient relates to points per game are:

  1. In the 2009/10 season, Norwich City scored 89 goals and conceded 47, resulting in a goal quotient of 1.89. They ended the season with 2.07 point per game.
  2. At the end of the 2013/14 season, Stevenage had scored 46 goals and conceded 72, resulting in a goal quotient of 0.64. This gave them a points per game average of 0.91.
  3. In the 2018/19 season, Barnsley scored 80 goals and conceded 39, resulting in a goal quotient of 2.05. They ended the season with 1.98 points per game.

You get the picture - good goal quotients lead to good points hauls and, ultimately, automatic promotion.

It’s nice to score lots of goals and it’s certainly more entertaining to watch than a solid defensive unit grinding out a narrow win, but ultimately, it’s not necessary. As long you can keep the opposition out of your own goal, then you might not need to score many goals at all.

If Sunderland were to continue on their current rate of scoring an average of one goal per game for the remainder of the season, then they could afford to concede an average of 0.48 goals per game and still achieve an expected 2.00 points per game.

If they were to continue the current rate of conceding an average of one goal every four games, then they could afford to score just 0.52 goals per game. But my goodness that would be dull!

Should Sunderland find the secret ingredient (psssst, it’s pace and early intelligent forward passes) to becoming a more potent attacking force - without compromising the defence too much - they will finally become the entertaining and unstoppable force we were all hoping we might see starting in August 2018.