Goal Difference, does it matter?

Historic Statistic Study

I'll start with the historical stats. I'm looking at all of the current 92 Football League (FL) teams and the time that they've spent in the FL since the 2003/04 season. 

So if they have spent any time outside the FL in that period then their goal statistics for those years have been discounted. I believe the last 11 seasons covered represent a good dataset as the modern era.

Football has changed a lot over the years and I believe that any time before the last decade could start to skew the data beyond its validity to be an indication for the potential outcomes of modern-day matches.

Season average GD & (average finish position) since 2003/04 in Football League tables.

• Accrington -10.63 (15)
• AFC Wimbledon -15.33 (17.33)
• Arsenal +36.00 (3.27)
• Aston Villa -4.18 (10.45)
• Barnsley -12.36 (17.09)
• Birmingham -3.36 (11.55)
• Blackburn -7.73 (12.45)
• Blackpool-2.27 (13.82)
• Bolton -7.55 (11.36)
• Bournemouth +5.00 (11.55)
• Bradford -4.91 (14.00)
• Brentford +2.36 (8.82)
• Brighton +2.64 (11.27)
• Bristol City +0.82(10.55)
• Burnley -1.27 (12.18)
• Burton -6.00 (12.00)
• Bury -2.91 (12.45)
• Cambridge -16.50 (17.50)
• Cardiff +5.91 (9.73)
• Carlisle -3.70 (12.90)
• Charlton -2.00 (11.82)
• Chelsea +43.00 (2.36)
• Cheltenham -9.55 (14.73)
• Chesterfield +2.91 (12.00)
• Colchester -1.09 (12.55)
• Coventry -7.36 (17.00)
• Crawley +5.67 (9.00)
• Crewe -8.91 (+16.64)
• Crystal Palace -0.64 (12.36)
• Dagenham -5.57 (15.14)
• Doncaster -0.45 (11.55)
• Everton +8.36 (7.36)
• Exeter -5.83 (13.17)
• Fleetwood +6.00 (8.50)
• Fulham -10.09 (12.18)
• Gillingham -7.55 (14.36)
• Hartlepool -4.18 (14.36)
• Huddersfield +6.82 (9.00)
• Ipswich +1.73 (10.91)
• Leeds +3.36 (11.09)
• Leicester +9.09 (11.09)
• Leyton Orient -2.27 (12.45)
• Liverpool +27.91 (4.64)
• Luton -3.50 (15.33)
• Man City +18.64 (7.64)
• Man Utd +43.27 (2.18)
• Millwall -1.64 (12.91)
• Milton Keynes +11.50 (9.00)
• Morecambe -3.00 (13.86)
• Newcastle -1.91 (10.27)
• Newport -3.00 (14.00)
• Northampton +1.18 (12.36)
• Norwich -4.55 (11.64)
• Notts County -3.00 (15.91)
• Notts Forest +6.45 (10.55)
• Peterborough +4.64 (12.55)
• Plymouth -3.91 (15.64)
• Portsmouth -11.73 (15.45)
• Port Vale -1.45 (12.09)
• Preston +2.73 (11.36)
• Oldham -3.36 (13.82)
• Oxford -0.57 (12.00)
• QPR -1.00 (12.00)
• Reading +11.55 (8.00)
• Rochdale +6.82 (10.45)
• Rotherham +1.18 (12.64)
• Scunthorpe -1.55 (13.73)
• Sheff Utd +6.91 (8.55)
• Sheff Wed -1.18 (13.64)
• Shrewsbury +6.55 (11.18)
• Southampton +7.64 (11.45)
• Southend +4.64 (10.36)
• Stevenage -0.14 (7.86)
• Stoke -3.73 (10.73)
• Swansea +11.64 (7.00)
• Swindon +8.64 (10.55)
• Sunderland -5.27 (11.18)
• Tottenham +9.36 (6.91)
• Tranmere -1.18 (12.55)
• Walsall -3.18 (14.27)
• Watford +0.27 (12.09)
• West Brom -0.27 (10.09)
• West Ham -3.45 (10.55)
• Wigan -9.45 (11.91)
• Wolves -2.18 (11.36)
• Wycombe -0.64 (13.18)
• Yeovil -4.64 (12.64)
• York -10.00 (16.00)

Now obviously some sides may have spent season after season near the top half of a division and failed to get promoted so they might have quite a high average league position and goal difference which needs to be taken in consideration if they have just been promoted to the higher division; for example, Brentford have a +2.36 GD average and an average league position of 8.82 but they are now in the Championship so have a tougher task on their hands (they seem to be taking to it very well though). 

Similarly, teams like Portsmouth have fallen from the Premier League to League Two so their average league position and GD are very poor but now they are in League Two you would expect that to improve amongst poorer quality teams.

Since 2003/04 the average goal difference of the top sides in the Premier League has been declining. Chelsea managed to concede just 15 goals in 04/05 but since then they've conceded more and this has pushed their average goals conceded up to 28.91. Similar trends appear with Manchester United, Arsenal and Liverpool. I would say this shows that the Premier League has become tougher than ever. Though some sides have bucked the trend, Man City's new influence has seen them score more than ever and concede less than ever. So we must take that in to account too. 

What I wanted this research to achieve was a source for reference when finding anomalies in the current season. Say I came across a team in the bottom half of the Premier League but with quite a good GD. I might want to know if history states that this GD is normal for them and that therefore their position in the table should improve or if they tend to average a poor GD and that their current GD should not mean they are worth backing in future weeks. Obviously this should all be taken with a pinch of salt. Things change from season to season that will have a far greater impact than GD but these stats could provide a good insight in to some fixtures and give us that edge we need in order to find winning bets.

In a regimented world goal difference would define how the table finishes; Teams would finish in numerical order of positive to negative GD. But as we know football and the world in general is quite chaotic so not everything goes to plan and quite often teams with a negative GD finish way higher than some teams with a positive GD. As an example, in 2003/04, Bolton finished 8th in the Premier League with a -8 GD but Man City finished 16th with a +1 GD. What is the reason for this? Was it just that Bolton were thrashed a couple of times and otherwise won narrowly or were Man City unlucky to lose my slim margins all season? 

GD tends to level itself out over the season and I believe that most of these anomalies diminish over the course of a season. In most cases, a team that scores more goals and concedes less will finish higher up the table. I think we can apply this theory to the current Football League tables; if we highlight the teams in the top half with the worst GD and teams in the bottom half with the best GD we could find good reasons to back or oppose these teams for the rest of the season. Sometimes this methodology will fail; as opposing Bolton and backing Man City would have in 2003/04 but I think more often than not it will provide a good indication that a team deserves to be doing better or worse than they are and therefore gives us clues from a betting perspective.

Current Stats

As we are only almost half-way through the Football League season some of these goal difference figures may be less accentuated than normal; a team at the foot of the table could expect their negative goal difference to be doubled come the end of the season if performances continue in the same vein and vice versa for teams at the top of the table. With that in mind, I will be mostly looking at teams in the middle of the table between 4th and 20th as their GD is less likely to change as dramatically and instead fluctuate around the same +/-. However, if teams catch my eye with a particularly low or high GD in relation to their standing then I will investigate out of interest.

The following is a list of the highest placed and lowest placed teams in relation to their GD of the current FL tables, and I think this demonstrates that these sides are potential over or underachievers, i.e. a team with a negative goal difference higher up the table is probably outperforming their potential and should be lower in the division and vice versa with clubs with a positive GD.

Premier League -

Overachievers;

Tottenham 7th GD -2

Newcastle 8th GD -4

Underachievers;

Crystal Palace 16th GD -5

Hull 19th GD -8

Championship -

Overachievers;

Brentford 5th GD +7

Wolves 12th GD -7

Underachievers;

Bolton 18th GD -6

Wigan 23rd GD -7

League One -

Overachievers;

Sheff United 5th GD +3

Notts County 7th GD +5

Underachievers;

Gillingham 18th GD -6

Leyton Orient 22nd GD -6

League Two -

Overachievers;

Accrington 11th GD -3

Cheltenham 18th GD -10

Underachievers;

Cambridge 12th GD +7

Tranmere 19th GD -6

Many of those teams are surrounded by sides with a far worse or far better GD. Does that indicate that they themselves are either underachieving or overachieving and could be likely to lose or win a lot of their upcoming games? I would say it's a pretty good indication that a team with a low positive or even negative goal difference in the top half is likely not to keep getting their results and vice versa with teams with a low negative or positive goal difference in the bottom half are likely to get a few good results in their upcoming fixtures.

In some cases after identifying these trends and finding a team it looks like is worth backing you might still be reluctant to back the team given they have injuries or suspensions or management issues etc. but at least you can see they should not be opposed and can, therefore, avoid losing bets.

Now let's look at these underachievers and overachievers this season and compare them to their previous seasons. Tottenham are 7th this season with a GD of -2. Spurs have averaged 6.91 in the final standing of the last 11 seasons of the Premier League. So currently sitting as close to their average position as possible but their GD is quite different. Spurs average +9.36 GD but their current GD of -2 is quite far from that and this suggests to me that their current position is ‘false'; they are likely to finish lower than 7th if they are scoring less and conceding more on average than they have in the past. 6 of their 7 Premier League victories this season have been by a one goal margin, with just the one 4-0 victory over QPR. 

Are they overachieving with their position in the table? Probably so. But this won't necessarily be reflected in the odds. In fact, this weekend they are 8/15 on to beat Burnley at home. That tells me that there is a lot of value in Burnley to win or draw the game. It is 10/3 for the draw and 6/1 for Burnley to win. Combine that with other trends like Tottenham's poor home record this season and you have a nice looking bet. Burnley double chance at 13/8.

Also potentially overachieving this season is Newcastle United. They are 8th with a GD of -4. Historically they have averaged a GD of -1.91 and finished 10th so we could expect a dip in their form a little in line with recent seasons. Crystal Palace are 16th with a GD of -5. At first sight, you would think that there GD indicates they are underachieving for their league position. 

Historically they've averaged 12th and a GD of -0.64. Though some of those seasons were spent outside the Premier League you would think they are likely to climb the table and pick up a few better results. Hull looked to underachieving at first look in line with their GD. They are 19th with a -8 GD but their historical stats show that they finish around 11th and average -1 GD. This is quite far away from their current form and so that stats aren't that helpful in this situation.

In the Championship, at first look, the potential ‘overachievers' are Brentford and Wolves. Brentford is 5th with a GD +7 and Wolves are 12th with a GD of -7. On this occasion, the data tells us that Brentford averages a GD of +2.36 and finish 9th and that their position is actually probably a true representation of their historical performance and therefore we should not oppose them as ‘overachievers'. 

Wolves have averaged a GD of - 2.18 and finished 11th. They are currently just one placed below that but their GD is quite a lot worse and this tells me that they should be opposed as they are probably higher in the table than their goal scoring record dictates. Underachievers at present take the shape of Bolton in 18th with a GD of -6 and Wigan in 23rd with a GD of -7. Wigan are surrounded by teams with a far worse GD and historically they have averaged -9.45 GD but finished 12th. If they continue with their current goal scoring record it is likely that their league position will improve as they are mostly losing narrowly. Bolton follow suit with an average GD of -7.55 and a finishing position of 11th. Therefore I'd expect them to move up the table in relation to their GD.

On to League One. Overachievers; at present Sheff United GD +3(5th) and Notts County GD +5 (7th). Sheffield have averaged a GD of +6.91 since 2003/04 but finished in 9th place on average. 

Their goal difference is currently worse than their average but their league position is higher; this makes me think they will drop down the table quite considerably. County average a GD of -3.00 and 16th place, this shows that they are having a better season than normal and that the historical stats are probably not a good indication of their potential in 2015. The ‘underachievers'; at present Gillingham GD -6 (18th) and Leyton Orient GD -6 (22nd). Gillingham average -7.55 GD, worse than their record this season, but a finishing position of 14th which is 4 places higher than their current standing. I would expect Gills to climb the table. Leyton Orient average -2.27 since 03/04 and a finishing position of 12th. 

Their current position of 22nd is probably a bit harsh on their goal-scoring record; most relegated sides record a GD of around -20 for that particular season. So I would also expect a few better results for Orient.

In League Two. At a glance the overachievers appear to be; Accrington GD -3 (11th) and Cheltenham GD -10 (18th). Stanley have average a GD of -10 and finished 15th in their last 8 seasons of FL football. When comparing this season to the historical stats Accrington appear to be pretty much in line with their standards so I would neither suggest backing or opposing them to any great length. Cheltenham Town average -10, the same as their present GD, but finish 15th on average so I would expect them to also float around the same position.

Underachievers; Cambridge GD +7 (12th) and Tranmere GD -6 (19th). Cambridge have only been in the FL for 2 of the last 11 seasons and those were 03/04 and 04/05 so their historical stats are not that relevant to the present day but their GD does suggest they should be higher in the table when you think of the average placing of a team with a +7 GD it is definitely higher than 12th. Tranmere average a GD of -1.81 and 13th place. 

This suggests that their GD of -6 this season should have them placed more around the 15th or 16th place not their current standing of 19th.

Teams to back

In light of all of this meticulous research, I would suggest that backing Cambridge, Tranmere, Gillingham, Leyton Orient, Bolton, Wigan and Crystal Palace could be quite a profitable avenue for the remainder of the season.

Without being foolish though, putting it in context, as backing them against the top sides would not always be wise but I would imagine if they perform to their potential and keep up their promising goal-scoring records they should get some positive results over teams around and below them.

Teams to oppose

On the other hand, the research identifies that betting against Sheffield United, Wolves, Tottenham and Newcastle could prove fruitful in the coming months. 

Similar to the teams I have suggest backing, it would be wise not to oppose these sides when they are playing teams from the bottom 4 or so but it could be worth opposing them when they play sides around or above them.

So all of the above is just an example of how you can use goal difference to pick out some quite nice information and this could give you some quite attractive betting angles. Of course, I will reiterate again, at the end of the day stats are just stats and anything can happen on any given day but in the long run these things tend to level themselves out and I would suggest that given the right system you will win more than you lose when adding more analysis to your repertoire.

We can also use these statistics to identify some other fun trends like who has the worst GD in relation to their finishing position, i.e. who have been the ‘luckiest' team in recent years and also the teams with the best GD in relation to a low finishing position, the ‘unluckiest'. Northampton, Peterborough, Rotherham, Chesterfield and Watford have all averaged a league position of lower than 12th place but registered a positive GD. Watford GD +0.27, Rotherham +1.18, Northampton +1.18, Chesterfield +2.91 but the unluckiest of them all, Peterborough, have averaged a GD of +4.64 but finished on average in 13th place. 

Aston Villa has a finishing average of 10th place but an average GD of -4.18 so they can be considered pretty ‘lucky' to have finished around 10th every year with a negative goal difference. Although I would add that in the Premier League the top teams have historically battered everyone else so much so that it's hard to get a positive goal difference if you don't finish in the top 6. That shows the volatility of this experiment; the value in these stats is only helpful when considered alongside a lot of other reasoning; stats should never be used blindly in my opinion.

We can also find the teams that are consistently good or consistently bad. Reading has the highest average positive GD outside of the current Premier League; the Royals average +11.55 and 8th place. Chelsea average a GD of +43.00 and Man Utd just a tad more at +43.27. Cambridge, York, Wimbledon and Barnsley have struggled in the period covered by the dataset. Barnsley is the only one of those sides that has been in the Football League every season since 2003/04 but they have finished on average in 17th place and the Tykes have an average GD of -12.36.

I hope this was interesting and will help you in some way when either looking at goal difference as an indication for 1X2's or in general when backing English sides this season.

Goal Difference Update

I will be tracking the progress of said teams in the 'betting blogs and systems' section of the forum.

To get detailed advice on studying football form please visit the olbg betting school.

Follow me on Twitter; https://twitter.com/w2Dyer and Merry Christmas to everyone on OLBG!