Goal Difference

[Win2Win]

I do not see why the second deciding variable is Goal Difference. The whole object of the League's is to WIN, that is how you get the points, by how much is irrelevant, otherwise they would not decide it on points!!

Surely AWAY wins are more important than GD, and that gives Chelsea the edge.

Now if they gave 4pts for an away win throughout the season, then used GD, that would be fine (Wouldn't help Derby ), it may also stop some teams playing for an away draw.

[mathare]

I have been reading some of my football betting books again recently and at least one of those talks about how points are awarded equally for outcomes that are not statistically equal, such as home and away wins which happen around 48% and 27% of the time respectively yet are both worth 3 points.

But FIFA and the sub-authorities are obsessed with goals and always have been. The rule changes have always been to bring about more goals so the idea of using goal difference as the second variable is supposedly to encourage more attacking play.

Jimmy Hill suggested a variable number of points for a win depending on how many goals you won by quite a few years ago now. Maybe that sort of idea should be considered once more

[scoobydoo]

IF Utd win the league today on goal difference, it would be total justice on the way Chelsea play the game and they will certainly be wishing they hadnt sat on thoses 1-0's throughout the season. With the players Chelsea have...its embarrassing the way they sit after going 1-0 up...and I really really hope Utd make them pay for it today. Football at this level has to be about entertainment as well surely? I have my prayer mat out!

[piggy]

IF Utd win the league today on goal difference, it would be total justice on the way Chelsea play the game and they will certainly be wishing they hadnt sat on thoses 1-0's throughout the season. With the players Chelsea have...its embarrassing the way they sit after going 1-0 up...and I really really hope Utd make them pay for it today. Football at this level has to be about entertainment as well surely? I have my prayer mat out!

after going 50 years without winning a title and at times playing very attractive footy, like most fans i don't care how we win as long as we do so c'mon wigan c'mon wigan :D

[peza2605]

piggy ,if you had the choice of one, european cup or league?

good luck by the way.

[GlosRFC]

If you used 4pts for an away win, the table would look like this:

Chelsea98
Man Utd97
Arsenal93
Liverpool85
Everton73
Aston Villa66
Portsmouth66
Blackburn65
Man City58
West Ham55
Tottenham49
Newcastle46
Wigan45
Middlesbrough42
Sunderland41
Bolton39
Fulham39
Reading38
Birmingham37
Derby11

The only significant change is that Chelsea would've won the title thanks to those 1-0 away wins.

Under the old 2 pts for a win system, then goal difference would have to be used to separate Man Utd and Chelsea:

Man Utd60
Chelsea60
Arsenal59
Liverpool55
Everton46
Aston Villa44
Blackburn43
Portsmouth41
Man City40
West Ham36
Tottenham35
Newcastle32
Wigan32
Middlesbrough30
Sunderland28
Bolton28
Fulham28
Birmingham27
Reading26
Derby10

Apart from a few minor positional adjustments (Birmingham above Reading for example), nothing would've changed as regards the title or relegation.

We can also adjust the table slightly to take account of Matt's views on the probability of different home/away outcomes. As the likelihood of winning at home seems to be twice as frequent as winning away, we can reflect this by giving 4 points for a home win, 2 points for a home win, and 1 point for a draw. And so the table looks like this:

Chelsea86
Man Utd80
Arsenal79
Liverpool73
Everton62
Portsmouth59
Blackburn57
Aston Villa56
Man City48
West Ham48
Tottenham41
Newcastle38
Wigan38
Middlesbrough34
Fulham34
Sunderland32
Bolton32
Birmingham31
Reading30
Derby10

So, once again, the only difference (apart from a few minor positional changes, e.g. Portsmouth/Aston Villa) is that Chelsea would've won by a canter thanks to their better ratio of away to home wins. Relegation, Champions League, and Eufa places would still not have been affected, although Fulham would've breathed a bit easier going into the last weekend.

So you could argue that the actual number of points awarded doesn't actually matter - which is why the emphasis has been placed on goal difference as a better way of separating teams.

[GlosRFC]

Okay, it stripped out the tabs, but I'm sure you can all get the basic drift :D

[scoobydoo]

Okay, it stripped out the tabs, but I'm sure you can all get the basic drift :D

Great post Glos. I see in Spain they go on head to head when on same points...this happened last year when Madrid won it against Barca. I think goal difference is the best way...cant see the FA going on 4 points for an way win can you...though I think it would be good...it could definitely encourage teams to go for it. & it wouldnt be that long I bet till you got 2 teams on the same points coming to the last game of the season and one of them would be away from home...with a win giving them the chance to win it. When you think about it, there should be something more in it for winning away from home shouldnt there

[vegyjones]

Footballs was a very simple game,

why do people want to overcomplicate it?

[mathare]

When you think about it, there should be something more in it for winning away from home shouldnt there

Where does it stop though? Do we award more points to Fulham for beating Man Utd than for beating Man City because Utd are a better side? If Spurs field an understrength side and still win should they get more points? I am deliberately exaggerating the point here but I'm sure you get what I mean.

Caution! Football cliches approaching.

At the end of the day all the teams within a division play each of the other sides home and away once (I am excluding the silly Scottish divisions at this point) so if they were all equal they should have equal opportunities to pick up points. Will changing the reward for an away win make things any different? Each team still has to play the other teams away from home once. If Man Utd are 30 points clear of another team at the end of a season under the current system could we not reasonably expect them to be 35 points clear of the same team if we were to award 4 points for an away win?
(35 points if we assume Man Utd are 10 wins better than the other team and that half the games are at home so worth 3 points and half are away and worth 4 points).

While it might be more statistically correct to award more points for an away win what will it change? It may actually change the trend on which it is based. If the FA reward away wins more then it may encourage teams to go for the away win at all costs - better to push for the win and gamble on getting the three extra points rather than settling for the draw. But it may also push some home teams to sit deeper and defend so as not to let their opponents get the full four points. So we may see a change in the home/draw/away percentages and thus potentially invalidate the whole reason behind rewarding away wins more than home wins.

[bigcumba]

Wasn't there a time when it was goal average rather than difference that was used? I seem to remember that from a few years ago...

[mathare]

Wasn't there a time when it was goal average rather than difference that was used? I seem to remember that from a few years ago...

Could have been, yeah. I know goals scored came into it at some point too. They have tinkered with it a few times but always to reward and encourage attacking play

[Win2Win]

Okay, it stripped out the tabs, but I'm sure you can all get the basic drift :D

You really need to learn to program

[Win2Win]

None of the points systems make Derby look any better

[vegyjones]

Could have been, yeah. I know goals scored came into it at some point too. They have tinkered with it a few times but always to reward and encourage attacking play

Goals scored survived one or two seasons in the fotbal league, I don't believe that it made it all the way to the Ptremiership!

[MattR]

Yes we missed out on promotion to West Ham on goals scored in 1993.


Mat, going back to your points for a win I am sure I read somewhere recently (not sure where) but the average number of goals has not changed from when it was 2pts for a win despite the idea of 3pts being that it would encourage teams to be more attacking to get an extra 2pts instead of one. I don't think any different arrangement of points in the Premier would make much difference now as they're all too scared to lose and get relegated now to worry about attacking!


Anyway all the time players on 50k a week plus can't hit a target 8 foot high and 24 feet wide from 18 yards away there's not much hope of more goals! And I make no apologies for not using metric measurements!

[GlosRFC]

Wasn't there a time when it was goal average rather than difference that was used? I seem to remember that from a few years ago...

Goal average was used up until the same season that yellow and red cards were introduced...1976/77. It was replaced because goal average leads to more negative games than goal difference. Imagine this highly-implausible scenario:

Leyton Orient beat Man Utd 3 - 0
Man Utd beat Liverpool 6 - 0
Liverpool beat Leyton Orient 1 -0

Each team has 3 points but the results are completely different depending on whether you use goal difference or goal average. With goal difference the table reads:

Man Utd GD +3
Leyton Orient GD +2
Liverpool GD -5

However, using goal average, the top two positions are reversed even though it's clear that Man Utd were the most attacking team over their two games (and, given who ends up finishing top, makes using goal average even sillier ):

Leyton Orient GA 3.0000
Man Utd GA 2.0000
Liverpool GA 0.1667

I'm not surprised you can recall it as Hearts were deprived of the Scottish championship in 1965 when they lost 2-0 at home on the last day of the season. They finished with a goal average of 1.84 and a goal difference of +41. The winners were Kilmarnock who, thanks to those two goals against Hearts, increased their goal average to 1.88 but had a much poorer goal difference of +29.

Twenty years later they lost by the same scoreline when Dundee scored 2 goals in stoppage time on the last day. This allowed Celtic to squeak the title on goal difference (+29 against +26) as both teams finished on 50 points...but if it had still been decided by goal average, Hearts would've been Scottish champions (1.788 against 1.763).

[mathare]

Almost exactly what Wikipedia says Glos, well done :)

[Win2Win]

He put Liverpool at the bottom :(

[scoobydoo]

Wasn't there a time when it was goal average rather than difference that was used? I seem to remember that from a few years ago...

Could have been, yeah. I know goals scored came into it at some point too. They have tinkered with it a few times but always to reward and encourage attacking play

Didnt Arsenal beat Liverpool on this in 89? They had to win 2-0 at Anfield...and they did... Oh how we laughed!

[jackster109]

What about this point format for the leagues:

Away Win = 4 pts
Home Win = 3 pts
Home Draw = 1 pt
Away Draw = 2 pt

3 or more goal scored = 1 extra bonus pt

or even getting points for the half time score. I know this would be complicated but would surely see more attacking football and see less teams parking the bus in front of the sticks.

Or even changing the goal difference format with an away goal worth double, this wouldn't effect the result but just the GD.

For example if Man Utd won 2-1 (GD = 1) at home and 2-0 away (GD = 4). This would then total +5 and not +3 like the current format.

[mathare]

I think at the end of the day there are more fundamental problems in the Premiership than how many points are awarded for each result. At the minute there are several mini-leagues within the one division. The top four have broken away to form their own league and the results of games between those sides tends to settle that one. Then you've got the UEFA Cup league where you have maybe 6-8 sides battling it out for one or two places in Europe. Then you have a slight blur between the bottom of that mini-league and the next one of 5 or so sides all battling it out for the highest midtable finish without troubling the UEFA spots or relegation, and then finally you have the battle to avoid relegation. And clubs rarely make the jump between these leagues, and if they do they tend to move up or down one league at a time. This season Spurs have slipped back into the midtable obscurity fight. Boro fight relegation most seasons, ditto Bolton. Unless these mini-leagues are broken down then the number of points for a win will be irrelevant as the top four will always walk away with it.

This season there was 22 points between 1st and 5th in the Premiership. Last season it was 29. In 2005/06 it was 26. In the first season of the Premiership (92/93) it was 20. 22 the year after and 16 the year after that. So this gap has always been there to an extent. Is it money-driven? This season the gap between 1st and 5th was 10 points in the Championship, 16 in League 1 and 17 in League 2. Not quite sure where this is leading now. Hmmm. I'll just stop.

[GlosRFC]

Almost exactly what Wikipedia says Glos, well done :)

Maybe I should charge 'em, huh? This is what I was looking at:
http://www.scottishleague.net/forum/...7606a911c14795
http://www.hmfckickback.co.uk/terrac...xt-friday.html

This goal average thing still seems to be an issue in tartan land as most of their footie sites still include it in the stats.

And some interesting statistical analysis here:
http://plus.maths.org/issue46/risk/index.html
The league simulation is impressive

Didnt Arsenal beat Liverpool on this in 89? They had to win 2-0 at Anfield...and they did... Oh how we laughed!

I think so too. Goals scored is the second decider. If the teams are still equal after goal difference and goals scored, then the FA would arrange a play-off game at a neutral ground. Presumably with extra time and penalties if needed.

In rugby it's slightly different. First it's total league table points, then match points difference, then tries scored, then most points scored, and finally try difference!!

[GlosRFC]

I'm not a fan of bonus points having seen how it can corrupt the Guinness Premiership tables.

Having said that, there might be some merit behind using goals as the deciding factor. Take a look at the final 1998 Swedish table which shows AIK as champions, even though every team below them scored more goals! They scored 2 goals in just 4 matches with the remaining games being either 1-0, 1-1, or losses:
http://www.rsssf.com/tablesz/zwed98.html

[mathare]

That 1998 Swedish table is quite a famous one and often crops up in discussions such as this.

Personally I don't see a need to change the system we currently adopt. Most seasons the league is decided on points and points alone. The other factors that will be taken into account if necessary are known in advance by all the teams and fans etc so everyone has all the relevant information so has no right to complain about it come the end of the season. After all, most teams build up their goal difference over a long series of matches so Chelsea would have been aware of Man Utd's superior goal difference for a while. It's not as though Man Utd won a game 40-0 towards the end of the season to massively change the outlook of things was it?

[bigcumba]

In essence I'd say goal difference rewards both good attacking football, and good defensive play, while goal average only really rewards the attacking play...

[MattR]

Having said that, there might be some merit behind using goals as the deciding factor. Take a look at the final 1998 Swedish table which shows AIK as champions, even though every team below them scored more goals! They scored 2 goals in just 4 matches with the remaining games being either 1-0, 1-1, or losses:
http://www.rsssf.com/tablesz/zwed98.html

I don't think you can ever go away from points being the first factor. AIK lost two games all season and clearly built their success on a good defence. This may not be fun to watch but they won the most points over the season and deserved to win the league.

[GlosRFC]

I don't recall saying that goal difference should be the primary method of determining who wins the league. However, it's certainly the fairest deciding factor in the event of a tie on points. After all, goal difference usually reflects both the attacking nature (goals scored) and the defensive capabilities (goals conceded) of each team.

For example Tottenham banged in 66 goals which makes them the fifth highest goalscorers and would imply that they're a more attacking team than Everton who scored 55. However, they also shipped 61 goals too, which compares unfavourably with Evertons's defensive record of 33 goals conceded.

So Everton were undoubtably the more balanced team which is reflected by their finishing 6 places above Spurs with 50% more points.

[MattR]

I know you didn't Glos. But the Swedish league you quoted for saying Goals to be used as a deciding factor was irrelevant because AIK won it by two points anyway.