Now that we’re down to the final two teams contesting the grand final, predicting who will win has become a lot easier – and not just because there are now only two teams left.
In sport, there are opinions, damned opinions, and then there are statistics. And statistics of a special kind suggest Sydney is very likely to win the 2014 grand final, and by a large margin.
Forget the home-and-away season, forget who has kicked the most goals throughout the season, inside 50s, Brownlow votes, scoring accuracy, who played who, and where who played who.
The critical statistics are to be found solely in the finals.
Of course, the home-and-away season is not irrelevant. After all, it’s a team’s performance during the season that determines whether they make the finals or not, and how high up the ladder. Still, the best way to determine who will win the grand final is to analyse the performances of the two contesting teams during the finals series.
Since 1991, reaching the grand final has meant that a team plays in at least two finals, sometimes three. For the statistician, this is a godsend, because it means more comparisons and numbers to play around with.
And, it is usually the team that has the better average points for-and-against differential, when compared against their rival, that wins the premiership.
Here is the points for-and-against table for Hawthorn and Fremantle in 2013 for the finals preceding the grand final.
Hawthorn
1. Qualifying final: Hawthorn 105 v Sydney 51
2. Preliminary final: Hawthorn 102 v Geelong 97
Total points for two finals: 207 v 148
Average points over two finals: 104 v 74
Fremantle
1. Qualifying final: Fremantle 87 v Geelong 72
2. Preliminary final: Fremantle 99 v Sydney 74
Total points for two finals: 186 v 146
Average points over two finals: 93 v 73
Hawthorn’s average points-for was 11 better (104 v 74) than Fremantle, and their average points-against was 1 point worse (74 v 73), for a net difference of 10 points in favour of Hawthorn.
Hawthorn was statistically the better team in the finals, and won the grand final by 15 points.
Since 1991, the team with the better net difference has won the premiership on 17 occasions (74 per cent) and, since 2000 when the current finals system was put in place, on 12 occasions (86 per cent).
Since 2000, the two teams that have bucked the trend and won with the worse net difference than their opponent are Port Adelaide (2004 against Brisbane, net difference of -9 points) and Brisbane (2003 against Collingwood, net difference -6 points).
Here is the premiers listing since 1991, with their net points difference in the finals matches preceding the grand final, compared to their opponents:
+56 – Essendon 2000 (defeated Melbourne in the grand final by 60 points)
+52 – North Melbourne 1999 (defeated Carlton by 39 points)
+46 – North Melbourne 1996 (defeated Sydney by 43 points)
+43 – Brisbane 2002 (defeated Collingwood by 9 points)
+38 – Collingwood 2010 (defeated St Kilda by 56 points, after a drawn grand final)
+28 – Geelong 2011 (defeated Collingwood by 38 points)
+26 – Geelong 2009 (defeated St Kilda by 12 points)
+19 – West Coast 1994 (defeated Geelong by 80 points)
+11 – Geelong 2007 (defeated Port Adelaide by 119 points)
+11 – Brisbane 2001 (defeated Essendon by 26 points)
+10 – Hawthorn 2013 (defeated Fremantle by 15 points)
+10 – West Coast 2006 (defeated Sydney by 1 point)
+9 – Hawthorn 2008 (defeated Geelong by 26 points)
+6 – Sydney 2012 (defeated Hawthorn by 10 points)
+4 – Essendon 1993 (defeated Carlton by 44 points)
+3 – Hawthorn 1991 (defeated West Coast by 53 points)
0 – Sydney 2005 (defeated West Coast by 4 points)
-3 – West Coast 1992 (defeated Geelong by 28 points)
-6 – Brisbane 2003 (defeated Collingwood by 50 points)
-9 – Port Adelaide 2004 (defeated Brisbane by 40 points)
-10 – Adelaide 1998 (defeated North Melbourne by 35 points)
-24 – Adelaide 1997 (defeated St Kilda by 31 points)
-48 – Carlton 1995 (defeated Geelong by 61 points)
There are two anomalies in this list, and both occurred prior to 2000, when the current finals system was installed: Carlton in 1995, with a worse differential of 48 points (defeated Geelong by 61 points); and Adelaide in 1997, with a worse differential of 24 points (defeated St Kilda by 31 points).
Statisticians like to compare consistent results, and the finals system between 1994 and 1999, and the one between 1991 and 1993 (a final six) skew these overall results slightly, but the pattern from 2000 is very clear. Twelve of the premiers had a better net points differential, so that’s 86 per cent clear.
The table for 2014
Sydney
1. Qualifying final: Sydney 93 v Fremantle 69
2. Preliminary final: Sydney 136 v North Melbourne 65
Total points for two finals: 229 v 134
Average points over two finals: 115 v 67
Hawthorn
1. Qualifying final: Hawthorn 104 v Geelong 68
2. Preliminary final: Hawthorn 97 v Port Adelaide 94
Total points for two finals: 201 v 162
Average points over two finals: 101 v 81
Sydney’s average points-for is 14 points higher than Hawthorn (115 v 101), and their average points-against is 14 points lower (67 v 81), for a net difference of 28 points in favour of Sydney.
Over the past 24 seasons, this is the seventh best result for a team going into the grand final, and only one team has lost from this position – Geelong in 1995, with a net difference in their favour of 48 points. Again, this was under the previous final-eight system, and the team to lose from the next most favourable position was Brisbane in 2004 (net difference of 9 points in their favour).
To further analyse these statistics, only one team since 2000 with a worse points-for differential of 14 or more has won the grand final (Sydney in 2012, with a 39 points differential). Since 2000, no team with a worse points-against differential of 14 or more has won the grand final.
For Hawthorn to win on Saturday, they will have to reverse a trend that has been consistent since 2000, and a statistic that has only been broken once since 1991, by Carlton in 1995, albeit under a slightly different finals system.
So while the statistics are strong for Sydney, this is never a guarantee. But there are other factors, aside from statistics.
Finals, just like politics, is all about the momentum. Sydney put in a consummate performance on Friday night against North Melbourne – clean, clear and precise for the entire four quarters.
Hawthorn are an excellent team, but they were out on their feet in the final quarter against Port Adelaide. To let in the final four goals and almost lose a preliminary final in the final four minutes of the match does not bode well for a grand final. It was shades of Geelong in 2008 against Western Bulldogs, or Collingwood in 2011 against Hawthorn in their respective preliminaries.
This is the reverse of Hawthorn in 2013 when, after being 20 points down against Geelong in the preliminary final, they managed to win by 5. They carried that momentum into the grand final.
But in 2014, the momentum is with Sydney.
There is also the residual sideshow of whether Cyril Rioli will play in the grand final, after 65 minutes in a VFL game and being absent from the top league for three months. Surely, this must be part of the pre-match mind games, because no coach in their right mind would consider such a proposition. Jordan Lewis is injured and uncertain to make the final team.
The Hawthorn team is slightly unsettled compared to the Swans, but will perform admirably.
However, the stats are in and the results suggest a very likely Swans victory on Saturday. It’s just a question of how the relative instability of Hawthorn affects the final margin.
Jon
Guest
So the all-powerful stats have once again shown us what an infallibly valid predictive tool they are! Bow down before the statistics, they know all. I hope you've learned something useful, EddyJ. Will you be so quick to remind others of your prediction as you are to remind them of your 2012 one?
EddyJ
Roar Rookie
Fremantle was a tough opponent for Sydney in the Qualifying Final, there's no question about that. It's true that Hawthorn 'clocked off' against Port, but that type of thinking almost cost them the game. Usually, the momentum from the finals series is carried over into the Grand Final, and that lapse in the final five minutes of the Preliminary Final will be costly to Hawthorn. Top line sport, including AFL, is based on small advantages here and there, momentum being one of them. I've placed great emphasis on finals in these statistics, for a number of reasons. They are played at a much higher level than other home-and-away matches, and they are the matches that teams have played most recently (why take into account Sydney's loss to GWS in Round 1? That was almost a season away!). The Swan's 'easy' draw idea doesn't stack up – they actually played more games against other top 8 teams than their opponent, and there is simply no evidence that they were 'gifted' an easy draw. Yes, they did land Tippett and Franklin through the assistance of the Cost of Living Allowance, but did they do this 'dodgily'? No, it was all above board and the COLA advantage has been seriously overblown The AFL signed off and approved the Franklin deal, Adelaide was behind any dodginess in the Tippett deal, and was appropriately sanctioned by the AFL. And with the treatment of Sanderson, who would want to stay at Adelaide anyway?
EddyJ
Roar Rookie
Totally agree with you. Using this formula, I predicted the Swans would win the 2012 Grand Final on The Roar. See this post: http://www.theroar.com.au/2012/09/27/swans-set-to-defy-the-odds-yet-again/ about half way down the page. Not blowing my own trumpet, but there are good statistical supports behind Sydney doing it again on Saturday.
EddyJ
Roar Rookie
I'm not doubting Hawthorn's abilities, only pointing out that these reliable statistics are overwhelmingly in favour of a Sydney victory on Saturday. Back-to-back flags is a serious accomplishment, as you suggest, but it's not likely to happen this year.
Graeme
Guest
This article is flawed! Swans played only Freeo & North, Hawks had Geelong & Port, much higher opposition!!! I tend to think Hawks clocked off against Port, thinking the game was over, otherwise they would've won by 4 goals. Swans were gifted an easy draw this season, plus dodgily got Tippett & Franklin. Hawks have played 11 matches this season against Swans, Port, Freeo & Geelong, plus finally have their full side. I think Hawks are battle ready & Swans may be underdone. Justice will be a Hawks win over the dodgy accounting Swans! They have unfinished business with these dodgy Swans!
Michael huston
Guest
And I also am a firm believer that Sydney shouldn't have been the underdogs in 2012. We had the better season IMO and were it not for a slack final month, we would have been the minor premiers. I think it's hard to get a grasp on how good Hawthorn have been this year. They haven't been as dominant as Sydney, but they haven't always had the list strength Sydney had. A bit of an unknown for me.
Michael huston
Guest
I cannot believe these people doubting the Hawks. Are people forgetting this is Hawthorn?! They just don't like to lose. Either team has what it takes to win, and I think Hawthorn have even more to play for, which can make them even more dangerous. Back to back flags is a serious accomplishment and they'll be even more confident with this underdog tag.
Wobbly
Guest
Lies. Damned lies. Stats! It'll be down to who kicks straight.
Angus Lowrie
Roar Rookie
Over the past ten meetings between the Hawks and the Swans the Hawks have an accumulated a +136 point net difference!
GazzaW
Guest
If the swans had played port it would have a lot closer and i think hawthorn would have done the same to north
The Protaganist
Guest
Statistics schmatistics..... yeah, perhaps Swans to win by lots, but that's not the story. Their good statistics have been set up by AFL who want them to win on Saturday. They get lots of support and funding from the AFL, so the AFL is getting the result they're after. COLA, Buddy, Tippett. etc. It's all a plot by the AFL to get the cup going to the the largest state in Australia.. Who cares what happened at the turn of the century. I want the Hawks to win an think they will. By about 6 goals. Roughead to kick about 5.
Winston
Guest
This is great! Of course it doesn't guarantee a result on a particular given day, but I think it does mean over a period of time there is a pattern. Guess this is how I will be betting from now on!!!!!!!!!!!!!!!!!!
EddyJ
Roar Rookie
Statistics are what AFL is all about and ARE very important. After all, the most meaningful aspect is to score the most points in a game – and that's a statistic. To paraphrase Julia Gillard, they don't mean everything, they don't mean nothing. But they mean something. Every coaches' box is replete with Apple Macs, analysing every aspect and every statistic of the game. Statisticians look for patterns in behaviour – I've analysed many other parts of the game: inside 50s, most disposals, the amount of time a team leads during a match (yes, there is a statistic for that too), accuracy in attack, inaccuracy in defence, but having a benefit in these areas isn't necessarily a marker of success, as Hawthorn found out in the 2012 Grand Final. By the way, the Swans have led for an incredible 77 per cent of game time this year, Hawthorn have led for 66 per cent of game time – but this is not relevant, as the most important statistic is who is ahead at the end of the game. After analysing the statistics of all of these results and available figures going back 24 years, the most meaningful was the performance of the two Grand Final teams DURING the finals. We'll see at 5.30 on Saturday afternoon. Anyway it's good to look at statistics, rather than the endless discussion about how the Cost Of Living Allowance is going to be the sole reason behind a likely Swans' victory on Saturday.
Olivia Watts
Roar Guru
Excellent first article Eddy, and a quite different take on assessing the game I detested statistics when I had to do it for my degree. Twin tailed sigmas and P values drove me even nuttier than I was before. In your case it is a straightforward comparison graph but I would think your N (total number of items assessed) was too small to draw any meaningful conclusions from. There's always the chance of an outlier which totally skews the data. All that said though, your stats have Sydney coming out on top, so suddenly I don't hate statistics anywhere near as much. Go Sydney!
doubledutch
Roar Pro
Stats are important but they don't mean anything. This game reminds me of the 2012 final, only in reverse. Can the under dogs win?
Glen
Guest
Stats mean absolutely nothing on Grand Final day,which team turns up to play,controls the nerves and settles into a hard game will probably come out on top.You can go read your stats but i'll watch the game.Roar rookies always seem to pull out the stats but they don't look at the game.
Glen
Guest
Stats mean absolutely nothing on Grand Final day,which team turns up to play,controls the nerves and settles into a hard game will probably come out on top.You can go read your stats but i'll watch the game.Roar rookies always seem to pull out the stats but they don't look at the game.
EddyJ
Roar Rookie
Yes, Hawthorn did have the better performance stats, compared to Geelong, during the finals in 2008 prior to the Grand Final. And it showed on the day.
EddyJ
Roar Rookie
It's primarily about statistics and assessing patterns over a 24-year period in the finals, especially since 2000. Statistics provide guidance, rather than exact predictions. By the way, I'm not a Swans fan (Collingwood is my team) – not that this matters – it's all about the statistics. I agree that Hawthorn should not be underestimated – I don't think many people would be – but it's not looking that good for them.
New York Hawk
Guest
So what you are saying is a lazy final ten minutes in a prelim, combined with the Swans playing a second tier team in their prelim, has cost us a premiership? Please. But I love all this anti-Hawks commentary. It is great we are the underdog which is a welcome change. Accuracy is king and gentlemen, we have that in spades. Not sure your Swans can say the same.... Maybe big Bud's girlfriend will come out and defend him and talk about how hurt his feelings were again after the Hawks fans boo him.... Imagine if he loses a GF to the team he left. Now that would have to crush your ego if you had a brain to think about it for a second....