<p class='western' style='margin-bottom: 0cm;]If I offered you 8:1 odds to guess which AFL team took home the most SuperCoach points in 2013 would you take the bet?
<p class='western' style='margin-bottom: 0cm;]Before reading any further have a guess. Which AFL team scored the most SC points in 2013?
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]Seem like easy money right? Surely Hawthorn would be your first guess, they were clearly atop the ladder with only 3 losses going into finals. If not them then surely one of the other top 4 sides Geelong, Freo or Sydney. Well if you guessed any of these you are wrong. In fact if you guessed any of the top 8 you were wrong. The team that scored the most SC points in the 2013 season was North Melbourne.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]Anyone who spends time on TooSerious quickly learns that SuperCoach is a game full of adages and maxims. Whether it's PODs, MPPs, or the scoring potential of KPPs, every day the comments are full of these little titbits of wisdom that we all take for granted.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]One such bit of wisdom that I've been thinking about recently is the link between a team winning and a team scoring well in SuperCoach.
<p class='western' style='margin-bottom: 0cm;]The theory goes that since there is a capped number of points in any match (the 3300 rule), and since SC is heavily weighted towards how a player impacts the final result of the match, it stands to reason that the team who wins should also get the majority of the SC points. Seems simple enough. So why the bloody hell did North (10 wins) get more points than Hawthorn (19 wins) in the 2013 season?

<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]Season trends
<p class='western' style='margin-bottom: 0cm;]First let's check if there is a correlation between winning and SC points. The graph below shows just this correlation, premiership points and total SC points for the 2013 do seem to be related.
<div>


<img alt='' width='600px' src='/Portals/0/User%20Images/i1.png[/img]
<p class='western' style='margin-bottom: 0cm;]However there is also a fair amount of &amp;lsquo;wriggle room' in this graph (adjust R squared 0.76). Perhaps it's not just the win, but the size of the win that matters. But a comparison with winning percentage shows an even worse correlation (R squared 0.67).

<img alt='' width='600px' src='/Portals/0/User%20Images/i2.png[/img]

<p class='western' style='margin-bottom: 0cm;]Game by game trends
<p class='western' style='margin-bottom: 0cm;]Analysis at the season level didn't really show us too much other than a weak relationship. So lets delve a little deeper and look at the game by game results.

<p class='western' style='margin-bottom: 0cm;]Below is a comparison of every game in the 2013 season. The graph shows the relationship between winning margin and the SC points margin (how many more, or less, SC points a team got compared to the opponent).

<img alt='' width='600px' src='/Portals/0/User%20Images/i3.png[/img]

<p class='western' style='margin-bottom: 0cm;]Straight away we see there is a pretty good relationship between the winning margin and the SC points awarded to each team.
<p class='western' style='margin-bottom: 0cm;]Fitting a linear regression, and looking at the prediction interval, we get a pretty neat little relationship:
<p class='western' style='margin-bottom: 0cm;]


<p class='western' style='margin-bottom: 0cm;]SC deviation = Winning Margin * 6 (&amp;plusmn; 288)
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]The ratio of final score to SC points margin is about 1:6, a team that wins by 10 points can expect 60 more SC points than their opponent. But unfortunately it ain't quite that simple. The &amp;lsquo;wiggle room' in the data means that this simple relationship comes with a pretty hefty margin of plus or minus 288 points.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]This creates some interesting results. The dots in the top left quarter of the graph represent teams that lost but were awarded more SC points. Vice Versa the bottom right shows teams that won the match but got less points than. (Given the 3300 rule these 2 groups must be from the same matches).

<img alt='' width='600px' src='/Portals/0/User%20Images/i4.png[/img]

<p class='western' style='margin-bottom: 0cm;]Back to North
<p class='western' style='margin-bottom: 0cm;]


<p class='western' style='margin-bottom: 0cm;]So if there is a relationship between winning and SC points then how did North, a team that won only 45% of their matches, end up on top?
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]Let's look at North's games and how they compare to everyone else's. Below is the same graph again, but with North's games highlighted in Blue.

<img alt='' width='600px' src='/Portals/0/User%20Images/i5.png[/img]

<p class='western' style='margin-bottom: 0cm;]It appears that somehow North managed to consistently tip that &amp;plusmn; 288 margin in their favour. Win or lose their games tend to be towards the top of the graph, indicating that compared to other teams they tended to get more points for the same final margin.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]Also, their series of narrow losses last season (see the clump around the centre of the graph) meant when they lost but didn't leak too many SC points - in fact in 4 of those losses North still managed to take home the majority of the super coach points.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]So how did North manage to do this? To be honest, I have no idea.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]The only stat that I could find which put North ahead of the pack was disposal efficiency. North had the best effective disposal percentage for the year with 74.2%, however a closer look showed almost no relation between DE% and SC points across the season.
<p class='western' style='margin-bottom: 0cm;]


<p class='western' style='margin-bottom: 0cm;]What does it all mean?
<p class='western' style='margin-bottom: 0cm;]It's no surprise really that the link between winning and SC points is a bit fuzzy. Championship data proudly state that they use over 50 stats to calculate SC scores, expecting to unravel them in a few quick graphs is probably asking too much.
<p class='western' style='margin-bottom: 0cm;]
<p class='western' style='margin-bottom: 0cm;]It's possible that North is just a statistical outlier, or perhaps their style of play is finely tuned to maximise their SC output. As a final thought I'll leave you with the same graph but this time showing a team that has a very different style of play: Freo.

<img alt='' width='600px' src='/Portals/0/User%20Images/i6.png[/img]