Touted as a top four contender before the season, the ‘Roos sit 5-5 at the start of the bye period, and are precariously positioned as a team that could miss the big dance all together.

With a seemingly unprecedented logjam around the middle part of the ladder, North Melbourne simply have to start winning.

While nothing is really going to plan for the blue and white – including in the coaches’ box – one part of the ground sticks out in particular.

An unbalanced portfolio
A well-worn financial investment strategy is diversification. Reducing your risk of loss by spreading your hard-earned over a range of different assets – blue chip stocks, bonds, property and the like – is the cornerstone of wealth creation for people the world over. It doesn’t matter if you’re Warren Buffet or the Wu Tang Clan. Everyone knows it.

North Melbourne, the AFL’s first major free agency experiment, are beginning to emerge as a case study in what happens when you put too many eggs in one basket.

North rank a putrid 13th on defensive efficiency thus far in 2015, with a Defensive Efficiency Rating (DER) of -3.7. If it weren’t for Gold Coast’s stunning implosion, and a solid assist from Mother Nature in the weekend’s game down in Tasmania (prior to this weekend, North were rolling with a DER of -13.4), the Roos’ stopping power would be bottom-four calibre to this point in the season. Putrid.

It’s something of a recent issue, with North Melbourne turning in relatively strong defensive performances after their first trip to September under Brad Scott in 2012.

In that first finals run, North built itself on a relentless, run-and-gun attacking style, jacking up a crazy 64 per cent scoring accuracy owing to a rampant midfield/forward group. But that left the Roos very exposed on the defensive end, and resulted in a poor DER of -4.3 (ranked 13th). They were commensurately blown away by West Coast in the second elimination final by almost triple digits.

In 2013, the Roos had a much improved DER of +5.4 (ranked eighth), without losing too much punch on the offensive end (ranking third in OER). A further improvement in 2014 followed (with North’s DER lifting to a fifth-placed +9.0), but it came at the expensive of some scoring punch, with the Roos less potent than the supposedly flaccid Fremantle in 2014.

As I’ve repeated ad nauseum, offence is the skillset that wins Premierships, and that’s certainly what the higher ups at Arden Street thought when they signed ex-Blue Jarrad Waite and ex-Dog Shaun Higgins as free agents in the off season. These complemented the 2013 additions of Nick Dal Santo and Robin Nahas, and with Daniel Wells starting the season off the injury list, and Brent Harvey still a very strong offensive threat, North Melbourne appeared to have assembled a very potent line up front of centre.

Their campaign has been derailed somewhat by with Dal Santo and Wells out of the line up since Round Two. Their OER has suffered as a result, coming in at just 0.6 so far in 2015. Given the injuries, and a relatively tough draw, its something I’m reasonably confident will correct itself over the rest of the year. North’s key forward set up is amongst the silkiest looking in the league – its actually the lack of midfield goals that’s letting them down a touch, with no ‘Roo midfielder averaging more than a goal a game this season.

But that’s not the issue. By the week, North Melbourne’s putrid defence, and abject inability to stop sides that are getting a run on, is putting them further and further out of the race for the 2015 flag.

This season, a DER of -13.4 (which was North’s rating before this weekend) corresponds to conceding almost 100 points per game, which is nowhere near where a premiership hopeful needs to be.

The signs of this were there for everyone to see in Round 1 against Adelaide, where North Melbourne conceded 67 straight points to the Crows from the 20-minute mark of the first quarter to essentially the end of the second quarter.

It happened again, although not quite so dramatically, in the third quarter of North’s loss to Collingwood a couple of weeks back; with the Roos conceding nine goals two to just three behinds as the defence looked insipid. Other instances include the end of the first quarter against Hawthorn (27 points to one) and the first quarter and a half against Fremantle (48 points on 13 shots to 14 points on four shots).

Giving up big runs of goals is a cardinal sin in today’s low scoring AFL, and as Champion Data noted last week there is a pretty strong correlation between the number of “streak goals” – goals kicked in a succession of five or more – and the wins and losses table on the ladder. For North, that’s one part of the story.

In breaking down the defence, it’s clear that the biggest problem North faces is stopping teams from scoring once they get inside the 50-metre arc. Until the weekend’s tornado-based game, North were conceding scores on more than half (50.3 per cent) of their opponents’ entries this season, which was more frequently than all but Carlton.

And Carlton are not travelling too well this season, so I’ve heard.

In five of their nine games thus far in 2015, North Melbourne have conceded scores on more than 50 per cent of their opponent’s inside 50 entries, including a staggering 66 per cent against Fremantle in Round Eight. Only Carlton has had more (six. Did we mention they’re terrible at football?), while Greater Western Sydney and Richmond are the other offenders.

The Giants’ issue in this respect is more to do with the way they attack – throwing the kitchen sink at their opposition once they get the ball leaves them a tad exposed if they turn the ball over.

Richmond, on the other hand, recorded their five poorest stopping games in the first six rounds, and have significantly tightened things up since.

So let’s see what they can do to fix this up, shall we?

Searching for defenders
This column has tended to try and shy away from picking on particular players when it comes to performance analysis. But it’s becoming quite apparent that one of the biggest issues facing North Melbourne’s defence is the players they have manning the post.

Robbie Tarrant has been recast as a defender in light of North’s overflowing talls in the forward line, and thus far he has looked out of his depth as centre half back. It is the job of centre half back to be the reference point with which the rest of the back six, and helpers from the midfield, position themselves when defending.

I spotted three instances in Collingwood’s third-quarter comeback where there were gaping holes in North’s defensive 50.

In all three of these instances, there are cavernous holes right down the centre of the 50, which simply must be filled in today’s modern, efficiency-focussed football.

Sure, you need to keep a close check on your direct opponent, but in all of these instances North Melbourne players are giving their man too much attention and forsaking their teammates. You look at how Fremantle defend inside 50 entries and its complete chalk and cheese.

While you can’t pin all of the blame on Tarrant – after all, there are upwards of 10 Kangaroos in some of these plays – Tarrant’s relative inexperience in defensive structures must be playing a role in this.

The other issue is with fullback Scott Thompson, fullback-in-waiting Lachie Henderson, and longstanding pillar Michael Firrito, who are beaten over and over again in even number contested situations. Here’s two such instances from that same horror quarter against the Pies.

In this first frame, we can see Taylor Adams looking to deliver the ball right on the 50 metre arc, with a two v two situation ahead of him. For Collingwood, its Travis Cloke and resting ruckman Jarrod Witts, while for North its Thompson (on Cloke) and Hansen (Witts).

Hansen has allowed Witts to get a jump on the lead and sticks with the chase, meaning Thompson now has a choice between trying to come in as a third man for the intercept or spoil, or keep to Cloke who is using Witts’ lead to overlap.

It does not end well, with Hansen ending up in no man’s land and Thompson caught having to try and double back to make a contest with Cloke. Cloke gathers and goals.

This second play involves Cloke and Jamie Elliott for Collingwood, and Hansen (marking Cloke) and Michael Firrito for North Melbourne. Jarryd Blair scrambles the ball forward to this two versus two situation. It’s not a particularly risky defensive set for North Melbourne, with Firrito an ideal match up for the small Elliot, and Cloke not a major threat in these situations.

But Firrito lets Elliot body up and push off, and its all unravels pretty quickly.

This time, Firrito’s inability to force a better contest leaves Hansen in no man’s land, with Eliott running around and kicking his second in the quarter.

Here’s one from Round 1. It’s a three on three situation in Adelaide’s left forward pocket, and Patrick Dangerfield has the ball for the Crows. In frame are Dangerfield, Eddie Betts (in the pocket, duh), and captain Taylor Walker.

Marking them are Thompson (there he is again), young midfielder Luke McDonald and the rookie listed Joel Tippett.

Dangerfield is literally hemmed in to the boundary, with the three Roos closing in on him. Again, a good defensive unit neutralises this kind of situation. What happens? Betts gets out the back, Dangerfield does Dangerfield things, Walker manages to keep Tippett out of the way…

…and it results in the easiest of uncontested goals.

We’ll leave it there. But you get the idea. It’s not just a scheme thing for North Melbourne – it looks to be the quality of the players they have playing down back. Us mortals don’t get ‘contests won’ stats (#freethestats), but I’d guess North Melbourne’s back six would be clearly the worst of the last year’s top eight.

And that brings us full circle: North Melbourne’s free agent strategy has been all about boosting their offensive potency, and it appears to be to the detriment of their side as a whole.

Teaching the players to be more conscious of their positioning inside 50 isn’t too difficult a task. It just requires commitment to a change in style, and would be something I’d expect to see fixed up in the coming weeks.

The last quarter of the weekend’s game showed they have recognised this problem. While the game situation demanded a +1 defender was put in place, the positioning of this player was precisely what’s required moving forward.

But personnel is a problem, and can’t be addressed until November this year. The Grand Final, quite obviously, is in October.

Fortunately for the ‘Roos, they have a ready-made fill in that could give them more structure and add a level of one-on-one ability that the existing players are, unfortunately, lacking.

Frank the Tank
This decade, Petrie has played mostly as North Melbourne’s number one key forward, averaging 2.2 goals per game on 3.6 shots, including two seasons where he was amongst the league’s most effective kicks for goal (a Luke Bruest-esque 58.18 in 2012 and 48.19 in 2013).

When Drew Petrie made his debut in 2001, the Roos won just nine games in finishing 13th on the ladder.

At 197cm, he was huge for his time, with many full forwards of that era still cut from Tony Lockett’s cloth. Petrie was sharing the forward 50 with a few guys you may have heard of: NFL Punter Saverio Rocca, all-around nice guy Corey McKernan, and a pretty-good-at-football Wayne Carey.

Except guess what? Petrie didn’t play as a forward. He played as a ruckman in nine games, and didn’t really have too much of an impact. It wasn’t until 2003 that Petrie started to hit his straps as a forward, as those players above started to move along.

It’s been a very strange, and somewhat unique career, for Petrie, with the stats suggesting he has played all around the ground throughout his time at the highest level.

In his fourth year, 2005, Petrie played as dual-primary ruckman with David Hale; while in 2006 Petrie was shifted down back as a promising young lad by the name of Hamish McIntosh took a greater share of the ruckwork for North. Then in 2007, Petrie played a utility-type role with his time split between the forward 50 (1.5 marks inside 50 per game, for 38 goals), centre of the ground (third-most hitouts behind McIntosh and Hale) and down back (more than one rebound and four spoils per game).

Petrie has been something of a whipping boy for AFL fans regardless of their affiliation, with number 20 never reaching the absolute upper echelons of other players that have reached the 250-game club.

Longevity is supposed to mean you’re a great player, apparently. Can’t you just be the ultimate role player, which is what I think Petrie is?

I think he’s a somewhat underrated commodity, and it’s because of the flexibility he’s shown throughout his career. In a lot of ways, Petrie may be the last of the old school footballers playing in the AFL – if you exclude the 400 gamers, of course.

He’s not one for being too showy, and when he’s forced by the sense of occasion to do a little bit of razzle dazzle, you can tell it’s not something that comes naturally to him.

And despite the advent of his 33rd birthday two weeks after this year’s Grand Final, Petrie might be the key to sorting out the Roos defensive woes.

Petrie has already been thrown back by North Melbourne coach Brad Scott a few times this season, most notably in their close win against Essendon a few Friday nights ago. Petrie spent much of the last quarter as a floating extra man in the back half, particularly in the back 50 around the centre half back position, and snuffed out any hope the Dons had of coming back. He ended that game with 12 marks.

I think the coaching set at North could do worse than shifting Petrie down back on a more permanent basis, at least for a few weeks, to see whether his 13 years of football nous and clear ability to mark the ball boost their defensive prospects. The problems are clear to see when you look at the tape, and one of the short-term solutions is right in front of North Melbourne’s nose.

Any good investment strategy has an eye on the short term and the long term. There’s no doubt North Melbourne’s defensive issues will require some long term attention, and a shift in the direction of player acquisition is probably a good place to start this year. But a putrid defence, if maintained, will count the ‘Roos out in the short term if the right investment isn’t made.

I hear there’s a key defender glut at Windy Hill.

For now, I still think North have the cattle and pedigree in the middle and up forward, and pretty stellar run of games to end the year, to make the four in 2015. Given the glut around the middle of the table, I expect a 13 win team to get into the four this year, and North could be the one.

Or it could be West Coast. Or GWS. Or Adelaide. Or Port Adelaide. Or Collingwood. Or Richmond. I mean, St Kilda could get a run on, right?

***

Adelaide’s rising midfield
Which side has the best midfield in the competition?

I’ve thought for a while its Hawthorn. Fremantle have a claim to the title in 2015, given the nine-headed Hydra they have assembled. You can’t count out Sydney, even if they employ the Golden State Warriors strategy of employing a core of players that all do the same thing.

The numbers say its Hawthorn – they tend to get a crazy 19 more inside 50s per game than their opponent, and to this point in the season have managed a clearance differential of +8.

Those are crazy numbers, just so you know. Particularly the inside 50s. If Hawthorn manage to maintain the rage, they’ll smash the modern day record of +11.6 set by the 2010 Collingwood side.

There’s a reason why I’ve still got the Hawks near the pointy end despite their distinct lack of Ws.

If Hawthorn is the midfield king, then Adelaide are the prince. Like the rest of the competition, Adelaide trail Hawthorn in both of the key midfield outputs. But they are fourth in inside 50 differential (+7.2) and a clear second in clearance differential (+7.9).

The triumvirate of Patrick Dangerfield, Rory Sloane and Scott Thompson are clearance machines, and the youngsters they have been forced to give some games to due to injury have acquitted themselves well.

Coach Phil Walsh has a very simple mantra painted on his office wall, from what we’ve heard, and doesn’t really like it when his players don’t play the way he wants them to. It’s all meshing together to create one of footy’s strongest midfield groups.

But how’s this for a stat? Hawthorn’s starting five in the middle have a median age of 29, and a total of 901 games of experience. Adelaide’s five have a median age of 25, and 689 games of experience – with Thompson’s age (32) and games played (271) pushing up the aggregates for the Crows.

Needless to say, number 32 has to stay a Crow for this group to build on their promise. Can you imagine a group headed by Dangerfield and Sloane, with Brad Crouch on the inside, with another hundred games or so under their respective belts? Jeebus.

Adelaide’s midfield dominance is helping propel their forward line into the conversation as the best in the competition. So far this season, Adelaide have recorded an Offensive Efficiency Rating of +8.8, putting them fourth in the league and ahead of Fremantle on the scoring charts for 2015.

It’s familiar territory for the Crows, who’ve had one of the league’s best offensive sets over the past five years. The only thing counting them out of returning to their regular number one or two post is a dip in scoring accuracy. It’s lost them one game, and almost cost them another, in the past fortnight.

From what I can tell it’s driven by those midfield clearance machines rather than the full-time forwards, with Adelaide’s regular forward line set running at a very hot 65 per cent scoring accuracy compared to a sub-50 mark for the rest of the side. One forward line specialist in particular is bolstering Adelaide’s radar, and scoring potency more broadly, is that maligned half a million dollar free agent.

Is Eddie Betts a sneaky Coleman chance?
Key forwards tend to kick the most goals over a 22 game season. The average height of Coleman medal winners in the AFL era is 192cm. The past 10 winners have an average height of 195cm.

Since the start of the decade, there have been just seven instances of a player less than 190cm tall scoring more than 50 goals during the home and away season:

Luke Bruest	2014	53
Eddie Betts	2014	51
Lindsay Thomas	2013	53
Stephen Milne	2012	56
Mark LeCras	2010	63
Steve Johnson	2010	57
Brad Greene	2010	55

At just about the halfway point of the year, Betts is putting in a season that will see him surpass all of these recent markers, and take the crown as the best small forward in the game. I mean, he could even beat the key forwards this year.

He’s kicking them from everywhere, too. I’d love to know how his 2015 season is stacking up on a +/- probability basis based on where he’s taking his shots (#freethestats).

I have a hunch that he’s the only player in AFL history to hit a general play goal from where the 50 metre arc intersects with the boundary line, off balance and from the opposite pocket for good measure. And yes, in the same manner that Han shot at Greedo first, Betts was most certainly having a ping at the big sticks.

Even though he’s kicking those sort of freakish goals, Betts is going at a crazy 76 per cent crude accuracy, ranked fourth amongst qualifying forwards (those with more than five games played or 10 goals kicked).

His winning the Coleman would be a fantastic story. Eddie’s move to Adelaide was one of the more hyped of recent memory, mostly because of the dollars (supposedly) involved.

Half a million dollars for a small forward sounds like a lot, and it is, but his performances for the Crows have thus far met expectations. Now in his second year, Betts has kicked 82 goals, 32 behinds, generated 68 inside 50s (including 31 this year, putting Betts in the top five at the Crows) and assisted on 38 more goals.

But I’m doubtful he’ll end up topping the goal kicking table. Sorry to be a party pooper.

So far this season, Adelaide have played the 11th, 12th, 14th, 15th, 17th and 18th ranked defensive sides. Betts has kicked four goals a game against those six poor defensive sides, and a more middling (but still very good) 2.7 a game against other, more highly ranked opponents.

Does it count him out? Not necessarily. But his scoring challenge will be amped up over the remainder of the year.

***

‘Tis but a flesh wound
Before we wrap this one up, just a quick word on this week’s Friday night surprise.

I wrote in my preview of the game that Fremantle’s defensive structure was on the verge of surpassing Ross Lyon’s St Kilda Squeeze side of 2009 in terms of efficiency, and that this was built on all three pillars of defensive efficiency: stopping the ball from going inside the arc, stopping score attempts, and limiting the accuracy of attempts that get off. Lyon’s scheme this year does it all, and well.

In the first quarter of Friday’s game, Richmond kicked eight goals and one behind on trips inside the stripe. For those who aren’t full bottle on mental maths, that means Richmond went forward with 100 per cent efficiency and at a crude accuracy rating of 89 per cent.

For the rest of the game, it was normal programming, with the Dockers conceding just 27 more inside 50s, eight more scoring shots on those inside 50s, and a more regular 57 per cent accuracy. The Dockers won the game from the first quarter onwards, albeit by three points.

The hyperbole surrounding Richmond’s victory is bordering on hysterical. Fremantle are still premiership favourites, and Richmond are still going to be competing for a spot in the eight as the season stretches into the 20s.

Let’s all take a chill pill, and look at ten rounds of football rather than one when assessing where teams are at, shall we?

Speaking of which, here’s how things are shaping up heading into the bye period.

Improper Projections: Round 10
It’s time to throw this Bayesian mumbo jumbo out the window and make some big calls in this week’s Improper Projections.

Alright, come at me.

So I’ve finally caved to the rest of the footy world, and now have West Coast making the eight this year. I still think it’ll be a pyrrhic victory, because the top three of Sydney, Hawthorn and Fremantle are streets ahead of the rest of the competition. But West Coast’s list, shiny new Hawthorn-esque strategy of positional flexibility and precision ball movement, is shaping up nicely for a tilt at the flag in the next couple of years. Imagine if they had another Liam Duggan running around for them, though.

Anyway, I’m not going to justify each move I’ve made here.

I’ll end on the addition of a new data nugget at the very right-hand side of the table, which I’m calling True Percentage.

Champion Data have been trolling football Twitter in recent weeks, posting a chart of “Premiership Standards” for offence and defence every Tuesday morning that shows West Coast’s output to this point in the season puts them into the Premiership quadrant. The Premiership quadrant is no Nat Fyfe Quadrant, but that’s another story for another time.

Its abundantly clear that West Coast have had a very soft draw to this point in the season, and that their percentage is inflated as such. So what I’ve done is a little bit of statistical jiggery pokery (it’s a little more scientific than that, trust me) to come up with what a side’s percentage should be, based on their strength of schedule. This method correlates more strongly with the number of wins a side has, by the way.

What it shows is that West Coast’s percentage, adjusted for their opponents, is closer to 130 per cent than 160 per cent.

Ok, so West Coast are good. And Essendon are not, despite what their list profile says. Is there anything else that’s stuck out to you over the past fortnight of AFL football?