Have the top five started to pull away from the competition already?

Currently the 2017 season of the AFL looks like a small number of really good teams and then an unusually long tail of ordinary but not terrible teams. Only six clubs currently sit above the competition average on our combined rating covering midfield, offensive and defensive strength. By contrast, the bottom ranked side (Carlton) is closer to the pack than any team has finished in our ratings for any season since 2009 (Melbourne).

Round 5 ratings

Adelaide, GWS and to a lesser extent Port Adelaide currently look well ahead of every other team in terms of their strength, with Geelong and Richmond also looking pretty solid on the back of great output in one part of the ground. The gap from Richmond (5th) to Melbourne (6th) is currently larger than the gap from Melbourne back to Sydney (15th). Likewise, the gap from Adelaide in first to Richmond in fifth is greater than the gap between Melbourne in sixth and Carlton in last.

We should note that it is early days and big movements can happen across a season as opponent strengths become more reliable and we get a larger sample of games. However, the question may simply be whether the lead group contains three, four, or five members and whether anyone can break out of the peloton composed of at least half the competition.

Last year HPN drew a line at 105% as the lowest a side could be and still be a viable premiership chance, which was based on the previous 20 years of ratings. The Bulldogs, a historical outlier, only just scraped by this barrier on an unadjusted basis in 2016, and were just under it on an opposition adjusted basis. On this basis all of the top five teams still remain viable premiership chances – but we note (yet again) that the season is very young, and a lot will change between now and September.

Some teams pop out of this as not really getting results commensurate with their apparent strength. The Demons sit 6th here but only have won two games, having let three close games slip, and been unlucky with structure-wrecking midgame injuries. The Bulldogs, now at 4-1, continue to be unimpressive.

By contrast, perhaps Richmond and Fremantle, popularly seen as over-performing or lucky so far, roughly deserve their respective 5-0 and 3-2 records.

This season shape as it stands right now is an intriguing one, as we can see comparing to past completed seasons. The dense cluster around 95% is pronounced:

annual ratings.PNG

It would be relatively unusual for the season to finish with no team below 90% of the competition average. We should expect some stragglers to fall further behind to where strugglers end up most years.

Similarly, Adelaide are currently tracking as strongly as Geelong of 2007 and nearly approaching Essendon of 2000. They do look very good right now, but if they can sustain these offensive, defensive and midfield efficiencies across a season, it would make them one of the all-time great teams and be a massive story in its own right, especially if the Giants don’t manage to stay with them.

In terms of specific line strengths, right now as we look at our team ratings, only the top three sides are rated above average in all three areas. Adelaide are continuing 2016’s historic performance in offensive efficiency but have added a greater dominance of inside-50 opportunities, meaning if they don’t have dominant midfielders then at least they are still dominating in the midfield area. Port Adelaide and Geelong both look like they’re strongest through the midfield as well. Richmond and Geelong also both have one line (Defence and Offense, respectively) sitting at least 10% above average.

Currently it looks as though Essendon have the worst midfield, Collingwood the worst forward line and Gold Coast the worst defence, but none of those sides shape as the worst team overall. Those three teams are counterbalanced by other lines – Essendon have defended well, Collingwood’s midfield looks strong, and the Suns have been strong between the arcs if not within them.

Only Sydney, Hawthorn, Carlton and Fremantle are below average across the board so far this year and of those, Sydney’s defence, Hawthorn’s defence and Fremantle’s midfield are above 99%, so pretty much on average.

Some sides with pretty uneven strength include North Melbourne, West Coast and Brisbane. North have exhibited a pretty good defence, meaning they haven’t conceded scores from inside-50s too often in spite of their strong opponents. Rather it’s the sheer quantity of inside-50 opportunities allowed (identified in the midfield strength rating) which have brought them undone. The Eagles’ offensive efficiency has been their saving grace, while Brisbane’s offence has been league average but undone by the sheer lack of opportunities delivered by their midfield.

Advertisements

Are North and Hawthorn bad, or have they just faced a tough draw?

Almost every credible (and non-credible) football journalist in the country has waxed lyrical about the decline of the Mighty Hawks, and the Considerably Less Mighty Roos. Both sides were finalists in 2016, but have gotten off to tough starts in 2017, a combined 0-8 through a month of footy.

Each rating is an inside-50 based efficiency measure, and we adjust each team based on their opponents. The following is our rating of the strength of the four opponents each team has so far played:Round 4 opposition ratings.JPG

North Melbourne and Hawthorn have played relatively tough draw sets so far. At this stage, this should be treated with some caution, because every game is a quarter of the sample for each club.

It just so happens that most teams have faced a spread of sides so a couple of tough or soft opponents can stand out.

Hawthorn for instance is rated as having had a difficult draw due mostly to playing a highly rated Adelaide and a pretty well rated Geelong. Those two alone push the average up because at this stage few other teams have faced two teams rated in the top third. The Gold Coast midfield and Essendon defence also push up the averages Hawthorn have faced, although it must be borne in mind that the Hawthorn games themselves contribute 25% of the weight to those teams’ ratings.

We deliberately factor in no historical data in these ratings, so the strength of each team can take a while to settle down. It also means the best 2016 sides get zero credit as tough opponents if their forward, midfield and defensive data in 2017 does not warrant it. This approach makes the early season jumpy, and it means we’ll have a better picture as the season progresses. Just like what happens in reality.

As we are now into our second week of tracking the HPN rating we can start to track the sides that are on the improve as shown above. We can see the week to week movements by teams, which is basically jet fuel for HOT TAKES.

Round 4 ratings

North’s draw of Geelong, GWS and West Coast is clearly the hardest set so far, which has a major impact on their standings in our ratings – their competitive performances against good sides mean they rate well. While Hawthorn do look abjectly terrible, we caution against pessimism regarding the Kangaroos until more form is exposed. They might be okay. Might.

While most would rate the Bulldogs as another tough opponent for North, we currently beg to differ as the above chart indicates. The reigning premiers (this still feels strange to type) have seemed pretty scratchy thus far. Nothing about them currently stands out as terrible, but their performances to date look below the league average in all three areas of the ground, once their soft opponent set is factored in. Sydney, Collingwood and Fremantle all sit bottom half, scaling the Dogs’ ratings down with them.

Fremantle bolt up the ratings this week after beating Melbourne, while Essendon continue to slump as the win over Hawthorn loses its shine and more results come in. GWS, the presumptive pre-season premiership favourite, have claimed top spot by virtue of a solid defeat of Port Adelaide. Whilst Adelaide beat GWS in round one, GWS’s performances since then have been enough to make up the deficit on the day. The two form teams of the competition don’t face off again during the season, so we will likely have to wait until finals for the (already) highly anticipated rematch.

Richmond have mostly beaten up on poor sides (Collingwood, Carlton, Brisbane) but have done it well, and their defeat of West Coast looks good. We’re still not sure they’re as good as third in the competition, but there’s no evidence to the contrary yet.

The Giants’ midfield is their strongest area relative to league averages, while Adelaide continue to reproduce their 2016 offensive prowess – they are simply an amazing side at converting inside-50s into scores. The best defence, currently, is somehow Port Adelaide, and we’ll be fascinated to see if this holds and to try to explain why if it does.

In order to win on the field, the AFLPA must win off it first

It’s hard to argue that rich young men deserve more money. The sympathies of the average person often go to the minimum wage battlers before it goes to footballers on a comparative motza. And that’s fair – no-one at HPN will argue for a second that you shouldn’t read up on Fair Work Commission decisions on the minimum wage or penalty rates.

But ultimately this is a sports (primarily AFL) website, and the AFL is currently locked in an industrial dispute with it’s players. And it shows no sign of ending. Despite a recent offer to the AFL Players Association to attempt to end the more than year long stand-off, it appears that a final deal is a while away yet. The issues as reported appear to many and varied, with one major sticking point.

Cash money.

Earlier in 2017 the AFLPA was pushing hard for a fixed percentage of total revenue – a contentious issue in the negotiations that seems to fallen off the radar of the most recent round of coverage. Instead, the most recent reported offer to the AFLPA has been a 20% increase immediately, with 1% increases for the next five years.

While this seems generous in isolation, the AFL recently negotiated the single biggest increase to broadcasting rights in the code’s history. This is a massive new income stream, and it makes the offer seem comparatively… well, have a look for yourself:

tpp

In this situation, AFL Broadcast Rights include both TV and Radio rights, which have both exploded in value in recent years. The most recent TV rights deal was for $2.508 billion dollars over six years, a roughly 60% increase on the previous deal. Whilst broadcast rights don’t represent the total revenue of the AFL, it provides a decent simulcra to the state of the AFL’s finances.

However, we know from our post on club revenues that the AFL distribution of funds (including broadcast rights) provides a small fraction of most clubs’ revenue.

A fair question would be to ask how the rise of TPP compares with the wages of the average Australian worker, which we luckily have in graph form:

tpp2.png

There is little doubt that compared to the average Aussie worker, footballers have had a pretty decent run over the last 20 years. However, when compared with the rapidly-increasing revenues of the AFL, it’s hard to mount the argument that the players don’t deserve a significantly larger share of the increased success that the AFL is seeing.

Some may point to extra effort being put into grassroots football, or the introduction of the AFLW competition, by the AFL as a mitigation for offering the players such a deal. This argument works at face value, until you find out that it cost the AFL about $4 million for the clubs’ contribution to the AFLW’s running costs – a small fraction of the money being negotiated above. And the AFL spent just $41 million on game development grants, and $16 million on developing new markets. The 2016 AFL Annual Report indicates that total AFL revenues have increased by over $200m in the last decade – a number that will increase by another $175m in 2017. In the same period player wages have increased by less than $100m across the board.

Previously, HPN has assumed that there would be a 50% rise in the salary cap to about $15m per club – a figure that would only amount to between $280 and $320 million of player payments across the entire competition (depending on the spread of increases per year). This would adequately compensate for the loss of the veteran’s allowance, and the removal of the rookie list (with a subsequent expansion of the main list).

While other issues exist on the bargaining table, one can’t help but feel that they would proceed quickly once the issue of pay is sorted out.

An extremely early look at HPN’s AFL Team Ratings

Three weeks of the AFL season are past us and…honestly, we don’t know who is good or bad yet. Are the Suns set for another long year, or is there light in the tunnel? Is Ross Lyon about to be sacked, or is he a genius for the Dockers upset win over the Bulldogs? Should the Tigers make a big offer to Dustin Martin, or for some unknown reason not offer one of the best players in football the money he deserves?

WE. DON’T. KNOW.

We are going to look at how the team strengths sit (according to the HPN Team Strength Ratings introduced last year) and see what rash judgements we can bring to the table.

In 2016, HPN did not run our team ratings this early in the year; unsure what affect the volatility of the small sample size would have over the data. However, we think there are some interesting trends emerging early, some which will hold and others that won’t.

Across a 22 game season, each week of football is worth roughly 4.5% of the total season. However, after round three each game for each team represents 33% of their sample – a very high amount. To attempt to counter this, we have (as we did last year) adjusted each Rating for the opposition strength as determined by Ratings in their opposition’s other games. In short, this should make the data less “noisy”.

Round 3 ratings

As you can see, the consensus best team so far in 2017 (Adelaide) holds the top spot on the HPN charts as well. Two of the three currently undefeated teams hold spots in the top four; namely Adelaide and Richmond. Joining the Crows and Tigers at the top end of the HPN Ratings is pre-season presumptive premeriship favourite GWS and early season surprise packets Port Adelaide. It is worth noting that the only loss so far by both the Giants and Power has been to the top ranked Crows. All four sides sit near the top with respect to the Mid Scores, and have scores in each of the three categories above the league average (after adjustment for opposition). These are good early signs for this quartet, but early signs nonetheless.

Also siding with popular opinion is the noted backward movement of two perennial favourites, Geelong and Hawthorn. The Cats and Hawks may have records that are polar opposites, but both sides seem to have slipped a little from last year. If it were not for Geelong’s incredible efficiency up forward, they could be sitting at 1-2. And the Hawks have barely fired a shot yet.

Perhaps the most surprising result of the early HPN ratings is the low grade given to the Bulldogs, who have a promising 2-1 record. The Dogs have taken a significant hit from the opposition adjustment, as they have only recorded close-ish wins (and a loss) against relatively low ranked sides. This should adjust over the coming weeks when they face tougher opposition – pending their ability to win those games. It is also worth noting that the Cats and Dogs record-breaking midfield performances from last year have declined severely in the season to date – something to keep an eye on.

Finally, the Essendon and Carlton ratings were skewed significantly by the extreme weather on display last weekend. This should normalise over the coming weeks.

The Arc/HPN Crossover: Don’t Stress If Your Team Is 0-2

This is a collaboration post between HPN and Matt Cowgill of The Arc/ESPN. Matt is an expert in finding beautiful ways to display footy data, in order to tell compelling stories. And we are HPN. Matt has kindly provided the graphs, and some of the general comments for this piece from Pakistan. Unfortunately, as Matt wrote some of his ESPN article from there this week, this is only the second best AFL Stats article from Pakistan this week.

Earlier this week Rohan Connolly wrote a pretty well balanced piece on sides that start a season 0-2. Connelly stated that since 2008 only one of the 46 sides to start winless from their first two games has gone on to make finals.

Note: Connolly doesn’t include Carlton’s 2013 finals appearance in his count, which we politely disagree with. Whilst they only made the finals due to Essendon’s penalties ruling them out of finals contention, they still played in two actual finals, winning one. Which is one of only two finals wins for the Blues in the last decade. What we’re trying to say is: cut Carlton some slack for once!

But this doesn’t tell the whole story. Since 1994 (the introduction of the final eight), 101 sides have started out 0-2, and 18 have gone on to make the finals. Here’s how it looks as a season progression:

HPNzerotwo.exclfinals-1

Using our superpowers of subtraction, that means that between 1994 and 2007, 54 sides started 0-2, with 16 making the finals from there. Here’s two different views of that information:

1994-2007: 16 of 54 = 29.6%

2008-2016: 2 of 47 = 4.2%

That’s a pretty striking difference, one explained by a few different factors. Firstly, for much of the early part of this decade the AFL had expansion sides present, and a Melbourne team that was barely better. That depressed the quality of the competition and increased the number of sides without a realistic hope of finals.

Secondly, the increase of the size of the competition has made it slightly harder for average sides to make the finals. In 1994, 53% of the AFL made the finals, decreasing to 50% between 1995 and 2010. When the AFL increased to 17 sides in 2011, 47% of the league played in September, and from 2012 on that share has decreased to 44%. In short, this means that it is harder to make up the ground after a poor start, because of the sheer number of teams ahead of them.

However, where there are a large number of teams getting off to poor starts (such as in 2013 and 2014), there seems to be an increased chance of a team making the finals after a 0-2 start. Since 2011, there has been one finalist who has started 0-2 where there have been at least six teams who have done so. It is also worth noting that in 2017 we have eight teams on 0-2; another great chance to test this theory.

But what clubs tend to get off to a poor start, and which of them have turned it around over time?

Pasted image at 2017_04_06 09_51 PM

There are two Port Adelaide dots at the head of that chart, representing the 2002 and 2003 Power seasons. In both years the Power recovered from 0-2 starts to finish as minor premiers, winning 18 games in each year.

The Carlton dot represents the 1994 season, where the Blues finished second and went out in straight sets in the finals.

HPNzerotwo.inclfinals-1

It’s quite possible to get out to a slow start and have a successful season, and the conditions seem to be as good in 2017 as they have been in any other recent year.

Who is your AFL club death-riding this year?

The introduction of future draft pick trades in 2015 has given us, the footballing public/nerds, the wonderful spectacle of clubs quite literally taking bets on each other and then “death-riding” them the following year in the hopes of securing a better pick.

In 2016, for example, Collingwood’s struggles resulted in the pick they traded for Adam Treloar turning into pick 8 (in addition to the pick 7 from 2015), significantly more than they would have hoped to have paid for him. By contrast, Melbourne took a bet on themselves last year and traded their 2016 first round pick for a 2015 one from Gold Coast, a deal that ended up in Melbourne’s favour due to their moderate rise up the ladder.

Thus, we give you, a death-riding chart for 2017:

deathride.PNG

We’ve kept this to the first two rounds as there’s probably not a lot to be gained from pick shifts in the third and fourth rounds.

St Kilda stand to win handsomely from the Hawthorn trade even if the Hawks end up premiers, but if the Hawks miss the eight it will turn from a bargain into an outright heist. Here’s what we said at the time of the trade:

stkhaw.png

St Kilda, for their part, effectively split their first rounder into two useful second round picks while retaining a first rounder of currently unknown quality. The pick split is a move we think works in isolation due to the greater expected output of two picks in this range vs a typical pick 10 but the addition of a 2017 selection is a massive bonus.

St Kilda gave up pick 10 for multiple good picks, and right now there’s a chance that Hawthorn’s first rounder by itself could land around the pick 10 they gave up.

Hawthorn, in a frenzied final day’s trade, were left holding just one pick in the first two rounds, and it’s tied to GWS’ finishing position. This was due to the Hawks swapping the equivalent of more than entire draft in multiple moves to obtain O’Meara. Currently GWS are one of two teams sitting 1-1 so on our death-ride chart the pick looks better, but the Giants are still one of the presumptive premiership favourites (Matt Cowgill at The Arc has them with a 15% chance of finishing in the top 2). If that pans out it leaves Hawthorn entering the draft at pick 36.

Richmond have ended up with Geelong’s first round pick after it changed hands no less than three times. First it moved to Carlton in the Tuohy-Smedts swap, then went to GWS in the deal that moved Marchbank and Pickett. Finally, GWS traded it to Richmond in the swap for Deledio. The end result is that the Tigers will be hoping Dangerwood fails to repeat last year’s preliminary final appearance.

Brisbane acquired Port Adelaide’s first round pick in the Pearce Hanley trade, where the balance of traded goods roughly values Hanley at whatever the Port pick nets them:

hanley swap.PNG

Port Adelaide had pick 9 in the first round last year after finishing 10th and popular opinion (although not that of HPN) had them getting worse and regretting giving up that first rounder this year. However, on early season form that looks less likely.

After a subsequent, perfectly weighted swap with Sydney, Port took four picks in the first 33 last year, including round 1 Rising Star Sam Powell-Pepper. The others, Todd Marshall, Willem Drew and Joe Atley haven’t been seen yet. If Port can get some use out of those other draftees, they’ll be pretty happy with their 2016 trades even before we consider a potential jump up the ladder devaluing their first rounder.

GWS ended up with St Kilda’s second round pick for Jack Steele, which as we noted last year, was a win-win, with the Giants getting extra value from a pick and the Saints getting more from Steele than the Giants would. A Saints preliminary final appearance (or complete collapse after their 0-2 start) would probably be required to alter the pick value enough to change this equation.

The Giants also got Collingwood’s second round pick for Will Hoskin-Elliott, and that bet remains one between Collingwood’s ladder expectations vs Will Hoskin-Elliott’s potential.

Gold Coast are sitting on a glut of second round picks, and should be variously tracking the fortunes of Fremantle, Hawthorn and Richmond this year. They got the Hawks’ second rounder as part of the swap for O’Meara and the Tigers’ pick for Prestia.

Most interesting is probably the Fremantle pick, which was a mundane pick swap and represents a pure bet against the Dockers improving:

freogc.png