St Kilda’s record going into the bye was just a little too cute. Against the bottom nine they hadn’t lost. Against the top nine they hadn’t won.
What better way to break a pattern than by going right to the top.
A thrilling win over Geelong gives them their first big-time scalp of the season, and a much-needed one at that, given margins against top opposition this season had ballooned into the 80+ and 100+ zones.
There were plenty of positives to go with the outcome. Seb Ross hurt Patrick Dangerfield the other way with 33 disposals. Jack Steven was superb when it counted.
They won the clearance count and were on top of the inside 50s most of the game – and then soaked up the pressure when that changed late.
It was a good win, but it’s also worth noting that when it comes to Geelong, we’ve been here before.
For all Saturday night did to break one team’s trend, it reinforced an odd trend of the other – the Cats seem to play to their opposition, excessively so when it’s an opponent outside the eight.
The challenge for St Kilda is to prove they weren’t just as much in the right place at the right time as Collingwood and Carlton.
The challenge for Geelong is to figure out why this is happening.
+1 // Ladder: 6th (9W, 4L)
The win over North Melbourne could’ve been more emphatic given the Crows scored the most behinds in a game since the Brisbane Bears in round 11, 1995. But let’s give credit, that’s now three wins over top eight sides in their last four matches. The other game in between was an 88-point demolition of the Saints. This team is flying.
+1 // Ladder: 4th (10W, 4L)
Speaking of flying, the Giants can’t do much wrong either. Against Carlton, the trio of Dylan Shiel, Callan Ward and Stephen Coniglio had 98 disposals, 23 clearances, eight rebound 50s, 14 inside 50s and three goals between them. Damaging all over. And how about Zac Williams gaining 732 metres at 92 per cent efficiency?
-2 // Ladder: 3rd (10W, 4L)
On the bright side for Cats fans, you won’t play any teams outside the eight in finals. So that’s something.
+1 // Ladder: 8th (8W, 5L)
Week off out of the way, it’s a home game against Essendon coming up this Thursday.
+1 // Ladder: 2nd (10W, 3L)
Swans are back in action with a Saturday twilight blockbuster against the Western Bulldogs at the SCG.
-2 // Ladder: 1st (11W, 3L)
The Hawks may have moved to the top of the ladder, but they’ve slipped here due to their shaky start against lowly Gold Coast. Hawthorn have the second-lowest percentage of the top eight sides, which is a tad concerning.
No change // Ladder: 7th (9W, 4L)
Bring on the Swans. Should be a good game. And you just have to love that we’ve got these type of fixtures every single week this season.
No change // Ladder: 5th (10W, 4L)
The last three weeks were a test of where North were at and while in-game injuries muddied the waters somewhat, we’ve probably got a better read on them. The Roos can compete with the best and are very much a threat, but we shouldn’t pretend there aren’t seven other teams you can describe in a very similar way.
+2 // Ladder: 12th (6W, 7L)
Now that’s done, let’s move on to this trend: it’s been over a year since St Kilda have won interstate. Gold Coast at Metricon Stadium is their chance to change that.
No change // Ladder: 9th (6W, 7L)
Are the Power ready to get serious? Then they absolutely have to beat Richmond at home on Friday night.
+1 // Ladder: 10th (6W, 7L)
Host the red-hot Adelaide at the MCG on Sunday. Will need to bring their A-game.
+3 // Ladder: 14th (5W, 8L)
Full credit to the Herald Sun for illustrating the historical significance of Friday’s low supporter turnout against Fremantle. The last time Collingwood played in such a small crowd at the MCG, “Australia was at war with Germany, but Japan was still 19 months short of bombing Pearl Harbour.” Bleak times for Pie fans, it would seem.
No change // Ladder: 13th (6W, 7L)
Bryce Gibbs had 29 disposals in his 200th game, but the Blues faced one of the toughest assignments in footy travelling to Spotless to play GWS. Arch rivals Collingwood at the ‘G will be a more comfortable environment.
-5 // Ladder: 15th (3W, 11L)
The Dockers were off from the opening bounce against Collingwood. To have one goal at half time was quite the contrast on the previous three weeks. Any air in that bubble didn’t survive the flight over.
-1 // Ladder: 11th (6W, 7L)
His coach wasn’t as impressed as everyone else, but Dustin Martin was the standout in the Tigers’ win over Brisbane. He’s averaging 2.5 more disposals a game this season and 2.4 of those are on target, going by the increase in his effective disposal numbers. Surprisingly, this team has won five of its last six, but it’s hard to read too much into that given who four of those wins were against.
No change // Ladder: 16th (3W, 10L)
Led at quarter time and half time (just) against the reigning premiers. A good first post-bye step. Oh, and Gary Ablett had his first 40-disposal game in two years (well, 729 days, but no one’s counting, right?).
No change // Ladder: 18th (1W, 12L)
Back in action on Thursday night but it’s a difficult ask: West Coast at Domain Stadium.
No change // Ladder: 17th (1W, 13L)
For a while there, it did seem Justin Leppitsch’s post-game press conference a fortnight ago bought some clear air around the Lions. But it’s difficult to see that being sustained if the underwhelming performances continue. Somehow, there were 10 players who didn’t get above 10 touches against Richmond. Bright spot was the three-goal haul of second gamer Eric Hipwood.
Macca
Guest
No Dalgety it is a determination based on what a Power Ranking is designed to do and the obvious disconnect between that and what this formula is producing. The weightings I would give measures may not be correct but that doesn't change the fact that this formula isn't producing plausible results. And you haven't answered the question - you have deliberately fudged because if you answered it your whole argument would fall like a house of cards. To help you answer here is what Michael said in the first article of this series was the purpose of the ranking; "Rather, the AFL Power Rankings are designed to answer one simple question for each team: Where are they at?" - So do you think Freo was "at" 9th?
Dalgety Carrington
Roar Guru
So it's an assumption then, based on what you think it "should" be. I could just as easily say the dimensions I give above are the way it "should" be measured. And I've answered that question Macca.
Macca
Guest
Dalgety - "Your list you’ve just reeled of would be fair if they are the measures Michael is using for power rankings" that is why I am able to use the word "should" - Michael should be using these factors if he wants his ranking system to work properly and give an accurate representation of the form teams of the comp. I am confident that Michael isn't placing enough weighting on these things because if he was we wouldn't be getting so many strange results. And what Michael is trying to measure isn't in question - it is the metrics and the weighting he uses - so I'll ask again do you think Freo was the 9th ranked form team of the comp last week?
Macca
Guest
8
Dalgety Carrington
Roar Guru
I've answered that before Macca. I can't take issue with the actual rankings themselves without knowing what the power rankings are measuring exactly. Your list you've just reeled of would be fair if they are the measures Michael is using for power rankings...or are you just assuming?
Macca
Guest
"Well if you know what you said why shift tack?" Where did I shift tack? "It’s a ridiculous idea to say a measure can’t be sensitive and worthwhile at the same time" True but it isn't a ridiculous idea to say something that is overly sensitive isn't worthwhile. On your critique of the Richmond Freo situation - yes Richmond were above Freo when they played them- that shouldn't impact the result of a power ranking when measuring them head to head - did Richmond beat Fremantle - Yes, was it by a considerable margin - yes - was the game within the last 6 weeks - yes - these factors point to Richmond being ranked above Freo. "Richmond beating Sydney is not a surprise" - maybe not but that shouldn't impact the power ranking - Sydney have been around Michaels top 4 all year - Richmond beating them withn the last 6 games should give them a significant bump and as they are vastly superior to anyone Freo have beaten should point to Richmond being above Freo. And Again Richmond have won 5 from 6 - Freo have won 3 from 6. " the context isn’t just who, but the how of their wins (size of margin). " I would inverse this - it isn't just the size of their margins but who - the beat up on 2 terrible sides and just beat an ordinary side in Perth and managed to climb to 9th - that is a sign of a formula not working. Answer this simple question - do you think Freo's ranking of 9th last week was too high?
Dalgety Carrington
Roar Guru
Well if you know what you said why shift tack? It's a ridiculous idea to say a measure can't be sensitive and worthwhile at the same time, which is understandable though if you're rigidly locked into a fixed idea of what a measure is supposed to be. Richmond were three places above Freo in ladder positions when they played (ladder position). Richmond beating Sydney is not a surprise, they've been a challenging side for the Swans for a number of years now (result precedent). Also for Freo, the context isn't just who, but the how of their wins (size of margin). One of those under 50 point games was in a near cyclone (external conditions). I don't know how much any of these are factored in, but the main point is perhaps you're overly fixated on the W/L.
Macca
Guest
Dalgety - I know what I said and I stand by it - t shouldn't have swings that big if it was working properly - the Freo result this week occurred because of errors from previous weeks, ie over rating wins against terrible opposition. As for Richmond V Freo - while your point re the Brisbane results would be a plus for Freo the fact Richmond beat Freo in Perth 11.17 - 83 to 6.9 - 45, has a win against second placed Sydney to its name and 2 more wins overall in the last 6 weeks should override that - especially as Freo has twice been held to under 50 points in the past 6 weeks while Richmond has been under 80 just once. Even if Michael is heavily weighted to percentage over the last 6 weeks Freo and Richmond are only fractions of a percent apart.
Dalgety Carrington
Roar Guru
What you said was it shouldn't record swings that big. My argument is more on the general principle of where we have an indicator that can be sensitive and that is perfectly valid if it's measuring what's intended. I guess I also like that is appreciably different from the ladder and places a different emphasis on W/L. Michael is under a constant attack on this, and I reckon it's often under the misconception of what the thing measures. I can understand where it might be a struggle for Michael to keep up with that. As for the Freo vs Tigers thing, you could look at other factors, like say Freo beating the Lions (a team that just the week before was above them on the ladder) by near to double the margin the Tigers did, away from home and with a crippling injury list, while the Tigers a close to full strength lineup at home when they got over the Lions, barely two weeks later.
Macca
Guest
Dalgety - "Who says a sophisticated model is not useful if it has a high degree of sensitivity from round to round? That might be the exact thing Michael is looking for" Michael said just a week or two ago that he tinkered with his model to remove results like teams moving plus 4 after beating a lowly side - I think it is safe to say he doesn't want wild swings. "open and observant to what did happen." I will tell you what did happen, Freo won 3 game in a row, 2 against sides that have 2 wins between them and a combined percentage of 58.7% and the third against a side that whose wins have been against sides currently sitting 10th, 11th, 12th, 14th, 17th & 18th on the ladder but because Michaels formula is too percentage heavy and doesn't put enough weight to number of wins and quality of opposition it in correctly had Freo sitting at 9th - above teams like Richmond who have only lost 1 game in the past 6 weeks.
Dalgety Carrington
Roar Guru
Who says a sophisticated model is not useful if it has a high degree of sensitivity from round to round? That might be the exact thing Michael is looking for. Shoulds & shouldn'ts when used in a past tense tend to fix attention on our preconceived notions of what we thought would happen, rather than being more open and observant to what did happen. For instance, form is relative and could be linked to relative positioning as well so lower teams could have greater jumps than higher teams on the ladder.
andyl12
Guest
"What exactly is there to fear?" That no matter how much a side thinks they're dominating, that Hawthorn's class and experience can change that very quickly. Selwood gets more undeserved frees than the entire Hawthorn side put together.
Richard
Guest
We can beat Geelong in late September 1989 and 2008 ..and the Swans 2014.. HAHAHA
Richard
Guest
September only matters. Who cares who you lose too during the season ,as long as you make the top four it`s anyones game..
Cat
Roar Guru
Why are Hawks rated so highly when they have been absolutely flogged by GWS, shown they can't beat Geelong, nor Sydney. I'd be more worried showing you can't beat the teams you'll be facing in finals than some also rans.
Cat
Roar Guru
What exactly is there to fear? That you will play a good quarter, maybe two? That you will need to have every bit of luck, the bounce of the ball and the umpires whistle to go your way just to squeak over the line by 3 points again? Or maybe we should fear that round 6 performance showing up again?
Richard
Guest
So why are the " premiership favourites " Geelong Cats power ranked at 3rd, when Saints ( 12th ) Caaaaarton (13th ) and Pies (14th) have beaten them ?? Keep pumping up Danger and the handbagger boys.... Us Hawks fans love sitting back and quietly watching .. ..
Richard
Guest
Should have , Would have, Could have.. blah blah blah ..Sorry Bill.. You didn't..
Richard
Guest
% is irrelevant if you win more games than the rest..hahaha
Macca
Guest
That's true PaulD - and that is my point - if this is a model sophisticated enough to be worth a weekly article and our time reading an d contributing to it it shouldn't have the results mentioned above - if however it is just Michael randomly assembling a list then the roar shouldn't be publishing it under the "expert" banner and we as readers should be pointing out the obvious flaws.