The Roar
The Roar

Gordon P Smith

Roar Guru

Joined February 2017







I'm a Yank who stays up Friday and Saturday nights from February through September to get my fix o'footy.



Yes – replicating last year’s All-Australian roster would be an admirable achievement for a team I see dropping several places on the ladder. And while beating Melbourne by slim margins during the season was impressive, it was a sign of the Demons’ growth as a team that they won handily in finals, rather than any change in Geelong’s play. I’m not suggesting the Cats are far off the mark as much as that other teams around them are growing closer to that mark. (Or at least, that’s what my numbers say. My late wife was a Cats fan and they’ve always been close to my heart. I’m hoping you’re right…)

How about an encore? Comprehensive AFL season previews (Part 1)

Love the discussion board. As with all pieces like this, the best reward is the discussion of where you disagree with the assessments provided, so thanks for the intelligent feedback, everyone. I’m sure my forecasts are just as wrong as everyone else’s will be.

And I’ll freely admit that in a few cases, my gut feeling is different than what the numbers in my research are predicting. I’m not personally bullish on either the Hawks or Swans, but I said that last year, too, and they ended up playing R23 for the last double chance. I have a hard time looking at Richmond’s depth and not asking why they’ll stay up while the Giants give up similar depth and fall by the wayside.

But I’m simply reporting what it appears to my numbers to be the most likely outcome in 2019, barring the unforeseen. It cost me nothing to share, it costs you nothing to refute, and in the end, we’ll all get to watch how the season unfolds together. Thanks for the forum, and glad you cared enough to pitch your pennies into the well, too.

How about an encore? Comprehensive AFL season previews (Part 2)

The meta-rankings, like the Brownlow, is a week-to-week cumulative scoring. As many games as Josh missed in 2018, he simply couldn’t tally as many points as he “deserved”. If you made me predict next year, he’d have to be a top 50 favorite. *Assuming good health*!

The comprehensive end-of-year review: In summation

Thanks to both of you. You’re right, of course.

The top 20 best and worst games of the 2018 season

I want to give credit where credit is due. There are a group of us who work with the ELO-Following Football system, not just for the AFL and AFLW but for the NCAA (my other specialty), the NFL, Canadian Football League and English Premier League. I write the articles for the Roar, but I can’t claim all the credit for the work. (That’s why all the original posts had “we” pronouns, instead of “I”.) Thanks to Stuart, Joe, and Dana, plus everyone else who’s helped with the testing over the years.

The comprehensive end-of-year review: Adelaide Crows

I really missed your game previews when you weren’t submitting them midseason, mastermind. You’re remarkably thorough with both player and team trends and habits, even to the point of the Olympic year correlations! Thanks for your hard work and well-written features. (And while I don’t have a horse in this race, I actually think West Coast wins by a goal or so.)

2018 AFL grand final preview

Excellent article on a fairly obvious point – neither team was assumed to even make finals, let alone the GF, and once you factor in the injuries that both team faced this year, it seems almost unimaginable that they made it this far in 2018.

And two 13ths in a row is interesting but not definitive, Bulldog fans. (Having said that, Footscray looks like they indeed have the tools to do precisely what’s being implicitly predicted for them.) It seems more reasonable to look at Adelaide’s disastrous season and assume they’ll bounce back next year, and to look at Brisbane’s internal progress and assume they’ll make the big leap next year. (Unabashed plug: I’ll put all that together in team analyses for The ROAR once the season’s over.)

Instead, let’s just enjoy these two storied franchises getting this unexpected shot at another cup upon the shelf. Let’s enjoy watching Steele Sidebottom get another shot at the ultimate prize towards the end of his storied career, and “Mason Freaking Cox” getting a shot at the “Super Bowl” early in his. Let’s enjoy watching the Eagles prove that you don’t need Gaff or Naitanui when you’ve got Darling and Kennedy on the field, and that playing on a field that more closely resembles the MCG at home will make their visit to Melbourne feel more familiar to them than it was three years ago.

Most of all, here’s hoping for a one-goal game, where the result isn’t obvious at halftime!

Nothing in the AFL makes sense anymore

Without taking a side on the roaring debate taking place in the comments about the merits or hypocrisy of the article itself, let me add a foreigner’s perspective to the topic. Or at least an American’s.

I can’t remember ever hearing any US player talk about being traded or retiring early or whatever because they were “homesick” for another part of the country. (I have heard futbol – “soccer” – players say that after living abroad for a while.) The closest I can think of is LeBron returning to Cleveland after four years in Miami to “bring a title to my hometown”.

The topic has fascinated me from afar for quite a while, and makes me wonder if that’s a fundamental difference in Aussie v Yank cultures. We also have boarding schools but what you hear universally is that it’s the “privilege of the rich”, not that it makes the kids homesick, although they undoubtedly are at times). I went to college about 800 km from home, and after a tough first couple of weeks loved it. (Full disclosure – I did have to come home a couple of years in, because of my mother’s impending death. Related but not the same.)

The comment that hit home to me, following professional sports from a non-participant’s perspective, is that they’re getting paid far more than I ever made to play a game for one of the top 18 clubs in the country. So, either they need to understand that’s the nature of the business they’ve chosen – you can’t live just anywhere to do your particular job; a marine biologist probably has to live by the ocean – or stop using it as an excuse to get out of a club situation they don’t like.

Homesick? Suck it up, be a man, and blame someone else instead

Maybe I’m not the first to point out that Mr. Dangerfield wasn’t mentioned in the article, although you included him in your list at the end. Not sure you want Dusty at full forward, either, but that’s up to you.

My predicted All Australian team

I love it. That’s what the mixed rules cup with Ireland’s Gaelic stars is supposed to be.

Who will play the All Australian team?

Here’s a highly minority viewpoint.

As an American, the evening games tip-off at 3:30 in the morning, and finish about 6, 6:15. Three of the four games start in the middle of the night for me – three nights in a row; the other one around midnight. None in my cherished “Saturday afternoon” fixture where I can take it in without needing medication to stay awake.

I hate adjusting my life around finals. (I do it anyway.) Suck it up. We only get nine more games this season; treasure them.

The Week 1 AFL finals fixture is a disaster

You didn’t address why Shannon Hurn is your choice for captain. He’s worthy of AA but I don’t see him over a few more experienced souls you named.

Here's who should be in the 2018 All Australian Team

If the Brownlow was strictly a “Most Valuable Player” award, I would back your opinion completely, although I’d also support a Max Gawn campaign this season, and a Lance Franklin campaign any season.

But the game-by-game nature of the Brownlow lends itself to supporting a candidate like Mitchell, who in the majority of Hawthorn’s games was the most significant player on the field. The Hawks have 13 wins (at the moment I write this), so if he got three votes eight times, it’s easy to think 24 votes might win the medal right there. That doesn’t mean he’s the most valuable Hawk, let alone the most valuable league-wide. But unlike last year, that’s a horse of a different color.

The Brownlow rewards the players who get recognized like the column that got cut out of the table in my article – who stands out in how many games? Mitchell and Gawn have been recognized by 80% of our sources in ten games each this season, and those usually translate to two or three votes per game. (Fyfe, Dangerfield, Grundy, and Oliver are next with seven each.) That, too, is why I’m fairly certain it’ll be one of those two men winning the Brownlow this season.

Two weeks left: Who's our player of the year? Who wins the Coleman?

Sorry. It’s the thesaurus lover in me. I got tired of “field”, “oval”, “grounds”… resorted to pitch. (And I’m not from Sydney – I’m from Idaho.)

Last weekend I witnessed something never before seen on an AFL pitch

I wish anon’s comment wasn’t necessary, but his point is well taken. So far I haven’t seen that this makeup of the tribunal has been “starstruck” – five weeks for Cameron was probably about right, especially since Andrews came back in four. And Justif01’s connection to Bugg’s ban is superb: six weeks is probably right. Maybe just “the rest of the season”, to be safe. It was unquestionably more heinous than Cameron’s crime.

Gaff set for big ban, Eagles win derby

It was a goal. No question. We constantly hear the phrase “he threw it on his boot” and praise the player involved for finding a way to get a kick through the sticks. This is no different, just more spectacular. (But it’s fun to argue about because we get to relive the goal of the year over and over while doing so.)

With four AFL rounds to go, every spot in the eight is up for grabs

I agree completely, James and Cat. I had the Suns pegged to finish “19th” this year because by the time they’d lost eleven weeks of road games, they’d be too toasted to be competitive regardless of the site. It was a pleasant surprise to see them go 3-2; Dew deserved some reward for his work, and not taking the spoon is significant.

How bad is too bad?

To be utterly frank with you, I have on several occasions tried to refine this system by accommodating top players in or out using a variety of valuations (player ratings, fantasy values, you name it), but it was so time intensive and (amazingly) had such a minimal result that it wasn’t worth the effort. Having just retired, it’s possible that I’ll take another shot at it, but my results have been good enough without doing so that just adding a bit of common sense (GWS got some players back, so hedge your tips that direction), you can be pretty accurate with your picks.

A math geek's guide to the true AFL ladder

I chose these four because every one of them exclusively use only the final score and treat the team as a whole – no breaking apart offense and defense, no player injuries accounted for, just the raw end results. What differentiates the results of the four systems is how they weigh immediate scores compared to older results. That basically decides how quickly the rating responds to one good or bad game – another way of saying how likely the system thinks one aberrant result is part of a trend, like the positive outcomes for the Magpies after about round three or so, or a one-off that should be taken with a grain of salt (think last year’s Saints rout of the Tigers in R17).

I don’t have a short answer as to what can be gleaned from these results that the ladder itself wont answer, although the predictive accuracy of all four systems significantly exceeds picking by ladder or percentage position. But here’s a tidbit that you CAN glean from the ladder that most people don’t realize – if there’s a team at season’s end whose percentage is much higher than their win/loss record would indicate, bet on them making a BIG jump next season. Collingwood was that team last year. (It doesn’t consistently work the other way, BTW.)

A math geek's guide to the true AFL ladder

FT, the quick answer is that doing the variances produced close enough to the same results as the much simpler (and less mathematically justifiable, you’re right) averaging of the four sigmas that for a mainstream site like this it made much more sense to go the simple route for public consumption.

As for your second question, I didn’t “allow” for the differences – I felt that those differences highlighted the different ways the four systems analyzed results. The “lumpy nature” of the draw is not completely immaterial, and both WF and FMI try to account for it more than the Arc and FF do, but the nature of all four systems is to compare expected results rather than raw scores. If the Giants only defeat the Saints by, say, twenty points, then they’ll drop in any of our system, whereas if the Crows beat Melbourne by twenty, they’ll move up. So it’s almost immaterial which teams play whom. (Now, accounting for playing hot teams is nearly impossible…)

A math geek's guide to the true AFL ladder

AD – Thanks for all your great work! Can’t imagine anyone hopping in to fill your void will be able to replace your pithy succinct style and opinion. I keep an eye on your rankings every week to see how they compare with my ELO-Following Football rankings, even though they’re completely different formats. You were always able to take a fresh look at each team each week and reacted without overreacting to the tosses and turns of the maelstrom that is an AFL season (see: Gold Coast over Sydney, for ex).

Good luck with your travels and your future plans – you will be missed!

AFL Power Rankings 2018: Round 18

Re: Format change.

My assumption is that the length of each segment, particularly the lead, will depend on the worthiness of a lead topic that weekend. Interesting subject = longer first portion. Nothing stands out this week = fairly equal segment lengths this week.

We’re going to read your Hot Takes regardless because we’re interested in what you have to say and we like your writing. How long each segment is will be irrelevant to me, and probably most others. Write however it works for you that week.

Just keep writing, please.

Ten hot takes from AFL Round 18

That opens up an interesting line of thinking that I also track: if you add up all the points for all the players on each team so far this year, here’s the ranking so far as of R17:

Melbourne – 1978
Collingwood – 1967
Richmond – 1907
West Coast – 1868
Geelong – 1780
Sydney – 1763
Hawthorn – 1705
North – 1703
GWS – 1691
Adelaide – 1585
Port – 1572
Brisbane – 1563
Fremantle – 1524
Essendon – 1466
Western BD – 1301
St Kilda – 1242
Carlton – 1219
Gold Coast – 1020

It’s not terribly out of line with the ladder, but it wiggles in curious places. Brisbane’s been floating near 12th most of the season, more so over the last six or eight weeks (they were 16th after R10 and 13th after R14). Melbourne was as low as 13th in April, but jumped from 11th (R6) to 6th (R8) to 2nd (R14) and took first two weeks ago. Geelong has several players up over 100 points but not nearly as many in that mid-range; eyeballing the list, I’d guess maybe 4-5 Cats would show up in the next 100 players (Menzel, Tuohy, Blicavs, Parfitt, Stanley). On the other hand (and injuries have helped this greatly), the Giants would probably have 13 players in that next 100, maybe 120 names. Normally, you wouldn’t think of Delidio, Haynes, Greene, Hopper, Patton and Shaw as “second tier” players, but this season they’ve had second tier numbers just from playing time alone. It can be a deceptive list when you have players out for awhile; that’s why I’m amazed Buddy’s still in the top ten. even missing three games.

The AFL's top 99 players in 2018

Thanks very much, Slane – you said it better than I could have myself! Peter, I did gloss over the list because it’s pretty long – besides some smaller sites from different areas (trying to balance that Vic-centric tendency), I tag the Age, the Roar,, and a host of other sites including the various fantasy scoring sites like Supercoach to get some of the stat-hounds credited as well. My biggest concern, less than the prejudice for certain clubs, is the prejudice to recognize midfielders over all other positions. So I’ll admit I do try to seek out polls that are less mid-biased and occasionally recognize defenders, for example.

The AFL's top 99 players in 2018

One minor error I caught after sending this article in: Fremantle’s current rating jumped up to 31.5 following their upset of Port – I recalculated their six remaining games with that number, but forgot to update the graphic at the top of their part of the chart.

I’m sure the bright-eyed among you will catch a few more mistakes, and then the eighteen teams will throw wrenches into all of our predictions and make the entire thing a waste of calculations!

Six rounds to go: Every club's run home