A Nerdy Analysis of the 2024 AFL Season

This last AFL season was insane - it was nearly impossible to tell if anyone was really doing well at any particular time. Did Essendon underperform, overperform or were they just led by a couple of stars? Were Sydney really performing on the level of 2000 Essendon? Did Brisbane deserve to win the flag? These are questions that are hard to answer by watching alone.
Luckily, I don't get out much, and so have performed a brief statistical analysis of every team this year, by metrics that I pretty much made up, so you can see for yourself who the studs and duds were this season through a glorified slideshow of Google Sheets graphs.
Like what you see? Subscribe here for more: / @thegaz-man
As for my other links, here they are below:
Patreon: www.patreon.co...
Business/enquiry email: [e-mail]
Twitter: / gazmansports
Want an Orange Team shirt? Get yours here: www.redbubble....
DISCLAIMER:
All match content is the property of the Australian Football League, with additional material from Fox Sports and Channel 7, and is used for entertainment and educational purposes under Fair Use.
All music and backing tracks were humbly, and clumsily, self-made on GarageBand and BandLab.

[@wrestlessradio]

Been a lapsed footy fan for the last 10 years. Was randomly recommended your videos and its been a fun binge over the past month. Can't wait for next season now! Thanks for all effort you put in, hope you continue to grow, you deserve it!

[@thegaz-man]

Thank you so much!!! Glad you've got the footy bug back, I'm hoping to really expand next year so really looking forward to it!

[@SmokeyBandit500]

Great vid as always. A lot of hard work put into this. We appreciate it Gaz 👏 👏

[@thegaz-man]

Thanks! Definitely a lot of data to pour through haha

[@gallifrey27]

12:27 I knew something was off here because there is no way that in any kind of analysis that Garcia could be St Kilda's least valuable player. You have listed some players here twice (Wood, Wilson, Garcia, Phillipou, Hill).

[@thegaz-man]

That's a fair concern - all player positions come from the AFL website stats and some have two - touched on it with JHF but it means some players have two VORPS, comparing the same stats to different positions. In this case it's best to ignore the lower VORP. I didn't notice Garcia doubled up, so that outlying minus score can be disregarded. Haven't engineered a way for the numbers to spit out the best VORP per player in my spreadsheet so apologies for the confusion caused with all that!

[@statchatsports]

As fellow stats lovers, we are OBSESSED with your content! What is the best way to get in touch with you Gaz, we have a video idea we'd love to run past you 😁

[@thegaz-man]

Ahhh thanks! Really like the work you guys do as well with trivia, love that sort of thing myself! The best line of communication would be my email at [e-mail] Keen to work something out with you!

[@lukevacca1537]

Awesome stuff as always. Slightly concerning that my mental health for the forseeable future is in the hands of a 19 year old with negative total vorp and considered the 385th best player in the comp.

[@thegaz-man]

Young mids almost always start out with low VORPS, especially in low-win teams. In terms of being a pure impact player he rates with Coniglio and Rozee which is very promising.

[@Winks3]

Great video. Might be the first long video I've seen in a while that I didn't fall asleep to. Are you going to release any of your spreadsheets publicly?

[@thegaz-man]

Thanks! Yeah the plan is to release the spreadsheets on my Patreon, just gotta clean them up bc right now there's graphs all over the place and data isn't sorted in the best way

[@Parmy7]

You should do a full History of the SANFL that would be great to see

[@thegaz-man]

I definitely want to tackle that someday. I've got something in the works about the league in the 1990s, genuinely think it's the most interesting time any league has had since the throw-pass era. Not sure when that'll be ready but it won't be for a while, I think

[@declanmoore75]

I wonder if blake hardwick doing his preseason as a forward and generally being a swingman throughout the season explains the drop off.

[@thegaz-man]

Swingmen can get hurt or be helped by this system - Tom McDonald had an insane season judged as a key back when he scored 20 odd goals one year. Hardwick probably got hurt by not spending enough time front or back

[@mitchelltidball6944]

My 45 minute slack time in the office is now sorted. Thanks legend❤️

[@thegaz-man]

Haha thanks!!!

[@Jaxtar25]

This is what I come to youtube for

[@thegaz-man]

Glad you enjoyed it!

[@alexanderlaird9344]

Although I do really like this video, it’s also proof stats don’t present the entire story. Dont support either team these players play for, but from the eye test and basic stats, Nick Daicos and Patrick Cripps were the best players this season just in-front of the likes of Neale and Bontempelli in terms of mids. Champions Data’s rankings of highest rated Brownlow winners has Cripps highly rated in the top 10 across a 20 year span. Daicos was just exceptional too.

[@thegaz-man]

My algorithm may look for different things compared to Champion, I think VORP is looking for something slightly different than what Ranking Points do - Daics and Cripps certainly had better starts to the season, though. For me, Bont had better whole ground impact, which is what VORP looks for, and why pure mid Brownlow types often find themselves outside my top 10, as well as my system necessarily giving more top spots to players of different positions. Absolutely agree that you can make stats tell different stories, and that Cripps and Daics had great seasons. I feel these numbers help prove that their vote tallies, as well as some others, were over inflated.

[@ajh22895]

We're in the 23 game era.

[@thegaz-man]

Apologies, my brain is still used to pre-Gather Round fixturing haha I can assure you that past me has locked in all season length adjusted metrics to a 23 game season length.

[@SportsReplays]

Can't believe we got rid of D'Ambrosio and kept kelly.

[@thegaz-man]

As a Crows fan, I'm shocked that Kelly has lasted this long. I hope he and McKay improve for Bombers' fans sakes but I can't see that becoming a solid core.

[@PeekayIP]

Not sure about your analysis of Geelong in this. Anyone who watched them through the year could see they had a massive weakness in the midfield - yet you rated them a top 6 midfield? Might need some tweaking.

[@thegaz-man]

These numbers are a little deceptive - they're based on player performance and don't have anything to do with what the opposition are doing. Geelong's clearance numbers for were high, as was their scoring, assisting and contested ball - hence why their individual mids posted high ratings relative to other teams. However, they conceded the most clearances of any team. My VORP numbers don't really have a way of tracking a midfielder's defensive accountability like that, hence the overperformance. If I create some team stat-based metrics (such as my points per live minute or attacking efficiency comparison) this will better illustrate these types of deficiencies in a team's game, but I'm hesistant to work that into VORP as I try to make VORP as team-independent as possible (i.e. for generating points created through score assists or saved through 1 on 1 contests, I use league average accuracy rather than the player's team). Hope that helps explain how the numbers came up as they did here.

[@PeekayIP]

@@thegaz-man Thanks for explaining. Really appreciate the work you are doing on statistical analysis for AFL. I've always loved the analytics of baseball and felt our sport was a bit lacking but it seems to be a growing area.

[@thegaz-man]

​@PeekayIP no worries! I've absolutely still got a lot to learn about data analytics, especially with team related stats. I would love to create a predictive model one day but I just don't know enough yet haha

[@alexjones9038]

8:37 this list is completley bogus as Brisbane in 2018 was easily the best 15th placed side of the modern era

[@thegaz-man]

I'm just judging by Elo ranking. Pythagorean wins are probably a better measure of a single season team, I was more referring to the overall form of teams taking the prior season or two into account, considering the Crows' 2023 season, and Elo is a better tool for that.

[@alexjones9038]

@@thegaz-man ah yeah fair enough, I thought it was just a single season measure and I was very confused lol

[@digsby7134]

Melksham hasn’t been a defender for 10 years minor nitpick

[@thegaz-man]

Ahhh yep apologies, haven't really seen too much of his game.

[@1nchwonder290]

Wow. I thought West Coast were known to suck at defending.

[@thegaz-man]

Their key backs were good - their small and medium defenders really weren't by my metrics and that's where the damage was done Plus McGovern is more of an intercept guy than a one on one guy which skews the numbers a bit

[@Codewordthecerealkiller0]

Tbf Quaynor only dropped mainlt because he was playing a role he isn't used to

[@thegaz-man]

Cheers for the clarification! I'm guessing a more defensive and less rebounding role this year?

[@Codewordthecerealkiller0]

@thegaz-man Believe it or not actually the opposite, he was always the small lockdown Defender with intercepting prowess, this year he has played Darcy Moore role if Moore was smaller

[@frarfarf]

Oh wow amazing insight to shit on Melbourne you should get a media gig

[@thegaz-man]

I might be a bit bitter after my Melbourne pickle juice wager haha However I have covered Melbourne favourably in the past, namely in my Demonic Ritual series

[@trentmahoney188]

Your analysis of Cripps is just hate. Everyone knows he was easily our best player!!! EASILY. To say he was just above Kennedy shows you have no clue what's going on and are just talking shit data out of your ass... If Daicos wasn't fed the ball in R24, Crippa would've been the Coaches MVP, he came 2nd in the Leigh Mathews Trophy, 2nd in the best captain award and led the league for clearance, stoppage clearance and CP. He ranked = 1st in AFL player rankings and was a top 5 player by Champion Data. So where's your biase data coming from???

[@thegaz-man]

My numbers compare each player to others relative to their position. This means, by necessity, that more players who aren't midfielders get better scores, unlike the systems you just mentioned. It's important to note that VORP isn't about 'the best' player - It's value over a replacement-level (average stats) player in the same position, relative to Time on Ground. Cripps had a TOG of 83.1%. Out of midfielders to have that level of minutes, only Bont, Daics and Neale had a higher positional VORP. However, guys like Heeney, Young and Petracca did as much, or slightly more, in less on field time which my system values. Cripps also didn't score as much, had lower defensive contributions, and had way more clangers than the other guys he is competing against. This is why his VORP is around 5.3, still a very good season, and a VORP that other Brownlow winners have come near. Kennedy, he benefitted from circumstance. AFL.com.au lists him as a forward and midfielder. His stats for both are the same, and listed forwards who rotate through the midfield always have somewhat inflated VORPs. His VORP as a FW was 4.84, compared to 1.73 as a pure mid. Because he split time at both, he seemed like a great forward due to boosted disposal numbers, but low numbers compared to top mids bc of time forward tank his midfield numbers. Tom De Koning, who also ranked higher than Cripps, is a simpler story - he was the second most efficient ruckman which is reflected in his VORP of 5.8, but since he sees less of the ball compared to Cripps, metrics like the ones you mentioned undervalue his contributions relative to other players in his own position. If he'd won a close count with votes in the high 20s, I wouldn't mind. I'd still feel Bont was robbed, but it would be more acceptable. Him and Daicos both breaking the record vote tally imply mind-bending, record-shattering seasons that aren't reflected in any stats that I've seen. My VORP stats, based on data, and medal voting - based on percieved impact and gut feel - are fundamentally different but runaway Brownlow winners (Ablett, Danger, Martin) generally do well in both. Cripps had a VORP similar to Wines and Priddis when they won. That's what I mean when I said what I said. It's not 'just hate'. It's a formula designed to put a single value to each player, regardless of team or position, for a basic comparison.