Watson11

Yes after every game the model is updated with those stats and the result. Because the model is based on many years of data, each game only changes it a small amount. The player ratings change far more after each game, as they are based on only 2 years and weighted to the most recent. The intent of all of this data is simply to make CD lots of money, because the professional clubs pay big $ for it. The professional clubs pay for it because it gives them unbiased insights into what the really important stats are that give teams an edge. Watch Moneyball if you have not seen it. It's where all this data stuff really started. Clarkson was the first in the AFL to use data and built his 4x premiership list using it.

One on one loss stats for 2018 are below showing the top 20 ranked key position defenders. Name 2018 Career Will Schofield 15.1% 19.9% Jake Lever 15.4% 30.6% Harry Taylor 16.7% 15.4% Lachie Henderson 16.7% 22.8% Sam Frost 17.6% 26.7% James Frawley 18.4% 25.0% Alex Keith 18.5% 32.8% Daniel Tahlia 18.9% 20.5% Heath Grundy 20.3% 22.6% Steven May 21.4% 23.2% Robbie Tarrant 21.7% 27.8% Scott Thompson 23.5% 26.7% Tom Jonas 23.6% 24.3% Alex Rance 25.0% 21.3% Phil Davis 25.3% 31.1% Lynden Dunn 26.1% 22.4% Oscar McDonald 26.7% 25.1% David Astbury 27.5% 24.7% Jeremy McGovern 27.6% 19.4% Jake Carlisle 27.6% 24.6% Michael Hurley 27.6% 29.1%

Can't predict injuries, modified game plans, improvements in players or teams, poor form or loss of confidence. So no one knows for sure what will happen in round 1 next year let alone the entire season. Where they are good is when the siren goes at the end of our round 1 game, you could put the stats into one of these models and it would predict the winner with 99% accuracy. We just can't predict what those stats will be before the game with any certainty. All this tells us is based on the data we have a very, very good list.

You are correct in that humans at the moment record the stats (and need to decide what is a 1%er, tap to advantage, effective kick etc). But no human decides how to weight the various stats. When you train these systems, you start with all the stats as inputs, and you know the result the model needs to give for each game (Team A won by X points). The system repeatably adjusts parameters (weights) until it gets the correct answer, and repeats that for every game in its database. What you end up with is a model that given a set of input stats will predict the result, and that is simply how the lists are ranked. I

Nope that's wrong. They just put all the data in and train the model. The training of the model weights the stats. Not a human. Then they put the stats of each list in and put that into the model to weight the list. It's just data. As was made clear it can't account for everything and is imperfect, but there is no assumptions put in.

Not sure who says we have had the best defence by a country mile. But definitely have never agreed with your view that the problem has been that our key defenders are hopeless one on one defenders.

CD just uses data and they make no assumptions. They use machine learning, specifically a convolutional neural network (https://en.wikipedia.org/wiki/Convolutional_neural_network). Consider it a "brain" that has no pre-concieved assumptions but hundred of millions of ways to combine information. They just train the network on the data they have (well over 20 years of comprehensive data), putting in the stats for each game and the result. The limitation is the stats don't cover everything, as that is impossible to do, and they don't take into account things like injuries, the draw, player improvements etc. They do measure a lot though. The list is just a ranking based on this information and simply tells us in an unbiased way that we have a very very good list. As Prodee says though the premier can come out of any of the top 10 teams on that list.

One good thing about stats is they don’t suffer from confirmation bias. Not all us supporters know our defence was our weakness. Some of us attribute our habit of leaking goals to our forward 50 and midfield pressure going missing at times.

Based on the stats West Coast's midfield was middle of the road for most of 2018. They had a 3 week purple patch that just happened to co-incide with the finals, helped hugely by Gaff replaced by Redden who averaged around 15 CPs through the finals. Adelaide will be top 4 this year.

Won the 20m sprint at the combine. 2nd in goal kicking comp and his skills look good. Had big stats in TAC cup. Memo to JT. Pick him up in the rookie draft and sort out whatever other problem made everyone pass. Can’t be any worse than Dusty and he went pick 3.

Has this kid not been picked up?

I was a huge Jesse fan, and argued we should keep him over May, but in hindsight I can see the logic of the trade and now think we are a much better chance to win a flag in 2019, so JM gets a big tick from me. IF we win a flag this year, imo it will be because of improving 3 things: 1. A forward line that can kick goals when the ball is bombed in. 2. Forwards and mids who consistently pressure. 3. A defence that holds up. Jesse out, Weid+May+Preuss in helps all 3 of the above immediately. For Jesse to be the difference we would need a midfield with elite inside 50 footskills to make the most of his speed. That is something we could develop, but we just don’t have it yet.

I’ll take a bath and drink some of the water while I dream about the next 4 flags in a row....

In the US the NFL draw is biased but not as badly as ours. They have groups where every team plays the others twice in that group which creates some equity in their group. The AFL could do something similar: 1. Two conferences of nine teams (5 Vic and 4 interstate) who play each other twice. To be fair each conference to have 1 team from each non vic state and split the smaller supporter bases (Saints, North, dogs, us). That is 16 games. 2. Other 6 games play against teams from the other conference including two games against a rival ie showdown, Coll v Mel if needed etc. These games are also based on ladder position from prior seasons with 2 games (one home and one away) against 1-3, 2 against 4-6 etc. means playing 5 of the teams in the other group with one double up. 3. Finals could just be top 4 from each conference and same as current series. Or ditch the pre finals bye and have the 4th from each conference play a wildcard team from the best performed 5th and 6th from both conferences to make the final 8. Would even up the travel for all teams, not lose rivalry games, and the draw difficulty would be much closer for everyone. Conference strength would change year to year but come finals it would be fair.

Better than most comparisons. Funny to look at the 2018 analysis that had us as the equal 3rd most difficult draw! Yet everyone else saying we had an easy draw by the end of the season. I don't mind our double ups and difficulty of the draw. In 2018 Richmond, Geelong, and Hawthorn were handed ridiculously good schedules. In 2019 all the top teams have tough schedules due to more Thursday night games. I think the top end of the draw is fairer this year. I also think it helps us as there is a lot of "veterans" in the other top teams that will be more injury prone than our younger list.

Now the draw is out time for ladder predictions Adelaide (soft draw) Melbourne (6/6/6 and learnt to pkay the G) GWS (soft draw) WC (Sheed has stepped up to solve their cp weakness) Richmond (Drop back because still weak around the contest 6/6/6 will affect them) Collingwood (Tough draw, backing Melbourne but could swap with 2nd if they have our measure) North (Ok draw and will improve) Essendon (May just scrape in) Hawthorn (Too old) Brisbane (Big improvers) Sydney (Drop back some more) Dogs (Soft draw) Geelong (Tough first 12 weeks. Season over by the bye) Fremantle (Improve over last year) Port (Cooked) St Kilda Carlton GC

Surely Champion data have stuffed up the analysis. Geelong are bracketed as one of the top 4 toughest double ups, not Collingwood or GWS. I'd put Geelong maybe 9th or 10th. If they fixed that then Richmonds draw is in the top 4 or 5 most difficult. Richmond also have 2 sets of back to back 6 day breaks this year, versus none in 2018, so they have a tougher schedule.

I agree Geelong, Sydney, and Hawks will slide. Suspect Adelaide, North and Bombers to come up. Don't think Port will. Is Gil from Adelaide? They have no back to back 6 day breaks. Have 12 games at Adelaide Oval and mostly very winnable away games (Sydney where they do well, North at Marvel, St's, Lions, Cats, Suns, Carlton, Dogs). They double up against 2 bottom 3 teams despite not finishing bottom 6. They won 12 games in 2018 having a shocker, and have been gifted an extra 2 wins this year. I can't see them winning less than 15 games and they will probably win 17 and finish top 2. Considering Essendon finished equal with Adelaide on points and percentage, the contrast could not be more different. They double up against 4 of last years top 9 (like us but we earned it) plus the Dockers who will improve and be tough in Perth. The first 8 weeks could cook Essendons season with 4 x 6 day turnarounds and 1 x 5 day turnaround in a 7 week stretch. Let's see how Danihers groin stands up to that. The back half they have two 6 day turnarounds leading into away games (Eagles at Optus, Adelaide), and head to the Gold Coast after Adelaide (kind of game they will drop). I expect them to be more hot and cold than they were in 2018 and no certainty to make the 8 despite looking good on paper. We have a tough first 5 weeks due to short turnarounds (5 games in 25 days), but after that we have a good draw and will not have any excuses if we don't finish top 4 (we should be aiming for top 2). We can beat any team anywhere in 2019.

Our record at the MCG last year (and the year before) was 50:50..... We have a better record elsewhere. Hopefully we sorted that out at the end of the year and this year will be better.

The NT helped us recover financially and imo now we are decent we should support the NT. We should have an advantage up there by now. We've played it enough. I can't see why the geniuses at the AFL can't arrange a fixture with no more than one 6 day break in a row. Two in a row starts to hurt teams. This year Collingwood, Port, Adelaide, and Sydney had back to back 6 day breaks. Adelaide and Collingwood had bad soft tissue injuries. Port ended up cooked. Three 6 day breaks in a row kills seasons and should never happen (ie us in 2017). It's why we should have opened on Friday night.

The annoying thing is we will have two days less turnaround over the first 4 weeks than Richmond before ANZAC eve. We’ll have 3 6 day breaks, Richmond 1. Same re Collingwood and Essendon. No surprise to see the AFL favour Richmond and Collingwood there.... They know revenue will be great if they have good seasons...

They are 2018 figures. I agree with you re Omac. If you look at career stats Omac is slightly ahead at 25.1% versus 26.7% for Frosty. Frosts 2018 one on one stats support your prior views (and Goody's) that the forward/midfield pressure was the big issue with our leaking goals at times. Frost didn't suddenly go from an average to great one on one player. He was able to improve so much because our forwards and mids lifted the pressure out of sight. If we maintain that in 2019 it really won't matter who is in between Omac and Frost, we'll be hard to score against and bolstering our defense is really icing on the cake. May and Lever are also great leaders, so will help the mids remain accountable.

I don't think so. I think the teams they play twice will be 3 in their 6 based on ladder positions, 1 in each of the other two groups of 6, with a showdown guaranteed against Port. Last years the teams Adelaide played twice were Port, Richmond, Melbourne, Carlton, and the Giants. They'll play Port twice and double up against 2 of Geelong, Sydney, North, and Essendon in 2019. They'll double up against one of the top 6 and have requested Hawthorn, Collingwood, and Richmond before us...

Read yesterday that Adelaide requested 6 home games against Melbourne teams (Richmond, Collingwood, Hawthorn, Geelong, Carlton, and Essendon). Surely they don’t want to play in Alice again!

I think binman was comparing Omac and May at the same age. Frost had 222.85m gained per game versus Omacs 200m so not too different. Omac had 2.08 turnovers per game versus Frost 3.08. Frost 2.77 score involvements versus Omac at 1.84. They are close. Different strengths and weaknesses.