Moved Thread maximum Discrepant Anomalies In Coach OpinionS Index(D.A.I.C.Os index)

[Meteoric Rise]

Any method would be just as stupid. There isn't a quantitative, objective way to do this.

We can compare what is known, and from that, draw any conclusions.

We don't hve to shy wy from the task in fear because someone is scared we will draw incorrect conclusions.

If you don't think it is worthwhile then thanks for sharing that opinion with us, others do, and nobody is holding you here against your will.

 

[sabso]

Any method would be just as stupid. There isn't a quantitative, objective way to do this.

This is as much a quantitative objective way to do it as there honestly can be. Your rash, unconsidered reaction to it is possibly a sign that MR is onto something.

…or it could be standard mr meow I dunno

 

[Meteoric Rise]

How does it measure integrity? If Scott and Clarkson disagreed about how good Dangerfield was in a game in 2016, one of them lacks integrity? Your "method" doesn't quantitatively assess integrity at all.

What we know for sure is 2024 both coaches agreed Isaac Heeney was worth at least 100 votes for the season. That is his minimum agreed votes. For Daicos, the corresponding figure was 94 votes. Yet Daicos won the award due to a much larger amount of votes in the potentially disputed area. And he is off to a flying start with his potentially disputed vote tally in 2025.

Where both coaches agree there is no real controversy. Even if we might have voted differently, most of us would accept the votes. Where coaches have voted differently there is by definition, controversy. Where those differences fall within a normal or understandable range, most of us would likely accept it. What the thread is about is working out what is a normal range and figuring out to the extent possible, which award wins and career tallies are more or less controversial.

From that we are free to speculate as to the integrity of the coaches involved, but I am sure we will learn other things along the way. Why I bet you are already looking at coach votes slightly differently, and that can only be a good thing.

 

[Mr Meow]

This is as much a quantitative objective way to do it as there honestly can be. Your rash, unconsidered reaction to it is possibly a sign that MR is onto something.

…or it could be standard mr meow I dunno

Yes, it's about as quantitative as me guessing cloud size by drawing my finger around them from my POV. Still, that's as quantitative as I can be from my backyard so maybe I'll release it as a study.

Explain where I've lacked consideration in my critique.

I suspect Daicos has received the rub of the green too, but it doesn't make MRs method a reliable way to determine voting integrity (so much as he is concluding player X had a questionable tally, while player Y had a tally based on sound voting integrity).

You do realise he makes all of these threads with his mind made up already, then simply reverse engineers whichever approach "proves" him right. He has done it countless times. It deserves to be called out as such.

 

[Mr Meow]

What we know for sure is 2024 both coaches agreed Isaac Heeney was worth at least 100 votes for the season. That is his minimum agreed votes. For Daicos, the corresponding figure was 94 votes. Yet Daicos won the award due to a much larger amount of votes in the potentially disputed area. And he is off to a flying start with his potentially disputed vote tally in 2025.

Where both coaches agree there is no real controversy. Even if we might have voted differently, most of us would accept the votes. Where coches have voted differently there is by definition, controversy. Where those differences fall within a normal or understandable range, most of us would likely accept it. What the thread is about is working out what is a normal range and figuring out to the extent possible, which award wins and career tallies are more or less controversial.

From that we are free to speculate as to the integrity of the coaches involved, but I am sure we will learn other things along the way. Why I bet you are already looking at coach votes slightly differently, and that can only be a good thing.

You've yet to explain how in all cases where there are discrepancies, that would give you a readout of integrity.

Norm Smith votes must be next on your agenda. If there were disagreements, then that particular game must have had low integrity judges, right?

 

[Mr Meow]

What we know for sure is 2024 both coaches agreed Isaac Heeney was worth at least 100 votes for the season. That is his minimum agreed votes. For Daicos, the corresponding figure was 94 votes. Yet Daicos won the award due to a much larger amount of votes in the potentially disputed area. And he is off to a flying start with his potentially disputed vote tally in 2025.

Where both coaches agree there is no real controversy. Even if we might have voted differently, most of us would accept the votes. Where coches have voted differently there is by definition, controversy. Where those differences fall within a normal or understandable range, most of us would likely accept it. What the thread is about is working out what is a normal range and figuring out to the extent possible, which award wins and career tallies are more or less controversial.

From that we are free to speculate as to the integrity of the coaches involved, but I am sure we will learn other things along the way. Why I bet you are already looking at coach votes slightly differently, and that can only be a good thing.

The issue is that you are generalising to all coaches from one situation. In other words you are spinning one controversial example into some sort of league wide conspiracy.

It doesn't make sense because people disagree on the top 5 players (and their order) of a match all the time. If you randomly took any 2 people's club board MVP votes from this forum, you would often get "controversies". If you take all 50, you get a consensus.

You seem to be struggling with the concept that n = 2 is bound to bring discrepancies in a "select 5 players in order of their performance" criteria.

The thread would have worked better to keep it purely opinion based rather than pretend you have sorted reliable and unreliable coaches votes tallies across all player seasons. You are basically shoehorning players like Oliver and Dangerfield into Macrae's adoration of Daicos. It is not objective like you are making it out to be.

 

[Meteoric Rise]

Jeez, can you not care for something but really care about it? If so, congratulations are in order for Mr Meow who doesn't seem to care for the thread but has found a way to care about it passionately.


Tomorrow hopefully I will have some time to collate a few more facts. And I suppose where people are trying to stop you dead in your tracks during the fact finding stage of an investigation, then that is all the more reason to carry on finding out the facts.

 

[Mr Meow]

Jeez, can you not care for something but really care about it? If so, congrtulations are in order for Mr Meow who doesn't seem to care for the thread but has found a way to care about it passionately.


Tomorrow hopefully I will have some time to collate a few more facts. And I suppose where people are trying to stop you dead in your tracks during the fact finding stage of an investigation, then that is all the more reason to carry on finding out the facts.

Now you're just making it personal instead of responding to any of my points. That's called deflecting.

Don't post terrible methods of analysis if you don't want constructive feedback on their flaws.

 

[Meteoric Rise]

Now you're just making it personal instead of responding to any of my points. That's called deflecting.

Don't post terrible methods of analysis if you don't want constructive feedback on their flaws.

If you want a response to any of your questions or points, then don't also post a whole lot of supposition about my intentions or methods along with any legitimate points you might have.

 

[sabso]

Yes, it's about as quantitative as me guessing cloud size by drawing my finger around them from my POV. Still, that's as quantitative as I can be from my backyard so maybe I'll release it as a study.

Explain where I've lacked consideration in my critique.

I suspect Daicos has received the rub of the green too, but it doesn't make MRs method a reliable way to determine voting integrity (so much as he is concluding player X had a questionable tally, while player Y had a tally based on sound voting integrity).

You do realise he makes all of these threads with his mind made up already, then simply reverse engineers whichever approach "proves" him right. He has done it countless times. It deserves to be called out as such.

 

[Mr Meow]

Looks like your "brash and unconsidered critique" statement was a little brash and unconsidered itself.

 

[Mr Meow]

If you want a response to any of your questions or points, then don't also post a whole lot of supposition about my intentions or methods along with any legitimate points you might have.

Explain where my critique of your methods and the validity of your conclusions was based on supposition.

You can start by addressing each specific point I made by highlighting where I said something without basis, or anything that does not have merit, or is in indeed made up.

 

[sabso]

Looks like your "brash and unconsidered critique" statement was a little brash and unconsidered itself.

I said it was rash, not brash. Burden of proof is on MR to demonstrate his findings, not on you or me talking about it.

You are aware what sealioning is aren’t you mate.

 

[doppleganger]

BTW … in Martin’s 2017 season, he got 2 x coaches votes from 1 x game when not in the 5 x top rated players on the ground.

So what did 2023 show for Naicos?

He got 26 coaches votes from 4 x games when not rated in top-5 players on the ground.

BTW, you may wish to take a look at Dusty's 2015-2016 seasons and you will find a player who polled 75+ coaches votes (multiple 10s) in games in which he was not rated in top-5 players on the ground according to PLaYER Ratings.

So Pies supporters should celebrate his ability to get votes when not really deserving

Not surprising that Fly is also copying how Dimma looked after by Dusty, who wasn't "deserving".

According to PLaYER Ratings, Dusty was a solid footballer, only had 4 seasons with an avg season rating above 15 (Daicos is 2 from 3 above 15).

PLayer Ratings career avg has Dusty below Rory Sloane, yet Dimma fluffing his votes meant the inferior and undeserving Dusty was gifted 300+ coaches votes.

Fly learning from the best at TigerLand.

 

[Mr Meow]

I said it was rash, not brash. Burden of proof is on MR to demonstrate his findings, not on you or me talking about it.

You are aware what sealioning is aren’t you mate.

If you submit anything then it is open to critique. I explained why I thought the method provided was not suitable to make the broad conclusions he did. Another poster offered a reasonable critique as well. It is not a reliable method for assessing "voting integrity". In my opinion of course. You are free to think it is and even explain why if you can be bothered. "It's the best method we have" doesn't make it a sound method.

 

[doppleganger]

Amazing correlation coaches votes v player ratings, except for one player.

You guessed it, Daicos the only player in the top 6 of Coach votes who was not also top 6 in the player ratings in 2024.

Nowhere near Dusty of 2016 levels though.

Dusty wasnt even in top40 according to PlaYer Ratings, but good old Dimma fluffed him to finish 3rd in coaches votes.

Not a once off either, 2015 was more of the same. Once again Dusty was outside top40 in player ratings (not unusual for Dusty) and Dimma fluffed him to come top 12 in coaches votes.

Doubt Daicos will ever be outside top40 in player ratings at seasons end like Dusty until well into his 30s.

 

[CharlieMortdecai]

AKA the max D.A.I.C.OS index. No idea how that happens to make an anagram exactly the same as last year's Coaches MVP, just one of life's weird coincidences I guess.

Thought it was time we had a thread about this simmering little talking point, as much to protect the reputations of the innocent as anything.

We can track this and discuss it here. To me the Coaches MVP used to be about the most prestigious award, certainly the most credible. Now I am less sure.

Votes being given to N Daicos by C McRae in particular have raised a few eyebrows. Said player actually won the award last year. He was the only player in the top 6 in the award last year who was not also in the top 6 for average player ratings.

The truth is we don't know which coach gives which votes and we also often don't know how a players votes are comprised, like a 5 vote game could possibly be 5-0, or 3-2 when you consider all the other votes. So how can we make objective sense of this to help us understand what could be happening?

My suggestion here is a new index, the maximum discrepant anomalies in coach opinions when they issue their votes weekly in season.

How this works:

We simply work out the maximum possible discrepancy between the 2 coaches votes for the player we are interested in.

Eg

2025 Rd 2 Bulldogs v Magpies we get the following votes -

View attachment 2260096

We can see here 3 players have 8 votes. So 1 of the 3 must have got 4+4 and the other 2 got 5+3.

So for Daicos, the maximum dicrepancy possible is 5-3 = 2.

So how did 2024 look for the accidentally eponymous hero of the thread?

He received 117 votes overall. His maximum discrepancy was by my calculations, 23 votes. Ie, the maximum Mcrae could have given him is 70 votes. The minimum opposition clubs could have given him was 47 votes. It seems quite a discrepancy, but we are comparing the maximum possible with the minimum possible to derive that discrepancy, so how does it compare with others?

2024 Isaac Heeney had a maximum discrepancy of just 12 votes. The minimum possible he could have received from opposition coaches was 50 votes, 3 more thanDaicos's minimum.

2017 Dusty got 122 votes, the highest ever. His maximum discrepancy was 10 votes. 66-56 was his max possible discrepancy between one coach and the other.

The really interesting thing here so far is Daicos 2025. After just 3 rounds he already has a maximum possible discrepancy of 7 votes, on world record pace.

Feel free to request other players maximum discrepancies, or contribute yourselves by submitting some.

By my calculations Daico's maximum discrepancy was 21 votes. His minimum discrepancy was 13 votes - that is in all bar 4 games we know exactly what the discrepancy was (12 votes), in one game his discrepancy could have been 1 or 3, and the other 3 games could have had a total discrepancy of 0 to 6.

By way of comparison, runner-up Patrick Cripps had a minimum discrepancy of 6 votes, and a maximum of 14 votes.

The round 21 game against Carlton was the most discrepant, where we know that Daicos got 5 votes from one coach and zero from the other. It is the most lopsided votes I've ever noted:



The voting must have been along these lines:

	Coach A	Coach B
Cripps	5​	4​
Pendlebury	4​	2​
N Daicos	0​	5​
Hewett	3​	0​
Cameron	0​	3​
Howe	2​	0​
Moore	0​	1​
Weitering	1​	0​

I think this is the game where McGovern had a kick after the siren to win, but didn't make the distance? From memory I had been out to dinner and only caught the last 5 or 10 minutes, so no idea who was best on. I do remember McGovern looking as nervous as hell, and Cripps cheering and whooping and getting up in his face grinning like an idiot which I don't think actually calmed his nerves in the slightest.

Brownlow votes for the game were Cripps 3, Daicos 2, Hewett 1.

 

[CharlieMortdecai]

Probable votes:

From oppo coaches 0 0 3

From Craven McRae 0 5 5

Or:
Oppo 0 2 4
McRae 0 3 4

Would Kenny have given Houston 5? If not he must have given either Daicos or McStay 5, with McRae then giving that player 0. I reckon it must have gone:

	Kenny	Fly
Houston	4	5
Sidebottom	3	4
Daicos	2	3
McStay	5	0
Rozee	0	2
Lipinski/Membrey	1/0	1/0

 

[Crankyhawk]

What we know for sure is 2024 both coaches agreed Isaac Heeney was worth at least 100 votes for the season. That is his minimum agreed votes. For Daicos, the corresponding figure was 94 votes. Yet Daicos won the award due to a much larger amount of votes in the potentially disputed area. And he is off to a flying start with his potentially disputed vote tally in 2025.

Where both coaches agree there is no real controversy. Even if we might have voted differently, most of us would accept the votes. Where coches have voted differently there is by definition, controversy. Where those differences fall within a normal or understandable range, most of us would likely accept it. What the thread is about is working out what is a normal range and figuring out to the extent possible, which award wins and career tallies are more or less controversial.

From that we are free to speculate as to the integrity of the coaches involved, but I am sure we will learn other things along the way. Why I bet you are already looking at coach votes slightly differently, and that can only be a good thing.

It may be more relevant to look for instances where you have 3-5 from one and 0 from other coach. I don’t see much eyebrow raising for a difference of 1 or even 2. Which means going back to weekly vote results

 

[Meteoric Rise]

It may be more relevant to look for instances where you have 3-5 from one and 0 from other coach. I don’t see much eyebrow raising for a difference of 1 or even 2. Which means going back to weekly vote results

What I am looking for really is the minimum possible opposition coach votes. We can locate that figure. The players who are highest in that figure, we know their overall vote tallies are likely amongst the safest from being fluffed by their own coach. When we hve compiled the figures from criticl mass of players, we will have a baseline established - a figure that natural unbiased disagreement is unlikely to fall below. Let's say that is 10% of total votes for eg. We can then reduce each player's discrepancy tally by 10% of total votes and what remains would be the amount of potentially biased votes. Not proven biased of course, just potentially so. But it would tell us for eg whether differences between the coaches are incresing over time or not, which is one of the things I am most interested in. We should also be able to see who the obvious outliers are amongst the coaches - in either direction. Of course we can then drill down deeper into those votes tallies for further clues, much the way CharlieMortdecai has done in his excellent post #117 above.

What I find strange is that some posters seem to think I already know the answers to all this and the exercise is just building up to declare Dusty the most genuinely awarded player, or Daicos the most fraudulent or whatever. If I had the answers now, I would present them now. But I don't, and we have no idea what we are going to find. We may well find the Daicos discrepancy figures fit within a very normal range of potential discrepancies.

I don't feel it is right to limit the potential discrepancies in any way. A coach should just as easily show bias by bumping a certain player from 2 votes to 3 votes regularly, as he could by giving the odd 5 votes when 0 may have been more fitting. But I have all my round by round tallies so can easily go back & identify for eg where the potential exists for one coch to hve given several votes to a player & the other coach to have given zero.

 

[Meteoric Rise]

Nowhere near Dusty of 2016 levels though.

Dusty wasnt even in top40 according to PlaYer Ratings, but good old Dimma fluffed him to finish 3rd in coaches votes.

Not a once off either, 2015 was more of the same. Once again Dusty was outside top40 in player ratings (not unusual for Dusty) and Dimma fluffed him to come top 12 in coaches votes.

Doubt Daicos will ever be outside top40 in player ratings at seasons end like Dusty until well into his 30s.

Let's have a look how much potential Dimma love Dusty got in 2015-16 then - the years his average ratings had him rated between the 40th & 50th highest, but he got a healthy amount of total coaches votes, finishing 14th & 3rd. In fact I will focus on just 2016 here where Dusty got 3rd in the coach MVP but was the 40th highest rated player to play at least 10 matches.

2016 Dusty 90 total votes, max from Hardwick = 53 v min from oppo = 37. Max possible discrepancy is 16 votes.

This seems fairly moderate compared to some others we have looked at.

-----------------------------

While I am at it with Dusty, let's have a look at his finals coaches votes.

2017 his max discrepancies were 5+4, 4+4 & 5+3 so 25 votes with a max discrepancy of 3 votes(14 v 11)

2018 4+3 so 7 votes with a max discrepancy of 1 vote

2019 5+4 & 5+5 19 votes with a max discrepancy of 1 vote

2020 5+3 5+5 5+5 so 28 votes with max discrepancy of 2.


Overall Dusty in finals 79 votes with a max discrepancy of 7 votes. Dusty seems to have pretty high levels of agreement between the coaches when he was in the votes.

I think it is pretty clear that in 2016 he played the type of high accumulation, lower impact style of footy that can catch the eye(& votes), but the ratings don't reward highly.

In finals where he voted, he played the sort of high impact footy the ratings really like:

2017 9 coach votes = #2 player ratings
2017 8 cv = #6 pr
2017 8 cv = #1 pr

2018 7 cv = #1 pr

2019 9 cv = #3 pr
2019 10 cv = #1 pr

2020 8 cv = #1 pr
2020 10 cv = #1 pr
2020 10cv = #1 pr

We can see very strong correlation here between the coches votes & ratings.

 

[PhatBoy]

I hope this can bring peace to all

 

[CharlieMortdecai]

Of course we can then drill down deeper into those votes tallies for further clues,

It might also show that your metric is not very useful, or can tell us two contradictory stories.

The 5-0 to Nick Daicos in Rd 21 blows out his discrepancy. Reading the match reports for that game, considering the votes awarded in the Brownlow (unfortunately I couldn't drill down to any other awards for their votes on this game), it appears to be a fairly universal view that Patrick Cripps was BOG, and Daicos was the next best player on the night, as per the Brownlow voting.

So maybe what we learned here was not that McRae over-rated Daicos - or not by much - but rather that Voss was being churlish in not giving him any votes.

If that discrepancy was 1 vote rather than 5, it would reduce Daicos's minimum discrepancy from 14 to 10, still higher than Cripps, but far less significant.

 

[KissKiss]

Im amused enough by this analysis to see it keep going. For consistency you may as well fill in some of the other winners: Gawn in 2018, Bont 2019 and Neale in 2020

 

[CharlieMortdecai]

I'd like to see it continue too, but CBF crunching the numbers.

I'd request the known minimum discrepancy, as I described above, and over how many games those votes were spread. I don't think there is any benefit using the maximum potential discrepancy, when it is unknown and can range from 0-4.

So using the 2024 top 2:

Daicos 13 votes over 13 games, neatly works out to 1 vote per game.
Cripps 6 over 13 games, something like 0.5 votes per game.

Do everybody for every game this year and see what it tells us Meteoric Rise