We need to change the way we calculate percentage.

[mcgarnacle]

% is a variable affected by a lot of things. The ultimate tie-breaker should be head-to-head, i.e., if say Richmond and Essendon are equal on points at 8th and 9th, the winner of their home and away match(es) gets the 8th position. If teams can't be separated by this, i.e., teams play twice home and away for 1-1 split, then go to %.

 

[BF Tiger]

The current way overweights percentage of games on dry decks with high scoring.

if you win 60 to 20 you should get a far greater percentage boost then a game where you win 130 to 80. Yet at the end of the season it’s the later game that provides the biggest percentage boost.

this is stupid.

we should calculate percentage on a per game basis and weight every game equally. I.e. take an average of all games percentages to derive the final percentage of your team.

this values defensive games and wet weather games the same as dry and attacking games.

it’s such a simple thing to implement.

why haven’t we done it?


edit - there is a flaw in the above approach. instead each game should only have 200 percent given out to the winners and losers. That way a team that wins 90 to 10 doesnt doesn’t get 900 percent. Instead they get 190 and the loser gets 10 percent. The overall percentage at the end of the season is the average of each teams game result percentages as suggested above.

If I’m understanding you right... let’s suppose Richmond play Geelong and win 150 to 50.

Richmond would be 1-0 at 167% (200-33).
Geelong would be 0-1 at 33%.

A week later Richmond lose to Melbourne 50 to 150. Their % for that match is 33%.
Geelong beat Collingwood 50 to 15. Their % for that game is 170%.

The ladder looks like...

Richmond 1-1 200f/200a 100%
Geelong 1-1 100f/165a 101.5%

Yes

Ok. I should have asked would it change the ladder for the better?

To support your case I suggest you come up with a hypothetical season. It can possibly just be four teams playing each other twice. Show me why and how your new ladder is a fairer result than the ladder that would normally be generated.

Even better would Be to come up with a real-life example. Yes, I’m asking you to trawl through 100+ seasons of football and do the calculations.

 

[Howard Littlejohn]

What am I missing and/or how do you think percentage is calculated?
Win 60 to 20 you get 300%
Win 130 to 80 you get 160ish%

I think what is being said is when you add them together you get 190 to 100, and 190 is much closer to 160 than to 300.
"Average match percentage" would be 230.

I actually think the current system is fine, single match percentage can result in much greater percentage variation in low scoring games. The SANFL system is also a good one, IMO.

 

[JohnZ]

Seeds this is an embarrassing thread.

Take the cats current F/A/% being 948/726/130.6%

If we win the next game against Port 60-20 that becomes: 1008/746/135.1%

If we win the next game against Port 130-80 that becomes: 1078/806/133.7%

So winning 60-20 increases your % more than a 130-80. % is heavily in favour of teams who can keep their opposition to a low score (for the same margin).

 

[Vikingnz]

Because a greater winning margin. What's the prob.

People were saying the first game gives a higher boost which is true in isolation but not in reality.

 

[KarmaFarm]

OPs concept is dumb, but this isn't true.

If I had 1000 points for and against for a percentage of 100 and I won a game 60-20, my % increases to 103.9%. If I won a game 130-80, my % increases to 104.6%.

Yes but in the second game you won by a larger margin. If the game was still a 40 point margin say 130-90 the % increases to only 103.67%. Meaning that a win in a higher scoring game is worse then the same margin in a lower scoring game.

 

[Seeds]

Seeds this is an embarrassing thread.

Take the cats current F/A/% being 948/726/130.6%

If we win the next game against Port 60-20 that becomes: 1008/746/135.1%

If we win the next game against Port 130-80 that becomes: 1078/806/133.7%

So winning 60-20 increases your % more than a 130-80. % is heavily in favour of teams who can keep their opposition to a low score (for the same margin).

Not heavily weighted. Equally weighted. gross margin is not a sign of how good a win is. It’s proportional margin. A 60-20 win is a better win then 130-80.


you would think a Scott fan would get this.

 

[JohnZ]

Not heavily weighted. Equally weighted. gross margin is not a sign of how good a win is. It’s proportional margin. A 60-20 win is a better win then 130-80.


you would think a Scott fan would get this.

Yes and that's how % works. a 60-20 win will increase your % more than a 130-80 win.

 

[Seeds]

If I’m understanding you right... let’s suppose Richmond play Geelong and win 150 to 50.

Richmond would be 1-0 at 167% (200-33).
Geelong would be 0-1 at 33%.

A week later Richmond lose to Melbourne 50 to 150. Their % for that match is 33%.
Geelong beat Collingwood 50 to 15. Their % for that game is 170%.

The ladder looks like...

Richmond 1-1 200f/200a 100%
Geelong 1-1 100f/165a 101.5%

Ok. I should have asked would it change the ladder for the better?

To support your case I suggest you come up with a hypothetical season. It can possibly just be four teams playing each other twice. Show me why and how your new ladder is a fairer result than the ladder that would normally be generated.

Even better would Be to come up with a real-life example. Yes, I’m asking you to trawl through 100+ seasons of football and do the calculations.

You have just provided the perfect example. Richmonds percentage under both approaches would be 100. But under the current approach geelongs percentage would be 60 something when it really should be 101.5.

geelongs win was slightly better then it’s loss when looking at the proportional margin. geelongs 50-15 win could of been in a massive downpour or because it’s opponent flooded defence to the extreme. There is no way geelongs percentage should be in the 60s. Yet under the current system it is because it overweights the higher scoring game compared to the lower scoring game. Geelongs actual percentage should be 101.5

That’s a massive difference. A difference that prevents the right side from making top 2, top 4 or the final eight.

 

[Private Hudson]

Who gives a shit?

 

[bullish23]

The current way overweights percentage of games on dry decks with high scoring.

if you win 60 to 20 you should get a far greater percentage boost then a game where you win 130 to 80. Yet at the end of the season it’s the later game that provides the biggest percentage boost.

this is stupid.

we should calculate percentage on a per game basis and weight every game equally. I.e. take an average of all games percentages to derive the final percentage of your team.

this values defensive games and wet weather games the same as dry and attacking games.

it’s such a simple thing to implement.

why haven’t we done it?


edit - there is a flaw in the above approach. instead each game should only have 200 percent given out to the winners and losers. That way a team that wins 90 to 10 doesnt doesn’t get 900 percent. Instead they get 190 and the loser gets 10 percent. The overall percentage at the end of the season is the average of each teams game result percentages as suggested above.

Why not just put a roof over KP or whatever it is they call it these days.

 

[Cleavy]

THE NUMBERS DON'T LIE, AND THEY SPELL DISASTER FOR YOU AT SACRIFICE!

 

[Brad Goodman]

Thinking is good. Sometimes though, thought shouldn’t progress beyond that.

 

[BF Tiger]

You have just provided the perfect example. Richmonds percentage under both approaches would be 100. But under the current approach geelongs percentage would be 60 something when it really should be 101.5.

geelongs win was slightly better then it’s loss when looking at the proportional margin. geelongs 50-15 win could of been in a massive downpour or because it’s opponent flooded defence to the extreme. There is no way geelongs percentage should be in the 60s. Yet under the current system it is because it overweights the higher scoring game compared to the lower scoring game. Geelongs actual percentage should be 101.5

That’s a massive difference. A difference that prevents the right side from making top 2, top 4 or the final eight.

I’m understanding what you’re saying but you have not convinced me. Show me a real-life example where your method has promoted a more deserving side.

 

[Seeds]

Who gives a sh*t?

Who gives a shit who wins? It’s just a game.

 

[Seeds]

Why not just put a roof over KP or whatever it is they call it these days.

That wouldn’t fix the problem one bit.

 

[Seeds]

Yes and that's how % works. a 60-20 win will increase your % more than a 130-80 win.

No it doesn’t work that way. That’s the problem.

if you are at round 21 and you have 1000 points for and 1000 points against then the 130-80 win adds more to your percentage then the 60-20 win.

 

[theyellowsash]

geelong person whingeing for a rule change.

well i never

 

[caveman72]

Do funny shaped grounds affect % more or less than dry decks?

 

[LightTower4]

Current system everyone knows how it works and have done for 125 years.
Even those like me who are crap at maths.

 

[CaptchaProofBot]

percentage concept hasn't changed since ancient Rome I don't think.

been fine and correct.

 

[Dirty Bird]

% is a variable affected by a lot of things. The ultimate tie-breaker should be head-to-head, i.e., if say Richmond and Essendon are equal on points at 8th and 9th, the winner of their home and away match(es) gets the 8th position. If teams can't be separated by this, i.e., teams play twice home and away for 1-1 split, then go to %.

But if Adelaide and Geelong are tied at the end of the year, why should Adelaide be higher by virtue of getting the home game?

I'm all for h2h if we play each other twice. But if not, dont bother.

give me pd instead.

 

[Call Me Cake]

Wet weather grounds cause a disadvantage to teams that are tall heavy, or rely on pure execution

We must find a way to help these teams

 

[gotterdamerung]

Fremantle beat GCS in a wet slog 40-20, then lose 100-50 against Richmond

Average percentage: (200+50)/2 = 125%

Geelong lose 90-100 against Richmond and beat GCS 120-90

Average percentage: (133+90)/2 = 112%

Good system IMO.

 

[HG Pennypacker]

This is quite the thread