The SANFL percentage system

[spell_check]

I read today that the SANFL (then SAFL) adopted the current % system in 1921:

28/7/21: The League on Monday adopted the following formula of working out the proportion of points for and points against: Add the points won to the points lost and divide the total into the points won multiplied by 100.

What was used before this? I don't recall seeing what the % was for each team in the newspapers before that.

[Hondo]

I grew up with the SANFL system and so have never understood how you can have a % > 100

To me, the SANFL's method, altho harder to work out, is technically correct.

Dividing PF by PA is not a %, more of a ratio. What do the mathematicians in our midst think ... Spelly, Ecky?

[spell_check]

That's why I prefer the SANFL % hondo, you just said it.

[bayman]

That's why I prefer the SANFL % hondo, you just said it.

spot on boys as you said how can a % be greater than 100 when % itself means out of a 100 to put it in basic terms

[Adelaide Hawk]

I grew up with the SANFL system and so have never understood how you can have a % > 100

To me, the SANFL's method, altho harder to work out, is technically correct.

Dividing PF by PA is not a %, more of a ratio. What do the mathematicians in our midst think ... Spelly, Ecky?

Actually, both systems are correct. "Per Cent" means "per hundred". The SANFL system measures the amount of points your team scores "per each hundred" for the match. The AFL system measures your team score "per each hundred" the opposition scores.

Both are correct, both are reliable, so I have no problem with either.

It's interesting Spelly claims the percentage system was introduced in 1921 because in Ern Kolosche's book "League Football in South Australia" published in 1976, he publishes premiership ladder for each year back to 1907, and includes percentages back to then.

Ern must have calculated the percentages himself and thrown them in.

[JamesH]

before introducing their current % system the VFL used to caluclate it in reverse..... PA / PF..... refer to 'Every Game every played'.....

maybe the SAFL used the same system they use now but in reverse.

[redandblack]

Both systems show a bias towards defence.

In 2 games where the margin is 20 points, both systems reward the winner of the lower scoring game over the higher scoring game.

[Wedgie]

Both systems show a bias towards defence.

In 2 games where the margin is 20 points, both systems reward the winner of the lower scoring game over the higher scoring game.

I have an issue with that too, Id prefer a points difference system but that has its mathematical drawbacks too.

[FlyingHigh]

Have always wondered whether there would be an instance of two teams could be ranked differently based on which percentage method is used. Has always sounded too hard to try to work out, but does anyone know if it could happen?

[spell_check]

I don't think it can, FlyingHigh.

I know that you can have more than 100% in business terms, but I'd prefer our method because I'm a proud South Aussie - we like to be different - red behind posts, two points for a win and one for a draw, teams named kicking up instead of down.

[FlyingHigh]

I don't think it can, FlyingHigh.

I know that you can have more than 100% in business terms, but I'd prefer our method because I'm a proud South Aussie - we like to be different - red behind posts, two points for a win and one for a draw, teams named kicking up instead of down.

And dancing girls.....but they all seem to be only about 14 these days

[Pseudo]

Have always wondered whether there would be an instance of two teams could be ranked differently based on which percentage method is used. Has always sounded too hard to try to work out, but does anyone know if it could happen?

The function mapping Victorian percentage to SA percentage is injective. i.e. if it's greater under one system, it's greater under the other.

in fact, let's have a go at it:

SA_%_1 > SA_%_2
-> F1/(F1+A1)*100 > F2/(F2+A2)*100
-> F1/(F1+A1) > F2/(F2+A2)
-> (F1+A1)/F1 < (F2+A2)/F2
-> F1/F1 + A1/F1 < F2/F2 + A2/F2
-> 1 + A1/F1 < 1 + A2/F2
-> A1/F1 < A2/F2
-> F1/A1 > F2/A2
-> F1/A1*100 > F2/A2*100
-> Vic_%_1 > Vic_%_2.

QED

[FlyingHigh]

Cheers Pseudo. Still like our system best!!

[Pseudo]

Cheers Pseudo. Still like our system best!!

oh, so do I.

(PS I edited my above post after you replied - added a proof )

[spell_check]

Cheers Pseudo. Still like our system best!!

oh, so do I.

(PS I edited my above post after you replied - added a proof )

By the way, I still haven't found the quarter scores 0.1 1.2 3.5 8.13 (61). It's had to have been done in a League somewhere.

[mal]

Why not used the MALcentage system

If a team wins by 20 points just make the % +20
It would make life easier for the predictors and not the Mathematicians

Regards
EINMAL

[Ecky]

Good work, Pseudo.

I agree with Mal that a +/- system would be much better as it would make it much simpler for the average person to understand, with almost always no difference to overall ladder positions over a season.

Plus you could argue that the differences in positions that it makes is beneficial, as it (slightly) favours higher scoring teams.

PS I remember getting into an interesting, heated discussion about this on Bigfooty a few years ago, but it all got deleted as they had a server crash.

[HeartBeatsTrue]

Both systems show a bias towards defence.

In 2 games where the margin is 20 points, both systems reward the winner of the lower scoring game over the higher scoring game.

I have an issue with that too, Id prefer a points difference system but that has its mathematical drawbacks too.

Yes but the loser of the low scoring game gets a lower percentage than in a high scoring game so it works both ways.
What about games in bad weather conditions? A team can dominate a match yet win by a smaller margin. A points difference system would favour a team like Sturt. Take the 2 last Norwood v Sturt games, both teams won their respective matches with the same dominance, but Norwood (in the wet) won by 44? points and Sturt (in the dry) won by 71 points.

The SA system is the best % wise and premiership points wise. When do you ever award 1 point or 3 points in VIC/AFL?

[Benchwarmer]

I reckon that the "for and against" differential is the best system as the margins are the better indicator of percentage - I am against a system where being defence oriented is an advantage (let it rain down goals!).

I reckon that if teams finished on the same premiership points in a season, the breakdown should be:
* For and Against Difference

If it remains equal, go for:
* Most Points For

If, somehow, they are still unable to be split, go for:
* Least Points Against

If they are still locked, take these calculations into their head to head record, starting at wins and go down the same list again.

[Psyber]

I think the SA system of percentage based on what percentage of the total points scored in your matches you kicked is better than they AFL one, and that despite the bias towards defence, it is easier to handle in your head than a point difference system in a game where it doesn't matter most of the time - after all it it not like the game is full of 0-0 draws or big 1-0 victories!
The KISS principle applies.