Determing NRL Ladder rankings

[Apey]

Does anyone remember what they use (in precise terms) after F - A to determine ladder rankings for the NRL?

According to the 2012 NRL Operations Manual, it's For/Against as a %.

However, that would have

13. Souths
14. Cowboys

whereas nrl.com has

13. Cowboys
14. Souths

Most other websites (including NRL) seem to have the former.

So has the method for determining rankings changed since 2012 or is NRL.com just wrong? It wouldn't surprise me in the least if they were.

 

[muzby]

nrl.com works on alphabetical order..

this is why the Adelaide Aardvarks are so keen to enter the comp..

 

[Apey]

Somewhat embarrassing if their 3rd criteria is indeed alphabetical order...

 

[Timmah]

My understanding is their first criteria is points (obviously), but it looks like the second criteria is actually matches played.

 

[madunit]

First criteria is points difference, second is tries scored, third is points conceded (least ranked higher)

 

[Apey]

First criteria is points difference, second is tries scored, third is points conceded (least ranked higher)

Hm they must have changed it somewhat recently then.

Thanks.

 

[Timmah]

Correct.. but what's been identified is that NRL.com's ladder engine must be all wrong. They haven't changed how the ladder's structured, its just a shit website backend.

 

[Tayjack]

Hm they must have changed it somewhat recently then.

Thanks.

Nope - never been changed, has always been ponts difference.

 

[Apey]

Nope - never been changed, has always been ponts difference.

then someone will have to tell them their 2012 Operations Manual was incorrect

http://www.abc.net.au/4corners/documents/concussion2012/NRL_Operations_Manual_2012.pdf

1.7.6 At the conclusion of matches in NRL/NYC Competition rounds as provided for in
the NRL/NYC Draws, if two (2) or more teams are equal in competition points,
rankings will be determined according to the following criteria from the total
results of all matches played in NRL/NYC Competition rounds:

a) greater positive difference between points scored for and against,
then if equal,

b) greater percentage of points scored for and against, being determined
by:
points scored for x 100
points scored against 1
then if equal,

c) most tries scored, then if equal,

d) most goals kicked, then if equal,

e) most drop goals kicked, then if equal,

f) by the toss of a coin.

b) didn't quote across properly but it's for/against x 100

 

[madunit]

It is regularly being fiddled with. For a season or two in the 80's or 90's it was pts diff then best defence. Now, defensive record is not as important which i'm not entirely a fan of.

This multiplying thing is nonsense imo. Use the raw data available, theres more than enough there to figure it out fairly.

But Tim is right, the ladder on the NRL site has bad mechanics and always has.

 

[madunit]

Hm they must have changed it somewhat recently then.

Thanks.

Yep you're right. I didn't realise they had. Last time i looked into this was 2006 i think. Had no ide it had been changed (or that it needed to be)

 

[madunit]

Nope - never been changed, has always been ponts difference.

No it hasn't always been points difference, that only started in the 80's/90's and was only used to order teams on the ladder, not to determine if they would miss the finals or not (as they still had play offs if teams were tied on competition points). It was tweaked several times in about a decade

 

[TheFrog]

1.7.6 At the conclusion of matches in NRL/NYC Competition rounds as provided for in
the NRL/NYC Draws, if two (2) or more teams are equal in competition points,
rankings will be determined according to the following criteria from the total
results of all matches played in NRL/NYC Competition rounds:

a) greater positive difference between points scored for and against,

then if equal,

b) greater percentage of points scored for and against, being determined
by:
points scored for x 100
points scored against 1
then if equal,

c) most tries scored, then if equal,

d) most goals kicked, then if equal,

e) most drop goals kicked, then if equal,

f) by the toss of a coin.

This is REALLY poorly written. For instance, in (a) only a positive points difference appears to be taken into account. (b) doesn't seem to make sense at all. The only thing that is clear is the toss of a coin. This method was in fact used to determine a pre-season semi final in 1968.

 

[Apey]

It is regularly being fiddled with. For a season or two in the 80's or 90's it was pts diff then best defence. Now, defensive record is not as important which i'm not entirely a fan of.

I agree that defence should probably be worth more than offence as far as rankings go.

1.7.6 At the conclusion of matches in NRL/NYC Competition rounds as provided for in
the NRL/NYC Draws, if two (2) or more teams are equal in competition points,
rankings will be determined according to the following criteria from the total
results of all matches played in NRL/NYC Competition rounds:



This is REALLY poorly written. For instance, in (a) only a positive points difference appears to be taken into account. (b) doesn't seem to make sense at all. The only thing that is clear is the toss of a coin. This method was in fact used to determine a pre-season semi final in 1968.

Yeah they chose their words fairly poorly in a), in attempt to explain that +3 is regarded as greater than -7 for example. Difference often has the implication of absolute difference in mathematical terms, that's probably why they've slipped positive in there.

b) is just a silly way of saying For/Against as a percentage.

 

[simmo1]

Pretty sure that if points difference is equal, then ratio of for/against will be higher for the team with the better defence anyway.

EDIT: Unless the PD is negative, in which case the better attacking record will give the higher ratio (I think).

 

[Bovrick]

Pretty sure that if points difference is equal, then ratio of for/against will be higher for the team with the better defence anyway.

EDIT: Unless the PD is negative, in which case the better attacking record will give the higher ratio (I think).

Pretty much, they're equivalent. Teams tied on positive PD are divided by best defence; team tied on negative PD are divided by best attack. Can't for the life of me think why they made that particular judgement call though.

 

[Apey]

All true, equal differentials will see the F/A ratio approach 1 as F & A increase (absolutely). For positive F-A it approaches from above 1 whereas for negative F - A it approaches from below 1, meaning the lower the F&A for the positive differential the higher the ratio will be and conversely for the negative differentials.

It does seem sillier the more you think about it. Better off just deciding whether you give more credence to for or against and rank them accordingly.

 

[TheFrog]

In 2005 Sydney Roosters and Penrith finished level on 26 points, one win outside 8th place. Roosters difference was +1 and Penrith's was 0. Had they both been 0 Penrith would have taken 9th place on most tries (101 v 86). Had both been +1 Roosters would have taken 9th on percentage (100.20 vs 100.18). This would have been for 8th place had Manly lost to Canberra in the final round.

This was a strange season in many ways. Two teams met in the Grand Final both of which had never reached the GF before, and the wooden spooners Newcastle had a 6 match winning streak near the end of the season yet still ran last.

 

[firechild]

I would love to see Souffs finish 9th as the result of a lost coin toss.

 

[goodplayer]

most grand finals played ok