# World Rankings - Method

For our latest rankings and links to match criteria, see World Rankings - Current.

There is currently no **official** system of World Rankings for Australian Football, as the international game is still mostly amateur, still maturing and historically there have been insufficient matches between the nations to reliably compare them. However, at worldfootynews.com, we have attempted to create a ranking which resulted in opinion based tables for 2006, 2008 and a statistically based table since 8th August 2010. Below is a discussion of the statistical method used.

- All eligible games since 1994 are included (criteria is discussed separately). We gratefully acknowledge the efforts of Cam Homes in compiling the initial data set.

- The system is based on a combination of changes by Cam Homes and World Footy News to the IRB method used for Rugby Union.

- All teams start on 40 points.

- Points will typically vary from 0 to 100 although they are not strictly limited to this range.

- The table is updated periodically.

- The rankings are not claimed to be official but we would welcome their use in tournament rankings and as a relatively reliable guide.

- Points are awarded or lost after each game based on a formula.

- The points available when two sides play each other is based on their current ranking points (in real time, not simply the last available published table). The calculation is as follows: the ranking points margin between the two teams, multiplied by a fixed gradient, then an intercept is added. Essentially there is a linear curve that shows how many points either team will get/lose if they win/lose. The further apart they are in ranking the more the lower team has to gain and the more the higher team has to lose. If the teams are equally ranked then they have the same amount to win or lose (the intercept). The gradient controls how quickly that difference changes as the teams separate in ranking.

- E.g. consider two imaginary sides, Footystan (starting with 50 ranking points) and Aussierulsia (46 ranking points). If Footystan hosts Aussierulsia in a normal match, then Footystan start with 50 + 3 points as home team, and Aussierulsia with 46. The difference is therefore -7 from the higher team's perspective, close to the maximum gap (10) where we consider them a long way apart and the higher team would be purely defending their points. So Footystan should have a lot to lose and only a modest improvement if they defeat the lower ranked side. The equation for the points available for the higher ranked team to lose is therefore

*difference * gradient + intercept = -7 * 0.15 + 1.5 = 0.45*. From the lower ranked team's point of view the gap is +7, so the equation gives*7 * 0.15 + 1.5 = 2.55*. If it is a draw then the equation exchange is simply the difference times the intercept, i.e.*7 * 0.15 = 1.05.*

A key to remember is that the points are exchanged. So here if Footystan win they gain 0.45 and Aussierulsia lose 0.45, if Footystan lose they lose 2.55 and Aussierulsia gain 2.55, if it's a draw Footystan lose 1.05 and Aussierulsia gain 1.05, a reasonable amount which recognises the underdog drew with the favourite.

- A lot of data and hypothetical situations were tested to tune the equation's parameters (the gradient and intercept). In the end, the proof will be in seeing the rankings reflecting the strengths of nations and moving approximately as expected.

- To reflect home ground / country advantage the home team has 3 ranking points added to their total for the purposes of the points exchange calculation. This is applied to one off or two team series, it generally is not applied to tournaments with multiple teams, as the home ground advantage is often diminished as teams experience neutral games, often stay for extended periods and the home side often has a heavy burden in organisation.

- The points are also weighted according to the importance of the tournament. Generally they will have a weighting of 1 and International Cup matches 2 (for IC11 the initial shortened seeding games were weighted as 1).

- A win by 40 or more points attracts an additional bonus weighting of 1.5.

- A team more than 10 ranking points ahead of their opponent is simply defending their points, they cannot win any more from that match.

- The maximum points transfer for any given game is 3, to limit overly dramatic changes in the system to potentially unusual circumstances.

- New countries are given provisional ranking points but not actually ranked until they have played 8 eligible games.

- If a country does not play for more than 4 years they lose their position but not their ranking points, i.e. they become provisional again. If a country does not play for 8 years they will also lose their ranking points altogether, starting from scratch when/if they resume competition, unless a special case is considered.

We're aware that the system is somewhat convoluted and difficult to follow, however the basic model appears to work in Rugby Union. This version is an accelerated model to allow more rapid movement of teams from less games. It does have some weaknesses, e.g. a team ranked 10th could keep beating a team ranked 11th and acquire points, pushing up past teams ranked higher, deliberately avoiding playing them. We hope this is not exploited and will monitor the situation but for now we have not produced a better alternative that meets all the desirable criteria.

Australia is given an honorary ranking of 1. This appears to be generally accepted although there is the occasional reader who disputes it. In essence our argument is that Australia is clearly capable of producing the best side in the world from existing players. No international side has yet shown itself likely to beat a state league side let alone Australia, which does not compete at internationally not remotely for fear of loss but as a genuine contender has not emerged. If at some time a contender emerges then the #1 rank will be reconsidered.

An additional argument has been that Ireland could assemble a competitive team. But the criteria is based on Australian football players, not "potential" players that could be drawn from other sports. Likewise no doubt a very good side could be produced by taking America's best athletes from the NBA and NFL and fast-tracking them in Aussie Rules, but rankings are never about such hypotheticals. At a more realistic level, yes, Ireland's existing players in the AFL could be considered, but professional players have never made themselves available for international Australian football, so it is too hard to judge, whereas Australia could comfortably draw a #1 side from the AFL, from the SANFL, VFL, WAFL, etc or indeed from numerous top semi-pro and amateur leagues. So for now, the rankings award Australia #1 and all other positions are based on actual results.