Since the introduction of free agency starting in the 2012 off-season a total of 26 delisted, 13 restricted and 17 unrestricted free agents have moved clubs with heavy scrutiny on the impacts of the system in regard to competitive balance of the league. It has been claimed that lower ranked clubs are negatively affected by the system as they are seen as a less attractive proposition by players whom qualify for free agency. The most prominent example of this being when James Frawley left Melbourne for Hawthorn in the 2014 off-season, moving from a then perennial cellar dweller to the reigning premiers. However is this example representative of player movement under free agency, or is it just a convenient vehicle for people to use as an argument against players moving between clubs with Premiership aspirations at the forefront of their minds?

In analyzing this, ignoring delisted free agents is necessary. This is because they essentially are players whom are deemed not worthy of a list spot at their current club and thereby will naturally tend to be picked up by lower ranked clubs where they may get more opportunity. Thus their inclusion would skew the results toward lower ranked clubs benefiting from this system, even though these players would be picked up with or traded for very late picks otherwise.

table-fa2.png

The 29 restricted and unrestricted free agents averaged moving to a team 0.43 places above them on the ladder during the season that had just concluded. This is also consistent with using the difference in win percentage as a metric for performance of a team, with these free agents averaging moving to a team with a 5.22% higher win rate than their former club. This 5.22% equates to roughly one extra win over a 22 game season, practically insignificant. However the median value for both these metrics was zero, thereby these numbers lead toward there being no real trend in the number of players moving to clubs higher or lower on the ladder.  The histograms below depict and relatively even spread of movement, with the histogram not including delisted free agents perhaps slightly leaning toward players moving toward higher ranked teams as a trend.

histo-all.png

Histo - No DFA

However the above histograms do not take into account the quality of player that is moving, that is the above essentially cancels out James Frawley moving 16 rungs up the ladder in 2014 and Tom Murphy moving 15 rungs downward in 2012 despite their clear difference in quality. In doing so I have used the past two years Supercoach average as a proxy for performance of an individual player, with any player whom has played 10 or less games over the period being not included due to lack of sample size and any individual season with 5 or less games being not included and their other sole season average used.

With an admittedly relatively small 27 player sample size, there was a very weak, negative linear association between the difference in win percentage and two year supercoach average. As can be seen in the below scatterplot, there is practically no pattern between these two metrics. However is the concern that in the long term lower to middle ranked clubs may become ‘feeder clubs’ for the high ranked teams through free agency a genuine possibility?

Scatterplot FA

The AFL has many different measures in place to attempt to keep the league as competitively balanced as possible, with the major ones being the salary cap and the reverse order draft. The former essentially controlling spending on total player payments so that clubs have equal opportunity monetary wise in retaining and luring talent, whilst the latter ensures that the lower ranked clubs have greater opportunity in securing young talent and have more currency for trading in the off season. Free agency hypothetically should not have a negative effect on competitive balance given the mentioned key implemented policies, allowing more free player movement and thus players are paid more closely to market value. However there is one key free agency policy specific to the AFL which undermine this.

The first being compensation picks being awarded for players whom leave and are deemed valuable enough via the AFL’s compensation formula to justify draft picks being awarded for their loss. These picks are tied to the clubs order in the draft, hence James Frawley leaving awarded Melbourne pick number 2 whereas Lance Franklin leaving awarded Hawthorn pick 19, despite both under the formula being worth a “first round draft pick”.

These compensation picks make it more difficult for the current club to justify further competing for the players signature, as they are in effect bypassing this pick as well as the salary cap space. This encourages clubs to allow players to move to thus open up cap space to use as they please, whilst being awarded a draft pick to use to secure young talent. The value the club will be willing to pay to hold onto their player is thus lower than the clubs attempting to poach them. The option to allow your restricted or unrestricted free agent to leave scales as being a more and more attractive option the lower you have placed on the ladder in that season.

Thereby this system, by design, incentivises the lower ranked clubs to allow players to leave via free agency as these more experienced players likely won’t be as useful for their next premiership tilt as a potential young gun. This is problematic from a competitive balance point of view, as in effect would cause teams to spend longer stints at both poles of the ladder.  Higher ranked clubs are given less incentive to allow their experience players to leave whilst having a higher willingness to pay to ‘buy’ talent off lower ranked clubs due to the better picks being awarded to these lower ranked clubs for equivalent losses in playing list talent. This creates a situation whereby players whom hold a greater ability to immediately impact matches tend toward the currently high ranked teams, whilst low ranked teams draft players whom typically have very limited impact early in their careers.  All typical forms of competitive balance suffer with this, those measures being game to game, within season and across seasons due pushing the league toward a more ‘tiered’ league in the short to medium term.

Keeping the reward of compensation picks within the AFL’s free agency regulations, even though the data so far has shown there to be negligible bias in the movement towards currently contending clubs, over time could have the side effect of an increased premiership window and greater time required to rebuild, all else equal. From the standpoint of the health of fan interest in variety of clubs being successful and uncertainty of outcome of individual matches, of which teams participate in finals and whom competes for premierships, the compensation pick should be scrapped so that the current competitive balance regulations can work more effectively alongside the free agency system.

Advertisements