What’s a Draft Pick Worth?

The trade period has come to a close. Like always, it was a time of squabbling and posturing by the clubs in a bid to maximise the talent on their list. Some of that talent was established, the rest is potential; either untried players or picks in the upcoming draft.

It is a difficult challenge to know what, exactly, a player is worth. This is because the draft is an inexact science.

It is possible to recruit a Brownlow medallist or premiership captain, such as Matt Priddis or Nick Maxwell respectively, via the rookie draft.

It is equally possible to recruit a Tim Walsh (1 game for the Bulldogs) with pick 4, or Scott Gumbleton (35 games for the Bombers) with pick 2.

The average AFL career is barely more than two full seasons, at a mere 55 games, but the far most common outcome for a new draftee is to never play a senior match. The average 55 game player barely exists. Instead, the draft contributes a handful 100 to 200 games players (14% of all draftees), a scant minority above that figure (7%), and a vast array of zeros (23%) and single digit careers (16%) that drag down the average.

The median value for an AFL career is 21 games. The median is another type of average which, rather than dividing the total number of games by the total numbers of players, seeks to find the middle of the sample. This means that half the recruits play less than 21 games, and half play more. That is to say around half of all new recruits will not play a full season worth of games in their entire career.

In short, the draft is very hit and miss, therefore a good measure of drafting success is longevity.

A 200 game player need not be an outright star. At the very least they will hold a position for 10 or more years which gives their club a chance to focus on recruiting and developing the rest of the team. So how do you get one?

As stated, the averages aren’t a reliable prediction, but the following graph gives an indication of how many games players achieve based on their draft position.

wpid-draft3.jpgDon’t worry if this graph appears messy or hard to follow, it is included merely for explanatory purposes. A cleaner version is below. Feel free to skip ahead if you are not interested in methodology.

The red line is the average for all players since the official formation of the national competition in 1990. Including current players draws down the graph as they are still within the system and can yet accrue more games.

The green line accounts for this shortcoming by only considering retired players. However this also is an underestimate as it emphasises failure. New draftees who have prematurely exited the system are registered while current 200 game veterans are omitted.

The yellow line neglects all players drafted within the last 10 years such that every player accounted for has had time within the system to fulfill their potential. Those who are still playing will gradually raise the average into the future. The drawback with this approach is that the smaller sample size produces a more erratic graph, as you can see.

The important thing to notice is that all three ways of considering the problem plot a remarkably similar trendline.

And now for the neat version using only the red line (ie. all players). As stated, it is an underestimate. This is a useful error as it reflects the conservative approach of list managers who value establish talent over potential talent, a reasonable stance given the hit and miss nature of the draft. It is the ‘bird in the hand’ theorem.

wpid-draft.jpgThere is a bit of an anomaly at pick 6 which appears to be cursed. Last years pick, Collingwood’s Matthew Scharenberg, injured his knee and is yet to debut. Sydney’s Gary Rohan has been limited to 43 games in 5 years due to a severe hamstring injury and that horrific broken leg which saw the introduction of the new sliding rule. Mitch Thorpe and Beau Dowler managed just 2 and 16 games respectively for the Hawks. West Coast’s Ashley Sampi flew to great heights on the field to claim Mark of the Year in 2004 but his career only lasted 78 games. Dylan Smith, Murray Vance, John Rombotis, Daniel Healy, Robert McMahon, Trent Cummings, Robert Pyman, Paul Burton and Ben Aulich round out a list of players you’ve probably forgotton, or never heard of in the first place. Any pick can flop. Inexplicably, position 6 has had more than its share.

The large spike at the end is West Coast’s Brayden Lyle who played 116 games after being drafted with pick 124 in 1992. It was the only time that pick, and the 11 preceding it, had been used making the end of the graph a relatively useless average from a sample of one. It can justifiably be ignored.

Overall, the trend is as one would expect. The first few picks tend to be standout players then the returns diminish.

However, by the third round of selections the trendline asymptotes. A pick 100 is no less reliable than a pick 60. This suggests that the jocking for position clubs engage in with the swapping of late picks is a relatively useless endeavour.

For example, the Jonathan Giles trade from GWS to Essendon involved the swap of picks 53 and 62. This swap moved GWS slightly forward in the order of the upcoming draft, but if history is any guide this will provide little advantage or disadvantage to the respective teams. In effect, Essendon has recruited Ryder’s replacement for free.

One of the most talked about trades of the period followed Dayne Beams’ decision to return home to Brisbane. In the end, Collingwood were compensated with picks 5, 25 and former rookie listed player, Jack Crisp. Was it enough?

A pick 5 averages 110 games, a pick 25 averages 62 games. Adding those expectations to Beams’ current tally of 110 games, and bearing in mind they are slight underestimates, he would need to enter the 300 club for Collingwood to be significantly disadvantaged. Thus it is unlikely they will lose from the deal.

The decision to on-trade pick 25 for North Melbourne’s Levi Greenwood (a young best and fairest runner up) will guarantee Collingwood a higher return than would be expected from the pick itself, improving their position.

Furthermore, if Crisp is able to force his way into the best 22, they could end up well in front.

Reluctantly farewelling a player many thought untouchable could turn out to be the most advantageous trade of the period, provided that pick 5 isn’t wasted on one of those many zeros.

Do you agree? Comments welcomed below…

Below are the exact values from the red graph. Try to add up who won or lost a given trade using this measure.

Pick 1 154 Pick 26 56 Pick 51 12 Pick 76 27 Pick 101 32
Pick 2 141 Pick 27 54 Pick 52 30 Pick 77 27 Pick 102 17
Pick 3 130 Pick 28 49 Pick 53 54 Pick 78 29 Pick 103 65
Pick 4 92 Pick 29 72 Pick 54 15 Pick 79 57 Pick 104 14
Pick 5 110 Pick 30 47 Pick 55 65 Pick 80 7 Pick 105 12
Pick 6 51 Pick 31 66 Pick 56 76 Pick 81 19 Pick 106 68
Pick 7 97 Pick 32 61 Pick 57 25 Pick 82 40 Pick 107 18
Pick 8 104 Pick 33 58 Pick 58 42 Pick 83 9 Pick 108 38
Pick 9 87 Pick 34 43 Pick 59 20 Pick 84 20 Pick 109 26
Pick 10 88 Pick 35 22 Pick 60 39 Pick 85 39 Pick 110 3
Pick 11 95 Pick 36 53 Pick 61 32 Pick 86 33 Pick 111 16
Pick 12 80 Pick 37 77 Pick 62 41 Pick 87 49 Pick 112 5
Pick 13 107 Pick 38 48 Pick 63 27 Pick 88 13 Pick 113 0
Pick 14 82 Pick 39 46 Pick 64 31 Pick 89 24 Pick 114 0
Pick 15 72 Pick 40 90 Pick 65 7 Pick 90 42 Pick 115 0
Pick 16 88 Pick 41 45 Pick 66 46 Pick 91 34 Pick 116 0
Pick 17 69 Pick 42 47 Pick 67 47 Pick 92 40 Pick 117 0
Pick 18 67 Pick 43 63 Pick 68 29 Pick 93 9 Pick 118 1
Pick 19 86 Pick 44 48 Pick 69 27 Pick 94 29 Pick 119 0
Pick 20 71 Pick 45 52 Pick 70 24 Pick 95 32 Pick 120 0
Pick 21 61 Pick 46 61 Pick 71 38 Pick 96 10 Pick 121 0
Pick 22 44 Pick 47 63 Pick 72 22 Pick 97 26 Pick 122 0
Pick 23 49 Pick 48 39 Pick 73 38 Pick 98 19 Pick 123 0
Pick 24 72 Pick 49 56 Pick 74 42 Pick 99 30 Pick 124 116
Pick 25 62 Pick 50 32 Pick 75 25 Pick 100 39

, ,

  1. Leave a comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: