Tuesday, April 30, 2013

Trading Established AFL Players For Draft Picks: Who Wins?

For the past few years I have thought of looking at all the AFL trades in which an established player was traded for a draft pick, and seeing who ended up playing the most games for their respective clubs after the trade took place. My hypothesis was that the established players, on average, ended up playing more games than the unestablished ones. It’s all in ‘Moneyball’ folks! It’s better to bank on a player that has done something rather than an untried player that might do something … because many don’t.
Investigating this question though wasn’t as clean as I had hoped. First I had to chuck out all the trades where it wasn’t just a trade of one or more players for one or more draft picks. That is, trades where a draft pick was traded for a player and a different draft pick were discarded, because the players involved couldn’t be neatly categorised on each side of the ledger. Second I chucked out all the trades where the club ended up on-trading the draft pick to somewhere else, because the player taken with the draft pick didn’t end up playing for that club, which turns out to be quite a few trades. Third the records on Wikipedia of AFL trades (from where I got the data) are confusing and look a bit dodgy. This left about 70 trades since 2001 that I could use. (2001 was the year of the trade referenced by the pictured above – Trent Croad and Luke McPharlin to Fremantle for the draft picks that Hawthorn would use on Luke Hodge and Sam Mitchell.)
After all that, the results were… unclear. In most years, it looks like the teams taking the established players come out ahead in terms of number of games played. But in 2001, the teams taking the draft picks won out by a large margin: apart from the Hawks getting 200-gamers Hodge and Mitchell, Melbourne took Brad Miller with the draft pick it received for Brent Grgic, and Essendon got Andrew Welsh with one of the draft picks it received for Damien Hardwick. That took the overall total to 3435 games for the established players to 3343 games for the draft picks.

So I’m not quite sure what to conclude in the end. The established players look to have ended up with more games, but not by much, and I’m not putting a lot of confidence in the calculations. I think the established players would also have played better quality games on average, because they would  be closer to the prime years of their career (or at least further from the initial, relatively unproductive years of their career), but again I’m not sure on that. At the least, we can say that a team giving up an established player for a draft pick shouldn’t be overly confident they are going to get more games out of the latter.

No comments: