I have charts on hit rates of picks (live picks only, father/sons etc are excluded):
This one shows the percentage of players taken in each band of picks who get to 100 games. 20 to 40 isn't a huge step down, and even 40 to 50 has a bit of a hit rate. Take several shots in these ranges and you're likely to get something out of it. Three picks in the 30s is probably going to have odds around that of two picks in the 20s.
Barry is right that a 1 in 3 hit rate of 100 gamers from picks 30-40 is a good outcome, but that's also true of all picks past pick 10.
survival.png
This one is the log-regressed value we use for each pick at HPN. The main noteworthy feature is, compared to the AFL's Draft Value Index (academy points) the picks in the back end of the second round have more value - the AFL undervalues them compared to what they actually produce. A general rule of thumb is if you've got say pick 22 from early in the second round, you should trade for any two picks before about pick 35 (Port's trade in 2016 that netted us 9 and 19 got them 14, 17 and 31 was a bet something like this that worked out for both parties - Florent and Hayward on one side, Powell-Pepper and Marshall on the other, jury still out on Atley and Drew).
2017DPVC-Dots.jpg
Example theoretical equivalent expected values are: pick 5 likely to give about what two pick 25s or three pick 43s do. The "sure thing" status of very high picks drops off a lot after about pick 3, and pick 5 usually goes well but has some notably ordinary recruits attached to it such as Xavier Clark, Matt Buntine, Brock McLean and Jarrad Grant.
Those values are all a bit theoretical, and clubs simply would not trade backwards just to get the equal odds suggested here. They'd want significantly greater odds from sacrificing the high pick. In reality everyone puts a premium on the early picks because they're surer things, they have prestige, you have a wider range of choices about who to take, and because list spots are finite. Carrying multiple guys as bets on getting at least one good best 22 player has a cost in terms of list spots. If you did it three years in a row, you might be carrying ten or twelve guys in the hope of getting what 3 high picks could give you.
(Plus I think list managers have a bias towards not trading early picks because of reputation risk).
But clubs do trade backwards occasionally, and to good effect. They do much more in US sports with drafts, and I reckon we're a bit too conservative about trading down to take multiple shots in the AFL. This is, after all an unusually team oriented game, and 22 good players is a lot of players, and hard to get quickly. Quantity has its own value, and spreading your risk over more players means less eggs in one high pick basket who might blow an ACL. I also think the "trade backwards and take multiple picks" approach probably needs to be done as a long term game. There's an unavoidable failure rate to taking a higher quantity of second and early third rounders, but if done consistently over several years it should smooth out the probabilities and ensure enough successful strikes to help build a team.
Upshot: If we get Blakey at 7 that leaves picks 37, 38 and 40, and at least one of those 3 later guys would be expected to work out as well, and sometimes a second guy might work out too. Especially if we choose to believe Sydney are, through Beatson and co, better at drafting and developing from the middle part of the draft than the average club is. These averages and probabilities do, after all, include players selected by the clubs who are worst at drafting and developing players.
blakey bid.PNG
Reply With Quote
Originally Posted by barry
Bookmarks