The True Value of Draft Picks

[Straight Sets Simon 23,113]

Does anyone know of any analysis which shows historically on average the quality of player that each draft pick will produce?

"Quality" is difficult to measure, therefore something along the likes of number of games played per draft pick could be a good start.

It would be interesting to see given how we argue over what a player is worth compared to a draft pick.

[beelzebub 23,390]

If I may :

I just think this is somewhat flawed as a concern. By this I suggest many wil look for a more classical , liner style, equation to answer the question. Reality is I think it's more akin to Chaos Theory.

Any algorithm you could hope to use would be so convoluted with so many variables as to render it a nonsense.

The value of draft picks is inherently what ever the holder's need are as compared the place in queue they find themselves at EACH juncture of the call.

I.e It morphs even as it it happens.

[DistrACTION Jackson 10,180]

Does anyone know of any analysis which shows historically on average the quality of player that each draft pick will produce?

"Quality" is difficult to measure, therefore something along the likes of number of games played per draft pick could be a good start.

It would be interesting to see given how we argue over what a player is worth compared to a draft pick.

I remember someone posted a graph showing the average games played per pick.

It was pretty much as expected with the the top 10 averaging the most and then it would decrease as it went into the later picks.

There were a few surprise results, I think there was a pick in the 30's that was a bit of an anomaly and had a very high return.

[Dr. Gonzo 24,435]

Any analysis would probably have to group the draft picks as it would be difficult to pin down on individual draft numbers. Either grouping them in lots if 5/10 or doing some kind of formula like Draft Pick +/- 2 picks and start looking from there.

[Adzman 2,154]

Honestly it is a lottery. Being at the pointy end of the draft helps, but their are never any guarantees.

I think a good football department is where you can cut out some guesswork and risk. Until recently we haven't had this. The true value of picks is getting as even return as possible on your picks. We haven't managed to do this in recent years.

Cool thread, I look forward to seeing some charts and stats around this topic.

[rpfc 28,895]

The AFL has made the points system based on salary which is a better measure of quality in my opinion than games played.

But the use of that system in all trades is fraught and I wouldn't recommend fans doing it to judge what a player is worth, or how their club views a player.

[Curry & Beer 5,444]

Honestly it is a lottery. Being at the pointy end of the draft helps, but their are never any guarantees.

Then it's not a lottery. A lottery implies that any number you pull out of a hat is just as likely as any other to be a winner. The history of the draft shows very clearly that there is a correlation between quality and higher number, but of course you are correct that there are both good and bad exceptions. However, the bottom line is your probability is higher and that's why clubs covet them. Don't you think it's a bit of a fluke that Hogan, Brayshaw, Tyson, Salem and Viney (realistically) are all high draft picks? Conversely, your odds of getting a Toumpas are much lower. Then of course there is the quality of the club developing them that factors in.

[ChaserJ 5,169]

I remember someone posted a graph showing the average games played per pick.

It was pretty much as expected with the the top 10 averaging the most and then it would decrease as it went into the later picks.

There were a few surprise results, I think there was a pick in the 30's that was a bit of an anomaly and had a very high return.

Pretty sure it was pick 40 that was the anomaly in that range. Think the average games played per player selected there was something like 56, which was quite high by comparison to picks around that mark.

[w00dy 146]

the chart on page 6 of this file on the points system gives you a good idea.

http://www.afl.com.au/staticfile/AFL%20Tenant/AFL/Files/Father-son-bidding-system.pdf

The R-squared of the regression indicates that around 75% of the player's salary can be explained by the draft position that they're taken. But looking at the chart, the raw data would indicate the points allocated to most picks in the top 35 tend to overestimate the value of those picks - i.e. they should be worth less, and picks 35-50 should be worth a bit more.

I'm surprised they weren't able to get a better fit to the data. It looks like the number 1 pick has dragged up the regression - unsurprisingly - it's probably he who I will not name and his millions causing that. I wonder if they adjusted for him - or perhaps a general bias for number 1's to get paid more on potential - no doubt we've paid Jack Watts more than his output reflects thus far.

[djr 1,605]

So does each afl player have a rating? eg what is Howe's rating in points?

[jnrmac 20,271]

Does anyone know of any analysis which shows historically on average the quality of player that each draft pick will produce?

"Quality" is difficult to measure, therefore something along the likes of number of games played per draft pick could be a good start.

It would be interesting to see given how we argue over what a player is worth compared to a draft pick.

Probably only the games played stat - everything else is pretty subjective.

[jnrmac 20,271]

I remember someone posted a graph showing the average games played per pick.

It was pretty much as expected with the the top 10 averaging the most and then it would decrease as it went into the later picks.

There were a few surprise results, I think there was a pick in the 30's that was a bit of an anomaly and had a very high return.

You are more likely to persevere with a high draft pick altho Melbourne has mad a habit out of prolonging careers like Danny Hughes and Rohan Bail.......

[Middymalt 244]

You are more likely to persevere with a high draft pick altho Melbourne has mad a habit out of prolonging careers like Danny Hughes and Rohan Bail.......

matty bate would've played 40 games at a decent club.