Analysis AFL Draft Pick Value Calculator- Do you think this is an accurate tool?

[Socrates2]

I like using this tool to evaluate the value of draft picks.
For example the Kelly trade -The Cats will receive picks 14, 24, 37 and a future first-round selection while the Eagles receive Kelly, Pick 57 and a future third-round selection.

Picks 14, 24 and 37 are equivalent to pick 3 according to the points system.

AFL Draft Pick Value Calculator - Draftguru

Are picks 14, 24 and 37 worth pick 3?
There are other examples eg the draft points calculator says picks 5 and 6 is worth pick 1 and pick 30.

 

[RobbieK]

It uses the official point system the AFL uses to determine bid for father/son and academy selections, so in that sense the calculator is accurate. But is that point system accurate? I guess that is the real question you are asking.

The problem is that the same pick can have a different value depending on the year it is being used and the needs of the club.

Take your pick 1 and 30 being equivalent to 5 and 6. In a year with a very even top 10 which then drops away a lot it might be a no brainer to trade away pick 1 and 30 to get two selections in the top 10. Likewise, if you are in a big rebuild you might favour two players who are very good but not quite as good as pick one than putting your eggs in the one basket. But there might be cases where the number one player is clear and fits a need and a club will decide that they would rather that player than two they don't rate as much.

In general, though, I think it is about right. Most criticism I see tends to be that it doesn't value higher picks enough and values lower picks too much, which I think is a bias that people have that doesn't really match the actual output of the picks and the risk of any pick bring a total bust.

Taking 1 and 30 vs 5 and 6, which combination of players would you prefer?

Walsh and Hammill or Rozee and B King?
Rayner and de Koenig or Cerra and Stephenson?
McGrath and McLarty or Setterfield and Petrevski-Seton?
Wietering and Redman or Parish and Francis?
McCartin and Maynard or De Goey and Marchbank?

In theory these should be of equal value according to the draft points. I think in general people would end up prefering to have 5 and 6 more often than 1 and 30, which maybe suggests pick 1 is still valued higher in the points system than it ought to be?

 

[RobbieK]

Your other one was 3 vs 14, 24 and 37.

Rankine vs Hately, Hill and Vandermeer
Dow vs Ling, Kelly and Petty
McCluggage vs Perryman, Cox and Rotham
Mills vs Hipwood, Keays and Bonner
Brayshaw vs Lever, Steele and Rose

Pretty even I think. With 3 you put all your eggs in one basket. Tends to be a good player, but with three later picks you have the advantage of taking multiple punts, so even with a few busts in there there are also a bunch of players teams would be very happy to have.

 

[footyfan1978]

It is accurate to figure out and follow the father son and academy bids in the drafts but beyond that to consider it a true accurate value of each draft pick or an array of draft picks it is only really a guide as to a sense of ball park value. If you thinking it something more you probably misleading yourself.

Kelly trade is not the same as Eagles trading Cats pick 3 for example because they do not have pick 3 to use at the draft. But it gives a sense the later array of picks in themselves has value to Geelong in what Eagles ended up trading them. The fact Eagles could not trade to get that pick exactly for Geelong tells you the pointy pick in itself worth more than DVI tells you it is for bidding purposes.

But the Draft Value Index itself for live bids on academy players I find probably does not give enough points to what the top 3 or 4 picks should be worth.
I think they should have more points given to them to need to match as they really are the cream of the crop and the clubs in bottom 3 are having their access to cream of the crop maybe pushed further back than it should be. I basically think it should be harder for clubs to match bids that early to protect the clubs down below that trying to work the system to usually build up from a mess. A club that in top half of ladder finding it too easy to get a real top 4 pick from matching a club from bottom rungs of ladder is not a good look for mine in terms of their own stated equalisation policies. I think for pick 5 and below I think they got the value index as good as you can expect. Just those first few picks I believe undersell their real value in terms of young talent at pointy end of draft.

The DVI kind of drops off giving any value to picks beyond 70 which is good guide as about 70 to 80 players drafted each year and probably about 30 players make a decent career out of and maybe on average each draft should produce two or three absolute champions and maybe a dozen elite players and the best 30 players of each age group is what you want to get access to. Once you get to third round picks you really are getting in lottery territory so if my club trading picks from outside that range I never get concerned. Picks in 20's should be valued and not traded away easily. Picks in top few should be treated as precious.
Gold Coast trading away pick 2 for Weller a few years back was just beyond belief stupid. Do not worry, AFL come along to make up for their mistakes with so much charity this off season.

Be interesting exercise to go back and use DVI to find how much the DVI says clubs got with priority picks when they handed out more freely in 1990's and first decade of this century.

 

[Franc de Borges]

 

[Mitchell54]

I'd say it's quite accurate but the real answer is that there isn't an answer as it changes from year to year, club to club.

 

[Richo83]

I'd be genuinely interested when the calculations were made. I think we're moving towards recruiting being more of a science than an art for teenagers, meaning that early picks are more valuable. This is blurred by academy picks and the like, but I find it unlikely that the sums should stay the same into the future. IMO we're seeing less of a need to pick first and more of a need to pick somewhere in the top 10-15 to get a real star, with an importance of somewhere in the single digits.

Take the picks one and 30 being equal to picks five and six. Does that stack up in recent years?

2018: Walsh and Hamill v Rozee and King - Walsh is better than Rozee, but Rozee is a gun and I'd take King over Hamill
2017: Rayner and De Koning v Cerra and Stephenson - take the latter to be honest
2016: McGrath and McLarty v Setterfield and SPS - take the latter to be honest, McLarty was delisted
2015: Weitering and Redman v Parish and Francis - this is fairly even, with Weitering shading Parish and Francis
2014: McCartin and Maynard v De Goey and Marchbank - easy winners to the latter

What stands out is that most if not all of the 30th pick players are role players at best and seconds depth at worst, whilst everyone of the picks five and six either plays in the AFL side or will about to. This is reflected in the fact that the difference between picks one and five is equal to the difference between picks five and 25. Yet let's have a look at who was taken at pick 25 in the last five years:

2018: Rowbottom
2017: Balta
2016: Long
2015: Dunkley
2014: Nielson

Dunkley is nice, Balta shows promise but Long and Nielson isn't and wasn't much. Rowbottom is a nice prospect but he's little more than that and I think the gap between Walsh and Rozee is bigger than the gap between him and Rozee. I know this is a limited sample size but the changes in the trends of player careers IMO requires a re-think.

 

[Richo83]

It uses the official point system the AFL uses to determine bid for father/son and academy selections, so in that sense the calculator is accurate. But is that point system accurate? I guess that is the real question you are asking.

The problem is that the same pick can have a different value depending on the year it is being used and the needs of the club.

Take your pick 1 and 30 being equivalent to 5 and 6. In a year with a very even top 10 which then drops away a lot it might be a no brainer to trade away pick 1 and 30 to get two selections in the top 10. Likewise, if you are in a big rebuild you might favour two players who are very good but not quite as good as pick one than putting your eggs in the one basket. But there might be cases where the number one player is clear and fits a need and a club will decide that they would rather that player than two they don't rate as much.

In general, though, I think it is about right. Most criticism I see tends to be that it doesn't value higher picks enough and values lower picks too much, which I think is a bias that people have that doesn't really match the actual output of the picks and the risk of any pick bring a total bust.

Taking 1 and 30 vs 5 and 6, which combination of players would you prefer?

Walsh and Hammill or Rozee and B King?
Rayner and de Koenig or Cerra and Stephenson?
McGrath and McLarty or Setterfield and Petrevski-Seton?
Wietering and Redman or Parish and Francis?
McCartin and Maynard or De Goey and Marchbank?

In theory these should be of equal value according to the draft points. I think in general people would end up prefering to have 5 and 6 more often than 1 and 30, which maybe suggests pick 1 is still valued higher in the points system than it ought to be?

Stole my thunder! But yea, we overvalue number one picks in comparison to picks in the top ten. IMO the index is a bit outdated and is still reflective of the fact that the likes of Riewoldt, Hodge, Deledio, Goddard, Gibbs etc. have played a lot of games. In recent years however, they've been slightly underwhelming.

Watts
Scully
Swallow
Patton
Whitfield
Boyd
McCartin
Weitering
McGrath
Rayner
Walsh

Walsh looks like the exception to the rule, but there's lot of goodish in that list. High picks matter, but I'm not sure the highest pick matters as much as it does. Capping the draft value at 3,000 also feels arbitrary.