Can Hawthorn succeed while ignoring the elite end of the draft?

[Ghost of Punt Road]

I think all clubs (maybe not Carlton) will pick a good late player or two occasionally.

I think the question is how many do you need? My best guess is that one decent draftee per year isn't enough to build a good team.

On general feel you'd want to bring in by any means about three decent players every year or the team will drift over time.

So one trade with a high draft pick or two plus a good late pick is probably still thinning the list I suspect.

 

[Carl Spackler]

For the scholarly minded amongst you Massy and Thaler, have researched the NFL draft strategies. They concluded that the best strategy is NOT to trade in first round draft picks as first round draft picks are overvalued. But instead accumulate second and third round picks as there is much more value to be found in later picks.

Interesting.

“Using archival data on draft-day trades, player performance and compensation, we compare the market value of draft picks with the surplus value to teams provided by the drafted players. We find that top draft picks are overvalued in a manner that is inconsistent with rational expectations and efficient markets and consistent with psychological research.”

The Loser's Curse: Decision Making & Market Efficiency in the National Football League Draft

A question of increasing interest to researchers in a variety of fields is whether the biases found in judgment and decision making research remain present in c

papers.ssrn.com

A little more googling reveals a few articles where they discuss their findings.

 

[cryptor]

Interesting.

“Using archival data on draft-day trades, player performance and compensation, we compare the market value of draft picks with the surplus value to teams provided by the drafted players. We find that top draft picks are overvalued in a manner that is inconsistent with rational expectations and efficient markets and consistent with psychological research.”

The Loser's Curse: Decision Making & Market Efficiency in the National Football League Draft

A question of increasing interest to researchers in a variety of fields is whether the biases found in judgment and decision making research remain present in c

papers.ssrn.com

A little more googling reveals a few articles where they discuss their findings.

I think most people intuitively know that first round draft picks are overvalued, but I also think that the hope that comes with them can make people fixate on the upside.

 

[KermitJagger]

Interesting.

“Using archival data on draft-day trades, player performance and compensation, we compare the market value of draft picks with the surplus value to teams provided by the drafted players. We find that top draft picks are overvalued in a manner that is inconsistent with rational expectations and efficient markets and consistent with psychological research.”

The Loser's Curse: Decision Making & Market Efficiency in the National Football League Draft

A question of increasing interest to researchers in a variety of fields is whether the biases found in judgment and decision making research remain present in c

papers.ssrn.com

A little more googling reveals a few articles where they discuss their findings.

It's probably weighted against the environment that they're drafted into, where a positive environment has a positive impact on the output of the player.

Have a look at all of the highly rated players who have left teams where the system made them thrive, only to become average at their new club - or conversely, average players at other clubs that come to thrive in the right system.

San Antonio Spurs and the New England Patriots are both fantastic (if oft-repeated) examples of the sum being greater than the parts.

Very interesting reading and would be interesting to apply it to the AFL draft.

 

[Carl Spackler]

I think most people intuitively know that first round draft picks are overvalued, but I also think that the hope that comes with them can make people fixate on the upside.

That’s pretty much what that paper argues. I found another detailed article that looked at the research again 10 years later under the new NFL cba which is more similar to the AFL agreement (fixed salaries/wages for picks by round) and it argued the bias remains. The sweet spot for value is mid-second to mid-third now (previously mid-second).

It doesn’t mean you get the best players from mid-second onwards but that you get the best bang for your buck. The best players are still at the top. So if you need a specific player then you need to move up, but if you’re shopping to fill out your lest with value then you want to load up in the second and third round.

 

[Carl Spackler]

So applying this to Hawthorn’s list strategy: use your first round picks to trade for known quantities and use your middle picks to fill the list. Seems reasonable.

 

[Carl Spackler]

It’s also important to remember that the NFL draft runs seven rounds deep rather than the AFL’s five, so presumably the sweet spot for AFL (if it’s exists as it does in the NFL) is scrunched up near the top a little.

 

[Pessimistic]

It’s also important to remember that the NFL draft runs seven rounds deep rather than the AFL’s five, so presumably the sweet spot for AFL (if it’s exists as it does in the NFL) is scrunched up near the top a little.

And they can be much more certain what they are getting as the draftees are more advanced due to college systems

 

[Pessimistic]

Where a hawk team reaches a high point, (2001, 2008, 2015...2022?) the mix seems to be 1/3 already listed: 1/3 draftees: 1/3 mature recruits.

They seem to be doing this simultaneously.

The alternate, the dive and rebuild strategy, seperates these into first drafting through draft picks hopefully better because your ladder position has cycled low, and only then daring to 'top up'
Thisin itself is a relatively new strategy, but somehow seems to be holy script as a starting point particularly for thiose in here arguing agaisnt hawthorn

It isnt that well supported by evidence of success. as an example here we have St Kilda having done a rebuild, yet struggling for a ruckman wile a perfectly good one they offloaded to the hawks, who ironically is in such a purple patch he is keeping the hawks from going lower on the ladder right now.

proverbially, the hawks are running and sucking eggs at the same time

 

[Furn2]

I did a thread on a similar theory a while ago, the basis was that picks in there 15-55 range gave similar odds of getting a good or better player and therefore you are better to trade down to that 40-55 range and take multiple picks than 1 pick at 22 for example.

It was based on this data at draftguru.com.



Now of you look at a 200 gamer (obviously not all 200 gamers are the same but its a factual measure as a general guide) you can see 4 distinct bands.

picks 1-5 where u have a great chance at a player about 50%
picks 6-20 decent chance about 25%
Then we see 20-30 and 31-55 about the same at 13%
picks 55-70 it drops off a bit to 8%

After that its very low and you'd be lucky to hit much, although the rookie draft is still about 5%.

getting an AA player, the odds are similar except for the 20-30 range where theres a outlier and only about 6% become AAs but that can be ignored id say.

So what this all seems to say is if you have a top 5 pick use it.
If you have a pick 6-20 you want to try to trade it for A grade talent but if not use it. (looking deeper into i found that 15-20 isn't as good as 6-15 so maybe this range is a little smaller)
Picks 20-30 are overrated and should be traded for either a decent player or more picks in the 40-55 range.
while the odds of hitting players after 55 are low, it still happens and its worth doing as picks are so easy to get in that region.

of course the past doesn't = the future, but there's something there and when you consider the other downsides to bottoming out with youth such as player confidence, culture, fanbase and revenue, there's plenty of evidence that this ap[proach has merit. Like any method it needs to be done well to succeed.

The other thing to consider is that this is only the case while some teams still highly covert picks and picks in that 15-30 range. If everyone took the Hawks/Cats approach it becomes way harder to succeed.

 

[cryptor]

I did a thread on a similar theory a while ago, the basis was that picks in there 15-55 range gave similar odds of getting a good or better player and therefore you are better to trade down to that 40-55 range and take multiple picks than 1 pick at 22 for example.

It was based on this data at draftguru.com.

View attachment 714401

Now of you look at a 200 gamer (obviously not all 200 gamers are the same but its a factual measure as a general guide) you can see 4 distinct bands.

picks 1-5 where u have a great chance at a player about 50%
picks 6-20 decent chance about 25%
Then we see 20-30 and 31-55 about the same at 13%
picks 55-70 it drops off a bit to 8%

After that its very low and you'd be lucky to hit much, although the rookie draft is still about 5%.

getting an AA player, the odds are similar except for the 20-30 range where theres a outlier and only about 6% become AAs but that can be ignored id say.

So what this all seems to say is if you have a top 5 pick use it.
If you have a pick 6-20 you want to try to trade it for A grade talent but if not use it. (looking deeper into i found that 15-20 isn't as good as 6-15 so maybe this range is a little smaller)
Picks 20-30 are overrated and should be traded for either a decent player or more picks in the 40-55 range.
while the odds of hitting players after 55 are low, it still happens and its worth doing as picks are so easy to get in that region.

of course the past doesn't = the future, but there's something there and when you consider the other downsides to bottoming out with youth such as player confidence, culture, fanbase and revenue, there's plenty of evidence that this ap[proach has merit. Like any method it needs to be done well to succeed.

The other thing to consider is that this is only the case while some teams still highly covert picks and picks in that 15-30 range. If everyone took the Hawks/Cats approach it becomes way harder to succeed.

I've had this idea in the back of my mind for a while about draft pick values pretty much since they brought in draft pick points. I like the beauty of math popping up naturally in things ("neeeerrrd!!", yes I know) so I've wondered if some sort of number pyramid would align closely to the relative value of draft picks. Something like the below where the picks in each row of the pyramid are effectively going to give you around the same chance of getting a good player.

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

I've coloured them based on the round each pick is in which helps illustrate some things that I think we all know to be true or at least intuitively make some sense.

Things like how the difference in relative value between picks within the first round change the most significantly. So the difference between pick 2 and pick 13 is only 11 spots and on the pyramid it is 3 levels different. Compare it to something later on like pick 23. 11 picks later is pick 34, but it's only 1 level down. When you get down to the base of the pyramid that same 11 pick difference between picks is effectively nothing.

This isn't to say pick 7 equals pick 10. Pick 7 is still better than pick 10. This is to say that the difference between pick 7 and pick 10 might not be worth losing sleep over when it comes to a trade.

It might not be perfect, and within the first round it might not ring true in a given year based on who is available and what competition there is for certain players or types of players. But in general it seems to be a pretty decent way of appraising the value of particular picks relative to each other.

 

[Scotland]

+1 for the pyramid.

At a glance it looks like yellow and blue towards the top and purple on the bottom.

 

[Carl Spackler]

And they can be much more certain what they are getting as the draftees are more advanced due to college systems

That’s sounds true but they still get a lot of busts in the skill positions. And the point of the study was that they overestimated the value (and hence their ability to pick it) of the top recruits.

 

[SSSSSS]

So applying this to Hawthorn’s list strategy: use your first round picks to trade for known quantities and use your middle picks to fill the list. Seems reasonable.

Bird in hand > two in the bush.

 

[Ghost of Punt Road]

Hawks supporters are falling into a trap here. You are all arguing for the strategy your club has adopted.

That is barracking for an outcome, it is not statistics.

The strategy might be right or wrong, but what we have here is confirmation bias.

 

[Randy Campbell]

Hawks supporters are falling into a trap here. You are all arguing for the strategy your club has adopted.

That is barracking for an outcome, it is not statistics.

The strategy might be right or wrong, but what we have here is confirmation bias.

All we and you have is hope.

 

[Pessimistic]

Can they succeed? yes its possible
Will they? most would want to see it (or at least better signs) to say yes or no

 

[KermitJagger]

Hawks supporters are falling into a trap here. You are all arguing for the strategy your club has adopted.

That is barracking for an outcome, it is not statistics.

The strategy might be right or wrong, but what we have here is confirmation bias.

Barracking for an outcome and statistics aren't mutually exclusive. You can barrack for an outcome and use favourable statistics to do so.

That said, understanding the rationale of why a club is making certain decisions is not the same as blindly accepting it. There has been nothing recently that has made me reconsider the logic behind the strategy that Hawthorn has taken.

What is the alternative to understanding and accepting the strategy? Sit on this thread and piss and moan with the Essendon supporters about how terrible it is?

 

[Simon_Nesbit]

I've had this idea in the back of my mind for a while about draft pick values pretty much since they brought in draft pick points. I like the beauty of math popping up naturally in things ("neeeerrrd!!", yes I know) so I've wondered if some sort of number pyramid would align closely to the relative value of draft picks. Something like the below where the picks in each row of the pyramid are effectively going to give you around the same chance of getting a good player.

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

I've coloured them based on the round each pick is in which helps illustrate some things that I think we all know to be true or at least intuitively make some sense.

Things like how the difference in relative value between picks within the first round change the most significantly. So the difference between pick 2 and pick 13 is only 11 spots and on the pyramid it is 3 levels different. Compare it to something later on like pick 23. 11 picks later is pick 34, but it's only 1 level down. When you get down to the base of the pyramid that same 11 pick difference between picks is effectively nothing.

This isn't to say pick 7 equals pick 10. Pick 7 is still better than pick 10. This is to say that the difference between pick 7 and pick 10 might not be worth losing sleep over when it comes to a trade.

It might not be perfect, and within the first round it might not ring true in a given year based on who is available and what competition there is for certain players or types of players. But in general it seems to be a pretty decent way of appraising the value of particular picks relative to each other.

Love the effort, and agree in principle, but I think the "standard" is much flatter, probably only 3-4 levels. Perhaps a factor of 3 would be closer.

1 (1)
2,3,4 (3)
5,6,7,8,9,10,11,12,13 (9)
14-41 (27)
42+ (the rest).​

 

[Carl Spackler]

Hawks supporters are falling into a trap here. You are all arguing for the strategy your club has adopted.

That is barracking for an outcome, it is not statistics.

The strategy might be right or wrong, but what we have here is confirmation bias.

Except for the rigorous models presented. Otherwise yes.

 

[cryptor]

Love the effort, and agree in principle, but I think the "standard" is much flatter, probably only 3-4 levels. Perhaps a factor of 3 would be closer.

1 (1)
2,3,4 (3)
5,6,7,8,9,10,11,12,13 (9)
14-41 (27)
42+ (the rest).​

Potentially. Might be somewhere between. It's hard to measure these things as there's so many variables and so much of it is subjective. Golden Ratio version would probably look pretty solid too.

 

[Scotland]

People tend to look at draft position vs games played which is interesting but misleading. Chris Masten was pick 3 and has played 200 games, Luke Shuey was pick 18 a year later and about to catch him and a far superior player. We've had 22 players play 200 and most from the draft era weren't high picks. Every club wants players who will have long careers but you don't win flags with 22 GOPs.

2018 AA side

B: Stewart (40) Rance (18) Laird (R)
HB Hurn (13) McGovern (R) Whitfield (1)
C: Gaff (4) Martin (3) Sidebottom (11)
HF: Dangerfield (10) Franklin (5) Gray (55)
F: Gunston (29) Riewoldt (13) Breust (R)
Foll: Gawn (34) Cripps (13) Mitchell (F/S 21)

IC: Grundy (18) Oliver (4) Higgins (11) Edwards (LOL 26)

2017 others

Hibberd (PSD)
Hurley (5)
Docherty (12)
Kelly (2)
Merrett (26)
Zorko (zone)
Daniher (F/S 10)
Kennedy (4)
Betts (PSD)
Ryder (7)
Crouch (23)
Yeo (30)
Selwood (7)
Shiel (GWS 17 year old)

There's a strong correlation between the pointy end of the draft and top quality midfielders. Drops off a bit in other areas but realistically about 3/4 of the players in the last two AA sides were drafted in the top 25-30 and more than half in the first round. The key forward posts have been dominated by first round picks for years. Best chance of getting a top quality player outside the first round is small forward, defender or ruckman.

Brownlow Medal (skewed towards midfielders) is another gauge. 1, 2, 3, 5, 7, 10 last year were all former first round picks.

All of which is entirely separate to drafting vs trading. Just one way of looking at where top players start out.

 

[HeHateMe]

Games played isnt perfect for best value.

But if you get a gun at pick 1 who never gets on the park plays 50 games over 7 years and retires. That picks a dud.

Or you get the 10th best player of that drat at pick 1 and they play 200 games.

Your better off with the durable player with less ability.

Its a Lottery so the whole thing is based on calculated bets/risks.

 

[btdg]

People tend to look at draft position vs games played which is interesting but misleading. Chris Masten was pick 3 and has played 200 games, Luke Shuey was pick 18 a year later and about to catch him and a far superior player. We've had 22 players play 200 and most from the draft era weren't high picks. Every club wants players who will have long careers but you don't win flags with 22 GOPs.

2018 AA side

B: Stewart (40) Rance (18) Laird (R)
HB Hurn (13) McGovern (R) Whitfield (1)
C: Gaff (4) Martin (3) Sidebottom (11)
HF: Dangerfield (10) Franklin (5) Gray (55)
F: Gunston (29) Riewoldt (13) Breust (R)
Foll: Gawn (34) Cripps (13) Mitchell (F/S 21)

IC: Grundy (18) Oliver (4) Higgins (11) Edwards (LOL 26)

2017 others

Hibberd (PSD)
Hurley (5)
Docherty (12)
Kelly (2)
Merrett (26)
Zorko (zone)
Daniher (F/S 10)
Kennedy (4)
Betts (PSD)
Ryder (7)
Crouch (23)
Yeo (30)
Selwood (7)
Shiel (GWS 17 year old)

There's a strong correlation between the pointy end of the draft and top quality midfielders. Drops off a bit in other areas but realistically about 3/4 of the players in the last two AA sides were drafted in the top 25-30 and more than half in the first round. The key forward posts have been dominated by first round picks for years. Best chance of getting a top quality player outside the first round is small forward, defender or ruckman.

Brownlow Medal (skewed towards midfielders) is another gauge. 1, 2, 3, 5, 7, 10 last year were all former first round picks.

All of which is entirely separate to drafting vs trading. Just one way of looking at where top players start out.

I think rather than draft position look at Tiers.
Tier 1 draftees = almost certain good AFL players (eg: Walsh). Tier 2 draftees = potential stars, a few risks/doubts (eg Charlie Curnow), but probably a good player. Tier 3 draftees = we like some things about him, but we're really just hoping.

Midfielders translate very well from juniors to seniors; so the best of them end up Tier 1 or Tier 2 and there is less risk. You can still find them later, but often if they are good they slot into the top tiers. KP players are a bit more risky and harder to project; Tier 1 KP players (ie; a big guy whose attributes make them an almost certain good player - like Hawkins, or even Weitering) are worth their weight in gold for that reason, and often even the best KP prospects are Tier 2 (risky). Rucks, small forwards, taggers/nuggety defenders - these are the hardest to predict, as they don't translate at all well from juniors. Lots of these available in Tier 2

Draft to draft varies in depth in each Tier. Last year I'd say there were 5-6 Tier 1 players, and as many as 15-20 Tier 2. This year I think there's 2-3 Tier 1 players and maybe only 7-10 Tier 2 players.

And within the Tiers, it is better to pick earlier, but mainly cos you get the best 'fit' for your team. EG: pick 1 almost certainly lets you pick a Tier 1 player (the 2014 draft is the only one I can think of with zero Tier 1 guys), AND you get the one you think fits. If you have the first Tier 3 pick, you get to take a punt on the guy you like the most. They aren't necessarily a prospect better than pick 70 overall, but you get your guy (and can back your recruiters). And once you get to about pick 30-50 it tends to be very flat and you can just guess at 'your guy' anyway, or hope some Tier 2 guys slide.

Hawthorn's approach is really interesting and forward thinking for a few reasons:
1. They've been good enough to not have any Tier 1 picks;
2. They've consciously chosen to trade away Tier 2 picks for established players. They've targetted players who are entering their prime (age 23-24) and who have shown they can play at AFL level. Essentially, they are eliminating the risk in Tier 2 (in some cases giving up 2x Tier 2 picks to do so).
3. In doing so, they have kept the team very competitive (essentially emphasising point 1)
4. They have recognised that basically all Tier 3 picks are speculative, and haven't valued them highly, focusing instead on identifying 'their guys' (Lewis) and Tier 2 players who slide (Worpel). They've been very happy to trade early Tier 3 picks away and use late ones

This contrasts with Carlton's approach rebuilding over a similar time period:
1. We've had several Tier 1 picks by virtue of low ladder position (Walsh, Weitering, and arguably SPS and Dow).
2. We've traded actively to get Tier 2 picks (Curnow, McKay, Cuningham, O'Brien, Stocker - all Tier 2 picks we traded to get).
3. We've traded to get even more young Tier 2 picks who haven't yet entered their prime or really proven themselves at AFL level (Marchbank, Pickett, Kennedy, Setterfield, Plowman)
4. We've basically ignored Tier 3 outside the very top end (Fisher and De Koning drafted in the 20s) other than a couple of F/S picks. We've also basically ignored the rookie list or used that as a way of finding stop-gap mature players.

I think 5 years ago Carlton's approach was the conventional wisdom - so Hawthorn are challenging it. I think Hawthorn's approach is tougher, but better if it works. To do it, you have to 'win' the battle to get free agents. You have to get 'your guys' right and also have a bit of luck with Tier 2 guys sliding.

IF Hawthorn get Coniglio, I think they'll look like geniuses. If they don't... who knows. But, it shoud probably be said, not everyone could do what Hawthorn did - there was only 1x Mitchell, 1x O'Meara, 1x Wingard to trade for (just like there's only 1x Shiel, 1x Coniglio to fill that last midfield spot).

Still, will be an interesting off season.

 

[Scotland]

There's also luck in being good/bad at the right time. Hawthorn did well out of the 2001 and 2004 drafts as everyone knows. If you look at 2001-2005 when clubs were doubling up with first round priority picks yo-yo Melbourne had their down year in 2003 when the talent crop was weakest. Every draft with hindsight you can say 'Har har you took some spud at pick 7 and we got a gun at pick 43' but some draft classes are much stronger than others. 2005 had Shaun Higgins, Nathan Jones, Shannon Hurn, Grant Birchall, Travis Varcoe, Richard Douglas go 11-16. You can't be unhappy with any of those guys in that range. Other years that's the sort of player you get from 5-10 or earlier.