How effective is each draft pick from (2000-2009)

[shotgun_blues]

More interesting stats for the Top 10 key forwards and half forwards this year and their relative draft position.

KEY FORWARDS
Roughead #2
Cloke (F/S)
Cameron (GWS conc)
Franklin #5
Kennedy #3
J.Riewoldt #12
N.Riewoldt #1
Schultz #12
Hawkins (F/S)
Petrie #23

Ave. draft pick = #8

SMALL/HALF FORWARDS
Thomas #53
Walters #53
Motlop #39
Garlett (Rookie Draft)
Wingard #6
Bruest (NSW zone sel)
Monfries #14
Gunston #29
Darling #26
Mayne #40

Ave. draft pick = #40

Is there any point drafting a key forward at all from pick 30 onwards? Only 2 in the last 10 years have regularly kicked 40+ goals per season - Bradshaw (pick 56) and Fevola (pick 37).

Why are the good small forwards going around at an average of pick 40?

 

[Tommycash]

So this is just based on your subjective rating of players? Rather than actual games played? Surely the best barometer is actual games played.

 

[Thurlow2Mackie]

So this is just based on your subjective rating of players? Rather than actual games played? Surely the best barometer is actual games played.

Kosi played 200 games

 

[gandaal]

just compiling a list of Champions from each

2003 - #1 Cooney, #3 Sylvia, #13 Stanton, #19 Mundy, #33 Adcock, #41 Dawson, #48 H.Shaw(f/s), #53 Jackson, #55 S.Fisher, #58 B.Hudson, #61 Rischitelli, #73 S.Tuck (2 x T10, 2 11-20, 3 x 21-50, 5 x 50+)

I appreciate the effort involved in doing this post but seriously where's Andrew Walker?
If we did the draft again I'd still want Carlton to pick Walker.

 

[liz]

More interesting stats for the Top 10 key forwards and half forwards this year and their relative draft position.

KEY FORWARDS
Roughead #2
Cloke (F/S)
Cameron (GWS conc)
Franklin #5
Kennedy #3
J.Riewoldt #12
N.Riewoldt #1
Schultz #12
Hawkins (F/S)
Petrie #23

Ave. draft pick = #8

SMALL/HALF FORWARDS
Thomas #53
Walters #53
Motlop #39
Garlett (Rookie Draft)
Wingard #6
Bruest (NSW zone sel)
Monfries #14
Gunston #29
Darling #26
Mayne #40

Ave. draft pick = #40

Is there any point drafting a key forward at all from pick 30 onwards? Only 2 in the last 10 years have regularly kicked 40+ goals per season - Bradshaw (pick 56) and Fevola (pick 37).

Why are the good small forwards going around at an average of pick 40?

Much of the explanation comes in terms of sheer numbers. In most draft years, there are far more small / medium forward options available than decent tall forwards. The good tall forwards get taken early at least partly because there are so few of them. On the other hand, if you're after a small forward and there are a dozen or so available in the draft that you can't split at age 18, you're more likely to wait until later in the draft to pick one.

 

[CaneGarner]

It is equal in value if the goal is 100 games only. That doesn't factor in the quality of games played though.

Exactly right. My rating system takes into account quality through AFLCA votes and All Australian selection. Based on all picks 2000-2009, here is a breakdown of how likely you are to get a "star" (someone who scores over 8.0 / 10, all 48 of them are listed in my previous post) and how likely you are to get a "solid player" (someone who scores over 5.0 / 10, basically someone who is best 22 over a long period of time but doesn't do anything too spectacular). The results are surprising.

According to the analysis, an 11-20 range pick is worth more than a 6-10 pick! Surely that is a statistical anomaly though, there are 50 data points for 6-10 (5 picks over 10 drafts). Picks 31-40 are also somewhat skewed by all the father son picks before the bidding system came in.

This again suggests that 1-5 picks are worth their weight in gold, with a much higher probability of picking both a star and a long term solid performer than later picks. This also certainly shows that a pick in range 21-30 is more valuable than a pick past 40, as there is a sixfold greater chance of picking up a star, and almost twice as much chance of getting a solid performer.


Range 1-5: 36% likely to get a star, 78% likely to get a solid performer.
Range 6-10: 8% likely to get a star, 34% likely to get a solid performer.
Range 11-20: 12% likely to get a star, 39% likely to get a solid performer.
Range 21-30: 6% likely to get a star, 26% likely to get a solid performer.
Range 31-40: 7% likely to get a star, 26% likely to get a solid performer.
Range 41-50: 1% likely to get a star, 19% likely to get a solid performer.
Range 51-60: 1% likely to get a star, 15% likely to get a solid performer.

 

[Tommycash]

Kosi played 200 games

And I dare say the coaches who picked him are better judges than you or I. St Kilda were a very successful team when Kossie was playing.

 

[Davo-27]

I appreciate the effort involved in doing this post but seriously where's Andrew Walker?
If we did the draft again I'd still want Carlton to pick Walker.

I might have missed a handful of players in the list

 

[Davo-27]

Range 1-5: 36% likely to get a star, 78% likely to get a solid performer.
Range 6-10: 8% likely to get a star, 34% likely to get a solid performer.
Range 11-20: 12% likely to get a star, 39% likely to get a solid performer.
Range 21-30: 6% likely to get a star, 26% likely to get a solid performer.
Range 31-40: 7% likely to get a star, 26% likely to get a solid performer.
Range 41-50: 1% likely to get a star, 19% likely to get a solid performer.
Range 51-60: 1% likely to get a star, 15% likely to get a solid performer.

thanx for doing the extra work here, saves me some time

 

[Boris the Contradiction]

The lower picks are the difference between the also rans and the contenders/champions. Unless you are a complete basket case of a club a littering of first rounders will get you into the eight but the lower picks/rookies and the ability to develop them into contributing members of a team are the difference between the Carltons and North Melbournes and the Sydneys, Geelongs and Collingwoods.

 

[gandaal]

I might have missed a handful of players in the list

You did a great job overall. I'm not sure how you can miss pick 2 though

 

[gandaal]

The lower picks are the difference between the also rans and the contenders/champions. Unless you are a complete basket case of a club a littering of first rounders will get you into the eight but the lower picks/rookies and the ability to develop them into contributing members of a team are the difference between the Carltons and North Melbournes and the Sydneys, Geelongs and Collingwoods.

I don't know about north but Carlton have done reasonably well out of late picks. It's the 2nd rounders that haven't kicked on for us.

Carrazzo, Kade Simpson, Eddie Betts, Dennis Armfield, Curnow (3rd in BOF this year), Jeff Gartlett, Michael Jamison, Mitch Robinson.

All 40+ or rookie selections. Not a terrible effort.

 

[Boris the Contradiction]

I don't know about north but Carlton have done reasonably well out of late picks. It's the 2nd rounders that haven't kicked on for us.

Carrazzo, Kade Simpson, Eddie Betts, Dennis Armfield, Curnow (3rd in BOF this year), Jeff Gartlett, Michael Jamison, Mitch Robinson.

All 40+ or rookie selections. Not a terrible effort.

They have done reasonably well, but not to the same extent. Collingwood used seven rookies to good effect in the premiership year. 6 played in the grand final and that doesn't even mention the low draft picks, of which there were a few. It's the factor that makes the biggest difference.

 

[Dr Tigris]

Exactly right. My rating system takes into account quality through AFLCA votes and All Australian selection. Based on all picks 2000-2009, here is a breakdown of how likely you are to get a "star" (someone who scores over 8.0 / 10, all 48 of them are listed in my previous post) and how likely you are to get a "solid player" (someone who scores over 5.0 / 10, basically someone who is best 22 over a long period of time but doesn't do anything too spectacular). The results are surprising.

According to the analysis, an 11-20 range pick is worth more than a 6-10 pick! Surely that is a statistical anomaly though, there are 50 data points for 6-10 (5 picks over 10 drafts). Picks 31-40 are also somewhat skewed by all the father son picks before the bidding system came in.

This again suggests that 1-5 picks are worth their weight in gold, with a much higher probability of picking both a star and a long term solid performer than later picks. This also certainly shows that a pick in range 21-30 is more valuable than a pick past 40, as there is a sixfold greater chance of picking up a star, and almost twice as much chance of getting a solid performer.


Range 1-5: 36% likely to get a star, 78% likely to get a solid performer.
Range 6-10: 8% likely to get a star, 34% likely to get a solid performer.
Range 11-20: 12% likely to get a star, 39% likely to get a solid performer.
Range 21-30: 6% likely to get a star, 26% likely to get a solid performer.
Range 31-40: 7% likely to get a star, 26% likely to get a solid performer.
Range 41-50: 1% likely to get a star, 19% likely to get a solid performer.
Range 51-60: 1% likely to get a star, 15% likely to get a solid performer.

Seems to indicate that top 5 picks are worth their weight in gold. After that first rounders are very good. Second rounders good value. And after that it slides to be a lottery. You can see why clubs (allegedly) tank. And trading into the quality end is worthwhile in terms of increasing the probabilities. But conversely, you can get talent anywhere.

The other thing is that there is no certainly. About 1 in 5 1-5 picks = nothing. And just over 1/3 = star. So each year's top 5 should roughly include 2 stars, 2 solid players and one bust. Getting good players out of your first rounders is really important because that is where the stars are (in general). Often people focus on the later rounds as the test of recruitment. But to me it looks like good early drafting provides the core of the side if done right. OK later round drafting and trading can fill in the holes.

After the top 5 you're about 10% likely to get a star in the first round. So 1.5 stars per year. About 5 solid performers and 8.5 busts. That is, more than half of all first rounders outside the top 5 are busts. Some clubs will just have a lot of 'lost' first rounders, whilst others will get hits. That means that it is very difficult to get a good squad out of late round players without good drafting in early rounds. Interesting in seeing the major value in 1-5 picks and the relative high chance of busts in the first round thereafter.

 

[bombermick]

Games played is a little misleading now as a lot of clubs are now taking limited mature aged recruits here. Someone like Kommer may be able to squeeze out 100 games, although he'll never be more than a foot soldier.

 

[CaneGarner]

They have done reasonably well, but not to the same extent. Collingwood used seven rookies to good effect in the premiership year. 6 played in the grand final and that doesn't even mention the low draft picks, of which there were a few. It's the factor that makes the biggest difference.

Here's a quick analysis that tests that theory.

A "hit" means drafting of a player that scores a 5.0 or more (solid long term player). These are then broken down into four categories - National Draft top 40 (ND 1-40), National Draft 40+ (ND 40+), Rookie Draft (RD) and Pre-Season Draft (PSD). Shown is the number of hits between 2000 and 2009. The bracketed percentage refers to the conversion rate - how may of those players went on to become "stars" (score more than 8.0).

Teams are ordered from top to bottom according to how many premiership points they racked up between 2010 and 2013.

I think this shows that most teams have about the same output from late National Draft, Rookie and Pre-Season picks. Even the total number of hits doesn't seem to matter. The thing that strikes me as standing out in the best teams is their conversion rate of top 40 picks into stars (i.e. Collingwood, Geelong, Hawthorn). I'm guessing this has more to do with development programs than recruiters, although surely some of both.

An interesting side note is that Sydney have the lowest number of hits out of any team, and also a poor conversion rate of top 40 picks. Obviously they have a different strategy, letting other teams do the drafting and development and then trading in mature players, to great effect. So while focussing on development of top 40 picks clearly works, it's not the only strategy that can work.

Geelong 21 hits ND 1-40: 11 (55%), ND 40+: 4 (25%), RD: 5 (0%), PSD: 1 (0%)
Collingwood 21 hits ND 1-40: 9 (78%), ND 40+: 5 (20%), RD: 7 (14%), PSD: 0 (0%)
Hawthorn 24 hits ND 1-40: 13 (54%), ND 40+: 3 (0%), RD: 7 (0%), PSD: 1 (0%)
Sydney 16 hits ND 1-40: 7 (29%), ND 40+: 4 (0%), RD: 4 (25%), PSD: 1 (0%)
Fremantle 21 hits ND 1-40: 7 (0%), ND 40+: 4 (0%), RD: 7 (14%), PSD: 3 (0%)
Carlton 21 hits ND 1-40: 8 (13%), ND 40+: 6 (0%), RD: 6 (0%), PSD: 1 (0%)
North Melbourne 20 hits ND 1-40: 11 (18%), ND 40+: 4 (0%), RD: 5 (0%), PSD: 0 (0%)
St Kilda 16 hits ND 1-40: 9 (56%), ND 40+: 4 (0%), RD: 3 (0%), PSD: 0 (0%)
West Coast 16 hits ND 1-40: 13 (23%), ND 40+: 2 (0%), RD: 1 (0%), PSD: 0 (0%)
Essendon 23 hits ND 1-40: 12 (17%), ND 40+: 5 (0%), RD: 5 (0%), PSD: 1 (0%)
Adelaide 23 hits ND 1-40: 10 (10%), ND 40+: 7 (0%), RD: 5 (20%), PSD: 1 (0%)
Richmond 23 hits ND 1-40: 12 (25%), ND 40+: 4 (0%), RD: 5 (0%), PSD: 2 (0%)
Western Bulldogs 17 hits ND 1-40: 6 (50%), ND 40+: 5 (20%), RD: 6 (17%), PSD: 0 (0%)
Brisbane 18 hits ND 1-40: 10 (0%), ND 40+: 6 (0%), RD: 1 (0%), PSD: 1 (0%)
Port Adelaide 19 hits ND 1-40: 9 (44%), ND 40+: 6 (0%), RD: 3 (0%), PSD: 1 (0%)
Melbourne 21 hits ND 1-40: 12 (17%), ND 40+: 5 (0%), RD: 4 (25%), PSD: 0 (0%)

 

[Tommycash]

Games played is a little misleading now as a lot of clubs are now taking limited mature aged recruits here. Someone like Kommer may be able to squeeze out 100 games, although he'll never be more than a foot soldier.

A lot less misleading then a subjective assessment of player worth. I note your point though. Perhaps a premium can be set on older players so each game is worth more? The whole point of the draft is finding players who will give your club a long tenure. May as well make the performance measure equal the desired firm outcome.

 

[Nomad12]

history only tells us what has happened in the past not what is about to happen in the future... don't know where I heard that but it seemed right lol

you never know whats going to happen in the draft ..work out all the stats you want but in my opinion its pointless figuring out a % that your club will get a star or 100+ gamer with certain round draft pick.. theres talent through the whole draft its proven almost every year when a first or second year player comes out and you wonder how did this guy get picked so late??..trouble is picking which of all these talented kids are going to work hard enough and improve enough to be a top player, and touch wood they wont be injury plagued.

 

[Damon_3388]

A lot less misleading then a subjective assessment of player worth. I note your point though. Perhaps a premium can be set on older players so each game is worth more? The whole point of the draft is finding players who will give your club a long tenure. May as well make the performance measure equal the desired firm outcome.

I suppose then you have to determine the value of "getting a game", and whether someone repeatedly "getting a game" is evidence of them doing their job, or whether they're simply being "gifted" games/experience.

 

[gandaal]

They have done reasonably well, but not to the same extent. Collingwood used seven rookies to good effect in the premiership year. 6 played in the grand final and that doesn't even mention the low draft picks, of which there were a few. It's the factor that makes the biggest difference.

Just because your're a premiership player that doesn't make you instantly a better player. I would doubt your six are better players as a group than that eight I mentioned.

 

[Tommycash]

I suppose then you have to determine the value of "getting a game", and whether someone repeatedly "getting a game" is evidence of them doing their job, or whether they're simply being "gifted" games/experience.

Not so, I'm a firm believer that if you're playing you deserve it (In one way or another). A club could gift a player 300 games easily, but to do so to a spud would be suicide for the club. The sheer fact I'm having this discussion with so many is mind-boggling. Of course games played is going to be the most accurate measure of a players draft worth. Simply of course. Over time you will have no other measure more accurate. As soon as you take it into qualitative terms, you suddenly lose any sense of legitimacy. Give me a coaches opinion (The one who gives games) over a bigfooty warrior's any day of the week.

 

[bombermick]

A lot less misleading then a subjective assessment of player worth. I note your point though. Perhaps a premium can be set on older players so each game is worth more? The whole point of the draft is finding players who will give your club a long tenure. May as well make the performance measure equal the desired firm outcome.

In the main I agree with you, as games played is the most objective criterion. However, if we're looking at where we are most likely to find stars, the games played by mature recruits can distort the figures. They will in all likelihood play more games than a project player, but are no more likely to become a star.

 

[Boris the Contradiction]

Just because your're a premiership player that doesn't make you instantly a better player. I would doubt your six are better players as a group than that eight I mentioned.

I didn't say they were. The seven names I mentioned however were all off the rookie list. I didn't even talk about players taken later in drafts. Your list was a combination. It is quite clear Collingwood has done better than Carlton in recent years with later picks and rookies. In fact, our success with rookies alone is unique and the sheer number in our premiership team is probably unprecedented.

As others in here have mentioned, it probably has less to do with drafting than it does development.

 

[Top_4_Tiger]

Massive difference between drafting in 2000 and in 2009.

The difference is so massive you really can't pile all those data together for any reason.

 

[Tommycash]

In the main I agree with you, as games played is the most objective criterion. However, if we're looking at where we are most likely to find stars, the games played by mature recruits can distort the figures. They will in all likelihood play more games than a project player, but are no more likely to become a star.

Yes indeed. That's a fair call. How do you define star though? I think we should just forget about putting in a concept like that. If a pick reaches x games then it's a good pick. If not xxxx. Perhaps you would need an entirely different rating system to work out which picks will give you a "star". Again that's just an annoying concept to decide upon. There's about 70-80 stars in the AFL at any given time. They drop in and out of that status as their form commands. If you go with AA's then you'll have some unlucky fellows and some very lucky fellows. I tend to dislike such assesments.

You are right though. More chance of a mature ager reaching x games then a project player, but less chance of them becoming a star.