With the 2015 AFL season finished and the trade period upon us, I thought it would be an interesting exercise to take a close look at the state of each club's list - using a different approach!
Recently, the AFL introduced a Draft Value Index (DVI) which allocated points to draft positions. Fundamentally, it was created to facilitate the new bidding system for the various academies.
As the AFL explanation sheet demonstrates (as an example of its potential other uses), it can also be used to evaluate trades involving draft picks. Comparisons seem fine for unrealised picks, however trading of just picks is relatively rare - we invariably seek ways to quantify an actual player's worth. Naturally, we can use the DVI to rate a player based on the draft position they were taken - and on average it may well be an accurate reflection. Unfortunately, in reality, many factors would affect a player's value over time, so maybe these numbers should represent potential or just opportunity. Regardless of interpretation, applying the DVI to team lists has the advantage of overcoming the usual categorical nature of list analyses.
Taking each player from an AFL list and assigning a value based on their draft position leads to the following totals.
Team
|
Points
|
Players
|
Points/Player
\GWS|46377|47|986.7\Carlton|33603|46|730.5\West Coast|32280|46|701.7\Gold Coast|30140|46|655.2\Richmond|29885|45|664.1\Melbourne|29708|44|675.2\Essendon|28963|45|643.6\Port Adelaide|28362|46|616.6\Geelong|28010|46|608.9\North Melbourne|27345|45|607.7\Hawthorn|27196|46|591.2\St Kilda|24960|45|554.7\Western Bulldogs|24568|44|558.4\Fremantle|22411|46|487.2\Collingwood|22099|45|491.1\Brisbane|19290|48|401.9\Adelaide|18783|45|417.4\Sydney|18449|44|419.3\
AVERAGE
|
27357.2
|
45.5
|
600.6
As expected, GWS are on top with Sydney on bottom. Carlton's list has a high number of quality draft picks, while Brisbane's total value is quite low - an indication of the recent talent exodus maybe. Draft penalties may be affecting Adelaide's value. Note that not much can really be read into this data unless you compare over time or know the age breakdowns of each list for example.
To get a better picture, we can divide the total points into age brackets. Arbitrarily, under 23, 23 to 27 inclusively, and over 27 years of age are chosen (for simplicity, age at start of 2015 is used). Firstly, the under 23 cohort:
Team
|
Points-U23
|
Players-U23
|
Pts23/Pl23
|
Pts23/Pts-%
|
Pl23/Pl-%
\GWS|35957|36|998.8|77.5|76.6\Gold Coast|28597|32|893.7|94.9|69.6\Western
-
Bulldogs|17029|27|630.7|69.3|61.4\Melbourne|16821|21|801.0|56.6|47.7\St Kilda|15688|26|603.4|62.9|57.8\Collingwood|13229|25|529.2|59.9|55.6\West Coast|11521|23|500.9|35.7|50.0\Richmond|10875|19|572.4|36.4|42.2\Geelong|10371|24|432.1|37.0|52.2\Brisbane|10107|26|388.7|52.4|54.2\North Melbourne|9541|17|561.2|34.9|37.8\Port Adelaide|9423|22|428.3|33.2|47.8\Essendon|9214|19|484.9|31.8|42.2\Carlton|9094|21|433.0|27.1|45.7\Adelaide|8529|21|406.1|45.4|46.7\Fremantle|6876|19|361.9|30.7|41.3\Hawthorn|6451|20|322.6|23.7|43.5\Sydney|6403|23|278.4|34.7|52.3\
AVERAGE
|
13095.9
|
23.4
|
534.9
|
46.9
|
51.4
To clarify, Pts represents points and Pl represents number of players. They represent the whole list unless coupled with a specification. For example, in the last column, Pl23 represents the number of players on the list under 23 years old while Pl represents the number of players on the whole list (76.6% of GWS's list is under 23, say).
Again, the order is as anticipated - GWS and Gold Coast have had massive draft concessions, while Sydney and Hawthorn are victims of their own successes.
To make clear my reasoning, I'll go through the columns step by step. The first few columns are self explanatory, and it's normal to compare to the AFL average (for example, North Melbourne have 17 players under 23 years of age, well below the AFL average of 23.4).
Column D: GWS have the most value per under 23 player (998.8 points per young player) while Sydney have the least value per under 23 player (278.4 points). Sydney's long time poor draft positioning makes that understandable.
Column E: Gold Coast's proportion of their list value from their under 23's is the highest (94.5% of their list value comes from their youth) while Hawthorn's is the lowest (23.7% of their list value comes from their youth). Of course, you have to be careful with these numbers, as will be demonstrated.
Column F: GWS has the highest proportion of their list as under 23 (76.6% of their list size is under 23) with North Melbourne the lowest (37.8%). I guess it shows North Melbourne have been going all out for a flag these last few years (they may eventually fall off the cliff for a while if unprepared).
Columns E and F: While the proportion of value the Gold Coast get from their under 23's is the highest (94.5% of the list value) that can be attributed to the fact most of their list is under 23 (69.6%). Notice the discrepancy between these two percentages. If the talent value was spread evenly across their list, these numbers should be equal. As it stands, it suggests that on average the quality of their under 23's is better than the quality of their older players (their mature recruits weren't great maybe).
Regarding Hawthorn, the proportion of value from their under 23's is 23.7%, but this time, it is not because of the cohort size (43.5% of the list is under 23). What we can say is that on average the quality of their youth is not as good as the quality of their older players. This doesn't necessarily say their youth is poor - their old players may be great. However, looking at the average value per under 23 player (332.6 points, well below the AFL average of 534.9), it definitely is a factor. Worryingly, Carlton exhibits a similar trend to Hawthorn.
Some other thoughts: GWS have a better spread of talent than Gold Coast. Bulldogs, St. Kilda, Collingwood and Brisbane have all invested in youth recently (the proportion of their lists under 23 is above the AFL average of 51.4%).
Brisbane worryingly have not done as well in this area as those other three, the average value is 388.7 per player (AFL average 534.9 points). Again, this shows the recent talent drain (note, in constructing these values, academy and father son picks and other concessions may not have their value truly represented, however, we'll assume they average out, or at least have minimal impact).
Melbourne haven't exactly flooded their list with youth (I guess Paul Roos has been trying to rebalance the list after Neeld's bloodletting) but they sure are quality (well, potentially, these numbers don't reveal poor selections or inadequate development).
Loosely, this table can be read as a gauge of potential, the higher on the table, the better the potential for the future (ideally should be the reverse of the current ladder).
Secondly, the middle age cohort (23 to 27 years inclusive).
Team
|
Points-Mid
|
Players-Mid
|
PtsMid/PlMid
|
PtsMid/Pts-%
|
PlMid/Pl-%
\Carlton|19080|19|1004.2|56.8|41.3\West Coast|17382|18|965.7|53.8|39.1\Richmond|17068|21|812.8|57.1|46.7\Port Adelaide|16443|20|822.2|58.0|43.5\Hawthorn|12658|19|666.2|46.5|41.3\North Melbourne|12130|19|638.4|44.4|42.2\Melbourne|12109|19|637.3|40.8|43.2\Essendon|10061|18|558.9|34.7|40.0\Fremantle|8717|15|581.1|38.9|32.6\Collingwood|8700|17|511.8|39.4|37.8\Geelong|8145|12|678.8|29.1|26.1\Sydney|8102|14|578.7|43.9|31.8\GWS|7690|7|1098.6|16.6|14.9\Brisbane|7508|17|441.6|38.9|35.4\Adelaide|6964|19|366.5|37.1|42.2\Western
-
Bulldogs|5415|11|492.3|22.0|25.0\St Kilda|2297|12|191.4|9.2|26.7\Gold Coast|878|9|97.6|2.9|19.6\
AVERAGE
|
10074.8
|
15.9
|
619.1
|
37.2
|
35.0
This cohort could represent the players in their prime and Carlton stand out as having one of the highest value per middle aged player (1004.2 points). Where did they go wrong? Recruitment? Development? Also, no surprise that West Coast are coming into their prime with Richmond also pushing (they certainly have recruited heavily in this age bracket recently, 46.7% of their list is between 23 and 27 years old - AFL average 35%).
It also shows Geelong and St Kilda have a gap in this age group, due to being up for so long from 2007-2010 for example. 29.1% and 9.2% are the proportion of their lists value from this age group respectively (AFL average 37.2%). The proportion of players in this age bracket for these two teams is similar (26.1 and 26.7%) so it indicates St. Kilda have done particularly poorly in this area (Ross Lyon's legacy?).
Again, this table shows GWS probably had a more balanced start-up strategy than Gold Coast. The proportion of Port Adelaide's list value in this age group is 58% - the highest in the AFL. It suggests Port should be pushing hard now (definitely underachieved this year).
Interestingly, the Bulldogs are relatively quite poor in this area (only a quarter of their list is in this age bracket with only 22% of their list value). They are certainly well ahead of time! (I wonder how they compare to the Baby Bombers say, or are they like Geelong in 2004....)
Finally, the over 27 age group:
Team
|
Points-Vet
|
Players-Vet
|
PtsVet/PlVet
|
PtsVet/Pts-%
|
PlVet/Pl-%
\Essendon|9688|8|1211.0|33.4|17.8\Geelong|9494|10|949.4|33.9|21.7\Hawthorn|8087|7|1155.3|29.7|15.2\St Kilda|6975|7|996.4|27.9|15.6\Fremantle|6818|12|568.2|30.4|26.1\North Melbourne|5674|9|630.4|20.7|20.0\Carlton|5429|6|904.8|16.2|13.0\Sydney|3944|7|563.4|21.4|15.9\West Coast|3377|5|675.4|10.5|10.9\Adelaide|3290|5|658.0|17.5|11.1\GWS|2730|4|682.5|5.9|8.5\Port Adelaide|2496|4|624.0|8.8|8.7\Western
-
Bulldogs|2124|6|354.0|8.6|13.6\Richmond|1942|5|388.4|6.5|11.1\Brisbane|1675|5|335.0|8.7|10.4\Melbourne|778|4|194.5|2.6|9.1\Gold Coast|665|5|133.0|2.2|10.9\Collingwood|170|3|56.7|0.8|6.7\
AVERAGE
|
4186.4
|
6.2
|
615.6
|
15.9
|
13.7
As expected, Geelong and Hawthorn are up the top - in fact, Hawthorn have done really well quality wise, which I guess is shown by the good form of their older players. Collingwood have clearly gotten rid of a lot of dead wood.
Fremantle's list has the highest percentage over 27 (26.1% compared to the AFL average of 13.7%). Interestingly, the average points per player for Fremantle is below average (568.2 versus 615.6) which supports the Ross Lyon stereotype of role players and not rejuvenating the list (might be an idea to look at St. Kilda's break down under Lyon).
Essendon seem to be coming to the end of a cycle (without a great peak, thanks to some poor decisions). A third of the value of their list comes from the older players (who make up only 17.8% of the list size). Their recent drafts have not been as strong.
There is undoubtedly lots more that can be gleaned (and disputed!), but it in general, these figures conform to expectations.
REFERENCES:
AFL Bidding System Explanation sheet
Recently, the AFL introduced a Draft Value Index (DVI) which allocated points to draft positions. Fundamentally, it was created to facilitate the new bidding system for the various academies.
As the AFL explanation sheet demonstrates (as an example of its potential other uses), it can also be used to evaluate trades involving draft picks. Comparisons seem fine for unrealised picks, however trading of just picks is relatively rare - we invariably seek ways to quantify an actual player's worth. Naturally, we can use the DVI to rate a player based on the draft position they were taken - and on average it may well be an accurate reflection. Unfortunately, in reality, many factors would affect a player's value over time, so maybe these numbers should represent potential or just opportunity. Regardless of interpretation, applying the DVI to team lists has the advantage of overcoming the usual categorical nature of list analyses.
Taking each player from an AFL list and assigning a value based on their draft position leads to the following totals.
As expected, GWS are on top with Sydney on bottom. Carlton's list has a high number of quality draft picks, while Brisbane's total value is quite low - an indication of the recent talent exodus maybe. Draft penalties may be affecting Adelaide's value. Note that not much can really be read into this data unless you compare over time or know the age breakdowns of each list for example.
To get a better picture, we can divide the total points into age brackets. Arbitrarily, under 23, 23 to 27 inclusively, and over 27 years of age are chosen (for simplicity, age at start of 2015 is used). Firstly, the under 23 cohort:
To clarify, Pts represents points and Pl represents number of players. They represent the whole list unless coupled with a specification. For example, in the last column, Pl23 represents the number of players on the list under 23 years old while Pl represents the number of players on the whole list (76.6% of GWS's list is under 23, say).
Again, the order is as anticipated - GWS and Gold Coast have had massive draft concessions, while Sydney and Hawthorn are victims of their own successes.
To make clear my reasoning, I'll go through the columns step by step. The first few columns are self explanatory, and it's normal to compare to the AFL average (for example, North Melbourne have 17 players under 23 years of age, well below the AFL average of 23.4).
Column D: GWS have the most value per under 23 player (998.8 points per young player) while Sydney have the least value per under 23 player (278.4 points). Sydney's long time poor draft positioning makes that understandable.
Column E: Gold Coast's proportion of their list value from their under 23's is the highest (94.5% of their list value comes from their youth) while Hawthorn's is the lowest (23.7% of their list value comes from their youth). Of course, you have to be careful with these numbers, as will be demonstrated.
Column F: GWS has the highest proportion of their list as under 23 (76.6% of their list size is under 23) with North Melbourne the lowest (37.8%). I guess it shows North Melbourne have been going all out for a flag these last few years (they may eventually fall off the cliff for a while if unprepared).
Columns E and F: While the proportion of value the Gold Coast get from their under 23's is the highest (94.5% of the list value) that can be attributed to the fact most of their list is under 23 (69.6%). Notice the discrepancy between these two percentages. If the talent value was spread evenly across their list, these numbers should be equal. As it stands, it suggests that on average the quality of their under 23's is better than the quality of their older players (their mature recruits weren't great maybe).
Regarding Hawthorn, the proportion of value from their under 23's is 23.7%, but this time, it is not because of the cohort size (43.5% of the list is under 23). What we can say is that on average the quality of their youth is not as good as the quality of their older players. This doesn't necessarily say their youth is poor - their old players may be great. However, looking at the average value per under 23 player (332.6 points, well below the AFL average of 534.9), it definitely is a factor. Worryingly, Carlton exhibits a similar trend to Hawthorn.
Some other thoughts: GWS have a better spread of talent than Gold Coast. Bulldogs, St. Kilda, Collingwood and Brisbane have all invested in youth recently (the proportion of their lists under 23 is above the AFL average of 51.4%).
Brisbane worryingly have not done as well in this area as those other three, the average value is 388.7 per player (AFL average 534.9 points). Again, this shows the recent talent drain (note, in constructing these values, academy and father son picks and other concessions may not have their value truly represented, however, we'll assume they average out, or at least have minimal impact).
Melbourne haven't exactly flooded their list with youth (I guess Paul Roos has been trying to rebalance the list after Neeld's bloodletting) but they sure are quality (well, potentially, these numbers don't reveal poor selections or inadequate development).
Loosely, this table can be read as a gauge of potential, the higher on the table, the better the potential for the future (ideally should be the reverse of the current ladder).
Secondly, the middle age cohort (23 to 27 years inclusive).
This cohort could represent the players in their prime and Carlton stand out as having one of the highest value per middle aged player (1004.2 points). Where did they go wrong? Recruitment? Development? Also, no surprise that West Coast are coming into their prime with Richmond also pushing (they certainly have recruited heavily in this age bracket recently, 46.7% of their list is between 23 and 27 years old - AFL average 35%).
It also shows Geelong and St Kilda have a gap in this age group, due to being up for so long from 2007-2010 for example. 29.1% and 9.2% are the proportion of their lists value from this age group respectively (AFL average 37.2%). The proportion of players in this age bracket for these two teams is similar (26.1 and 26.7%) so it indicates St. Kilda have done particularly poorly in this area (Ross Lyon's legacy?).
Again, this table shows GWS probably had a more balanced start-up strategy than Gold Coast. The proportion of Port Adelaide's list value in this age group is 58% - the highest in the AFL. It suggests Port should be pushing hard now (definitely underachieved this year).
Interestingly, the Bulldogs are relatively quite poor in this area (only a quarter of their list is in this age bracket with only 22% of their list value). They are certainly well ahead of time! (I wonder how they compare to the Baby Bombers say, or are they like Geelong in 2004....)
Finally, the over 27 age group:
As expected, Geelong and Hawthorn are up the top - in fact, Hawthorn have done really well quality wise, which I guess is shown by the good form of their older players. Collingwood have clearly gotten rid of a lot of dead wood.
Fremantle's list has the highest percentage over 27 (26.1% compared to the AFL average of 13.7%). Interestingly, the average points per player for Fremantle is below average (568.2 versus 615.6) which supports the Ross Lyon stereotype of role players and not rejuvenating the list (might be an idea to look at St. Kilda's break down under Lyon).
Essendon seem to be coming to the end of a cycle (without a great peak, thanks to some poor decisions). A third of the value of their list comes from the older players (who make up only 17.8% of the list size). Their recent drafts have not been as strong.
There is undoubtedly lots more that can be gleaned (and disputed!), but it in general, these figures conform to expectations.
REFERENCES:
AFL Bidding System Explanation sheet
