Draft Review 2023 - Re-do the draft

[JHF1870]

If its for 2026, then its an even worse take. Watson is elite AA lock so far this year.

Because its an impact model. Not role specific. I suspect Watson scores more in midfield. Its explained in his stat model thread and I would expect he would release role specific ratings

 

[lordbob]

Because its an impact model. Not role specific. I suspect Watson scores more in midfield. Its explained in his stat model thread and I would expect he would release role specific ratings

Except watson is one of the highest impact per disposal players in the afl this year. Impact is his whole game. Almost 50% of his disposals result in a goal. and has been involved in over 35% of the Hawk's sscores. So if its an impact model then its even more flawed.

Despite about twice the disposals, roberts has only been involved in 22% of his teams scores.

 

[peak17]

I probably should have framed the original table more carefully, because I think a few people are reading it as a redraft ranking, which it wasn’t intended to be. Admittedly I was a little too eager to publish my results and I should have either been clearer, or waited until I produced the relevant models.

The table I posted was not saying:

“Archie Roberts is a better prospect than Harley Reid”

or

“Nick Watson is only the 15th best player from the 2023 draft.”

It was a 2026 AFLIA output table for players from that draft cohort, using one version of one model.

That is a much narrower question.

It was asking: based on the recorded actions in 2026 so far, who has generated the strongest AFLIA output in this sample?

That is different from asking:

Who would you take first in a redraft?
Who has the highest ceiling?
Who has the most trade value?
Who is most likely to be elite in five years?
Who has the best role-adjusted output?
Who has the best peak game?
Who has the best repeatable high-end output?
Who is most consistent?

Those questions can, and probably should, produce different answers. It's also important to note that it's virtually impossible to compare a forward like Nick Watson to a territory player like Bailey Smith

3

Bailey Smith

Age

25

Ht

185cm

Wt

86kg

Pos.

M/F

Career

Season

Last 5

D

18.3

5star

K

8.5

3star

HB

9.8

5star

M

2.3

2star

T

3.8

5star

CL

2.7

4star

D

19.8

5star

K

9.6

4star

HB

10.2

5star

M

1.8

2star

T

3.2

4star

CL

3.2

4star

D

14.8

4star

K

7.6

3star

HB

7.2

5star

M

2.0

3star

T

3.4

5star

CL

1.6

4star

via an overall ranking.

That is also why I’ve said a few times that AFLIA is still a developing project. The current main table is an absolute output model. It is not yet a complete role-relative model, and it is not a draft projection model. It will naturally favour players who produce repeatable, measurable actions across the four categories: Gain, Territory, Forward and Defence.

That means some players will look better in this model than they do by eye, and some high-talent players will look lower if their current output is narrower, more role-specific, less consistent, or more dependent on things the public stats do not capture well.

Nick Watson is probably the obvious example in this discussion. His forward output is strong, but the current all-category AFLIA table does not treat a medium forward as a separate universe from a half-back, midfielder or key forward. That is a limitation if the question is “who is best for their role?” It is less of a limitation if the question is “who has produced the most total modelled output?”

That is why the next stage is not to pretend one table answers everything. It is to build out more layers.

Roberts may rate very well in an absolute 2026 output table. Reid may rate better in a peak or projection view. Watson may rate much better in a medium-forward-specific or finishing model. Those are not contradictions. They are different questions.

That is generally how I think modelling should work. In GIS/spatial analysis, you rarely build one model and declare it reality. You build models for different questions, compare outputs, test the assumptions, and then decide what the result is actually telling you.

Same idea here.

I also think it is worth remembering that all models have strengths and weaknesses. Other model builders have been open about changing weights, adjusting outlier handling, adding age/potential components, and re-running things as issues appear. That is normal. It is part of model development.

So I’m happy to cop criticism of the original table, especially around role context and forwards. That criticism is fair. But I wouldn’t judge the whole project off one 2026 absolute-output table, and I definitely wouldn’t treat it as a redraft board.

The better way to use it is:

“This is one lens. What does this lens show? Where does it disagree with the eye test? And what extra model layer do we need to explain that difference?”

That is what I’m working toward and if I'm being honest identifying a best of '23 rank I'd be looking at multiple models that wouldn't rely on one overall rating 'especially' for a re-ranking.

In reality, most clubs have different needs at draft time so what fits one club at the time, may not fit another. FWIW, Archie Roberts was exactly what Essendon needed over a high impact player and while that's my opinion I think they needed to get runs on the board and Roberts fit their needs well.

Here are some tables that DO show how well Watson is playing relative to his position and perhaps this is an opportunity to look at the relative rank to position over the last 3 years in contrast to the other players. In that case Watson ranks very highly.

As I mentioned earlier with single player ratings, no single statistician can come up with one player rating that can effectively rank every player fairly and anyone claiming to do so is lying.

The best thing we can do is rank the players overall on regression weights, of which I've done in blocks, and then explore the data further to compare players within their player role. I can say that the results show that Nick Watson has been the best small forward this year by a mile. I can comparatively say that Bailey Smith

3

Bailey Smith

Age

25

Ht

185cm

Wt

86kg

Pos.

M/F

Career

Season

Last 5

D

18.3

5star

K

8.5

3star

HB

9.8

5star

M

2.3

2star

T

3.8

5star

CL

2.7

4star

D

19.8

5star

K

9.6

4star

HB

10.2

5star

M

1.8

2star

T

3.2

4star

CL

3.2

4star

D

14.8

4star

K

7.6

3star

HB

7.2

5star

M

2.0

3star

T

3.4

5star

CL

1.6

4star

has been the best midfielder significantly.

Comparing the two become somewhat difficult because they both have completely different roles and impact different areas of the game.

MEDIUM_FORWARD — Position-Specific Forward Model
Minimum games: 5 in each season.
This compares players only against MEDIUM_FORWARD players within the same season.
The model has three equal blocks: Forward Creation, Diminishing Goal Volume, and Finishing Efficiency.
Scoring Shots = goals + behinds. Expected points uses 3.93 points per scoring shot.

While Nick doesn't have the biggest forward game, he's the most consistent and he's performing better now than anyone in the past two seasons.


Top 10 MEDIUM_FORWARD Seasons — 2026

Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
1	nick watson	Hawthorn	11	28	11	39	25.73	0.67	1.35	0.92	2.94	0.43	1.02	4.76	6.73
2	charlie cameron	Brisbane	11	25	7	32	31.24	0.67	1.14	0.98	2.79	-0.18	1.02	5.54	6.72
3	jack higgins	St Kilda	8	17	6	23	17.61	0.34	0.94	0.95	2.23	-1.54	0.71	4.04	8.23
4	brent daniels	GWS	6	10	1	11	17.77	0.68	0.72	0.82	2.23	1.67	1.03	3.00	4.27
5	shaun mannagh	Geelong	11	19	6	25	21.75	0.65	0.54	0.80	2.00	0.57	1.01	5.13	9.96
6	paul curtis	North Melbourne	11	22	14	36	4.52	0.90	0.94	0.14	1.98	0.15	1.24	4.54	11.32
7	kai lohmann	Brisbane	11	21	10	31	14.17	0.52	0.86	0.46	1.84	-0.54	0.88	3.99	5.93
8	max hall	St Kilda	11	14	6	20	11.40	0.97	0.31	0.29	1.57	1.72	1.31	3.90	7.89
9	jamie elliott	Collingwood	11	17	7	24	14.68	0.56	0.58	0.35	1.49	-1.31	0.92	3.59	5.06
10	harry sharp	Melbourne	11	16	7	23	12.61	0.33	0.66	0.20	1.19	-0.36	0.70	1.89	2.62

Top 10 MEDIUM_FORWARD Seasons — 2025

Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
1	jamie elliott	Collingwood	23	52	29	81	22.67	1.08	1.05	0.36	2.49	-0.45	1.37	8.14	9.42
2	paul curtis	North Melbourne	19	38	15	53	34.71	0.80	0.93	0.63	2.37	-0.05	1.10	5.13	8.18
3	ben long	Gold Coast	22	43	24	67	18.69	1.08	0.92	0.32	2.31	-0.27	1.37	6.12	7.12
4	jack higgins	St Kilda	23	46	17	63	45.41	0.26	0.90	0.79	1.95	-1.73	0.56	5.98	8.40
5	rhylee west	Western Bulldogs	23	39	19	58	25.06	0.71	0.75	0.29	1.76	-0.17	1.01	5.25	7.35
6	shaun mannagh	Geelong	20	28	19	47	2.29	1.09	0.44	-0.07	1.47	1.54	1.38	3.70	5.86
7	izak rankine	Adelaide	22	31	21	52	2.64	0.80	0.53	0.04	1.36	2.58	1.09	4.41	7.72
8	jake melksham	Melbourne	19	33	22	55	3.85	0.72	0.60	0.01	1.33	-0.57	1.01	5.23	8.04
9	cameron zurhaar	North Melbourne	22	38	24	62	8.34	0.46	0.73	0.08	1.27	-0.94	0.76	5.23	6.92
10	jy farrar	Gold Coast	6	9	7	16	-1.88	0.69	0.61	-0.03	1.27	-1.03	0.98	1.81	4.28

Top 10 MEDIUM_FORWARD Seasons — 2024

Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
1	izak rankine	Adelaide	15	29	9	38	33.66	0.59	0.94	0.53	2.05	1.85	0.95	4.71	6.81
2	dylan moore	Hawthorn	23	35	16	51	25.57	1.17	0.47	0.39	2.02	1.61	1.52	6.18	8.85
3	tyson stengle	Geelong	23	42	16	58	40.06	0.58	0.81	0.59	1.98	-0.12	0.94	4.50	6.22
4	jack higgins	St Kilda	20	36	20	56	15.92	0.39	0.73	0.37	1.49	-0.95	0.76	4.79	6.82
5	ben keays	Adelaide	23	34	20	54	11.78	0.61	0.48	0.20	1.29	0.51	0.97	5.45	8.30
6	tom papley	Sydney	18	30	29	59	-22.87	0.94	0.66	-0.32	1.28	0.61	1.30	5.05	7.83
7	ben long	Gold Coast	17	26	17	43	4.01	0.62	0.57	0.06	1.25	-1.00	0.98	4.21	6.07
8	bayley fritsch	Melbourne	23	41	23	64	17.48	0.23	0.70	0.27	1.20	-1.19	0.59	5.00	7.26
9	cody weightman	Western Bulldogs	16	27	16	43	9.01	0.32	0.58	0.26	1.16	-0.92	0.68	4.75	9.14
10	sam sturt	Fremantle	13	21	7	28	22.96	-0.01	0.67	0.50	1.16	-1.82	0.36	3.25	3.95

Nick Watson — 2026 Position Forward Profile

season.x	Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
2026	1	nick watson	Hawthorn	11	28	11	39	25.73	0.67	1.35	0.92	2.94	0.43	1.02	4.76	6.73

Top 30 MEDIUM_FORWARD Seasons — 2024 to 2026

Overall_Rank	Season	Season_Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg
1	2026	1	nick watson	Hawthorn	11	28	11	39	25.73	0.67	1.35	0.92	2.94	0.43	1.02
2	2026	2	charlie cameron	Brisbane	11	25	7	32	31.24	0.67	1.14	0.98	2.79	-0.18	1.02
3	2025	1	jamie elliott	Collingwood	23	52	29	81	22.67	1.08	1.05	0.36	2.49	-0.45	1.37
4	2025	2	paul curtis	North Melbourne	19	38	15	53	34.71	0.80	0.93	0.63	2.37	-0.05	1.10
5	2025	3	ben long	Gold Coast	22	43	24	67	18.69	1.08	0.92	0.32	2.31	-0.27	1.37
6	2026	3	jack higgins	St Kilda	8	17	6	23	17.61	0.34	0.94	0.95	2.23	-1.54	0.71
7	2026	4	brent daniels	GWS	6	10	1	11	17.77	0.68	0.72	0.82	2.23	1.67	1.03
8	2024	1	izak rankine	Adelaide	15	29	9	38	33.66	0.59	0.94	0.53	2.05	1.85	0.95
9	2024	2	dylan moore	Hawthorn	23	35	16	51	25.57	1.17	0.47	0.39	2.02	1.61	1.52
10	2026	5	shaun mannagh	Geelong	11	19	6	25	21.75	0.65	0.54	0.80	2.00	0.57	1.01
11	2026	6	paul curtis	North Melbourne	11	22	14	36	4.52	0.90	0.94	0.14	1.98	0.15	1.24
12	2024	3	tyson stengle	Geelong	23	42	16	58	40.06	0.58	0.81	0.59	1.98	-0.12	0.94
13	2025	4	jack higgins	St Kilda	23	46	17	63	45.41	0.26	0.90	0.79	1.95	-1.73	0.56
14	2026	7	kai lohmann	Brisbane	11	21	10	31	14.17	0.52	0.86	0.46	1.84	-0.54	0.88
15	2025	5	rhylee west	Western Bulldogs	23	39	19	58	25.06	0.71	0.75	0.29	1.76	-0.17	1.01
16	2026	8	max hall	St Kilda	11	14	6	20	11.40	0.97	0.31	0.29	1.57	1.72	1.31
17	2026	9	jamie elliott	Collingwood	11	17	7	24	14.68	0.56	0.58	0.35	1.49	-1.31	0.92
18	2024	4	jack higgins	St Kilda	20	36	20	56	15.92	0.39	0.73	0.37	1.49	-0.95	0.76
19	2025	6	shaun mannagh	Geelong	20	28	19	47	2.29	1.09	0.44	-0.07	1.47	1.54	1.38
20	2025	7	izak rankine	Adelaide	22	31	21	52	2.64	0.80	0.53	0.04	1.36	2.58	1.09
21	2025	8	jake melksham	Melbourne	19	33	22	55	3.85	0.72	0.60	0.01	1.33	-0.57	1.01
22	2024	5	ben keays	Adelaide	23	34	20	54	11.78	0.61	0.48	0.20	1.29	0.51	0.97
23	2024	6	tom papley	Sydney	18	30	29	59	-22.87	0.94	0.66	-0.32	1.28	0.61	1.30
24	2025	9	cameron zurhaar	North Melbourne	22	38	24	62	8.34	0.46	0.73	0.08	1.27	-0.94	0.76
25	2025	10	jy farrar	Gold Coast	6	9	7	16	-1.88	0.69	0.61	-0.03	1.27	-1.03	0.98
26	2024	7	ben long	Gold Coast	17	26	17	43	4.01	0.62	0.57	0.06	1.25	-1.00	0.98
27	2025	11	patrick dangerfield	Geelong	20	27	22	49	-8.57	0.93	0.41	-0.11	1.23	0.65	1.22
28	2024	8	bayley fritsch	Melbourne	23	41	23	64	17.48	0.23	0.70	0.27	1.20	-1.19	0.59
29	2026	10	harry sharp	Melbourne	11	16	7	23	12.61	0.33	0.66	0.20	1.19	-0.36	0.70
30	2026	11	corey durdin	Port Adelaide	11	17	8	25	11.75	0.28	0.74	0.16	1.17	-1.86	0.65

 

[lordbob]

I probably should have framed the original table more carefully, because I think a few people are reading it as a redraft ranking, which it wasn’t intended to be. Admittedly I was a little too eager to publish my results and I should have either been clearer, or waited until I produced the relevant models.

The table I posted was not saying:

“Archie Roberts is a better prospect than Harley Reid”

or

“Nick Watson is only the 15th best player from the 2023 draft.”

It was a 2026 AFLIA output table for players from that draft cohort, using one version of one model.

That is a much narrower question.

It was asking: based on the recorded actions in 2026 so far, who has generated the strongest AFLIA output in this sample?

That is different from asking:

Who would you take first in a redraft?
Who has the highest ceiling?
Who has the most trade value?
Who is most likely to be elite in five years?
Who has the best role-adjusted output?
Who has the best peak game?
Who has the best repeatable high-end output?
Who is most consistent?

Those questions can, and probably should, produce different answers. It's also important to note that it's virtually impossible to compare a forward like Nick Watson to a territory player like Bailey Smith via an overall ranking.

That is also why I’ve said a few times that AFLIA is still a developing project. The current main table is an absolute output model. It is not yet a complete role-relative model, and it is not a draft projection model. It will naturally favour players who produce repeatable, measurable actions across the four categories: Gain, Territory, Forward and Defence.

That means some players will look better in this model than they do by eye, and some high-talent players will look lower if their current output is narrower, more role-specific, less consistent, or more dependent on things the public stats do not capture well.

Nick Watson is probably the obvious example in this discussion. His forward output is strong, but the current all-category AFLIA table does not treat a medium forward as a separate universe from a half-back, midfielder or key forward. That is a limitation if the question is “who is best for their role?” It is less of a limitation if the question is “who has produced the most total modelled output?”

That is why the next stage is not to pretend one table answers everything. It is to build out more layers.

Roberts may rate very well in an absolute 2026 output table. Reid may rate better in a peak or projection view. Watson may rate much better in a medium-forward-specific or finishing model. Those are not contradictions. They are different questions.

That is generally how I think modelling should work. In GIS/spatial analysis, you rarely build one model and declare it reality. You build models for different questions, compare outputs, test the assumptions, and then decide what the result is actually telling you.

Same idea here.

I also think it is worth remembering that all models have strengths and weaknesses. Other model builders have been open about changing weights, adjusting outlier handling, adding age/potential components, and re-running things as issues appear. That is normal. It is part of model development.

So I’m happy to cop criticism of the original table, especially around role context and forwards. That criticism is fair. But I wouldn’t judge the whole project off one 2026 absolute-output table, and I definitely wouldn’t treat it as a redraft board.

The better way to use it is:

“This is one lens. What does this lens show? Where does it disagree with the eye test? And what extra model layer do we need to explain that difference?”

That is what I’m working toward and if I'm being honest identifying a best of '23 rank I'd be looking at multiple models that wouldn't rely on one overall rating 'especially' for a re-ranking.

In reality, most clubs have different needs at draft time so what fits one club at the time, may not fit another. FWIW, Archie Roberts was exactly what Essendon needed over a high impact player and while that's my opinion I think they needed to get runs on the board and Roberts fit their needs well.

Here are some tables that DO show how well Watson is playing relative to his position and perhaps this is an opportunity to look at the relative rank to position over the last 3 years in contrast to the other players. In that case Watson ranks very highly.

As I mentioned earlier with single player ratings, no single statistician can come up with one player rating that can effectively rank every player fairly and anyone claiming to do so is lying.

The best thing we can do is rank the players overall on regression weights, of which I've done in blocks, and then explore the data further to compare players within their player role. I can say that the results show that Nick Watson has been the best small forward this year by a mile. I can comparatively say that Bailey Smith has been the best midfielder significantly.

Comparing the two become somewhat difficult because they both have completely different roles and impact different areas of the game.

MEDIUM_FORWARD — Position-Specific Forward Model
Minimum games: 5 in each season.
This compares players only against MEDIUM_FORWARD players within the same season.
The model has three equal blocks: Forward Creation, Diminishing Goal Volume, and Finishing Efficiency.
Scoring Shots = goals + behinds. Expected points uses 3.93 points per scoring shot.

While Nick doesn't have the biggest forward game, he's the most consistent and he's performing better now than anyone in the past two seasons.


Top 10 MEDIUM_FORWARD Seasons — 2026

Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
1	nick watson	Hawthorn	11	28	11	39	25.73	0.67	1.35	0.92	2.94	0.43	1.02	4.76	6.73
2	charlie cameron	Brisbane	11	25	7	32	31.24	0.67	1.14	0.98	2.79	-0.18	1.02	5.54	6.72
3	jack higgins	St Kilda	8	17	6	23	17.61	0.34	0.94	0.95	2.23	-1.54	0.71	4.04	8.23
4	brent daniels	GWS	6	10	1	11	17.77	0.68	0.72	0.82	2.23	1.67	1.03	3.00	4.27
5	shaun mannagh	Geelong	11	19	6	25	21.75	0.65	0.54	0.80	2.00	0.57	1.01	5.13	9.96
6	paul curtis	North Melbourne	11	22	14	36	4.52	0.90	0.94	0.14	1.98	0.15	1.24	4.54	11.32
7	kai lohmann	Brisbane	11	21	10	31	14.17	0.52	0.86	0.46	1.84	-0.54	0.88	3.99	5.93
8	max hall	St Kilda	11	14	6	20	11.40	0.97	0.31	0.29	1.57	1.72	1.31	3.90	7.89
9	jamie elliott	Collingwood	11	17	7	24	14.68	0.56	0.58	0.35	1.49	-1.31	0.92	3.59	5.06
10	harry sharp	Melbourne	11	16	7	23	12.61	0.33	0.66	0.20	1.19	-0.36	0.70	1.89	2.62

Top 10 MEDIUM_FORWARD Seasons — 2025

Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
1	jamie elliott	Collingwood	23	52	29	81	22.67	1.08	1.05	0.36	2.49	-0.45	1.37	8.14	9.42
2	paul curtis	North Melbourne	19	38	15	53	34.71	0.80	0.93	0.63	2.37	-0.05	1.10	5.13	8.18
3	ben long	Gold Coast	22	43	24	67	18.69	1.08	0.92	0.32	2.31	-0.27	1.37	6.12	7.12
4	jack higgins	St Kilda	23	46	17	63	45.41	0.26	0.90	0.79	1.95	-1.73	0.56	5.98	8.40
5	rhylee west	Western Bulldogs	23	39	19	58	25.06	0.71	0.75	0.29	1.76	-0.17	1.01	5.25	7.35
6	shaun mannagh	Geelong	20	28	19	47	2.29	1.09	0.44	-0.07	1.47	1.54	1.38	3.70	5.86
7	izak rankine	Adelaide	22	31	21	52	2.64	0.80	0.53	0.04	1.36	2.58	1.09	4.41	7.72
8	jake melksham	Melbourne	19	33	22	55	3.85	0.72	0.60	0.01	1.33	-0.57	1.01	5.23	8.04
9	cameron zurhaar	North Melbourne	22	38	24	62	8.34	0.46	0.73	0.08	1.27	-0.94	0.76	5.23	6.92
10	jy farrar	Gold Coast	6	9	7	16	-1.88	0.69	0.61	-0.03	1.27	-1.03	0.98	1.81	4.28

Top 10 MEDIUM_FORWARD Seasons — 2024

Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
1	izak rankine	Adelaide	15	29	9	38	33.66	0.59	0.94	0.53	2.05	1.85	0.95	4.71	6.81
2	dylan moore	Hawthorn	23	35	16	51	25.57	1.17	0.47	0.39	2.02	1.61	1.52	6.18	8.85
3	tyson stengle	Geelong	23	42	16	58	40.06	0.58	0.81	0.59	1.98	-0.12	0.94	4.50	6.22
4	jack higgins	St Kilda	20	36	20	56	15.92	0.39	0.73	0.37	1.49	-0.95	0.76	4.79	6.82
5	ben keays	Adelaide	23	34	20	54	11.78	0.61	0.48	0.20	1.29	0.51	0.97	5.45	8.30
6	tom papley	Sydney	18	30	29	59	-22.87	0.94	0.66	-0.32	1.28	0.61	1.30	5.05	7.83
7	ben long	Gold Coast	17	26	17	43	4.01	0.62	0.57	0.06	1.25	-1.00	0.98	4.21	6.07
8	bayley fritsch	Melbourne	23	41	23	64	17.48	0.23	0.70	0.27	1.20	-1.19	0.59	5.00	7.26
9	cody weightman	Western Bulldogs	16	27	16	43	9.01	0.32	0.58	0.26	1.16	-0.92	0.68	4.75	9.14
10	sam sturt	Fremantle	13	21	7	28	22.96	-0.01	0.67	0.50	1.16	-1.82	0.36	3.25	3.95

Nick Watson — 2026 Position Forward Profile

season.x	Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg	Peak5 Position Forward Avg	Max Position Forward Game
2026	1	nick watson	Hawthorn	11	28	11	39	25.73	0.67	1.35	0.92	2.94	0.43	1.02	4.76	6.73

Top 30 MEDIUM_FORWARD Seasons — 2024 to 2026

Overall_Rank	Season	Season_Rank	Player	Team	Games	Goals	Behinds	Scoring Shots	Pts +/- Exp	Forward Creation Z	Goal Volume Z	Finishing Z	Position Forward Avg	AFLIA Avg	Raw Forward Avg
1	2026	1	nick watson	Hawthorn	11	28	11	39	25.73	0.67	1.35	0.92	2.94	0.43	1.02
2	2026	2	charlie cameron	Brisbane	11	25	7	32	31.24	0.67	1.14	0.98	2.79	-0.18	1.02
3	2025	1	jamie elliott	Collingwood	23	52	29	81	22.67	1.08	1.05	0.36	2.49	-0.45	1.37
4	2025	2	paul curtis	North Melbourne	19	38	15	53	34.71	0.80	0.93	0.63	2.37	-0.05	1.10
5	2025	3	ben long	Gold Coast	22	43	24	67	18.69	1.08	0.92	0.32	2.31	-0.27	1.37
6	2026	3	jack higgins	St Kilda	8	17	6	23	17.61	0.34	0.94	0.95	2.23	-1.54	0.71
7	2026	4	brent daniels	GWS	6	10	1	11	17.77	0.68	0.72	0.82	2.23	1.67	1.03
8	2024	1	izak rankine	Adelaide	15	29	9	38	33.66	0.59	0.94	0.53	2.05	1.85	0.95
9	2024	2	dylan moore	Hawthorn	23	35	16	51	25.57	1.17	0.47	0.39	2.02	1.61	1.52
10	2026	5	shaun mannagh	Geelong	11	19	6	25	21.75	0.65	0.54	0.80	2.00	0.57	1.01
11	2026	6	paul curtis	North Melbourne	11	22	14	36	4.52	0.90	0.94	0.14	1.98	0.15	1.24
12	2024	3	tyson stengle	Geelong	23	42	16	58	40.06	0.58	0.81	0.59	1.98	-0.12	0.94
13	2025	4	jack higgins	St Kilda	23	46	17	63	45.41	0.26	0.90	0.79	1.95	-1.73	0.56
14	2026	7	kai lohmann	Brisbane	11	21	10	31	14.17	0.52	0.86	0.46	1.84	-0.54	0.88
15	2025	5	rhylee west	Western Bulldogs	23	39	19	58	25.06	0.71	0.75	0.29	1.76	-0.17	1.01
16	2026	8	max hall	St Kilda	11	14	6	20	11.40	0.97	0.31	0.29	1.57	1.72	1.31
17	2026	9	jamie elliott	Collingwood	11	17	7	24	14.68	0.56	0.58	0.35	1.49	-1.31	0.92
18	2024	4	jack higgins	St Kilda	20	36	20	56	15.92	0.39	0.73	0.37	1.49	-0.95	0.76
19	2025	6	shaun mannagh	Geelong	20	28	19	47	2.29	1.09	0.44	-0.07	1.47	1.54	1.38
20	2025	7	izak rankine	Adelaide	22	31	21	52	2.64	0.80	0.53	0.04	1.36	2.58	1.09
21	2025	8	jake melksham	Melbourne	19	33	22	55	3.85	0.72	0.60	0.01	1.33	-0.57	1.01
22	2024	5	ben keays	Adelaide	23	34	20	54	11.78	0.61	0.48	0.20	1.29	0.51	0.97
23	2024	6	tom papley	Sydney	18	30	29	59	-22.87	0.94	0.66	-0.32	1.28	0.61	1.30
24	2025	9	cameron zurhaar	North Melbourne	22	38	24	62	8.34	0.46	0.73	0.08	1.27	-0.94	0.76
25	2025	10	jy farrar	Gold Coast	6	9	7	16	-1.88	0.69	0.61	-0.03	1.27	-1.03	0.98
26	2024	7	ben long	Gold Coast	17	26	17	43	4.01	0.62	0.57	0.06	1.25	-1.00	0.98
27	2025	11	patrick dangerfield	Geelong	20	27	22	49	-8.57	0.93	0.41	-0.11	1.23	0.65	1.22
28	2024	8	bayley fritsch	Melbourne	23	41	23	64	17.48	0.23	0.70	0.27	1.20	-1.19	0.59
29	2026	10	harry sharp	Melbourne	11	16	7	23	12.61	0.33	0.66	0.20	1.19	-0.36	0.70
30	2026	11	corey durdin	Port Adelaide	11	17	8	25	11.75	0.28	0.74	0.16	1.17	-1.86	0.65

First time watson has been referred to as a "medium" forward.

 

[Wild Bill]

Robertsl
10th in metres gained
1st in total disposals
11th in total rebound 50s
12th total marks
1st in effective disposals (per game)


he is going a bit better than most think or wish to acknowledge. I wouldn't be taking him number 1 however he is still clearly a top 5 player form this draft

High possession low impact players - it’s why he can get 42 possessions and not get a single coaches vote.

 

[peak17]

High possession low impact players - it’s why he can get 42 possessions and not get a single coaches vote.

He's good overall, and what Essendon have needed so they got value from him. Role specific there are better players.

 

[peak17]

First time watson has been referred to as a "medium" forward.

It's just what footywire defines his position as.

 

[Lars27]

High possession low impact players - it’s why he can get 42 possessions and not get a single coaches vote.

Roberts has got 40+ possessions three times this season. He got the 10 votes in one of them, and was equal second in the coaches votes in another. The third one he didn't get votes was in a complete flogging vs the pies (he was still one of our best players).
Sheezel got 40 disposals this year vs the Eagles and didn't get a vote.

 

[RCAB]

First time watson has been referred to as a "medium" forward.

It's LLM slop. It doesn't really know what it's saying.

 

[Wild Bill]

Roberts has got 40+ possessions three times this season. He got the 10 votes in one of them, and was equal second in the coaches votes in another. The third one he didn't get votes was in a complete flogging vs the pies (he was still one of our best players).
Sheezel got 40 disposals this year vs the Eagles and didn't get a vote.

Yesterday is a perfect example - the leading possession winner out of both teams - yet his opponent was one of the best on ground. Roberts has only polled coaches vote in 2 games in 2026.

 

[peak17]

First time watson has been referred to as a "medium" forward.

I thought I'd show you the output on Rstudio. Medium forward and defender are defined in the footywire data scrape and not AI Slop like a certain poster here is for some reason hell bent on discrediting the thousands of hours of work I've put into my models.

I think it's strange he's so focused on discrediting me on a public forum.

 

[peak17]

Yesterday is a perfect example - the leading possession winner out of both teams - yet his opponent was one of the best on ground. Roberts has only polled coaches vote in 2 games in 2026.

Hypothetical redrafts, even drafting itself isn't as simple as looking at a player rating and my overall rating was never intended to be a redraft and I should have explained it better.

Clubs won't all be in the same window when it comes to using their first draft pick but it's no surprise that territory players and ball accumulators are often first round picks along with generational, high impact and strong KPPs. Roberts landed at pick 54 so Essendon's value from that pick was relatively high for little investment.

My thoughts so far, both opinion and data based are and by no means is this an ordering:

• Harley Reid's peak potential is still unknown and his last month has been incredible. Had it not been for the heavy attention and poor turnover rate he would have had an even bigger game. IMO he's passed Roberts peak, and has the greater peak potential.
• Nick Watson, as I showed earlier has had one of the best role based seasons and has outplayed his role in both 2024 and 2025. While the OVERALL model isn't overly friendly to forwards and to a lesser extend defenders he's outplaying his position.
• Logan Morris as above. He's more than a proven key forward and I rate slightly higher than Georgiades having greater efficiency inside 50. He's definitely one of the best young key forwards.
• Archie Roberts has already peaked imo and while he ranks higher in the overall model due to his territory, rebound and clean accumulation. While Essendon picked him up with minimal investment his peak potential moving forward won't be as great imo as the others. As of round 11, I had his Peak rating at 45.
• The rest have all shown signs of peak games but haven't offered enough consistency to confidently rank.

I think Reid's unknown potential and more than capable of being the games best player would probably land him at #1 for most teams in a redraft.

Watson and Morris would be dependent on club needs and both being deserving of the 2, and 3 spot. Archie Roberts if you're looking for consistency and transition.

I don't think any of that is overly groundbreaking but I think the point is one player rating doesn't paint the whole picture.

In terms of Port Adelaide, I'd be trading Reid, Watson, Morris, Humphries, McKercher. With Rozee at half back, and Farrell I don't think we need a Roberts type player.

 

[Lars27]

Yesterday is a perfect example - the leading possession winner out of both teams - yet his opponent was one of the best on ground. Roberts has only polled coaches vote in 2 games in 2026.

I disagree. Murdock is a good player and it is no shame to be beaten by him (and Prior gave away a free to gift one of his goals). Archie still had a good game. And he is only 20 years old in a side that has won a single game in the last 25. Not really fair to rate him based on coaches votes

 

[Wild Bill]

I disagree. Murdock is a good player and it is no shame to be beaten by him (and Prior gave away a free to gift one of his goals). Archie still had a good game. And he is only 20 years old in a side that has won a single game in the last 25. Not really fair to rate him based on coaches votes

Murdock was playing his 9th game.

 

[Lars27]

Murdock was playing his 9th game.

And Roberts has played under 40 and is only 20 years old. Murdock is nearly 26 with a lot of WAFL experience. And It's only one game and it wasn't even a bad one for Archie.

 

[hawkman]

It's Harley #1, Watson #2, Morris #3 and then daylight right now

Would have O'Sullivan close behind in 4

 

[txbyhull]

Would have O'Sullivan close behind in 4

Gun