Shoot or Pass?

When a player marks the ball, should they take the set shot at goal or pass it to a teammate closer to goal? At 20 metres the answer is obvious. Around 50 or 55 metres, teams often look for the pass instead. This model tests where that trade-off changes by comparing the expected points from shooting with the expected points from passing, marking, and finishing the next shot.

Possession outcome

Pass the shot

Based largely on the current goal accuracy and pass mark rate set, the player should pass for the mark. It is the more efficient play, and making this decision over similar set-shot possessions is expected to result in more points scored per possession.

Goal accuracy needed

41.4%

The player needs to be about a 41.4% goal kicker or better from this spot for shooting to match the pass option.

Mark rate needed

59.5%

The team needs to mark about 59.5% of these passes or better for passing to match the direct shot.

Set shot accuracy41%The chance the player kicks a goal if they take the original set shot from where the mark or free kick was paid. For example, if this is 40%, then 40 of 100 direct set shots become goals before counting behinds.Set shot behind rate65%When the direct shot misses the goal, this is how often it still scores a behind instead of no score. For example, if this is 70%, then 70 of 100 missed direct shots still score one point.Pass mark rate60%The chance the pass to inside the 50 is successfully marked by a teammate. How often does a team get a mark inside the 50 from a passed set shot. For example, if this is 30%, then 30 of 100 passes lead to a mark.Shot accuracy after successful mark76%The goal accuracy of the teammate after the pass is marked. The dropdown below can seed this from chart zones, then you can override it. For example, if this is 75%, then 75 of 100 successful marks are converted into goals.Behind rate after marked pass90%If the teammate marks the pass and misses the goal, this is how often that miss still becomes a behind. For example, if this is 80%, then 80 of 100 misses after a successful mark still score one point.Shot accuracy after successful mark76.2%Choose where the follow-up shot is caught after a successful pass. This is seeded from the 2012 chart data and sets the after-mark accuracy slider above. For example, selecting caught center 15-40m uses that aggregate chart accuracy as the starting point.

After-mark shot

Caught center, 15-40m

After-mark accuracy

76.2%

Sample size

1,202 shots

Use this selected range as the real-world reference for the shot a teammate receives after a successful pass. For example, if the pass is designed to create a central mark inside 50, this chart-based accuracy helps estimate how valuable that follow-up shot is before you adjust for your own player, team, or match situation.

Adjust the assumptions to test a specific scenario: the player's direct goal chance, how often the pass is marked, and how accurate the next shot is after the mark. The side cards then show which choice is expected to produce more points over repeated possessions.

Direct shot EV

2.84

The direct shot is worth 2.84 points per possession. If this shot happened 100 times, that is about 284 total points from goals and behinds.

Pass-for-mark EV

2.87

The pass option is worth 2.87 points per possession. Across 100 similar possessions, that is about 287 total points after accounting for the mark rate and the shot after the mark.

Shot edge

-0.02

Shot edge compares those two EVs. This is roughly 2 points fewer than passing over 100 similar possessions, based on the assumptions above.

100-possession gap

-2

This turns shot edge into a season-style total. If this exact decision happened 100 times, the pass option is expected to score about 2 more points than the alternative.

After-mark shot EV

4.78

After a teammate marks the pass, their shot is worth 4.78 points on average. Across 100 successful marks, that is about 478 points before reducing it by the pass mark rate.

2012 Set Shot Data

Total sample: 4,599 shots

View source PDF

Set-shot goal-kicking accuracy by ground location

Source: “Set Shot Goal Kicking” figure from Anderson, Breed, Spittle and Larkin (2018), “Factors Affecting Set Shot Goal-kicking Performance in the Australian Football League,” via Victoria University Research Repository.

This chart gives the model a real-world reference point for judging what a reasonable set-shot accuracy might look like from different parts of the ground. When adjusting the sliders above, the chart can be used as a guide: a central shot from 30-40 metres should not be treated the same as a wide shot from 50 metres, and a pass that creates a mark closer to goal should be compared with the accuracy normally expected from that new location. The figure helps turn the calculator from a purely theoretical EV exercise into a practical tool for testing realistic AFL set-shot scenarios.

EV - Expected Value

Expected value is the long-run average score of a possession. It does not predict one kick; it estimates what that choice is worth if repeated many times.

For example, imagine a player marks the ball 53 metres from goal, straight in front. They can take the set shot, or they can chip it inside 50 and hope a teammate marks it closer to goal.

If the team chooses the pass 100 times, and the pass is marked 30 times, then the best-case follow-up is that every one of those 30 marks becomes a goal. That is a 100% goal rate after the mark: 30 goals, 180 points total, or 1.8 points per attempt.

So the set shot does not need to be a high-percentage kick. With no behinds at all, 31 goals from 100 direct shots is 186 points. That means a player who is a 31% goal kicker or better from that spot would beat the pass in this example. If some misses still score behinds, the required goal accuracy drops even lower.

In a nutshell: compare the points you expect from taking the shot with the points you expect from passing. The higher EV option is the more efficient football possession because, across 100 similar possessions, it is expected to produce more total points. That is the decision-making logic this calculator is testing.

Data Table

These tables are based on the Set Shot Goal Kicking figure and the supporting numbers gathered from the workbook, including distance splits, overall accuracy, and caught-center pass targets.

Distance	Left side		Center		Right side
Distance	Accuracy	Attempts	Accuracy	Attempts	Accuracy	Attempts
0-15 m	57.1%	7	98.2%	169	90.0%	10
15-30 m	63.9%	97	87.4%	540	46.2%	130
30-40 m	42.5%	167	67.1%	662	42.1%	228
40-50 m	38.3%	334	53.8%	1,127	37.6%	415
50 m+	25.3%	87	40.9%	501	20.8%	125

Overall accuracy

Distance	Attempts	Accuracy
0-15 m	186	96.2%
15-30 m	767	77.4%
30-40 m	1,057	57.8%
40-50 m	1,876	47.5%
50 m+	713	35.5%
Mod-high zone	3,137	63.5%

Total attempts: 4,599

Target area for a pass in

Distance

15-40 m center

Attempts

1,202

Accuracy

76.2%

Center accuracy rates

Distance	Accuracy	Attempts
0-15 m	98.2%	169
15-30 m	87.4%	540
30-40 m	67.1%	662
40-50 m	53.8%	1,127

Caught center mark accuracy

These are the dropdown options for the shot after a successful mark, aggregated from the center bands above.

Distance where caught	Accuracy	Attempts
15-40 m	76.2%	1,202
0-30 m	90.0%	709
0-40 m	78.9%	1,371
0-50 m	67.6%	2,498

Overall set shot accuracy: 55.0%

All shots

Left

41.5%

Center

63.1%

Right

38.2%

Inside 50: 58.5%

Left

43.8%

Center

67.6%

Right

41.0%

Inside 40: 68.9%

Left

50.5%

Center

78.9%

Right

44.9%

Inside 30: 81.1%

Left

63.4%

Center

90.0%

Right

49.3%

Formulae

These are the formulae used by the calculator. The main idea is to compare the expected points from taking the original set shot with the expected points from passing, marking, and then taking the next shot.

Simple set shot EV

EV = Accuracy x 6 + (1 - Accuracy) x 1

The basic expected value of a set shot. A goal is worth 6 points. If the shot misses, this simple version assumes the miss becomes a behind worth 1 point.

Behind-rate adjusted EV

EV = Accuracy x 6 + (1 - Accuracy) x behindRate

The adjusted version used by the sliders. Instead of every miss being worth 1 point, the behind rate controls how often a miss still scores one point.

Original set shot

EV1 = acc1 x 6 + (1 - acc1) x behindRate1

The expected value of taking the original shot from where the mark or free kick was paid.

Shot after the mark

EV2ndshot = acc2 x 6 + (1 - acc2) x behindRate2

The value of the shot after a successful pass is marked. This is calculated before applying the chance that the pass is actually marked.

Pass-for-mark EV

EV2 = markRate x EV2ndshot

The pass option only gets the follow-up shot value when the pass is successfully marked, so the after-mark shot EV is multiplied by the pass mark rate.

Shot edge

Shot edge = EV1 - EV2

Shot edge is the direct-shot EV minus the pass EV. Positive means the shot is ahead. Negative means the pass is ahead.

Mark rate needed

markRate = EV1 / EV2ndshot

The mark rate needed tells you how often the pass must be marked for passing to equal taking the original shot.

Goal accuracy needed

acc1 = (markRate x EV2ndshot - behindRate1) / (6 - behindRate1)

The direct goal accuracy needed tells you how accurate the original shot must be for shooting to equal the pass option under the current mark rate and behind assumptions.

Conclusion

Long set shots may be under-used

The model supports the core argument that AFL teams may be passing up too many shots when a player marks the ball around 50-55 metres from goal, especially when the player is central, because central set shots are more accurate. If the pass mark rate is low, the pass option has to overcome a major hurdle before the team even gets the closer shot.

The model also does not fully account for messy outcomes after a missed pass, such as a ground-ball scramble, play-on score, repeat stoppage, turnover pressure, or a defender spoiling the mark but the attacking team still scoring. Those outcomes are real considerations, but the base EV comparison shows why the automatic pass from long range deserves to be questioned.