Rd 3 vs Western Bulldogs @ Marvel - Match Thread

[neilfws]

Significant in what sense? If it leads to significantly more scoring for the favoured team, then yes. Was that the case? I don't know yet, but we will be able to figure it out when Josh updates his match data after Round 3.

I almost hesitate to revisit the game, now that the thread has calmed down...but the R3 match chain data has arrived, so let's do it.

What is this data? A footy stats nerd's dream is what it is. A chain is every sequence of play on the field: what happened, when it happened, who did it, where they were positioned. Here's a small example showing the chain that led to the Dogs first goal:

You can see that this is an example where a free kick (to Ugle-Hagan) did lead to a goal.

So we can select all the chains that end in a goal and count up how many of them included a Free For. The answer is:

In total then, 3 of the Dogs 9 goals came from a play that included a free kick: 2 in the first quarter and 1 in the second. The Swans - 2 of their 9 goals from a free kick: 1 in the first quarter, 1 in the third.

So the Dogs got 17 extra free kicks, but scored only one more goal from a free kick. This is why I'm always saying that the free kick differential just doesn't have a very big effect. If it did, we would see a strong relationship between the free kick differential and the score margin. Statistically, looking at lots of games, we don't see that - there is no essentially no relationship.

Now if you want to argue that missed free kicks should have been given which might have changed the result - and you probably do - well, that's a different issue to the differential.

Just to wrap up, some of you might be interested in the assessment from @hasumpstuffedup on Twitter. Overall he rates the umpiring for the game as "poor".

Has the umpire made A Bad decision? on Twitter: "FULL NOTES - Last night's controversial #AFLDogsSwans. Every FK decision analysed"

[RogueSwan]

Interesting work Neil. What about goals and behinds? Does that show anything different?
Edit:
If I'm looking at it right it was 10 scores (3.7) involved a free for. 10 out of 26, a similar ratio to 3 out of 9.

[i'm-uninformed2]

I almost hesitate to revisit the game, now that the thread has calmed down...but the R3 match chain data has arrived, so let's do it.

What is this data? A footy stats nerd's dream is what it is. A chain is every sequence of play on the field: what happened, when it happened, who did it, where they were positioned. Here's a small example showing the chain that led to the Dogs first goal:

You can see that this is an example where a free kick (to Ugle-Hagan) did lead to a goal.

So we can select all the chains that end in a goal and count up how many of them included a Free For. The answer is:

In total then, 3 of the Dogs 9 goals came from a play that included a free kick: 2 in the first quarter and 1 in the second. The Swans - 2 of their 9 goals from a free kick: 1 in the first quarter, 1 in the third.

So the Dogs got 17 extra free kicks, but scored only one more goal from a free kick. This is why I'm always saying that the free kick differential just doesn't have a very big effect. If it did, we would see a strong relationship between the free kick differential and the score margin. Statistically, looking at lots of games, we don't see that - there is no essentially no relationship.

Now if you want to argue that missed free kicks should have been given which might have changed the result - and you probably do - well, that's a different issue to the differential.

Just to wrap up, some of you might be interested in the assessment from @hasumpstuffedup on Twitter. Overall he rates the umpiring for the game as "poor".

Has the umpire made A Bad decision? on Twitter: "FULL NOTES - Last night's controversial #AFLDogsSwans. Every FK decision analysed"

Fair enough for what it is. But does it, for example, include a play where they got an easy release from a free in D50 to the wing, then there was stoppage or two, then they make their way forward for a goal. Or, free in middle, banged forward, ball bounces in and out of f50 once or twice, but ultimately back in for a goal because they were able to set up behind the ball.

What I'm basically thinking is field position matters a lot these days. Being released or positioned into a part of the ground that gives one side a competitive advantage is a substantial plus in the way the game is played.

Then, as you said, there were the absurd non-decisions that disadvantaged us. The Rowbottom trip and the below the knees, the Heeney high, the Libba HTB, etc etc

[neilfws]

What I'm basically thinking is field position matters a lot these days.

The complete data (I didn't show all the columns, there are 26 of them) includes coordinates on the field. It might be interesting to know where the frees that led to goals occurred, but it's probably not very relevant for this game, where goals from free kicks is only 3 versus 2.

[stevoswan]

I almost hesitate to revisit the game, now that the thread has calmed down...but the R3 match chain data has arrived, so let's do it.

What is this data? A footy stats nerd's dream is what it is. A chain is every sequence of play on the field: what happened, when it happened, who did it, where they were positioned. Here's a small example showing the chain that led to the Dogs first goal:

You can see that this is an example where a free kick (to Ugle-Hagan) did lead to a goal.

So we can select all the chains that end in a goal and count up how many of them included a Free For. The answer is:

In total then, 3 of the Dogs 9 goals came from a play that included a free kick: 2 in the first quarter and 1 in the second. The Swans - 2 of their 9 goals from a free kick: 1 in the first quarter, 1 in the third.

So the Dogs got 17 extra free kicks, but scored only one more goal from a free kick. This is why I'm always saying that the free kick differential just doesn't have a very big effect. If it did, we would see a strong relationship between the free kick differential and the score margin. Statistically, looking at lots of games, we don't see that - there is no essentially no relationship.

Now if you want to argue that missed free kicks should have been given which might have changed the result - and you probably do - well, that's a different issue to the differential.

Just to wrap up, some of you might be interested in the assessment from @hasumpstuffedup on Twitter. Overall he rates the umpiring for the game as "poor".

Has the umpire made A Bad decision? on Twitter: "FULL NOTES - Last night's controversial #AFLDogsSwans. Every FK decision analysed"

If the Swans get a flogging in the free kick department against the Fairypups the next time we play them.....we're still going to whinge.....and yes, the umpiring was poor. Beyond poor.

[mcs]

I almost hesitate to revisit the game, now that the thread has calmed down...but the R3 match chain data has arrived, so let's do it.

What is this data? A footy stats nerd's dream is what it is. A chain is every sequence of play on the field: what happened, when it happened, who did it, where they were positioned. Here's a small example showing the chain that led to the Dogs first goal:

You can see that this is an example where a free kick (to Ugle-Hagan) did lead to a goal.

So we can select all the chains that end in a goal and count up how many of them included a Free For. The answer is:

In total then, 3 of the Dogs 9 goals came from a play that included a free kick: 2 in the first quarter and 1 in the second. The Swans - 2 of their 9 goals from a free kick: 1 in the first quarter, 1 in the third.

So the Dogs got 17 extra free kicks, but scored only one more goal from a free kick. This is why I'm always saying that the free kick differential just doesn't have a very big effect. If it did, we would see a strong relationship between the free kick differential and the score margin. Statistically, looking at lots of games, we don't see that - there is no essentially no relationship.

Now if you want to argue that missed free kicks should have been given which might have changed the result - and you probably do - well, that's a different issue to the differential.

Just to wrap up, some of you might be interested in the assessment from @hasumpstuffedup on Twitter. Overall he rates the umpiring for the game as "poor".

Has the umpire made A Bad decision? on Twitter: "FULL NOTES - Last night's controversial #AFLDogsSwans. Every FK decision analysed"

Very interesting stuff NeilFWS - would love to have a year to properly comb the data across a whole season, especially when you say data has co-ordinates associated with it to. Would be fascinating to see what insights it would tell.

[Auntie.Gerald]

Many thanks Neil

Just to add to that analysis :


Two-thirds of all points usually come from turnovers in AFL.

Well that has been the stat's since around 2017 season.

Teams like Richmond 3 premierships last 5 seasons tended to also take the least marks so dont rely as heavily on forwards taking marks?

Using turnovers as a weapon was Richmonds success and they did it again in the 2020 GF with the least amount of marks in any GF since the 1960s.

The number of turnovers per AFL game has increased by 13 per cent over the past 5 seasons, from 126 per match in 2015 to an average of 143 season 2021.

No side was better at forcing its rivals into mistakes whilst minimising its own.

On average, the Tigers gave up the ball nine times fewer per match than their opponents in their 3 premiership years.

[TheBloods]

Gee some excuses in here . We played like tripe , nearly won , didn't deserve to . Move on

[Aprilbr]

If the Swans get a flogging in the free kick department against the Fairypups the next time we play them.....we're still going to whinge.....and yes, the umpiring was poor. Beyond poor.

Its amazing how many people I know that don't follow the Swans, but know that I'm a fanatic, have specifically mentioned the poor umpiring in this game to me even though I never raised it with them! Its uncanny, so clearly there was an issue although I agree with TB that we played like @@@@e on the night.

[neilfws]

Interesting work Neil. What about goals and behinds? Does that show anything different?

An excellent question, and I'm struggling a bit with the data to answer it. I think I've got the kicked behinds, but having trouble with the rushed. I'll report back when I understand it!

People might argue that an extra 6 points from frees (and maybe some behinds), equals a fair chunk of the margin of 11. And they might be right. I guess my point is that if the differential were very important, the margin should be a lot more. And were the frees that led to scores the "good ones" - those that were there, as judged by an impartial observer? A fan might say no frees against are good, or course

[Maltopia]

What the stats can’t really show is the opportunity cost. Our possession chains or attacks interrupted through bad calls. Of course not every chain or inside 50 leads to a score, but it is obvious some calls kill a drive or the momentum.

Conversely, uneven officiating where the other team is not called for HTB or incorrect disposal (when we get called for similar or less infringement) that then allows the other team to maintain their attack when it should have been stopped.

Not saying we deserved to win last week, but we might not have been so far behind early on and even gotten the lead a couple of times.

We were certainly outplayed and they had more energy and attack in the first half, but the officiating was just sickening and made the game so I enjoyable to watch.

[The Great One]

Perhaps when we next play this mob, we enquire as to whether Daryl Hair is available to umpire the game, At least he has the grapefruits to call players for throwing.

[Bexl]

Perhaps when we next play this mob, we enquire as to whether Daryl Hair is available to umpire the game, At least he has the grapefruits to call players for throwing.

????????????????YES. CAN NOT BE WORSE THEN WHAT IS CURRENTLY UMPIRING!!