![\"visual-follow-up\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F8523156%2F53e19f8ac67e8df5fde4133e5265b5ee%2Fheatmap.png?generation=1665608893211304&alt=media\")
The following plot shows the average ball position for individual events at different moments during the game. Each point in the heatmap represents the average position of the ball for a specific event in a ten-second period (not necessarily before scoring).
\nMy idea: Analyze how the ball's position change at different moments during an event.
\nHow did I make this plot? Here you can find the notebook with all the details.
\nTL;DR -
\nevent_time, I split the event into chunks of 10 seconds
The following plot shows the average ball position for individual events at different moments during the game. Each point in the heatmap represents the average position of the ball for a specific event in a ten-second period (not necessarily before scoring).
\nMy idea: Analyze how the ball's position change at different moments during an event.
\nHow did I make this plot? Here you can find the notebook with all the details.
\nTL;DR -
\nevent_time, I split the event into chunks of 10 seconds
I forgot to mention X and Y are inverted, so the scoring zone is at left and right (not up/bottom).
", "rawMarkdown": "I forgot to mention X and Y are inverted, so the scoring zone is at left and right (not up/bottom).", "votes": null }, { "id": "1985341", "postDate": "10/13/2022 08:59:41", "content": "Beautiful work. Would it be easy to animate this heatmap for team A or B only? (for the 10s leading to a goal)
\nI'd be interested to see if there is any obvious \"handedness\" bias to plays. (I've removed it and I suspect that I shouldn't have! :) ).
\nI looked at the probability surfaces - I'd be interested to see that kind of plot animated too :)
", "rawMarkdown": "Beautiful work. Would it be easy to animate this heatmap for team A or B only? (for the 10s leading to a goal)\n\nI'd be interested to see if there is any obvious \"handedness\" bias to plays. (I've removed it and I suspect that I shouldn't have! :) ).\n\nI looked at [the probability surfaces ](https://www.kaggle.com/code/paddykb/tps-2022-10-r-gam-probability-surface) - I'd be interested to see that kind of plot animated too :)", "votes": null }, { "id": "1987821", "postDate": "10/14/2022 20:51:59", "content": "Hi Patrick, that is a good question. I was using distances in my model but it seems to be more useful define sections.
\nMaybe this contour plot can help you. Using data (team scoring ==1) only, ignoring kick-off ( ball pos (x,y,x)=(0,0,0) )
\n![](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F10600930%2Fde5ed906f0b9c474fb8bb33ee7dadd1f%2FA_B_score.jpg?generation=1665780530578867&alt=media\")
",
"votes": null
},
{
"id": "1987894",
"postDate": "10/14/2022 22:00:25",
"content": "Amazing work @bryandarquea! When I saw my plots, I thought about using field zones too. I learned the Rocket League field has special sections where players can increase their booster and ultimately help them to make a power shot to score a goal.
", "rawMarkdown": "Amazing work @bryandarquea! When I saw my plots, I thought about using field zones too. I learned the Rocket League field has special sections where players can increase their booster and ultimately help them to make a power shot to score a goal.", "votes": null }, { "id": "1987909", "postDate": "10/14/2022 22:17:24", "content": "Hi @iancastillo, yes this is a good idea as players seem to follow the 'boost path' even when there is no scoring, but affecting their boost that seems to be a representative feature.
\nThis 2dHistogram of players position can help you, using events when Team A is scoring.
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F10600930%2F4050273f03e63eb15f707e9f34bd7b12%2Fplayers.jpg?generation=1665785770647771&alt=media\")
I forgot to mention X and Y are inverted, so the scoring zone is at left and right (not up/bottom).
", "votes": null, "replies": [] }, { "id": 1985341, "author_name": "paddykb", "author_url": "", "post_date": "10/13/2022 08:59:41", "content": "Beautiful work. Would it be easy to animate this heatmap for team A or B only? (for the 10s leading to a goal)
\nI'd be interested to see if there is any obvious \"handedness\" bias to plays. (I've removed it and I suspect that I shouldn't have! :) ).
\nI looked at the probability surfaces - I'd be interested to see that kind of plot animated too :)
", "votes": null, "replies": [ { "id": 1987821, "author_name": "bryandarquea", "author_url": "", "post_date": "10/14/2022 20:51:59", "content": "Hi Patrick, that is a good question. I was using distances in my model but it seems to be more useful define sections.
\nMaybe this contour plot can help you. Using data (team scoring ==1) only, ignoring kick-off ( ball pos (x,y,x)=(0,0,0) )
\n
Amazing work @bryandarquea! When I saw my plots, I thought about using field zones too. I learned the Rocket League field has special sections where players can increase their booster and ultimately help them to make a power shot to score a goal.
", "votes": null, "replies": [] }, { "id": 1987909, "author_name": "bryandarquea", "author_url": "", "post_date": "10/14/2022 22:17:24", "content": "Hi @iancastillo, yes this is a good idea as players seem to follow the 'boost path' even when there is no scoring, but affecting their boost that seems to be a representative feature.
\nThis 2dHistogram of players position can help you, using events when Team A is scoring.
\n
",
"1987894": "Amazing work @bryandarquea! When I saw my plots, I thought about using field zones too. I learned the Rocket League field has special sections where players can increase their booster and ultimately help them to make a power shot to score a goal.",
"1987909": "Hi @iancastillo, yes this is a good idea as players seem to follow the 'boost path' even when there is no scoring, but affecting their boost that seems to be a representative feature.\n\nThis 2dHistogram of players position can help you, using events when Team A is scoring.\n\n"
},
"source": "meta"
}