Fancy Stats for 2023

[Deleted member 2543]

Round 21 vs Manly




... and another Kaufusi update.

Boring, but consistent.

 

[bort]

Could you do two colours of Kaufusi dots - start and bench?

 

[Deleted member 2543]

Could you do two colours of Kaufusi dots - start and bench?

Yeah - was planning on doing that for everyone for their full season graph at the end. I’ll do it for the next Kaufusi update.

 

[Sparkles]

The GPT analysis of the analysis we never knew we needed. Looking forward to point 3 in the future, @BurgoShark

The forum thread titled "Fancy Stats for 2023" on "sharksforever.com" provides a detailed analysis of player statistics for the 2023 NRL season. Here's a summary of the content:

---

Thread Starter: BurgoShark Start Date: Sep 14, 2023

Main Points:

1. Refined Model: BurgoShark has refined his model for player involvement to reflect where runs take place rather than just how many meters they ran. The model now includes a small weighting for positive and negative plays in both attack and defense.
2. Data Sources: The model uses nrl.com stats for "tackles" and "missed tackles." All other data, except tackles, is captured manually by BurgoShark.
3. Off-Season Project: BurgoShark plans to start with round 1 and work through the season, aiming to represent all game data in one graphic.
4. Interpreting the Data: BurgoShark provides a guide on how to interpret the information, using Oregon Kaufusi's Round 1 performance as an example. The data includes metrics like "Attack vs Defence," "Involvement vs Ball in Play," and "Attacking Contribution."
5. Yardage Runs Pie Charts: These charts provide a cumulative representation of where players are taking runs on the field.

User Reactions and Questions:

• Tatus: Asks about differentiating between a carry 40m out or 15m out from the opposition try line.
• apezza: Comments on the model's reliance on nrl.com stats and appreciates the effort put into the analysis.
• Jaz: Seeks a summary of key insights for those who might find the graphs challenging to interpret. Also inquires about the data collection process and how it might be used for SuperCoach analysis.
• Sparkles: Thinks about how the data can be used for SuperCoach next year. Suggests identifying trending rises in involvement over a few weeks and predicting when a player returning from injury might hit optimal performance.

---

Analysis: The thread offers a deep dive into player statistics, with BurgoShark's model aiming to provide a more nuanced understanding of player involvement in games. The community is engaged, asking questions, and suggesting potential applications for the data, especially in the context of SuperCoach.

Future Suggestions Based on Poster Feedback:

1. Expand Data Collection: Given the interest in the data's application for SuperCoach, there might be value in expanding the data collection to include more metrics or even collaborating with other data enthusiasts.
2. Interactive Platform: An interactive platform or app where users can play with the data, make predictions, and see trends might be beneficial.
3. Workshops or Tutorials: Given the complexity of the data and the graphs, organizing workshops or tutorials to help users understand and interpret the data could be beneficial.

 

[Deleted member 2543]

The GPT analysis of the analysis we never knew we needed. Looking forward to point 3 in the future, @BurgoShark

The forum thread titled "Fancy Stats for 2023" on "sharksforever.com" provides a detailed analysis of player statistics for the 2023 NRL season. Here's a summary of the content:

---

Thread Starter: BurgoShark Start Date: Sep 14, 2023

Main Points:

1. Refined Model: BurgoShark has refined his model for player involvement to reflect where runs take place rather than just how many meters they ran. The model now includes a small weighting for positive and negative plays in both attack and defense.
2. Data Sources: The model uses nrl.com stats for "tackles" and "missed tackles." All other data, except tackles, is captured manually by BurgoShark.
3. Off-Season Project: BurgoShark plans to start with round 1 and work through the season, aiming to represent all game data in one graphic.
4. Interpreting the Data: BurgoShark provides a guide on how to interpret the information, using Oregon Kaufusi's Round 1 performance as an example. The data includes metrics like "Attack vs Defence," "Involvement vs Ball in Play," and "Attacking Contribution."
5. Yardage Runs Pie Charts: These charts provide a cumulative representation of where players are taking runs on the field.

User Reactions and Questions:

• Tatus: Asks about differentiating between a carry 40m out or 15m out from the opposition try line.
• apezza: Comments on the model's reliance on nrl.com stats and appreciates the effort put into the analysis.
• Jaz: Seeks a summary of key insights for those who might find the graphs challenging to interpret. Also inquires about the data collection process and how it might be used for SuperCoach analysis.
• Sparkles: Thinks about how the data can be used for SuperCoach next year. Suggests identifying trending rises in involvement over a few weeks and predicting when a player returning from injury might hit optimal performance.

---

Analysis: The thread offers a deep dive into player statistics, with BurgoShark's model aiming to provide a more nuanced understanding of player involvement in games. The community is engaged, asking questions, and suggesting potential applications for the data, especially in the context of SuperCoach.

Future Suggestions Based on Poster Feedback:

1. Expand Data Collection: Given the interest in the data's application for SuperCoach, there might be value in expanding the data collection to include more metrics or even collaborating with other data enthusiasts.
2. Interactive Platform: An interactive platform or app where users can play with the data, make predictions, and see trends might be beneficial.
3. Workshops or Tutorials: Given the complexity of the data and the graphs, organizing workshops or tutorials to help users understand and interpret the data could be beneficial.

Any time mate

 

[Deleted member 2543]

@bort I'll make them all look similar to this.

Blue for a starter. Black for a benchie.

 

[Deleted member 2543]

@Sparkles it's a shame Tricky isn't on the graphs. He could have joined in to the tutorial as part of his fan forum engagement.

 

[Proud Shark]

Any time mate

I'm more looking forward to your detailed analysis of how it pertains to supercoach

 

[Sparkles]

I'm more looking forward to your detailed analysis of how it pertains to supercoach

Chat GPT has spotlighted that as one of the more important takeaways, so I'm sure we'll see it for the new year

 

[Sevshark]

Not a fancy stat but was having a look at NSW cup stats leaders today and noticed that the Puru brother's managed to finish 1st and 2nd for missed tackles this year.

That has to be a first surely haha

 

[Deleted member 2543]

Round 3 vs Raiders
- Probably the least attacking involvement of the season for Brailey
- Very high involvement inside 20 for the back 5
- Big involvement from both second rowers this game. Wilton played 80 and is up there with the some of the middles.
- Quieter game from McInnes
- Rudolf and Finucane consistent as usual
- Bigger game from Kaufusi (highest involvement of the games I've looked at so far)

 

[Deleted member 2543]

... and another running Kaufusi update.

 

[Deleted member 2543]

Round 15 vs Storm
- Finucane and Williams the stand-outs here in terms of involvement
- Talakai a total workhorse as usual



... and another Kaufusi update.

With 7 / 25 games done now, I've added a season average to each graph.

 

[Deleted member 2543]

With almost 1/3 of games into the data set, I thought I'd throw up some very early analysis of the outside backs.

Remember that with all of this, it is a measurement of involvement. It says nothing about performance/impact.

This is data from 7 games (Rd 1, 2, 3, 15, 16, 21 and FW1). Tracey only played in 2 of them. Everyone else played at least 6.

Yardage contribution average for all outside backs.

150 is considered a solid level of involvement.

A few observations here:
- Outside of kick returns Kennedy is much less involved then the rest
- Talakai on average will do ~25% more work than everyone else, with all of that work coming inside his own 20
- Possibly a slight trend here for longer kicks going to Mulitalo's side of the field. Katoa still fields plenty of kicks, but maybe he is getting more shorter/contested ones which don't give him kick return metres.

Talakai takes twice as many runs inside 40 as any other Sharks outside back. Are oppositions deliberately kicking to that side to make Talakai do more work in his own end, is Talakai just really hungry for work, or does Fitz prefer Sifa getting back and taking this run no matter what?



Most Involvement from Outside Backs

Interestingly, 4 of the busiest performances came in Rd1 (Rabbitohs) and Rd15 (Storm). This tells me that these two teams did a very good job of pinning the Sharks into corners, and that while the Sharks backs were very heavily involved, they may not have moved the ball as effectively this game.

Those two big blue sections for Talakai are pretty crazy. A score of 200 in blue says that Talakai did the equivalent attacking work of a 40-50 minute middle forward (like Rudolf, McInnes or Finucane), but for 80 minutes straight, and all inside his own 20m line. Obviously not having the same workload as a middle forward helps here.

 

[Sparkles]

Are oppositions deliberately kicking to that side to make Talakai do more work in his own end, or is Talakai just really hungry for work?

I'll throw in a third option that it could just be his assignment. With how effective he is I wouldn't be surprised.

Those two big blue sections for Talakai are pretty crazy. A score of 200 in blue says that Talakai did the equivalent work of a starting middle forward, for 80 minutes straight, all inside his own 20m line.

So skip the second row altogether? Bold!

 

[Deleted member 2543]

I'll throw in a third option that it could just be his assignment. With how effective he is I wouldn't be surprised.

Yeah - I actually thought of that after clicking "post" and added this...

or does Fitz prefer Sifa getting back and taking this run no matter what?

So skip the second row altogether? Bold!

Yep. He either succeeds as a centre or he goes back to being a bench middle.

He's shown in the past that as a middle he can be an attacking force, but he doesn't have the defensive work rate to play long minutes. He could be effective in the role played by the other bench middles (high energy, get behind the ball early, run as often as possible, but never play more than 15 minutes straight) while also having a bit of utility value to cover injuries.

 

[Sparkles]

Yeah - I actually thought of that after clicking "post" and added this...

or does Fitz prefer Sifa getting back and taking this run no matter what?


Yep. He either succeeds as a centre or he goes back to being a bench middle.

He's shown in the past that as a middle he can be an attacking force, but he doesn't have the defensive work rate to play long minutes. He could be effective in the role played by the other bench middles (high energy, get behind the ball early, run as often as possible, but never play more than 15 minutes straight) while also having a bit of utility value to cover injuries.

I guess the workrate as centre suits his capacity. I do think if there's a sniff of an exciting centre that can hold their own defensively that we'll see him back on the bench. He could be part of our version of a dynamic forwards dou, Panffers style.
I have no idea who the Batman/Robin would be in this analogy

 

[bort]

- Possibly a slight trend here for longer kicks going to Mulitalo's side of the field. Katoa still fields plenty of kicks, but maybe he is getting more shorter/contested ones which don't give him kick return metres.

Anecdotally I'd say we do see a slight trend in long kicks to Ronnie (slightly more likely to **** up) and more contestable kicks to Sione (shorter, not previously known for his aerial prowess).

I say previously because he had some cracking contested catches late in the year.

Yep. He either succeeds as a centre or he goes back to being a bench middle.

He's shown in the past that as a middle he can be an attacking force, but he doesn't have the defensive work rate to play long minutes. He could be effective in the role played by the other bench middles (high energy, get behind the ball early, run as often as possible, but never play more than 15 minutes straight) while also having a bit of utility value to cover injuries.

Leniu style impact only prop - you can easily carry one in our side with guys like Dale and McInnes fit, depending who else is also fit I guess

 

[Deleted member 2543]

I guess the workrate as centre suits his capacity. I do think if there's a sniff of an exciting centre that can hold their own defensively that we'll see him back on the bench. He could be part of our version of a dynamic forwards dou, Panffers style.
I have no idea who the Batman/Robin would be in this analogy

I guess the question is whether there is a player who can do that work rate stuff at a similar level, while also being an improvement elsewhere. A lot of the options seem like robbing Peter to pay Paul.

 

[Deleted member 2543]

Leniu style impact only prop - you can easily carry one in our side with guys like Dale and McInnes fit, depending who else is also fit I guess

This is what we have been getting from Hunt and Hazelton. Low minutes with a very high level of involvement, and a willingness to come back inside their own end rather than wait for the ball. It's obvious they are under different instructions. Hunt says his job is to "fly off the bench".

I think we know what we get with Hunt and Williams, but if one of Hazelton, Colquhoun or Kaufusi can step into a bigger role, imo Sifa would be a very good option to take over one of those two low-minute bench impact spots.