2018 Official Offseason Thread

[Odin 137]

I was trying to figure this out. What do you think qualifies as an adequate sample size for line ups? I was using 50 minutes, but in reality, that is only a little over a game's worth of data.

[HeHateMe 697]

24 minutes ago, ?4thewin said:

The post all star numbers are kind of too noisey to use. Augustin-Simmons-Fournier-Gordon-vucevic and Augustin-Mack-Mario-Gordon-vuc were elite level offenses. But there's a lot of iwundu/Artis/purvis minutes that poison the data. 

I’m just going off the eye test. Admittedly Nothing factual. I just remember nearly every game post all star break thinking “what the *****” is AG doing? Selfish shots. Bad shots. Lack of passing. It was just annoying to me. But I think the team completely folded too. 

[Fultz4thewin 2,464]

1 minute ago, Odin said:

I was trying to figure this out. What do you think qualifies as an adequate sample size for line ups? I was using 50 minutes, but in reality, that is only a little over a game's worth of data.

It kind of depends on what you're trying to point out. For lineup data there's 3,936 total minutes in a season. So a 20% sample of that is 787 minutes. But there's only 3 lineups that exceeded that benchmark last year. 

I usually use 100 minutes to observe trends in major lineups because thats usually at least 6 or 7 games and that's when enough for outliers to quiet down. For minor lineups you might be able to take something from 50 minutes but you really need to drill down into the data to see where it came from. Did Birch enter the game down 35 at the start of the fourth quarter and cut the lead to 27 against scrubs? That's going to be noisy. 

[Odin 137]

1 minute ago, ?4thewin said:

It kind of depends on what you're trying to point out. For lineup data there's 3,936 total minutes in a season. So a 20% sample of that is 787 minutes. But there's only 3 lineups that exceeded that benchmark last year. 

I usually use 100 minutes to observe trends in major lineups because thats usually at least 6 or 7 games and that's when enough for outliers to quiet down. For minor lineups you might be able to take something from 50 minutes but you really need to drill down into the data to see where it came from. Did Birch enter the game down 35 at the start of the fourth quarter and cut the lead to 27 against scrubs? That's going to be noisy. 

That sounds fair.

[Murphy13 185]

57 minutes ago, Odin said:

I am confused on your argument now. Are you now arguing that Gordon was detrimental to our offensive production as well?

It means he is not a productive offensive player right now, he is not currently good at offense.  Thus his impact on our scoring is detrimental for the most part, whereas Isaac’s lack of offense isn’t detrimental due to its minimal impact on  available possessions.

Its less about AG and more about Isaac.  

[Murphy13 185]

To be clear, I wasn’t stating that our offense is worse with AG necessarily, just that there’s more wasted offensive possessions when he’s featured.  This is particularly evident post ASB but our team as a whole caved by that point so it’s difficult to apply external validity to what would otherwise be compelling.

[Murphy13 185]

17 minutes ago, ?4thewin said:

It kind of depends on what you're trying to point out. For lineup data there's 3,936 total minutes in a season. So a 20% sample of that is 787 minutes. But there's only 3 lineups that exceeded that benchmark last year. 

I usually use 100 minutes to observe trends in major lineups because thats usually at least 6 or 7 games and that's when enough for outliers to quiet down. For minor lineups you might be able to take something from 50 minutes but you really need to drill down into the data to see where it came from. Did Birch enter the game down 35 at the start of the fourth quarter and cut the lead to 27 against scrubs? That's going to be noisy. 

That’s a good baseline explanation but to reach true statistical significance it would probably require far more than 100 minute lineups.  Generally you would have to include minute intervals that cover 95% of available minutes in order to be powered appropriately.

[Nyce_1 311]

is there any way we can get Fultz?

[Fultz4thewin 2,464]

The more I do the math, the more I think we need to keep Fournier.

And the more I think FVV should be the preferred option. He's a fantastic shooter and I think that matters with our lineup

[Malik Buttface 57]

2 minutes ago, Nyce_1 said:

is there any way we can get Fultz?

Depends on their price for him, I'd be interested in him for sure. The 6ers are in a convenient place, they can ultimately hold onto him and see if he can shake off whatever it is that he went through last season or they can dangle him in a trade because he's kind of the odd man out.

[A Better DJ & Photographer 456]

6 minutes ago, Nyce_1 said:

is there any way we can get Fultz?

Philly won't trade him if he is good again and we don't want him if he's not. No more PGs that can't shoot please!

[Odin 137]

4 minutes ago, ?4thewin said:

The more I do the math, the more I think we need to keep Fournier.

And the more I think FVV should be the preferred option. He's a fantastic shooter and I think that matters with our lineup

I don't know if we need to keep Fournier, but I struggle to think of what would be an adequate return that is actually feasible, unless he is part of a package.

I think I am in your camp with FVV if we want to be trying to get a potential starter for the future.