[ClutchFans] Rockets to begin season with Beverley, Asik as starters; Jeremy Lin off bench

I'm a bit lost here.

May I ask again why exactly did Beverley start and Lin being benched?

 

*sigh*

Once again, you completely miss the point.

The reason why the +- differential holds *some* value is because the differential is SO large, and the complementary data *also* points in that direction.

If the differential was smaller, and did NOT across 2 different line ups (Casspi + Garcia), then you could easily consider it highly likely that it falls within the margin of small sample size/variation.

However, the differential was NOT small... it's Very, very large. And Very large across BOTH line ups. And secondary data seems to support that the data is at least valid *in part*

And THIS was the secondary data I was referring to at the time...

So, look at the line up differentials. They tend to fall in small ranges. 2, 3, 5, etc.

What are the odds of a single variation over 1 player being 10? What about 15?
What are the odds of it happening over TWO different line ups?
What are the odds of that difference being purely coincidental or incorrect when Lin generates more Assists per minute?
What are the odds of that difference being purely coincidental or incorrect when Lin is also shooting better?
What are the odds of that difference being purely coincidental or incorrect generates more of their own shots versus being assisted?

This is what I mean by multiple data sets all pointing in the same direction. It's a very simple notion when you have multiple data sets and models.

Let's take a hypothetical.

Player A and Player B is identical.

Player A now shoots better. Should the line up improve?
Player A now creates more assists. Should the line up improve?
Player A now can generate their own offense. Should the line up improve?

Now consider that within the context of +- line ups.

Make the sample size small, but create a very large variance (well outside of normal 1 or 2 sigma). Does the probability of the variance being statistically indicative or correlative increase as the variance increase?

Now do it across 2 different line ups, but create a very large variance (well outside of normal 1 or 2 sigma). Does the probability of the variance being statistically indicative or correlative increase when you see similar impact in a 2nd line up?

In other words, how many coincidences can you have before it becomes more than a coincidence?

This is what you're missing. The fact that the difference is Monstrous makes the sample more meaningful. That it happens across 2 line ups with similar player styles substituted in 1 position makes it more meaningful (Casspi vs Garcia, 3 point perimeter shooters).

That other statistics point towards the data being meaningful improves the argument for it's validity on some level.

Is it 100% accurate? Of course not. No statistical interpretation is 100% accurate or meaningful, and as any statistical interpretation it's merely a "snapshot" in time.

Could this all be moot with TJones now logging minutes? Possibly.

Sorry for the length. I wasn't intending to talk so much about statistical interpretation, however that is what I meant by taking a single data set and not looking at it within statistical context.

Anyway, one more set of statistics to add the the context.

Jeremy Lin
117 ORtg, 108 DRtg, 1.2 OWS, 0.2 DWS, 1.3 WS, 1.57 WS/48

Patrick Beverley
112 ORtg, 108 DRtg, 0.4 OWS, 0.2 DWS, 0.6 WS, 0.99 WS/48

Of course, they're all just single stats and data points and the sample size is small. It's all just purely coincidence...

 

Do you go to sleep thinking about me?

Try this post again in English btw. It's painful to read

 

torocan, you are definitely one of the most positive presence in this forum. Appreciate your effort on all the analysis!

 

Torocan is worthless. He's a Knicks fan.

 

Pfft. Now you're going to criticize me for NOT being a band wagon fan? Like it's MY fault I grew up with the Knicks...

If anything you should feel sorry for me for having to deal with the POS that's been the Knicks organization for multiple decades. You Rockets fans have had it easy... at least you've gotten a championship in living memory.

Knicks fans have to live on the memories of the Ewing play offs.

Not only tasteless, but low class. I would never wish the Knicks' record on any fan base... not even my worst enemy.

 

Nah. If you're scoring at home I'm actually tired of the whole back and forth. You're on a Rockets site dude. Roll your eyes all you want; (write three paragraphs trying to explain you obvious perspective) it's all barely worthy of mention.

Don't care.

You're intelligent. Do you realize the stuff you post is only does damage to Jeremy? The dude is a consummate professional; playing great as a 6th man, and can be a part of a championship team. Who cares if he starts? There's a ton of people going back and forth about one bad game when he has kicked ass this season. That falls soley on the shoulders of LOFs.

Wake up man.

 

Agreed. His posts are always interesting and informative regardless if you agree with him or not.

Torocan, don't let the TOH (Torocan Only Haters) get to you. Keep posting!

 

Thank you, Torocan!

 

Why is everybody all dick sucking? Sorry to be crude, but that's what it looks like...

 

There is a ton of stuff going back and forth about the one bad game. I didn't even need to log onto clutchfans to know it would be happening. There is a subset of JLin fans who see conspiracies everywhere they look. I look on that with a sort of bewildered amusement, but I understand being pissed off about it too.

As to the question of who cares who starts, it clearly does matter. If it didn't, you wouldn't have a consistent set of starters on most teams for every single game. It is a perfectly valid and worthwhile debate to question whether Beverley or Lin should start. Lin brings a valuable offensive skillset, and whether it can be used best on the bench or with the starting group is clearly a reasonable question to ask.

The ultimate result you want is a larger point differential at the end of the game. Is Lin off the bench right for that? There is no way to prove it one way or the other, but you can at least look at real evidence to try and figure it out rather than just saying it is so. That people can actually argue about the numbers he puts up is miles better than people just yelling at each other.

And lastly, I would hope that stuff posted on internet has no particular impact on JLin at all.

 

torocan catching bodies every damn day

 

Looks like it's game on here.

Torocan vs Larsv8

 

Oh larsv8 wins hands down...



Only because he's anti Lin. They get special privileges here. Look at all the high rep points Lin haters get

 

You are so full of **** it's not even funny. For god's sake, get an enema before you explode.

 

There's one thing missing, and I'm not going to get into Lin starting or not and that is Lin spending time playing with Dwight Howard and developing a chemistry. Even if Dwight does post moves he still needs to do some PnR with Dwight, Lin is just as good at lobbing passes to a big man as anyone on the team, he just needs to get the chemistry with Dwight and from what I'm picking up, Dwight wants to play with Lin and get chemistry. So somewhere in McHale's minute combination he needs to figure out how to have Lin and Dwight play together and also he needs to work that into practice.

 

Thus wrote Friedman

But Justaguy, as a fellow member of the Friedman fan club, the question is: Can stuff posted on the Internet have an impact on Jason Friedman? Did you notice how D12BabyMamas (the head of the club) quoted Nietzsche, and later in the same day Friedman mentioned Nietzsche in his game story! It's almost as if he were responding to his fans. I like to think he was wearing his red Nikes as he filed the story.