Jeremy Lin out 2 weeks with grade one knee sprain (UPDATE: Will return vs. Portland)

Torocan is Morey's tech team

 

meh. I hope I am interpreting this exchange correctly, butTorocan has a very good point. I personally never take offense if a person makes a statement with a prequel of "I believe" or "I think" or the like. They are stating an opinion, and everyone is entitled to one. However the bald statements based on conjecture presented as fact drive me up the proverbial wall. Those statements present opinion as fact, and as such should be actively discouraged by the community.

You are a very intelligent individual. On some level you agree. Or so I believe. LOL.

 

At least in the MMA the ref will jump in to protect the downed fighter. But here I'm beginning to get worried that Meh will suffer brain damage from the continual pounding. He shows a lot of heart in continuing to get up but he's going to get killed.

 

Reading Torocan's post just puts a smile on my face. If I were to have been blessed by a pope, I would imagine it to be this feeling.

 

Not to pile on, but what? Again, you seem like you're responding to other posts. My post was in direct response to YOUR direct response of a QUOTED torocan post. Please don't start drawing examples from elsewhere to muddle the issue.

Torocan was pointing out that there is a difference between "Lin is bad on defense right now off the bench" and "Lin would be better on defense if he were a starter, after his improved offseason." Statement of current situation vs. hypothesis. Torocan was pointing out the difference. I don't know what you're doing.

 

This thread is now about what should be meh's new signature.

 

So basically, after all these long diatribes that you need scientific research for any online statement(which I personally find ludicrous but whatever), you will not spend the time to actually do the scientific research. Yet you will criticize others for also not doing it.

Funny how the above diatribe against eFG% didn't appear when I first made my Beverley's opponent eFG% post. Why did you instead use the sample size argument?

Is $30 tipjar not good? Fine. I'll use paypal and put the money directly into your account. How about that? I'll pay you to do the research for me. Prove to me that Beverley's sample size so far is statistically insignifcant.

 

Hoping torocan would do what he preaches. He says my data from 82games is statistically insignificant and I shouldn't post stuff that are insignificant. I wish for him to prove it.

 

Torocan. Your post did not go into the mechanics of verifying sample size significance. Which I completely support given that it would be addressed to an audience of about four, with Morey being one of them.

I believe that the more important point is that, for this season, we are now getting to a sample size that can be used to discuss stats like drives, TS%, isolation defense, shot selection, etc. However stats based on PM like xRAPM, ASPM, and lineups have to be taken with a pound of salt.

Unassailable sample size studies depend on terabytes of data. Unfortunately NBA seasons only yield megabytes of data, two orders of magnitude less than unassailable. This is the life we choose to live in.

 

This will be fun:grin:

 

*sigh*

Writing the basics of calculating and explaining the meaning of sample size, standard deviation and confidence level is the equivalent of writing a significant portion of a college level Statistics 101 class.

In University/College, it usually consumes multiple days of lecture to cover Sample sizing, standard deviation and confidence level. It will take some time to explain the mathematics and precise reasoning of why the sample size is insufficient. So, basically I will be taking the equivalent of several days of lecture material, condensing it and simplifying it into a single post.

And the Tip Jar comment was because I don't want an online discussion to cost *anyone* REAL LIFE MONEY that you can put towards things like a Rockets cap/Jersey or a CSN Cable bill. Heck, even going out and taking the wife to a movie so you can watch more Rocket games in peace is money better well spent than losing it over an online debate.

The reason I made the sample size argument is I didn't want to write a long treatise on the failings of eFG% as I wrote above, and the sample size was *OBVIOUSLY* too small to anyone who has the most basic background in statistics and statistical analysis.

We have barely hit 20 games in the season, of which Beverley missed the entire first TWO weeks... a usual sample size demands 30-50 base samples unless you're arguing extraordinary statistical variance (more 3 sigma or more) and/or have corroborative statistical evidence/models to add to your argument.

Seriously man... turn the rage machine off.

 

I have no idea why I kept up the exchange. Probably because of the Rockets 3 days off and slow work days. Gotta kill time somehow. lol

It really just started because I posted a random Beverley opponent eFG% and somehow out of nowhere got called for my lack of "sample size" for using that stat. Which felt kind of weird because GARM has generally been all about small sample size overeactions. :grin:

 

If you post the 82games data I will attempt to do a statistical significance study using Minitab. But I am not guaranteeing anything. PhD statisticians have knockdown fights about that subject. And I certainly am not in their league.

 

Well, if your idea of two weeks is 7 days, I can see how you may have trouble with math. And not wish to do it.

And FWIW, I'm not raging. If I were, I'd be writing essays instead of stupid one-liners.

 

Hey meh, how about you proving that the sample size is statistically significant? That would show you are right and end this debate.

 

God you fake fans are annoying.

 

lmao. As someone who has been watching MMA since '06; no, just no.

 

Aren't the mods supposed to step in when they see a wounded-animal thread like this and just shoot it? Put it out of our misery?

 

Ok. Now I'm being totally serious here. But it would be kind of neat to see how to go about solving this. Unfortunately it's been a while since my last math class, so I don't really know the answer myself. I actually would like to know the answer for future references.

Basically the set up is the following.

According to 82games.com, Beverley this year has held opposing PGs to .430 eFG%. The question is, is this number statistically significant? The bar to judge this against would be .494 eFG%, which is league average. Perhaps not the most reliable bar(if someone has the average PG eFG% it would be better). But best from public data I know.

You can use 82games.com to find several stats. It does have Beverley's minutes played, as well as opponent FG attempts. Which you can extrapolate to the total number of attempts. Unfortunately, from there I'm not very sure what's the best next step.

Of course, there's no need for you to actually do this unless you really want to. I don't want to burden you for something that I myself am too lazy to do myself. But to me it is an interesting question. And I'll donate to the tip jar too. That was more me wishing to know the answer to a question more than anything else.

 

I'm not the one who made the claim that my stats were statistically significant. Someone else said it was insignificant.