Jeremy Lin stats against west top 9 this season

Some aspects of NBA "data" are not as cut and dried as making or missing a shot. Assist, steals, and turnovers are subject to some interpretation by official scorers. If two people try to grab an egg off the table, and it gets broken, there may be some dispute over who's fault it is.

 

Correct, it is subject to human interpretation at the point of the scorers table, however since the definition of a steal, assist or turnover is fundamentally arbitrary and only exists once recorded by an official, then the issue of it being subjective is really irrelevant.

Think of it like Schroedinger's cat.

The data is the number of assists. Assists are defined in the rulebook. The official determines whether it is an assist. At the point of a received pass and score, it is both an assist and not an assist until the official makes the decision (observes). So until it is actually recorded, it is not actually data. It is merely a series of events with the potentiality of becoming an assist or not becoming an assist.

And then you just hope it all averages out over time. :grin:

 

You didn't say it. But it was said.

 

For the record, what you said exactly was this...

 

I think per36 numbers can help with comparison, but they need to be paired with actual numbers to provide context.

For example, Troy Daniels is the Rockets leading scorer per36, and Robert Covington is third.

 

Agreed. This is why it's traditional to provide game averages + per 36/48, as well as individual games if possible. With very small sample sizes (< 10 games) you want to provide as much context as possible otherwise it's not only statistically of minimal value, but highly misleading.

You would never provide per game average without per 36/48 in a serious discussion simply due to how much uneven minutes can distort those numbers, especially over smaller samples (< 1/2 - 2/3 of a season).

In very small samples like 4-6 games, you would provide game logs in addition to averages + per36 were you so inclined.

And to provide a game average and game statistics without actual shot counts and breakdowns is egregious as there's no way to confirm the data without actually looking it up yourself.

It's like saying, 50% FG% without noting that they shot 1/2, or 33% FG% without noting that they shot 1/3.

 

It's not irrelevant if you plan on using the data rather than just watching how prettily it sparkles in the sunlight.

Think of it like knowing you have bad sensors recording your physics data, and futhermore you don't even know which sensor a given piece of data came from.

(I hate bad data. Makes everything harder).

Hoping it averages out over time. A fine one line definition of statistics, for certain values of the word statistics.

 

Never said I liked it, I just force myself to rationalize it away otherwise it would drive me nuts.

 

toro how do u have the patience to entertain all these ppl

 

You forgot to mention you also left out SHOT ATTEMPTS. FG/FGA.

You presented a series of FG%, but gave us no idea how much he shot during those games. If he had taken NO shots, would you have considered putting in 0/0 irrelevant as well? What if he had only taken 1 shot and shot 1/1?

So you posted FG%, but didn't include FG, FGA, 3p, 3pa, FT, FTA, Minutes, blocks, steals or rebounds which you "determined not directly relevant to the comparison I want to make" which was what exactly? That you could copy and paste part of a box score?

Your thread title is "Jeremy Lin stats against west top 9 this season". Yet you omit most of the traditional box score stats, and omit numerous stats relevant to the stats you did post.

For what purpose? Other than to create a thread that you admit to knowing would incite part of the CF poster base?

Sorry, but your purposeful omission of not only traditional statistics but relevant box score statistics sure doesn't seem like you're trying to construct a meaningful comparison...

 

You did not post that until I called you out on it.

And instead of immediately posting it, you claimed I was putting forth an agenda.

Had you just said, "My bad." and filled out the missing data or offered a rational explanation for why you were excluding that data (I left it out because I wanted to examine and compare these items specifically...", it would have stopped right there, instead you tried to argue how my analysis was attempting to make Lin look better instead of understanding what I was trying to explain to you.

This is why we end up in these back and forths. You post something, I try to explain why your post is inconclusive and misleading, and you try to frame it in a narrative instead of addressing the criticism of your presented data and analysis.

If I post an analysis and someone criticizes the methodology, I either explain my reasoning behind my use of that assumptions, or I say that I didn't see my oversight and thank them for their observation.

That's called peer review.
That's also called a discussion.

The real question is, why did You initiate this thread in the first place? Were you trying to create a meaningful discussion? To make a comparison? To demonstrate a stated point? To raise questions?

You certainly didn't provide any particular conclusions or raise any questions...

 

Did you even read my post? Maybe my post was too long, so I'll break it down for you again, bullet point style.

- Per 36 numbers over a sample are more reliable than average box score stats over that same sample if a player plays with uneven minutes
- Lin probably played with uneven minutes in those 9 games (last time I checked he had the highest mpg variance out of all Rockets besides bench players)
- Example, Lin played 44 min in one Spurs game, 33 min in another Spurs game
- Hence, per 36 numbers more accurately reflect Lin's performance

- You said because Lin's average mpg for those 9 games is similar to his season average, you left out minutes entirely, or per 36 is irrelevant
- Minutes can still vary wildly in those 9 games even if they produce an average that is similar to the season average
- You presented vs team stats, averaging the box score over a sample of minutes that may be uneven. For example, the vs Spurs stats, as I had already mentioned earlier, Lin played 33 min in one game, and 44 min in another. You averaged those box scores in your original post without taking into account the variance in minutes. That is relevant.

- torocan's per 36 example for Lin's stats vs Spurs showed that his scoring numbers actually DECREASED by this metric. So how is that in any way an attempt to "favor" Lin?
- Factoring in minutes is necessary IN GENERAL, especially when a player has uneven minutes (for reference, see first set of bullet points)

Huh? I already said Torocan posted Lin's per 36 stats for the Spurs, and it did not "favor" him. And noting that "there were games that he played much more minutes and perform more and games he played much less with poor stats" was my explanation for why per 36 should be used. So please explain to me how that is an attempt to "look from another angle", or trying to "switch" to another new procedure?

Direct quote from you:

"See?

That's exactly the response I expected from a LOF, and that's exactly the reason I left out the minutes of play and FT % and waiting for your feedback."

Maybe you can clarify... because the above seems to be saying you expected a certain response from an LOF, which is the exact reason why you left out the information. If provocation was not the goal, then what specifically was the "[exact] reason [you] left out" that information? And stay within the quote please.

 

Its clear now. LOFs and LOHs need to get a room! You guys obsess over Lin and each other way too much.

 

Texanasiafan directly accused me of saying he was "hiding facts". Still waiting on him to find that magical quote.

That's true. One thing I'm trying to figure out is at what point does per 36 become unreliable (i.e., how low do the minutes played have to be before it begins to mess things up)

 

You guys complain about a player's personal life only when they don't play well on the court.

 

You think i give two craps about a players personal life?? Unless it affects how he plays on the court i could care less...iam a Houston Rocket fan and anything i say here has everything to do with Basketball and ONLY basketball....

I just stated my opinion and said the reasons behind it, it has nothing to do with his personal life, i would trade lin straight up for any starting pg in the nba right now except for chalmers, felton,marshall, and whatever milwaukee is throwing out there.
The main reason for us keeping him after the trade deadline, i dont know chemistry?? or no takers??...but if the summer ends and hes still on this team then MARKETING is the only reason in my eyes ....and that has nothing to do with lins personal life, because thats the best thing he has going for him

 

That's not a very good analogy. In the case of Schrodinger's cat, it's the act of looking that causes the cat to be either dead or alive rather than a superposition of the two possibilities. It the case of assists, given the definition in the rule book and the actual event: a received pass and score, what's uncertain is whether the official will make a good decision or a bad one, not whether there was an assist or not.