So you're saying we have a great chance to get to the Western Conference Finals with Jeremy Lin as our starting PG? ..I'll take it!!!
Who said he's great? I've maintained all season that Lin at this point of his development is mediocre. He has skill sets that are very good, some that are average, and some that are below average. And clearly I'm pushing the Rockets agenda the way I call out Harden's defense, Dwight's tendency to turn over the ball in the post, pointed out Asik's excellent defensive footwork, and my personal view that McHale is an average coach. And stats don't "mask" what the eyeballs see, what they do is re-affirm what the eyes see, or show things that the eyes do NOT see. For example, how many posts have there been over the past 2 years saying "Lin's slow" or "Not quick". During one of his earliest games on ESPN, Brent Barry said of Lin during one of his games, "It's not like he's fast and can just blow by guys..." The eye is deceptive. Our memory is deceptive. We distort our memories to fit our preconceptions and self-constructed narratives. The idea that human observation and memory is anywhere near reliable is a lie that has been perpetuated since time immemorial. Look at studies of criminal eye witnesses, or studies of the ability of humans to focus on multiple events. or studies of accuracy when human recall of detail is tested. The human mind is NOT a computer. It is not precise. And since we typically analyze based upon our memories, fundamentally flawed. Numbers are cold and unforgiving. Data is cold and unforgiving. Do we sometimes look at the wrong numbers? Sure. However that's not the fault of the numbers. That's OUR fault for not fully understanding the nature of the numbers that lie before us. Numbers and statistics are tools to help us examine and understand what is happening when our eyes fail us. To call them bad is like saying a hammer is bad because one person uses it to hit you over the head instead of hitting a nail. No, numbers and statistics are not the definitive truth. They are however useful in helping us expose the lies that we tell to ourselves.
Nope. I think our chances of getting to the WCF is low regardless of who the starting PG is going to be. And our chances drop dramatically if Lin OR Beverley are out. Canaan can't replace either of their skill sets. In terms of a per match up basis, that would be debatable, however that's a subject for an entirely different kind of thread.
Hi DD, I am curious what you mean specifically when you say "skewing stats"? Torocan, and a few others are simply making an argument that TS% and Per 36 stats provide a clearer picture than just FG% and PPG. I think that is a valid claim. And I believe this is mostly true when it comes to evaluating nba players in general, and regardless of whether TS%/Per36 happens to make Lin "look better" or not. Do you dispute this argument and can you provide an explanation why? If there is a stat that can provide an even clearer picture than TS% and per36 stats, I'm all ears. And I'm sure Torocan and other ppl interested sports statistics would be interested as well, even if it may make Lin "look worse"
I dont understand how giving stats per 36 minutes gives more context. If anything, I think it decreases context. If Lin was taken out of the game early and not forcing his way into 35+ minutes a night in those big games (against top 9 teams in the West) I think that adds to the context and tells you everything you need to know about that player. The Stats would tells you that if X player scores Y amount of points in 20 minutes, he would score Z amount of points in 30 minutes. Nothing could be further from the truth in basketball because X player is doing something to limit his minutes in the first place. The picture the OP paints about Lin is wrong, and the counter argument was wrong because with Lin especially the actual eye & basketball analysis should almost always at least be a foot note to support or discount whatever argument you are trying to make statistically. There are reasons that Lin is playing few minutes in big games that go beyond the stats. The stats are interesting, but I'd like to see more commentary on actual basketball to support or discount Lin arguments or we will just go around in circles modulating Lin stat arguments.
Sometimes I like to tell myself you don't actually believe what you're typing and that you're trolling. It helps me feel better about humanity.
This is very true as well, which is why for a lot of these debates, it gets very complicated and so many different things have to be looked at. It's all just supporting evidence while some like to throw a hammer around and say look! I've found the clear 100% answer! Some may think that's what torocan does too, but I find that torocan tends to be careful with his arguments and has plenty of disclaimers. It's not his fault most people are simply too stupid to realize it. When he says he does not believe the sky is green, people often turn around and say YOU BELIEVE THE SKY IS RED?! It's absurd and I see it time and time again.
I think this description of statistics was unnecessary. His point--and your point--is that numbers can be "true" but extremely misleading. Stats are similar to guns--you use stats in order to mislead. Stats are not cold in the manner in which you describe them. Where do these stats come from? Someone must watch the game in order to determine when a turnover or rebound occurs. You yourself have made attempts to justify stats (turnovers) as actually not what they appear by watching clips and describing them. What is astounding is that the stats are never useful in exposing the lies that you tell yourself. Lin fans quote all of your posts with approval because they fail to notice the logic in which you erase Lin's poor play with statistical analysis.
Just having some fun. Clearly Torocan has an agenda to make Lin look as good as possible. Lin is an ok player - and is not above criticism - honestly - he needs to go to another team to get better, his style will not fit in with Harden ever - and Harden is a much better player. I like Lin, I am just not viewing him through rose colored glasses - and I honestly believe a Beverly & Brooks combo would have been better overall for this team. And NUMBERS are cold - but they don't account for everything either, and cherry picking them can make a case look a lot more solid than it does. DD
Per 36 is less about giving specific context as opposed to giving general context. Over a sample, per 36 gives a more equal weighting to impact over time than taking highly disparate time samples and doing a 1:1 comparison. Per 36 isn't supposed to be used to Project future performance as there are multiple variables. What it does do is that it tells us what the ACTUAL performance per minute was in whatever aspect you choose to examine. Taking a player that plays on the bench and using per 36 to justify that they should be playing starting minutes is a questionable use of Per36, especially without additional context and data. For example... XXX scores Y per 36. (Useful in that it tells us what he did with his time, but not relevant to what he WILL do in a different environment role) XXX scores Y per 36 against 2nd units, but Z per 36 against starting units. (More useful due to additional context in terms of opposition) XXX scores Y per 36 against 1st units when on the floor with the starting unit. (More useful as it provides context in terms of opposition and supporting players) Per36 is just a way of parsing the time in a way that is base lined. IE, you're comparing something closer to apples to apples and oranges to oranges (40 minute games vs 20 minutes games). Regardless of additional context, Per36 is useful when trying to determine IF a player contributed heavily or didn't contribute at all when they were actually on the floor. As for making an argument for/against Lin, that wasn't the point of my post. I just saw a set of statistics presented that were provided without any statistical context and an analysis that was procedurally flawed. The main point being that if you're going to make a statistical argument, then at least provide statistical analysis that is methodologically sound.
You're confusing data and statistics. Data is the raw information. That is what it is. You acquire data through recording observation (video counts, box score counts that we generally assume are correct, etc) Statistics is the mathematical method through which you process and interpret that data, which is subject to human error, bias, being overly narrow or overly expansive or completely irrelevant. If there are a bunch of eggs on a table and I count them, that's not statistics. That's data. A player shoots 5/17. That's not statistics, that's data. No amount of statistical manipulation will change that the ball went in the bucket 5 times and didn't go in 12 times. However, statistics allow us to examine that data and determine exactly what the impact of those 5 buckets were to the team. Was it 5 x 3 pointers? 5 x 2 pointers? 5 x 2 pointers with 3 x and1's? And how do we value those? What does it tell us about the performance of that player? Data itself is not subject to bias unless the person counting intentionally miscounts or discards parts of the count. And in that case the flaw is in the collector, not the data itself. If the ball goes in the bucket 5x and you write down 3x or 6x, that doesn't change the fact that the ball went into the bucket 5x. Data and statistical methods are not the same thing.
I agree with the above post from DM. I actually believe playing with Harden until his departure next year will help Lin in the long run, it exposes his weakness and make him work on it. Hopefully he will get better as a result.
Its pretty simple, no sabermetrics needed... when lin is the number one guy who has to score and make plays... he plays well (ie NY when melo was out)... when he has to be the spot up shooter like he is in our offense... he sucks... if he can be the single playmaker with the 2nd group... we will get the most out him... otherwise... PLEASE SHIP HIS ASS OUT OF HERE IN THE OFFSEASON!!!
Ask yourself this, what do you hope to accomplish with your posts. The people who like Lin will still like him, the ones that hate him will still hate him. It is much easier to ignore the posters you do not like than going into endless debates with them.
Okay--well, your statistical method in the past has been to deny that the data is data. Your method was to look at the turnovers and admit that they were turnovers but really not representative of poor performance. Technically you aren't denying that the data still exists but you are claiming that it is irrelevant. If the egg is a turnover you are denying that an egg is equivalent to another egg. Statistical analysis is not conducted by observing clips of a game. Regardless of the definition of data and statistics, your statistical method is to deny that the data should be interpreted in the most intuitive way whenever it involves Jeremy Lin. If his TOs are high or FG% low--or if he just seems to be playing like crap--you find a way of justifying his play with "statistical methods." Houston Texans fans are familiar with this line of attack (Kubiak's focus on time of possession/Wade's focus on our "#1 defense"). Losing teams are much more likely to rely on statistical analysis to show that "well, things are not as bad as we think." This is precisely how Socrates pissed off so many Athenians (not to boost anyone's ego). He would start from a premise that is implausible to his audience and try to convince them through the elenchus that they believe the reverse of what they thought they believed or that the "truth" is the opposite of their presumptions (immortal souls exist, etc).
Why don't you want to use PER or WS/48 or xRAPM or really any other advanced stats that show Lin is mediocre? Stop having a bias and just look at what the statistics tell you. When Jeremy plays well he gets to play more minutes. So in the games where he plays a lot of minutes he does well. You are trying to cherry pick stats to make a point. The point of stats is to get a large sample size and let the numbers tell you the story and throw out the outliers, but you don't want to do that you want pick specific points to make your point. You say avoid the eye test and also avoid the stats when they don't back up your point.
I appreciate the work you put into these threads, but I think your missing the point of what I was getting at. Not with the per 36(which I do find to be much more useful than per 48 which skews arguments far too often), but the overall use of stats to argue for or against Lin. -OP states that Lin's stats are bad in big games -counter argument is in those games, he played few minutes, and per however many minutes, he's actually good. -What I'm asking for is for that argument to also contain a foot note about the basketball side to support or discount... in this case it would be, well why isn't he playing more than 35 minutes than if his production was really good, and how do we know how that production impacts the team play? Then from that point on we can look at how match-ups came into play, and then further dissect where Lin's strengths and weaknesses really are based on what he does or doesn't do to get or not get PT & production. I see this argument all the time with Lin in the GARM. If you same out the time played, his stats look much better, and it gives Lin fan's something to argue in their case that he is actually good, and then allows them the gateway to bash on McHale (another topic for another thread), but I think we aren't paying enough attention in the first place to supporting stats with the basketball side, and vice versa. So I dont have a problem with looking at Per 36, or your argument to counter, but the backing in the OP's case & the counter wasn't there to support either side. I agree. You did a good job of making that point. OP was trying to make a point with stats that didn't give detailed stats, and it wasn't backed up by actual basketball analysis other than the obvious statement which is "Lin hasn't been a major factor in most big games vs West rivals".... yeah we got that... where you going with that argument?
