"No team has a drop off like Portland when Lillard, Aldridge sit"

[RoyToy]

Yup. It's been so long since I paid any attention to it, because I found it to be garbage - so I no longer know precisely what it is. That's often how the memory works - if it's not in active rotation the details slowly fade away. I try not to waste too much time on stuff that I perceive to be garbage.

Regarding how I use stats - I use them in conjunction with what I see when I watch games. First I get a feel for a player - good or bad. Then I look at stats to see if they're supporting what I'm seeing. If they don't then I either watch some more to see if I'm missing something, or try to see what is making the stat appear to tell something that isn't quite true. This process lead me to rethink the importance of RPG - it can be a very misleading stat and it certainly doesn't measure a player's ability when it comes to rebounding.

When it comes to FG% I like to look at a team's FG% and their OppFG% to help determine whether my opinion of that team's offense/defense is "correct". I find FG% & OppFG% to be more meaningful than OffRtg & DefRtg - it measures how well they are at putting the ball in the basket or keeping the opposition from putting the ball in the basket and weeds out stuff like how good a team is at shooting foul shots, or how many FTs a team shoots which is largely based on whether they have a superstar on their roster. It also weeds out stuff like steals, which is part of defense - but I'm interested in the actual ability of making a team shoot low percentage shots...steals are a gamble, bad defense can lead to a steal. I also like to compare individual's FG% by position, and their FG% by shot location. In this way you can see how good a player is compared to other people who score in similar ways.

No individual stat is going to tell you the whole story, and trying to make one stat by doing a bunch of math with a whole bunch of stats is just blurring an already blurry line - at least when it comes to fan talk. Advanced stats may come in useful for teams who are employing people to do nothing but pore over volumes of stats - but when those stats are used by fans as stand-alone measurements they lose all meaning.

But I more than anything it's the eye test - seeing is believing.

Yes! Yes! Yes!

Too many people here turn to these advanced stats when talking about a player they've seen play probably less than 5 times and think they're an expert. It's stuff like this that leads people to say that Travis Outlaw and Rudy Gay are equal talents(yes, that actually happened on this forum). Not many people here actually break a player's game down and tells you about his on-court play. It's just numbers

Some people crunch numbers

Other people watch the game

 

[e_blazer]

Yes! Yes! Yes!

That's what she said.

 

[BBert]

Uh-oh. Someone used a statistical analysis to suggest that LaMarcus Aldridge is a useful player. Of course that statistic is therefore garbage. What are you guys thinking?

 

[MickZagger]

Some people crunch numbers

Other people watch the game

Oh and thats not full of subjectivity...

 

[THE HCP]

I'd say 95% of people on here watch the games...... And half of them STILL don't know what they're talking about!

 

[blazerboy30]

Yes! Yes! Yes!

Too many people here turn to these advanced stats when talking about a player they've seen play probably less than 5 times and think they're an expert. It's stuff like this that leads people to say that Travis Outlaw and Rudy Gay are equal talents(yes, that actually happened on this forum). Not many people here actually break a player's game down and tells you about his on-court play. It's just numbers

Some people crunch numbers

Other people watch the game

And some people don't understand the point of statistical analysis.

I'm going to bad-mouth every statistic and call them garbage because they don't have a perfect r^2 value, perfectly predicting the performance of every player!!

 

[blue9]

How do you quantify "good things" versus "bad things" while a player is on the court?

Seems to me a rather basic stat is to see if a team scores more or less points while any given player is on the floor. It's not ideal over a large sample, but it seems foolish to call it and PER "garbage". Is it just a coincidence that the players with the highest PERs seem to be on All-NBA teams and All-Star teams?

If you're going by FG%, then Przy the year he almost shot 70% should have been MVP.

Regarding good/bad things, I realized I was thinking of yet another composite stat - EFF. And rather than type out all of the individual stats that go into is I simplified it by using "good" and "bad" - good things being points, rebounds, etc, and bad being TOs, missed shots, etc.

And I think I better explained in another post in this thread my feelings on PER. It's a stupid stat when used in isolation - and that's the only way message board fans can really use that stat. Most don't even know what it means - it's just a number to them and if it's higher that's better than if it's lower. And I'm guessing that most NBA teams have far better metrics than PER - so, really, it's stupid all the way around.

And as I explained in another post - FG% isn't the be-all end-all measurement of a player, it's just the purest measurement of a statistical category. Other statistical measurements have other factors that affect the meaningfulness of the stat, but FG% directly measures a players ability to put the ball in the basket. I don't think the player with the highest FG% is the best player in the league - that's stupid, and I'm sure you realize that I was never making that argument. But comparing FG% amongst players who play the same position is useful and can add meaningfulness when trying to determine which is the better player at a given position. Also, breaking down FG% by shot location is a very useful tool - shows you where a player is most and least effective.

 

[blazerboy30]

Regarding good/bad things, I realized I was thinking of yet another composite stat - EFF. And rather than type out all of the individual stats that go into is I simplified it by using "good" and "bad" - good things being points, rebounds, etc, and bad being TOs, missed shots, etc.

And I think I better explained in another post in this thread my feelings on PER. It's a stupid stat when used in isolation - and that's the only way message board fans can really use that stat. Most don't even know what it means - it's just a number to them and if it's higher that's better than if it's lower. And I'm guessing that most NBA teams have far better metrics than PER - so, really, it's stupid all the way around.

And as I explained in another post - FG% isn't the be-all end-all measurement of a player, it's just the purest measurement of a statistical category. Other statistical measurements have other factors that affect the meaningfulness of the stat, but FG% directly measures a players ability to put the ball in the basket. I don't think the player with the highest FG% is the best player in the league - that's stupid, and I'm sure you realize that I was never making that argument. But comparing FG% amongst players who play the same position is useful and can add meaningfulness when trying to determine which is the better player at a given position. Also, breaking down FG% by shot location is a very useful tool - shows you where a player is most and least effective.

All of this is being said with absolutely nothing showing that FG% has a higher correlation to "goodness" of a player than PER.

 

[blue9]

I'm not attempting to make that correlation - those are three separate statements, responding to PapaG.

 

[mook]

It's much, much easier to identify a good player, mediocre player or a bad player by PER than FG% or pretty much any other stat.

I will say, though, that PER is imprecise. A guy with a PER of 24 is not necessarily any worse than a guy with a PER of 26. Things like defense and intangibles come into play. PER can overvalue guys like Kevin Love, just because he puts up all kinds of great stats and gives you so little on defense.

Obviously, the best way to evaluate talent is to watch games. If you have time to watch all 1230 games in an NBA season and also have a great memory and a critical mind, you'll have a leg up on PER.

 

[blazerboy30]

I'm not attempting to make that correlation - those are three separate statements, responding to PapaG.

You're saying that PER and +/- are garbage, and that FG% is more useful than PER or +/- for determining which player is better. I'd like to see at least a little proof of that statement.

 

[blazerboy30]

It's much, much easier to identify a good player, mediocre player or a bad player by PER than FG% or pretty much any other stat.

I will say, though, that PER is imprecise. A guy with a PER of 24 is not necessarily any worse than a guy with a PER of 26. Things like defense and intangibles come into play. PER can overvalue guys like Kevin Love, just because he puts up all kinds of great stats and gives you so little on defense.

Obviously, the best way to evaluate talent is to watch games. If you have time to watch all 1230 games in an NBA season and also have a great memory and a critical mind, you'll have a leg up on PER.

PER doesn't ever claim to say that something is "necessarily" true. It's just a stat that was derived using correlations. It won't perfectly fit the performance and "goodness" of every player, but in general it will give a good indicator of the better player.

 

[blue9]

Obviously, the best way to evaluate talent is to watch games.

Agreed. I just think that other stats are more useful than PER to add more details and layers of understanding of a particular player.
PER is kinda like sticking your finger into water to measure the temperature - it'll tell you if it's hot, but unless you have a lot of experience measuring water temperature with your finger you probably won't know whether it's 100* or 120*.

 

[Nikolokolus]

Agreed. I just think that other stats are more useful than PER to add more details and layers of understanding of a particular player.
PER is kinda like sticking your finger into water to measure the temperature - it'll tell you if it's hot, but unless you have a lot of experience measuring water temperature with your finger you probably won't know whether it's 100* or 120*.

It's just like that ... except that it's 100% not.

 

[blue9]

It's just like that ... except that it's 100% not.

Sure it is - it gives you a vague idea of whether a player is "good", but absolutely no detail on WHY they're good or what it is that they're good at.

 

[Wizard Mentor]

Harvey Grant, yes Harvey Grant. I want stats to tell me about Harvey Grant.

When the Blazers signed Harvey Grant, I was excited - He put up a lot of points on a bad team. However, he failed the eye test after we signed him even before he was ever on the court. I knew he would suck when during his presser he was talking about going to the finals - the guy had never been to a playoff game. He indeed sucked as a blazer.

I want stats to tell me not to sign Harvey Grant.

Here's his stats: http://www.basketball-reference.com/players/g/grantha01.html

18.6 ppg on .487 shooting sounds like borderline all-star (the year before joining the blazers)
16.1 PER says he's above average, not borderline all-star.

the year after we signed him: 10.4 pts on 46%, with a PER of 12.3.
(for purposes of comparison, Meyers Leonard as a rookie has a PER of 13 after 29 games).

The problem with PER isn't that it's too advanced, it's that it's not advanced enough. It should be modified to reflect a team's win/loss record, and whether the skills translate to a better team (rebounding, FT translate, FG% doesn't).

 

[RoyToy]

And some people don't understand the point of statistical analysis.

Oh, I certainly understand it.

I just don't misuse and rely on it to form my opinion of players like many people here do.

Nice quote. Boring.

 

[Charcoal Filtered]

blue9, I agree that every stat has to be defined and cannot be taken at face value. However, to say +/- should be just thrown out while fg% is a true measure of a players effectiveness is wrong.

I will not debate how many factors need to be looked at for +/- since there are many. It could also be said fg% also has many other factors that need to be looked at. If there are no other offensive threats on the floor, the person shooting the ball is going to be guarded better. That has nothing to do with his ability. If a player creates no offense and the majority of his shots are putbacks and dunks, his fg% will be better than the person tasked with creating offense. A team with a point guard that can create and get the ball to the open man will also have a team with a better fg %. Again, the other players ability does not change even though their fg% does.

 

[maxiep]

It seems to me that Olshey has a decison to make pretty quickly. The Blazers are presently sitting in the 8th spot in the West. I suspect that's a surprise to Blazers' management as compared with preseason expectations. The starting lineup is competing fairly well and probably is playoff-worthy. In order to stay in the thick of the playoff hunt, especially given that the schedule gets much tougher from this point on, is going to take an infusion of talent in the bench unit. Whether through trade or even just a signing of an older vet or two, the bench could be beefed up. Guys like Bibby or even Derek Fisher, are available and have playoff experience. By now, Olshey has seen enough of the bench players to know who he wants to keep and develop and who can be dumped. Sitting out of the playoffs one more year and getting a decent pick wouldn't be the worst thing, but I don't see it as likely to really move the dial towards contending either. The Blazers are likely to get no higher draft pick than the 11th or so. The absolute worst thing would be to finish out of the playoffs but with the 13th or 14th pick and have to hand it over to Charlotte.

You touch on a post I was going to make: Given the dropoff due to the suckiness of our bench, in which direction do we go? I think--for better or worse--our core is Lillard, Batum and Aldridge. My guess is that's a core that can get you at best to the second round, so we'll need to attract a player better than one of those three. The question is, how do we get them?

It would be nice if we could sign Hickson as a backup PF or 20 mpg center, but he's going to command a big salary next year. Perhaps keep Wes; his contract was frontloaded. If he's your starting SG, however, you're in trouble. Other than that, we have rookie contracts and cheap vets. Harness cap space and draft picks and try to get lucky. Otherwise, sit back and enjoy average NBA basketball.

 

[blue9]

Fair points, Charcoal. And I hear where you're coming from. But in the way that I think of the game the recipient of the PG's passes is generally 50% responsible for creating the passing lane for the easy scoring opportunity - they must know where to go at a given point in time in order to receive the pass for an easy bucket. Same with put-backs - the player earns their higher FG% through the hard work of offensive rebounding. They are creating, or partially creating, easy scoring opportunities for themselves. If Player A works a pick/roll for a shot with a 70% success rate and Player B works a pick/pop for a shot with a 40% success rate, then I'm going to go with Player A.
But I get your point - FG% isn't as "pure" as I've made it out to be. Though I still think it's the most pure statistical measurement. And again, that's not to say it's the most important measurement. I just like to know what my stats are telling me and most stats have other factors that muddy the waters, or they just don't tell the full story within that statistical category - ie, not all assists are the result of good passing, not all rebounds are "equal", not all made shots are good shots, etc.

 

[PapaG]

Originally Posted by mook
Obviously, the best way to evaluate talent is to watch games.

...said the old-school scouts in Moneyball.

 

[blazerboy30]

Oh, I certainly understand it.

Nope.

I just don't misuse and rely on it to form my opinion of players like many people here do.

You certainly do misuse it by tossing it out because it doesn't have an r^2 value of 1.

 

[RoyToy]

The thing with FG% is you have to know what kind of shots that player is taking.

Batum and Lillard pretty much have identical numbers from beyond the arc when it comes to attempts and 3pt%.

Difference is Lillard creates most of his opportunities, while Batum mostly has to be set up just to get a good shot off.

 

[RoyToy]

Nope.



You certainly do misuse it by tossing it out because it doesn't have an r^2 value of 1.

K .

 

[VanillaGorilla]

The thing with FG% is you have to know what kind of shots that player is taking.

Batum and Lillard pretty much have identical numbers from beyond the arc when it comes to attempts and 3pt%.

Difference is Lillard creates most of his opportunities, while Batum mostly has to be set up just to get a good shot off.

It doesn't say specifically for 3pt shots, but I thought these stats might be relevant.

Batum:

Lillard:

I think it shows statistically how much harder Lillard has to work for his shots.

 

[RoyToy]

It doesn't say specifically for 3pt shots, but I thought these stats might be relevant.

Batum:

Lillard:

I think it shows statistically how much harder Lillard has to work for his shots.

Thanks for making it easier for everyone to see. I'll give you rep just for the effort.

And being the focus of the defense these days Lillard is definitely working much harder for his shots. He's still producing, though, and that's what makes him special.

 

[PtldPlatypus]

Couple interesting things I notice in those stats, that I would not have expected:
--Lillard's eFG% late in the clock is horrible, yet Batum's is about twice Lillard's.
--Lillard's "dunk" numbers indicate that if it's not blocked, it's going in. Batum apparently misses 11% of his dunk attempts that aren't blocked.
--Batum appears to get his shot blocked a lot more often when he's creating for himself.

 

[e_blazer]

--Lillard's eFG% late in the clock is horrible, yet Batum's is about twice Lillard's.

I'd say that's likely a result of their differing positions. Nic's often a safety relief valve, spotted up in the corner and being fed the ball at the last second for an open look at a 3 pt shot. When Lillard has to jack up a last second shot, he's usually got the ball in his hands and the wheels have come off of the offense so he has to put up a contested jumper.

 

[BlazerCaravan]

I'd say that's likely a result of their differing positions. Nic's often a safety relief valve, spotted up in the corner and being fed the ball at the last second for an open look at a 3 pt shot. When Lillard has to jack up a last second shot, he's usually got the ball in his hands and the wheels have come off of the offense so he has to put up a contested jumper.

Exactly. Lillard shooting late in the clock means the picks have failed, he's double teamed, the pass out came back to him, and he's gotta heave it up.

 

[RoyToy]

Liking the last 2 posts in this thread

Basketball talk!