General 76ers Thread

[Eyeamok]

Is there an advanced stat on enthusiasm and energy?

All I know is that, almost every game I have watched that Embiid has played in, the team as a whole seems to have a lot more energy and the ball movement is a lot better. A lot more enjoyable brand of basketball to watch. Minus the obvious egg that was laid against the Timberwolves.

[LloydFree]

Is there an advanced stat on enthusiasm and energy?

All I know is that, almost every game I have watched that Embiid has played in, the team as a whole seems to have a lot more energy and the ball movement is a lot better. A lot more enjoyable brand of basketball to watch. Mimus the obvious egg that was laid against the Timberwolves.

When a team puts a good defensive big on the floor it always makes the team look as though they are playing with more enthusiasm. Intuitively, the guards pressure the perimeter more, just because they know the other team isn't going to score inside. If you are getting stops defensively, that leads to more fast breaks on Offense. Much of that goes out the window when you play with a Center that doesn't protect the basket.

[Sixerscan]

Advanced stats or traditional stats I think have the same strengths and weaknesses, and require the same contextual analysis.

With Vorp measuring production, that would mean that Holmes is more productive than Embiid, which is absurd. I understand that Embiid's turnovers are depressing his efficiency/effectiveness.

So take the vorp stat between them with a grain of salt. If Holmes was handling the ball as much as Embiid he would have a much lower vorp as well.

People on this board all the time will say that player X has a certain vorp number, and therefore he is better than a player that has a worse Vorp number.

It would be a disaster if we tried to use Holmes like Embiid or Okafor. The stats are not reflecting the true disparity in skills.


Sent from my iPhone using Tapatalk

Well they don't have the same strength and weaknesses. You're not going to tell me that DRPM and Blocks per game are equally relevant.

Again, you're not getting the basic idea of small sample size or statistical significance if you think someone having 0.3 VORP over 200 minutes is a "better player" than someone else with a 0.2 VORP over 200 minutes. You're not in a position to say what "grain of salt" needs to be taken if you don't get that.

I also think it's interesting that your conclusion is that "well Holmes would do worse if he handled the ball more" when I think the much more interesting question is whether the other two are handling the ball too much. (Unless your goal is to just give them reps, efficiency be damned). There has been mounting evidence over the last year plus that this team is more efficient when it moves the ball and involves everyone rather than relying on one guy having a huge usage rate, and now you have a small but growing sample of the team having some success offensively with a center with very limited skills playing out there with basically the same teammates. Maybe using Embiid as a roll man more and posting up less would make a more efficient offense.

Of course there are quibbles, he's played some in blowouts (so have the other guys though) but I think it's at least something worth considering rather than just blindly assuming that whatever grains of salt you can come up with balance it out and going back to the same tired arguments we've been having for the last year.

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

For example, Number of Interceptions(analogous to blocks per game) for Deon Sanders won't tell you the truth about him that he was probably the greatest cover corner in the history of football. A player has strengths and weaknesses that influence the behavior of the teammates and opponents on the floor. The nba team with the best shot blocker may not lead the league in blocks per game because they dissuade people from taking blockable shots.

Also, I FULLY understand sample size, and that's why you won't see me ever use statistics over the short term to make sweeping conclusions. My point about advanced stats is that, correcting for sample size, they STILL don't measure who is better between two different players completely. They are PART of the equation, but not the ANSWER. Okafor played I think 59 games in his rookie season, but how many time have we seen people on this board use his advanced stats from those 59 games to make conclusions about him?

I think that advanced stats in context can be used to measure skills, but by themselves can not, and I CERTAINLY don't think that they have a lot of predictive value for saying how much potential someone has. For example, the classic idea that age factors in to potential(which it does in general) would make people today try to put Embiid's numbers into context by comparing him to other big men at 22 years of age. Someone on the general board I think was trying to compare Embiid at 22 to Cousins at 22, as if that is a valid appraisal of Embiid. In reality, the context of Embiid's current performance should be compared to someone who is 18 or 19 because the whole age thing usually is comparing players that have played basketball their whole lives, and he hasn't.

How do you account for a player that is playing his 4th year of organized basketball at age 22, who hardly played in his first season, and then got injured half way through his third season, and then missed over 2 years due to injury, and is 12 games into his 4th? The formula used to project players is totally flawed because it doesn't take into account this kind of back story. Embiid has probably played fewer than 82 official basketball games in his entire life!

I mean, you're wrong. Blocks per game and DRPM do not have the same strengths and weaknesses. Blocks per game is freaking blocks divided by games, DRPM is a regression that models how much better your team does defensively with you on the floor taking into account your teammates and oppponents. They aren't even measuring the same thing.

You keep talking about how advanced stats have outside factors that they don't consider, but it's pretty obvious that you have no idea what factors some of these do consider. RPM DOES attempt to consider the quality of your teammates and your impact beyond how how many shots you block. Thats kind of the point of it. It's not perfect, but I don't see how you can "fill in the blanks" when you don't even know what the blanks are.

And are you under the impression that blocks per game is an advanced stat? It's not. Neither is interceptions per game for that matter.

None of these stats I've mentioned purport to project how a guys career will go. Every projection model I've actually seen says that embiid will be a future hall of famer if he stays healthy. Like I get that he has a unique background but the guy has been nothing but productive at Kansas and here that shouldn't be a shocker.

Again I'm not sure exactly what we're arguing about here. And you keep talking about "advanced stats" like it's one thing when it's not. Some are more valuable than others depending on the situation and the player. It really just seems like you're arguing against your idea of what "advanced stats" are, which I can't help you with.

[Sixerscan]

Is there an advanced stat on enthusiasm and energy?

All I know is that, almost every game I have watched that Embiid has played in, the team as a whole seems to have a lot more energy and the ball movement is a lot better. A lot more enjoyable brand of basketball to watch. Mimus the obvious egg that was laid against the Timberwolves.

I agree on the energy. And additionally the team is something like 12 points per 100 possessions better with Embiid on the court than off. Even factoring in quality of teammate and opponent it's still noticeable. He actually has a positive RPM right now which Holmes does not.

People shouldn't get so hung up on a 0.1 VORP difference between two guys that have played 200 minutes. Still I think the fact that holmes has been successful can be instructive going forward. And fwiw I think he brings a certain energy himself.

[Ericb5]

Well they don't have the same strength and weaknesses. You're not going to tell me that DRPM and Blocks per game are equally relevant.

Again, you're not getting the basic idea of small sample size or statistical significance if you think someone having 0.3 VORP over 200 minutes is a "better player" than someone else with a 0.2 VORP over 200 minutes. You're not in a position to say what "grain of salt" needs to be taken if you don't get that.

I also think it's interesting that your conclusion is that "well Holmes would do worse if he handled the ball more" when I think the much more interesting question is whether the other two are handling the ball too much. (Unless your goal is to just give them reps, efficiency be damned). There has been mounting evidence over the last year plus that this team is more efficient when it moves the ball and involves everyone rather than relying on one guy having a huge usage rate, and now you have a small but growing sample of the team having some success offensively with a center with very limited skills playing out there with basically the same teammates. Maybe using Embiid as a roll man more and posting up less would make a more efficient offense.

Of course there are quibbles, he's played some in blowouts (so have the other guys though) but I think it's at least something worth considering rather than just blindly assuming that whatever grains of salt you can come up with balance it out and going back to the same tired arguments we've been having for the last year.

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

For example, Number of Interceptions(analogous to blocks per game) for Deon Sanders won't tell you the truth about him that he was probably the greatest cover corner in the history of football. A player has strengths and weaknesses that influence the behavior of the teammates and opponents on the floor. The nba team with the best shot blocker may not lead the league in blocks per game because they dissuade people from taking blockable shots.

Also, I FULLY understand sample size, and that's why you won't see me ever use statistics over the short term to make sweeping conclusions. My point about advanced stats is that, correcting for sample size, they STILL don't measure who is better between two different players completely. They are PART of the equation, but not the ANSWER. Okafor played I think 59 games in his rookie season, but how many time have we seen people on this board use his advanced stats from those 59 games to make conclusions about him?

I think that advanced stats in context can be used to measure skills, but by themselves can not, and I CERTAINLY don't think that they have a lot of predictive value for saying how much potential someone has. For example, the classic idea that age factors in to potential(which it does in general) would make people today try to put Embiid's numbers into context by comparing him to other big men at 22 years of age. Someone on the general board I think was trying to compare Embiid at 22 to Cousins at 22, as if that is a valid appraisal of Embiid. In reality, the context of Embiid's current performance should be compared to someone who is 18 or 19 because the whole age thing usually is comparing players that have played basketball their whole lives, and he hasn't.

How do you account for a player that is playing his 4th year of organized basketball at age 22, who hardly played in his first season, and then got injured half way through his third season, and then missed over 2 years due to injury, and is 12 games into his 4th? The formula used to project players is totally flawed because it doesn't take into account this kind of back story. Embiid has probably played fewer than 82 official basketball games in his entire life!

I mean, you're wrong. Blocks per game and DRPM do not have the same strengths and weaknesses. Blocks per game is freaking blocks divided by games, DRPM is a regression that models how much better your team does defensively with you on the floor taking into account your teammates and oppponents. They aren't even measuring the same thing.

You keep talking about how advanced stats have outside factors that they don't consider, but it's pretty obvious that you have no idea what factors some of these do consider. RPM DOES attempt to consider the quality of your teammates and your impact beyond how how many shots you block. Thats kind of the point of it. It's not perfect, but I don't see how you can "fill in the blanks" when you don't even know what the blanks are.

And are you under the impression that blocks per game is an advanced stat? It's not. Neither is interceptions per game for that matter.

None of these stats I've mentioned purport to project how a guys career will go. Every projection model I've actually seen says that embiid will be a future hall of famer if he stays healthy. Like I get that he has a unique background but the guy has been nothing but productive at Kansas and here that shouldn't be a shocker.

Again I'm not sure exactly what we're arguing about here. And you keep talking about "advanced stats" like it's one thing when it's not. Some are more valuable than others depending on the situation and the player. It really just seems like you're arguing against your idea of what "advanced stats" are, which I can't help you with.

You are simply not understanding what I am writing, and maybe that is my fault.

I am not calling blocks per game an advanced stat OBVIOUSLY. It is a counting stat like Interceptions per game.

You have again tripped over the same point which I tried to clarify here:

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

DRPM is an advanced stat, and Blocks per game is a conventional stat, but that both CATEGORIES(not talking about specific stats, but the categories in general) suffer from the same weaknesses in that they BOTH require context to interpret. Yes, one individual stat may be stronger predicatively than another one in a given area, but that is specifically why I am arguing that you need to look at multiple things in order to come to a conclusion.

Just because an advanced stat PURPORTS to add context doesn't mean that they are able to successfully do it. Hence the nonsensical conclusions that posters draw from them. My beef is with the strident adherents to analytics that think that you can learn everything you need to know about a player by reading the stats, and not with the stats themselves.

The position that you seem to be taking is exactly the one that I am decrying, which is that if statistics have the value that you claim that they have then it is unnecessary to watch the games. You can learn everything that you need to know about a basketball player by looking at the box scores.

If player A has a better DRPM than player B then you can't turn that into the statement of Player A is a better defender than player B, but scores of people do just that, and THAT is what I'm complaining about. I am not arguing against what advanced stats are. I'm arguing against the sweeping conclusions that people draw from advanced stats.

[Sixerscan]

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

For example, Number of Interceptions(analogous to blocks per game) for Deon Sanders won't tell you the truth about him that he was probably the greatest cover corner in the history of football. A player has strengths and weaknesses that influence the behavior of the teammates and opponents on the floor. The nba team with the best shot blocker may not lead the league in blocks per game because they dissuade people from taking blockable shots.

Also, I FULLY understand sample size, and that's why you won't see me ever use statistics over the short term to make sweeping conclusions. My point about advanced stats is that, correcting for sample size, they STILL don't measure who is better between two different players completely. They are PART of the equation, but not the ANSWER. Okafor played I think 59 games in his rookie season, but how many time have we seen people on this board use his advanced stats from those 59 games to make conclusions about him?

I think that advanced stats in context can be used to measure skills, but by themselves can not, and I CERTAINLY don't think that they have a lot of predictive value for saying how much potential someone has. For example, the classic idea that age factors in to potential(which it does in general) would make people today try to put Embiid's numbers into context by comparing him to other big men at 22 years of age. Someone on the general board I think was trying to compare Embiid at 22 to Cousins at 22, as if that is a valid appraisal of Embiid. In reality, the context of Embiid's current performance should be compared to someone who is 18 or 19 because the whole age thing usually is comparing players that have played basketball their whole lives, and he hasn't.

How do you account for a player that is playing his 4th year of organized basketball at age 22, who hardly played in his first season, and then got injured half way through his third season, and then missed over 2 years due to injury, and is 12 games into his 4th? The formula used to project players is totally flawed because it doesn't take into account this kind of back story. Embiid has probably played fewer than 82 official basketball games in his entire life!

I mean, you're wrong. Blocks per game and DRPM do not have the same strengths and weaknesses. Blocks per game is freaking blocks divided by games, DRPM is a regression that models how much better your team does defensively with you on the floor taking into account your teammates and oppponents. They aren't even measuring the same thing.

You keep talking about how advanced stats have outside factors that they don't consider, but it's pretty obvious that you have no idea what factors some of these do consider. RPM DOES attempt to consider the quality of your teammates and your impact beyond how how many shots you block. Thats kind of the point of it. It's not perfect, but I don't see how you can "fill in the blanks" when you don't even know what the blanks are.

And are you under the impression that blocks per game is an advanced stat? It's not. Neither is interceptions per game for that matter.

None of these stats I've mentioned purport to project how a guys career will go. Every projection model I've actually seen says that embiid will be a future hall of famer if he stays healthy. Like I get that he has a unique background but the guy has been nothing but productive at Kansas and here that shouldn't be a shocker.

Again I'm not sure exactly what we're arguing about here. And you keep talking about "advanced stats" like it's one thing when it's not. Some are more valuable than others depending on the situation and the player. It really just seems like you're arguing against your idea of what "advanced stats" are, which I can't help you with.

You are simply not understanding what I am writing, and maybe that is my fault.

I am not calling blocks per game an advanced stat OBVIOUSLY. It is a counting stat like Interceptions per game.

You have again tripped over the same point which I tried to clarify here:

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

DRPM is an advanced stat, and Blocks per game is a conventional stat, but that both CATEGORIES(not talking about specific stats, but the categories in general) suffer from the same weaknesses in that they BOTH require context to interpret. Yes, one individual stat may be stronger predicatively than another one in a given area, but that is specifically why I am arguing that you need to look at multiple things in order to come to a conclusion.

Just because an advanced stat PURPORTS to add context doesn't mean that they are able to successfully do it. Hence the nonsensical conclusions that posters draw from them. My beef is with the strident adherents to analytics that think that you can learn everything you need to know about a player by reading the stats, and not with the stats themselves.

The position that you seem to be taking is exactly the one that I am decrying, which is that if statistics have the value that you claim that they have then it is unnecessary to watch the games. You can learn everything that you need to know about a basketball player by looking at the box scores.

If player A has a better DRPM than player B then you can't turn that into the statement of Player A is a better defender than player B, but scores of people do just that, and THAT is what I'm complaining about. I am not arguing against what advanced stats are. I'm arguing against the sweeping conclusions that people draw from advanced stats.

i guess I just have a problem with you saying they have the SAME strengths and weaknesses. It's like saying Michael Jordan and tj McConnell have the same strengths and weaknesses as scorers because they both make and miss shots sometimes and have to go against defensive players.

of course watching the games is still relevant. i could flesh all this out more but at this point we've made our points and I actually agree with you to a lesser extent that sometimes people get too wrapped up in some fancy stat they discovered.

Anyway, I don't think people should get too wrapped up about holmes having some minuscule advantage over a small sample size in one advanced stat over embiid, either that neither shows that holmes has been better than embiid nor does it show some sort of cataclysmic issue with all advanced stats.

never been a big fan of basketball VORP anyway.

[Ericb5]

I mean, you're wrong. Blocks per game and DRPM do not have the same strengths and weaknesses. Blocks per game is freaking blocks divided by games, DRPM is a regression that models how much better your team does defensively with you on the floor taking into account your teammates and oppponents. They aren't even measuring the same thing.

You keep talking about how advanced stats have outside factors that they don't consider, but it's pretty obvious that you have no idea what factors some of these do consider. RPM DOES attempt to consider the quality of your teammates and your impact beyond how how many shots you block. Thats kind of the point of it. It's not perfect, but I don't see how you can "fill in the blanks" when you don't even know what the blanks are.

And are you under the impression that blocks per game is an advanced stat? It's not. Neither is interceptions per game for that matter.

None of these stats I've mentioned purport to project how a guys career will go. Every projection model I've actually seen says that embiid will be a future hall of famer if he stays healthy. Like I get that he has a unique background but the guy has been nothing but productive at Kansas and here that shouldn't be a shocker.

Again I'm not sure exactly what we're arguing about here. And you keep talking about "advanced stats" like it's one thing when it's not. Some are more valuable than others depending on the situation and the player. It really just seems like you're arguing against your idea of what "advanced stats" are, which I can't help you with.

You are simply not understanding what I am writing, and maybe that is my fault.

I am not calling blocks per game an advanced stat OBVIOUSLY. It is a counting stat like Interceptions per game.

You have again tripped over the same point which I tried to clarify here:

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

DRPM is an advanced stat, and Blocks per game is a conventional stat, but that both CATEGORIES(not talking about specific stats, but the categories in general) suffer from the same weaknesses in that they BOTH require context to interpret. Yes, one individual stat may be stronger predicatively than another one in a given area, but that is specifically why I am arguing that you need to look at multiple things in order to come to a conclusion.

Just because an advanced stat PURPORTS to add context doesn't mean that they are able to successfully do it. Hence the nonsensical conclusions that posters draw from them. My beef is with the strident adherents to analytics that think that you can learn everything you need to know about a player by reading the stats, and not with the stats themselves.

The position that you seem to be taking is exactly the one that I am decrying, which is that if statistics have the value that you claim that they have then it is unnecessary to watch the games. You can learn everything that you need to know about a basketball player by looking at the box scores.

If player A has a better DRPM than player B then you can't turn that into the statement of Player A is a better defender than player B, but scores of people do just that, and THAT is what I'm complaining about. I am not arguing against what advanced stats are. I'm arguing against the sweeping conclusions that people draw from advanced stats.

i guess I just have a problem with you saying they have the SAME strengths and weaknesses. It's like saying Michael Jordan and tj McConnell have the same strengths and weaknesses as scorers because they both make and miss shots sometimes and have to go against defensive players.

Ok, a better way of saying what I meant is this. The flaws that I'm complaining about exist in both advanced and non advanced stats. That is different than saying that both categories have the same weaknesses.

Those flaws(requirement of context) exist to greater and lesser extents for different stats. Obviously counting stats need more context than advanced stats, but they both require context to have meaning in a general sense. When people say that Player A is a better defender than player B, and then cite an advanced stat as proof, I contend that the advanced stat is a single INDICATION of an attribute, but that several attributes are necessary to prove the conclusion.

Some of what is required comes from watching the games, and not just reading the box scores.

Trust me. I'm a huge nerd, and love trying to quantify things that are basically qualitative in nature. It just frustrates me that people become slaves to the numbers.

I'm assuming that Analytics could prove that Scottie Pippen was a better player than Charles Barkley because he was great at offense and defense, whereas Charles wasn't a great defender. That being said, Barkley was a franchise player and Scottie wasn't, and I don't care what the numbers say. Barkley could carry a franchise, and Scottie was a second star. Hell maybe the numbers don't even say that, but it wouldn't surprise me if they did given Barkley's weaknesses. That is the kind of argument that makes me have less confidence in the numbers.

[Sixerscan]

You are simply not understanding what I am writing, and maybe that is my fault.

I am not calling blocks per game an advanced stat OBVIOUSLY. It is a counting stat like Interceptions per game.

You have again tripped over the same point which I tried to clarify here:

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

DRPM is an advanced stat, and Blocks per game is a conventional stat, but that both CATEGORIES(not talking about specific stats, but the categories in general) suffer from the same weaknesses in that they BOTH require context to interpret. Yes, one individual stat may be stronger predicatively than another one in a given area, but that is specifically why I am arguing that you need to look at multiple things in order to come to a conclusion.

Just because an advanced stat PURPORTS to add context doesn't mean that they are able to successfully do it. Hence the nonsensical conclusions that posters draw from them. My beef is with the strident adherents to analytics that think that you can learn everything you need to know about a player by reading the stats, and not with the stats themselves.

The position that you seem to be taking is exactly the one that I am decrying, which is that if statistics have the value that you claim that they have then it is unnecessary to watch the games. You can learn everything that you need to know about a basketball player by looking at the box scores.

If player A has a better DRPM than player B then you can't turn that into the statement of Player A is a better defender than player B, but scores of people do just that, and THAT is what I'm complaining about. I am not arguing against what advanced stats are. I'm arguing against the sweeping conclusions that people draw from advanced stats.

i guess I just have a problem with you saying they have the SAME strengths and weaknesses. It's like saying Michael Jordan and tj McConnell have the same strengths and weaknesses as scorers because they both make and miss shots sometimes and have to go against defensive players.

Ok, a better way of saying what I meant is this. The flaws that I'm complaining about exist in both advanced and non advanced stats. That is different than saying that both categories have the same weaknesses.

Those flaws(requirement of context) exist to greater and lesser extents for different stats. Obviously counting stats need more context than advanced stats, but they both require context to have meaning in a general sense. When people say that Player A is a better defender than player B, and then cite an advanced stat as proof, I contend that the advanced stat is a single INDICATION of an attribute, but that several attributes are necessary to prove the conclusion.

Some of what is required comes from watching the games, and not just reading the box scores.

Trust me. I'm a huge nerd, and love trying to quantify things that are basically qualitative in nature. It just frustrates me that people become slaves to the numbers.

I'm assuming that Analytics could prove that Scottie Pippen was a better player than Charles Barkley because he was great at offense and defense, whereas Charles wasn't a great defender. That being said, Barkley was a franchise player and Scottie wasn't, and I don't care what the numbers say. Barkley could carry a franchise, and Scottie was a second star. Hell maybe the numbers don't even say that, but it wouldn't surprise me if they did given Barkley's weaknesses. That is the kind of argument that makes me have less confidence in the numbers.

This is kind of what I was talking about earlier though. Go through their respective basketball reference pages. When it comes to all of the advanced stats Barkley blows Pippen out of the water. Barkley is 10th all time in win shares per 48 while Pippen isn't in the top 100. Even just comparing their peaks Barkley is clearly a lot better. If anything I would say that the advanced stats criminally underrate Pippen and somewhat overrate Barkley because most of them could do a better job accounting for defense.

Again, your issues with advanced stats seem to more stem from your preconceived notions rather than actually bothering to see what's under the hood... You can't provide the missing context to something if you don't know what the context is that's missing.

[Ericb5]

i guess I just have a problem with you saying they have the SAME strengths and weaknesses. It's like saying Michael Jordan and tj McConnell have the same strengths and weaknesses as scorers because they both make and miss shots sometimes and have to go against defensive players.

Ok, a better way of saying what I meant is this. The flaws that I'm complaining about exist in both advanced and non advanced stats. That is different than saying that both categories have the same weaknesses.

Those flaws(requirement of context) exist to greater and lesser extents for different stats. Obviously counting stats need more context than advanced stats, but they both require context to have meaning in a general sense. When people say that Player A is a better defender than player B, and then cite an advanced stat as proof, I contend that the advanced stat is a single INDICATION of an attribute, but that several attributes are necessary to prove the conclusion.

Some of what is required comes from watching the games, and not just reading the box scores.

Trust me. I'm a huge nerd, and love trying to quantify things that are basically qualitative in nature. It just frustrates me that people become slaves to the numbers.

I'm assuming that Analytics could prove that Scottie Pippen was a better player than Charles Barkley because he was great at offense and defense, whereas Charles wasn't a great defender. That being said, Barkley was a franchise player and Scottie wasn't, and I don't care what the numbers say. Barkley could carry a franchise, and Scottie was a second star. Hell maybe the numbers don't even say that, but it wouldn't surprise me if they did given Barkley's weaknesses. That is the kind of argument that makes me have less confidence in the numbers.

This is kind of what I was talking about earlier though. Go through their respective basketball reference pages. When it comes to all of the advanced stats Barkley blows Pippen out of the water. Barkley is 10th all time in win shares per 48 while Pippen isn't in the top 100. Even just comparing their peaks Barkley is clearly a lot better. If anything I would say that the advanced stats criminally underrate Pippen and somewhat overrate Barkley because most of them could do a better job accounting for defense.

Again, your issues with advanced stats seem to more stem from your preconceived notions rather than actually bothering to see what's under the hood... You can't provide the missing context to something if you don't know what the context is that's missing.

True.

I don't know the numbers and what they stand for as well as you do, but I know that I disagree with them in many cases when I see them when they don't comport with what I have seen with my own eyes. The stats are better today than they were 10 years ago though, and I'm sure that they will continue to get better.

Is there a stat that measures the attention from the defense that a given player receives? Something that will show the difference between what Iverson faced in 2001 and what Kobe faced?

I watched every Sixers game that year, and lots of Lakers ones too, and Kobe was able to float around and was hardly ever double teamed, or even concentrated on, whereas Iverson was the complete focus of the defense for whichever team he played every single game.

Every Sixers opponent knew that you had to stop Iverson, and every Lakers opponent knew that you had to stop Shaq.

If there is some statistical measure that would show that discrepancy it would go a long way to convincing me that statistics can represent reality. Even back then people would cite Kobe's higher field goal percentage as evidence that he was a better shooter than Iverson. I'm still disgruntled from those battles. Lol


Sent from my iPhone using Tapatalk

[Sixerscan]

Ok, a better way of saying what I meant is this. The flaws that I'm complaining about exist in both advanced and non advanced stats. That is different than saying that both categories have the same weaknesses.

Those flaws(requirement of context) exist to greater and lesser extents for different stats. Obviously counting stats need more context than advanced stats, but they both require context to have meaning in a general sense. When people say that Player A is a better defender than player B, and then cite an advanced stat as proof, I contend that the advanced stat is a single INDICATION of an attribute, but that several attributes are necessary to prove the conclusion.

Some of what is required comes from watching the games, and not just reading the box scores.

Trust me. I'm a huge nerd, and love trying to quantify things that are basically qualitative in nature. It just frustrates me that people become slaves to the numbers.

I'm assuming that Analytics could prove that Scottie Pippen was a better player than Charles Barkley because he was great at offense and defense, whereas Charles wasn't a great defender. That being said, Barkley was a franchise player and Scottie wasn't, and I don't care what the numbers say. Barkley could carry a franchise, and Scottie was a second star. Hell maybe the numbers don't even say that, but it wouldn't surprise me if they did given Barkley's weaknesses. That is the kind of argument that makes me have less confidence in the numbers.

This is kind of what I was talking about earlier though. Go through their respective basketball reference pages. When it comes to all of the advanced stats Barkley blows Pippen out of the water. Barkley is 10th all time in win shares per 48 while Pippen isn't in the top 100. Even just comparing their peaks Barkley is clearly a lot better. If anything I would say that the advanced stats criminally underrate Pippen and somewhat overrate Barkley because most of them could do a better job accounting for defense.

Again, your issues with advanced stats seem to more stem from your preconceived notions rather than actually bothering to see what's under the hood... You can't provide the missing context to something if you don't know what the context is that's missing.

True.

I don't know the numbers and what they stand for as well as you do, but I know that I disagree with them in many cases when I see them when they don't comport with what I have seen with my own eyes. The stats are better today than they were 10 years ago though, and I'm sure that they will continue to get better.

Is there a stat that measures the attention from the defense that a given player receives? Something that will show the difference between what Iverson faced in 2001 and what Kobe faced?

I watched every Sixers game that year, and lots of Lakers ones too, and Kobe was able to float around and was hardly ever double teamed, or even concentrated on, whereas Iverson was the complete focus of the defense for whichever team he played every single game.

Every Sixers opponent knew that you had to stop Iverson, and every Lakers opponent knew that you had to stop Shaq.

If there is some statistical measure that would show that discrepancy it would go a long way to convincing me that statistics can represent reality. Even back then people would cite Kobe's higher field goal percentage as evidence that he was a better shooter than Iverson. I'm still disgruntled from those battles. Lol


Sent from my iPhone using Tapatalk

I would say that advanced stats both grade Iverson and Kobe as both great players with neither one being conclusively better than the other in 2001. Either way going on a case by case basis sort of misses the point...

Theoretically if drawing attention like that if it is in fact valuable would result in your team doing better when you are on the floor, but we've been over the issues with that. The SportVU cameras that are installed in most (all?) arenas now are still just scratching the surface on how much attention a player draws and how that opens up opportunities for other players. (Stuff like Iverson, or how someone like John Wall can drive and set up teammates in the corners for an open shot, or how a player much cover Kevin Durant in the corner more tightly than Harrison Barnes which opens up the lane for others) People call it "gravity." It's going to be the next great advancement IMO.

[TTP]

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

For example, Number of Interceptions(analogous to blocks per game) for Deon Sanders won't tell you the truth about him that he was probably the greatest cover corner in the history of football. A player has strengths and weaknesses that influence the behavior of the teammates and opponents on the floor. The nba team with the best shot blocker may not lead the league in blocks per game because they dissuade people from taking blockable shots.

Also, I FULLY understand sample size, and that's why you won't see me ever use statistics over the short term to make sweeping conclusions. My point about advanced stats is that, correcting for sample size, they STILL don't measure who is better between two different players completely. They are PART of the equation, but not the ANSWER. Okafor played I think 59 games in his rookie season, but how many time have we seen people on this board use his advanced stats from those 59 games to make conclusions about him?

I think that advanced stats in context can be used to measure skills, but by themselves can not, and I CERTAINLY don't think that they have a lot of predictive value for saying how much potential someone has. For example, the classic idea that age factors in to potential(which it does in general) would make people today try to put Embiid's numbers into context by comparing him to other big men at 22 years of age. Someone on the general board I think was trying to compare Embiid at 22 to Cousins at 22, as if that is a valid appraisal of Embiid. In reality, the context of Embiid's current performance should be compared to someone who is 18 or 19 because the whole age thing usually is comparing players that have played basketball their whole lives, and he hasn't.

How do you account for a player that is playing his 4th year of organized basketball at age 22, who hardly played in his first season, and then got injured half way through his third season, and then missed over 2 years due to injury, and is 12 games into his 4th? The formula used to project players is totally flawed because it doesn't take into account this kind of back story. Embiid has probably played fewer than 82 official basketball games in his entire life!

I mean, you're wrong. Blocks per game and DRPM do not have the same strengths and weaknesses. Blocks per game is freaking blocks divided by games, DRPM is a regression that models how much better your team does defensively with you on the floor taking into account your teammates and oppponents. They aren't even measuring the same thing.

You keep talking about how advanced stats have outside factors that they don't consider, but it's pretty obvious that you have no idea what factors some of these do consider. RPM DOES attempt to consider the quality of your teammates and your impact beyond how how many shots you block. Thats kind of the point of it. It's not perfect, but I don't see how you can "fill in the blanks" when you don't even know what the blanks are.

And are you under the impression that blocks per game is an advanced stat? It's not. Neither is interceptions per game for that matter.

None of these stats I've mentioned purport to project how a guys career will go. Every projection model I've actually seen says that embiid will be a future hall of famer if he stays healthy. Like I get that he has a unique background but the guy has been nothing but productive at Kansas and here that shouldn't be a shocker.

Again I'm not sure exactly what we're arguing about here. And you keep talking about "advanced stats" like it's one thing when it's not. Some are more valuable than others depending on the situation and the player. It really just seems like you're arguing against your idea of what "advanced stats" are, which I can't help you with.

You are simply not understanding what I am writing, and maybe that is my fault.

I am not calling blocks per game an advanced stat OBVIOUSLY. It is a counting stat like Interceptions per game.

You have again tripped over the same point which I tried to clarify here:

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

DRPM is an advanced stat, and Blocks per game is a conventional stat, but that both CATEGORIES(not talking about specific stats, but the categories in general) suffer from the same weaknesses in that they BOTH require context to interpret. Yes, one individual stat may be stronger predicatively than another one in a given area, but that is specifically why I am arguing that you need to look at multiple things in order to come to a conclusion.

Just because an advanced stat PURPORTS to add context doesn't mean that they are able to successfully do it. Hence the nonsensical conclusions that posters draw from them. My beef is with the strident adherents to analytics that think that you can learn everything you need to know about a player by reading the stats, and not with the stats themselves.

The position that you seem to be taking is exactly the one that I am decrying, which is that if statistics have the value that you claim that they have then it is unnecessary to watch the games. You can learn everything that you need to know about a basketball player by looking at the box scores.

If player A has a better DRPM than player B then you can't turn that into the statement of Player A is a better defender than player B, but scores of people do just that, and THAT is what I'm complaining about. I am not arguing against what advanced stats are. I'm arguing against the sweeping conclusions that people draw from advanced stats.

Dude everything requires context. The eye test requires context too. You see someone perform poorly and conclude that they're a bad player. However, you do further investigation and find out they had an undisclosed injury. You see someone perform extremely well and then upon further investigation, find out that their game benefits from playing at higher altitudes (more evident in baseball).

Your conclusion should just be that the person that is most informed can come to the best conclusions. The person who doesn't watch the games is less informed than someone who does. The person who doesn't use statistics is less informed than someone who does. Someone who only does one has less context and less information than someone who utilizes both.

Who is claiming that you don't need to watch the games if you use stats? It seems like a ridiculous strawman against advanced stats to imply that most of the people that are using them lack context from other sources. The people utilizing advanced stats are generally less casual than fans who don't. They've often watched more basketball and done more investigation than the people who reject them.

[76ciology]

My gut feel is this team has a "catalyst" every month:

Nov: minute restriction
Dec: jah-biid front court
Jan: Ben Simmons
Feb: Noel Trade
March: Sixers deathball (combination of our young players)
April: 2017 draft hype

http://www.nba.com/sixers/news/seltzers-notebook-brown-liking-okafors-contributions-still-focused-fundamentals

Just in time

[HartfordWhalers]

Is it possibly to interpret anything without context?

How is this a discussion? I definitely need more background to understand this being a topic of debate.

[Ericb5]

I mean, you're wrong. Blocks per game and DRPM do not have the same strengths and weaknesses. Blocks per game is freaking blocks divided by games, DRPM is a regression that models how much better your team does defensively with you on the floor taking into account your teammates and oppponents. They aren't even measuring the same thing.

You keep talking about how advanced stats have outside factors that they don't consider, but it's pretty obvious that you have no idea what factors some of these do consider. RPM DOES attempt to consider the quality of your teammates and your impact beyond how how many shots you block. Thats kind of the point of it. It's not perfect, but I don't see how you can "fill in the blanks" when you don't even know what the blanks are.

And are you under the impression that blocks per game is an advanced stat? It's not. Neither is interceptions per game for that matter.

None of these stats I've mentioned purport to project how a guys career will go. Every projection model I've actually seen says that embiid will be a future hall of famer if he stays healthy. Like I get that he has a unique background but the guy has been nothing but productive at Kansas and here that shouldn't be a shocker.

Again I'm not sure exactly what we're arguing about here. And you keep talking about "advanced stats" like it's one thing when it's not. Some are more valuable than others depending on the situation and the player. It really just seems like you're arguing against your idea of what "advanced stats" are, which I can't help you with.

You are simply not understanding what I am writing, and maybe that is my fault.

I am not calling blocks per game an advanced stat OBVIOUSLY. It is a counting stat like Interceptions per game.

You have again tripped over the same point which I tried to clarify here:

I'm not saying two stats like DRPM, and Blocks per game are equally relevant at all. I'm saying that advanced stats as a category, and traditional stats as a category, share the same strengths and weakness in terms of their value to ascribing merit to a player. They both are susceptible to outside factors obscuring the truth.

DRPM is an advanced stat, and Blocks per game is a conventional stat, but that both CATEGORIES(not talking about specific stats, but the categories in general) suffer from the same weaknesses in that they BOTH require context to interpret. Yes, one individual stat may be stronger predicatively than another one in a given area, but that is specifically why I am arguing that you need to look at multiple things in order to come to a conclusion.

Just because an advanced stat PURPORTS to add context doesn't mean that they are able to successfully do it. Hence the nonsensical conclusions that posters draw from them. My beef is with the strident adherents to analytics that think that you can learn everything you need to know about a player by reading the stats, and not with the stats themselves.

The position that you seem to be taking is exactly the one that I am decrying, which is that if statistics have the value that you claim that they have then it is unnecessary to watch the games. You can learn everything that you need to know about a basketball player by looking at the box scores.

If player A has a better DRPM than player B then you can't turn that into the statement of Player A is a better defender than player B, but scores of people do just that, and THAT is what I'm complaining about. I am not arguing against what advanced stats are. I'm arguing against the sweeping conclusions that people draw from advanced stats.

Dude everything requires context. The eye test requires context too. You see someone perform poorly and conclude that they're a bad player. However, you do further investigation and find out they had an undisclosed injury. You see someone perform extremely well and then upon further investigation, find out that their game benefits from playing at higher altitudes (more evident in baseball).

Your conclusion should just be that the person that is most informed can come to the best conclusions. The person who doesn't watch the games is less informed than someone who does. The person who doesn't use statistics is less informed than someone who does. Someone who only does one has less context and less information than someone who utilizes both.

Who is claiming that you don't need to watch the games if you use stats? It seems like a ridiculous strawman against advanced stats to imply that most of the people that are using them lack context from other sources. The people utilizing advanced stats are generally less casual than fans who don't. They've often watched more basketball and done more investigation than the people who reject them.

I never said that anyone is claiming that they don't need to watch the games. I'm saying that the purely statistical arguments make it seem as if they think that.

I agree completely that context is required for all of this to make sense. Statistics, and watching the games together give you the most complete picture. That is sort of my point.

You need to take everything in context to evaluate players. You can't simply say that one player is better than another player by citing a statistic.


Sent from my iPhone using Tapatalk

[Ericb5]

This is kind of what I was talking about earlier though. Go through their respective basketball reference pages. When it comes to all of the advanced stats Barkley blows Pippen out of the water. Barkley is 10th all time in win shares per 48 while Pippen isn't in the top 100. Even just comparing their peaks Barkley is clearly a lot better. If anything I would say that the advanced stats criminally underrate Pippen and somewhat overrate Barkley because most of them could do a better job accounting for defense.

Again, your issues with advanced stats seem to more stem from your preconceived notions rather than actually bothering to see what's under the hood... You can't provide the missing context to something if you don't know what the context is that's missing.

True.

I don't know the numbers and what they stand for as well as you do, but I know that I disagree with them in many cases when I see them when they don't comport with what I have seen with my own eyes. The stats are better today than they were 10 years ago though, and I'm sure that they will continue to get better.

Is there a stat that measures the attention from the defense that a given player receives? Something that will show the difference between what Iverson faced in 2001 and what Kobe faced?

I watched every Sixers game that year, and lots of Lakers ones too, and Kobe was able to float around and was hardly ever double teamed, or even concentrated on, whereas Iverson was the complete focus of the defense for whichever team he played every single game.

Every Sixers opponent knew that you had to stop Iverson, and every Lakers opponent knew that you had to stop Shaq.

If there is some statistical measure that would show that discrepancy it would go a long way to convincing me that statistics can represent reality. Even back then people would cite Kobe's higher field goal percentage as evidence that he was a better shooter than Iverson. I'm still disgruntled from those battles. Lol


Sent from my iPhone using Tapatalk

I would say that advanced stats both grade Iverson and Kobe as both great players with neither one being conclusively better than the other in 2001. Either way going on a case by case basis sort of misses the point...

Theoretically if drawing attention like that if it is in fact valuable would result in your team doing better when you are on the floor, but we've been over the issues with that. The SportVU cameras that are installed in most (all?) arenas now are still just scratching the surface on how much attention a player draws and how that opens up opportunities for other players. (Stuff like Iverson, or how someone like John Wall can drive and set up teammates in the corners for an open shot, or how a player much cover Kevin Durant in the corner more tightly than Harrison Barnes which opens up the lane for others) People call it "gravity." It's going to be the next great advancement IMO.

Cool.

That has always been a big missing piece in my mind.


Sent from my iPhone using Tapatalk

[Ericb5]

Is it possibly to interpret anything without context?

How is this a discussion? I definitely need more background to understand this being a topic of debate.

It's a discussion from my standpoint because people will say that player A is a better defender than player B, and then will cite a single statistic as proof, and I'm arguing that there needs to be context for any of these statistics, and there is no single way to distill something like this down to a single number.

It is like using IQ to determine intelligence. IQ is one aspect of intelligence, and there are many different forms of intelligence.


Sent from my iPhone using Tapatalk

[Sixerscan]

Is it possibly to interpret anything without context?

How is this a discussion? I definitely need more background to understand this being a topic of debate.

Well my end was that you're not in a position to really give context to a stat if you don't know what goes into the stat. Otherwise it's just a blind critique based on your anecdotal biases. (Which sort of defeats the purpose)

[Ericb5]

Is it possibly to interpret anything without context?

How is this a discussion? I definitely need more background to understand this being a topic of debate.

Well my end was that you're not in a position to really give context to a stat if you don't know what goes into the stat. Otherwise it's just a blind critique based on your anecdotal biases. (Which sort of defeats the purpose)

That's fair, but whether or not i know all of the things that a given stat is trying to incorporate, doesn't change the fact that the stats are imperfect in practice and often disagree with observation.

Stats are digital, and observations are analog. Stats are basically like automated approximations of real world things. They are getting better though.

The "gravity" thing that you mentioned above is a huge thing, and if they can figure that out then the accuracy will go up significantly. Currently, advanced stats are trying to be scientific, but in reality they are not following the scientific method because the controls aren't consistent(variance in gravity between players), and therefore the conclusions have a large margin of error that equate to an area of probability, instead of an actual data point.

[76ciology]

32.7% of our points are 3pters. That is third overall in the league behind Cavs and Rockets.

[Sixerscan]

Embiid has a higher VORP than Holmes now. Our long national crisis is over.