Top 25 peaks of the 2001-25: #11-#12 Spots

[f4p]

and this is something that tends to get held against harden, who admittedly is not great in game 7's. but people act like the initial games were just destined to end up at game 7 no matter what happened. you see it in something like the 2023 boston series, where harden wins game 1 without embiid, on the road in boston, with 45 points and a game-winning 3. a completely unexpected win. and then in game 4 has another 40 point game and hits the shot to force OT and the game winner in OT. then he plays poorly in game 7 and it's like "can't count on harden". except literally the only reason it got to game 7 is harden having 2 high end games no one expected at that point in his career. it's like it would have been better to just suck in game 1 and lose in 5 or 6 and avoid the game 7 criticism (kind of the "it would be better if lebron lost in the 1st round than in the finals" argument). or last year, harden is great in game 1 against denver but loses b/c kawhi is terrible and then is the best player on either team in game 3 but then he's bad in game 7 so you can't count on harden. it's essentially treated like every series is fated to be in game 7, no matter what harden does in the beginning, so then only game 7 counts.

or something like the 2019 series against the warriors. steph plays terribly all series but then is great in the 2nd half of game 6 while harden is great all series but merely "good" in the 2nd half of game 6, so it's harden's fault and you can't count on harden. apparently when other people play well in the back half of a series, it leads to their teams winning, but by that logic harden should be winning all his series 4-0 or 4-1 if all you have to do is play well and then you win. but like you said, it's treated like the first 6 games are irrelevant, no matter how well harden plays, he then has to also play well at the end of the series, while other people can be bad for large stretches of the series as long as they are good in game 6 or 7 because apparently their teams will just reach game 6 or 7 regardless of how they actually play.

I don't think we should spend too much time on this but I think there's many forms of context within a playoff series. Much like how I don't like the way the 08 finals ended for the Lakers. Even though a lot of people will use that as his peak, the way the Lakers lose a 25 pt lead then get blown out in game 6 by like 40 to me isn't what I'd want to see in a player's peak. It's not really about game 7's or losing 3-2 leads so much as that there is a lot of context that can be looked at and seen as a + or a -. Harden's last Sixers series is one of the more up and down series a player has ever had but again I doubt anyone will use that as his peak so its sort of irrelevant. He also tends to get a bit of a pass for losing in 2018 due to CP3 being out.

what do you mean he gets a pass. a pass for what exactly? for only almost knocking off a top 10 all time team that crushed everyone else? and "only" averaging 29/7/6 against the #1 playoff defense?

[Outside]

That there's always context is absolutely true.

I'm not sure 4-0, 4-1 is directly relevent as I'm really just discussing order rather than a different final series score. But even I would say (I think I'm agreeing here) that 4-0 doesn't necessarily mean unwinnable (again further context required).

I don't think we're disagreeing here as far as I can tell (I may be wrong) ... you seem to be primarily explaining the "what is" regarding the perception of (a generic?) "we". I just don't see the merits of why that should be the case. I'll make clear why I think it shouldn't be.

Players played as well as they did over the series.

We could estimate "winnablility" before the series. We could do so again after with more information (though perhaps, typically, many might do so with some outcome inevitability bias) with accounts for what we are holding constant and what we aren't. I don't see the merit of evaluating a series by taking a moment in time as if a given (say 3-0 rather than 0-3 ... whilst knowing in each hypothetical case that it goes to 3-3 and then 3-4) ...

Of course, it is true at that point Team A (that's up 3-0 against Team C) is in a "more winnable" (higher win probability) series than Team B (down 0-3 against Team D) ... but even here just evaluating the teams (not even the player) on the series ... they had to get to that point ... winning those games was something they did. Treating part of a series as a given because it did happen and then saying a latter part probably shouldn't have ... doesn't make much sense to me.

Personally if it's a player evaluation I'd prefer to look at how the player played though.

I think even if we just narrow it down to game 7's that a game 7 always carries more weight when we look at how a player performed in a series. So if a guy plays poorly in a game 7 loss it tends to matter more than how he played in a game 1 or 2 loss. Of course you can say that maybe it shouldn't but everyone knows going into a game 7 that this is for all the marbles while that doesn't hold true for game 1 or 2. I think anyone who's competed in sports knows the difference. Competition is meant to bring out the best in those competing. So personally ya, I am going to look a bit harder at how players did in those last 1-2 games in a series for better or worse. To me that's when the pressure is turned up and I want to see who did the best.

The error here, and I argue it is an error, is in failing to acknowledge that game 7 is only ever played ... because games 1-6 turned out as they did. If we were to assume that all series were predestined to go 3-3, that that were a given ... it would make sense to weight game 7 more. I don't mind the argument that game 7 is a different atmosphere/experience. It's just not more valuable. Given the choice, holding all else equal, I'd marginally prefer a squad (or player) more likely to be good early such that we're more likely to sweep or 4-1 and save each player on the team extra pounding and extra travel. Of course in an ideal world I'd like my players to compete in every competitive game they play.

In simplified terms if you promise to give me one of games 1-4, just don't turn up ... I'll give you the same promise for game 7. And you'll never gain from taking that (at best, net 1-1) and I have a pretty good chance of a free win. We could get more complex factoring in homecourt and perhaps other things, but the fundamental point stands.

"All the marbles" only works if you take everything prior as a given. Some sports will bake in rules and structures to make late count more to (artificially?) keep the outcome in doubt ... it's what F1 did in 2014 and you could argue it's why North American sports have a playoff system that doesn't celebrate or champion the team that is best over the totality of a season but a small part of it (and - where all know this is the case, agree to it then in that sense this is fine, "fair" etc and teams/participants should strategize accordingly) ... but if it isn't the case (G7 doesn't count as, say, 3 wins) why act act as though it is?

Apologies, not part of this project, but this is interesting to me, so I'm chiming in.

Your point is valid, but it leaves the impression that performance in game 7s is no more valuable than any other game, and while many may over-mythologize performance in 7s, I don’t think it's correct to say it's no different from any other game.

Game 7s are different in that the pressure is more intense, much like it is in the fourth quarter versus earlier ones even though a basket counts the same. I would expand the game 7 pressure aspect to include any closeout game in a closely contested series. Players who perform well in the fourth quarter, especially down the stretch, are considered on a higher tier from those who don't, and that applies to those who perform well in closeout games, particularly game 7s. The ability to perform under that heightened pressure is a nebulous quality, but it's something. Some players thrive under those conditions, and some shrink from the moment.

Players like Harden and Chris Paul may be unfairly judged by that standard considering how well they play in other games, or even in closeout games when they play well but their team loses, but it's fair to credit players who break through those barriers by performing well and winning. LeBron fairly deserves criticism for his performance in the 2011 finals when he didn’t perform well overall, especially in fourth quarters, despite having played well in prior rounds. He also deserves credit for coming back from that experience and playing well in subsequent finals.

I do think you acknowledged this in your post ("I don't mind the argument that game 7 is a different atmosphere/experience"), so we're likely on the same page for the most part. I agree that a player who is exceptional enough in earlier games to close out a series before it reaches game 7 is just as valuable as one that helps you win game 7. Winning in dominant fashion so that a star can be on the bench during the last five minutes is as valuable as one that performs well in crunch time. But eventually, players will face a situation with enhanced pressure, such as a close game in crunch time or a game 7, and that is another test of their greatness. Players who perform well under those circumstances deserve recognition for it.

[Cavsfansince84]

what do you mean he gets a pass. a pass for what exactly? for only almost knocking off a top 10 all time team that crushed everyone else? and "only" averaging 29/7/6 against the #1 playoff defense?

What I am saying is he tends to get credit for leading or co-leading them to a 3-2 lead in that series but those last two games aren't really used against him much due to CP3 being out. It's not the same as leading them to a series win but people are free to make w/e they want out of those last two games which has been hashed and rehashed numerous times so let's not into that right now. That was my point though. You're taking it in a direction that it wasn't intended to go with what I said.

[Owly]

I think even if we just narrow it down to game 7's that a game 7 always carries more weight when we look at how a player performed in a series. So if a guy plays poorly in a game 7 loss it tends to matter more than how he played in a game 1 or 2 loss. Of course you can say that maybe it shouldn't but everyone knows going into a game 7 that this is for all the marbles while that doesn't hold true for game 1 or 2. I think anyone who's competed in sports knows the difference. Competition is meant to bring out the best in those competing. So personally ya, I am going to look a bit harder at how players did in those last 1-2 games in a series for better or worse. To me that's when the pressure is turned up and I want to see who did the best.

The error here, and I argue it is an error, is in failing to acknowledge that game 7 is only ever played ... because games 1-6 turned out as they did. If we were to assume that all series were predestined to go 3-3, that that were a given ... it would make sense to weight game 7 more. I don't mind the argument that game 7 is a different atmosphere/experience. It's just not more valuable. Given the choice, holding all else equal, I'd marginally prefer a squad (or player) more likely to be good early such that we're more likely to sweep or 4-1 and save each player on the team extra pounding and extra travel. Of course in an ideal world I'd like my players to compete in every competitive game they play.

In simplified terms if you promise to give me one of games 1-4, just don't turn up ... I'll give you the same promise for game 7. And you'll never gain from taking that (at best, net 1-1) and I have a pretty good chance of a free win. We could get more complex factoring in homecourt and perhaps other things, but the fundamental point stands.

"All the marbles" only works if you take everything prior as a given. Some sports will bake in rules and structures to make late count more to (artificially?) keep the outcome in doubt ... it's what F1 did in 2014 and you could argue it's why North American sports have a playoff system that doesn't celebrate or champion the team that is best over the totality of a season but a small part of it (and - where all know this is the case, agree to it then in that sense this is fine, "fair" etc and teams/participants should strategize accordingly) ... but if it isn't the case (G7 doesn't count as, say, 3 wins) why act act as though it is?

Apologies, not part of this project, but this is interesting to me, so I'm chiming in.

Your point is valid, but it leaves the impression that performance in game 7s is no more valuable than any other game, and while many may over-mythologize performance in 7s, I don’t think it's correct to say it's no different from any other game.

Game 7s are different in that the pressure is more intense, much like it is in the fourth quarter versus earlier ones even though a basket counts the same. I would expand the game 7 pressure aspect to include any closeout game in a closely contested series. Players who perform well in the fourth quarter, especially down the stretch, are considered on a higher tier from those who don't, and that applies to those who perform well in closeout games, particularly game 7s. The ability to perform under that heightened pressure is a nebulous quality, but it's something. Some players thrive under those conditions, and some shrink from the moment.

Players like Harden and Chris Paul may be unfairly judged by that standard considering how well they play in other games, or even in closeout games when they play well but their team loses, but it's fair to credit players who break through those barriers by performing well and winning. LeBron fairly deserves criticism for his performance in the 2011 finals when he didn’t perform well overall, especially in fourth quarters, despite having played well in prior rounds. He also deserves credit for coming back from that experience and playing well in subsequent finals.

I do think you acknowledged this in your post ("I don't mind the argument that game 7 is a different atmosphere/experience"), so we're likely on the same page for the most part. I agree that a player who is exceptional enough in earlier games to close out a series before it reaches game 7 is just as valuable as one that helps you win game 7. Winning in dominant fashion so that a star can be on the bench during the last five minutes is as valuable as one that performs well in crunch time. But eventually, players will face a situation with enhanced pressure, such as a close game in crunch time or a game 7, and that is another test of their greatness. Players who perform well under those circumstances deserve recognition for it.

I’m not really looking to go any further with this, and we're off topic enough but I’ll engage here since you’ve bothered to have the conversation and do so reasonably.

I’ll try do this carefully because there’s stuff here I agree with and things where I’m not sure and maybe don’t.

Your point is valid, but it leaves the impression that performance in game 7s is no more valuable than any other game

It is, it does, because (from my arguments) it isn’t. Per the post you can give double points in the last race and you could choose to make game 7 a “moneyball” game worth 2 or 3 games. Right now it isn’t.

I don’t think it's correct to say it's no different from any other game.

I do think you acknowledged this in your post ("I don't mind the argument that game 7 is a different atmosphere/experience"), so we're likely on the same page for the most part.

So yes, per my post if the case is “different” I don’t particularly object. If one is arguing more valuable as you, I think, may start to allude to with “it leaves the impression that performance in game 7s is no more valuable than any other game” then that’s where my issue is. The argument isn't made explicitly but this reads to me like implied disagreement.


At the margins, on some of the general discussion I might discuss sample sizes and how carefully we’re tracking these “late” things and how much we’re influenced by team level end result and building a blame/credit story around that and then in turn a media narrative (perhaps about a “killer instinct” or stories of “calling shots”) rather than close tracking of the individual but I don’t have a problem with the idea of increased pressure.

I agree that a player who is exceptional enough in earlier games to close out a series before it reaches game 7 is just as valuable as one that helps you win game 7

Per my post … this isn’t quite agreement. Your next sentence sort of alludes to why but continues with the equal value theme.

Winning in dominant fashion so that a star can be on the bench during the last five minutes is as valuable as one that performs well in crunch time.

I expressed a marginal preference for good early. It’s marginal but being good early means the same about the same chance of advancing in a given series* except … more chance of doing it quickly … thus saving on wear, injuries, travel for all my team and thus boosting my chances in future rounds (and in theory compounding - also boosting my chances of another quick finish). It’s marginal enough that I don’t apply it (for instance to player evaluation) in practice. But I should point out the slight distinction.
* Though – as I’ve highlighted - depending how we conceive of it manifesting itself, the “good late” might not get to play at a series level.

But eventually, players will face a situation with enhanced pressure, such as a close game in crunch time or a game 7, and that is another test of their greatness. Players who perform well under those circumstances deserve recognition for it.

Everyone who has made it to this level has already, I would venture, faced a situation with “enhanced pressure”. But okay. And all else equal you’d prefer good under pressure than not.
Is going 7 an inevitability? It’s probable to sometimes happen over a career. Games 1-4 are locked in though as things that will be played.
But sure, recognition for how they perform. I don’t see anyone disputing that. I’m just not sure it requires a multiplier or extra credit or whatever.

Players like Harden and Chris Paul may be unfairly judged by that standard considering how well they play in other games, or even in closeout games when they play well but their team loses, but it's fair to credit players who break through those barriers by performing well and winning.

It depends what you mean here. And I want to be fair here and the path for this train of thought isn’t exactly clear to me here.
One the one hand you grant team performance judgement might be unfair … “but it's fair to credit players who break through those barriers by performing well and winning”.
And it’s not clear to me the subtext of including the last two words (not entirely sure what “the barriers” are but the choice to credit the performance and then, separately, given performance already noted, team-level outcome also requires credit is where my radar is triggering to potentially sound the alarm). Or to what extent this is mitigated the prior stuff about not punishing people who incidentally lost but didn’t “deserve” to.

And for me it comes down to how well you think you can assess individuals. If someone thinks everyone on the winning team is better than everyone on the losing one or that the players on the winning team get some bonus or that winning makes the performance better (rather than better performance make winning more likely) … I’m going to disagree with them. I prefer a model where you attempt to honestly, seriously, to the best of your abilities analyze and credit the performance. And that’s difficult and not without substantial dispute.


Fwiw, I’ll just add, not "at" anyone but because I’ve seen it thrown out in the past as if some killer point that everyone missed that “winning matters” … which (depending somewhat what you mean by it) is true … and (depending what you mean by it) not contested … and misses the point. It’s imprecise. Winning is the end goal of the team in a team sport. Hopefully all players are pulling for that goal. Players contribution towards making that goal possible is what we are seeking to measure. But that doesn’t mean team level binary outcomes are good measure of individual performance.

[Outside]

Spoiler:

The error here, and I argue it is an error, is in failing to acknowledge that game 7 is only ever played ... because games 1-6 turned out as they did. If we were to assume that all series were predestined to go 3-3, that that were a given ... it would make sense to weight game 7 more. I don't mind the argument that game 7 is a different atmosphere/experience. It's just not more valuable. Given the choice, holding all else equal, I'd marginally prefer a squad (or player) more likely to be good early such that we're more likely to sweep or 4-1 and save each player on the team extra pounding and extra travel. Of course in an ideal world I'd like my players to compete in every competitive game they play.

In simplified terms if you promise to give me one of games 1-4, just don't turn up ... I'll give you the same promise for game 7. And you'll never gain from taking that (at best, net 1-1) and I have a pretty good chance of a free win. We could get more complex factoring in homecourt and perhaps other things, but the fundamental point stands.

"All the marbles" only works if you take everything prior as a given. Some sports will bake in rules and structures to make late count more to (artificially?) keep the outcome in doubt ... it's what F1 did in 2014 and you could argue it's why North American sports have a playoff system that doesn't celebrate or champion the team that is best over the totality of a season but a small part of it (and - where all know this is the case, agree to it then in that sense this is fine, "fair" etc and teams/participants should strategize accordingly) ... but if it isn't the case (G7 doesn't count as, say, 3 wins) why act act as though it is?

Apologies, not part of this project, but this is interesting to me, so I'm chiming in.

Your point is valid, but it leaves the impression that performance in game 7s is no more valuable than any other game, and while many may over-mythologize performance in 7s, I don’t think it's correct to say it's no different from any other game.

Game 7s are different in that the pressure is more intense, much like it is in the fourth quarter versus earlier ones even though a basket counts the same. I would expand the game 7 pressure aspect to include any closeout game in a closely contested series. Players who perform well in the fourth quarter, especially down the stretch, are considered on a higher tier from those who don't, and that applies to those who perform well in closeout games, particularly game 7s. The ability to perform under that heightened pressure is a nebulous quality, but it's something. Some players thrive under those conditions, and some shrink from the moment.

Players like Harden and Chris Paul may be unfairly judged by that standard considering how well they play in other games, or even in closeout games when they play well but their team loses, but it's fair to credit players who break through those barriers by performing well and winning. LeBron fairly deserves criticism for his performance in the 2011 finals when he didn’t perform well overall, especially in fourth quarters, despite having played well in prior rounds. He also deserves credit for coming back from that experience and playing well in subsequent finals.

I do think you acknowledged this in your post ("I don't mind the argument that game 7 is a different atmosphere/experience"), so we're likely on the same page for the most part. I agree that a player who is exceptional enough in earlier games to close out a series before it reaches game 7 is just as valuable as one that helps you win game 7. Winning in dominant fashion so that a star can be on the bench during the last five minutes is as valuable as one that performs well in crunch time. But eventually, players will face a situation with enhanced pressure, such as a close game in crunch time or a game 7, and that is another test of their greatness. Players who perform well under those circumstances deserve recognition for it.

I’m not really looking to go any further with this, and we're off topic enough but I’ll engage here since you’ve bothered to have the conversation and do so reasonably.

I’ll try do this carefully because there’s stuff here I agree with and things where I’m not sure and maybe don’t.

Your point is valid, but it leaves the impression that performance in game 7s is no more valuable than any other game

It is, it does, because (from my arguments) it isn’t. Per the post you can give double points in the last race and you could choose to make game 7 a “moneyball” game worth 2 or 3 games. Right now it isn’t.

I don’t think it's correct to say it's no different from any other game.

I do think you acknowledged this in your post ("I don't mind the argument that game 7 is a different atmosphere/experience"), so we're likely on the same page for the most part.

So yes, per my post if the case is “different” I don’t particularly object. If one is arguing more valuable as you, I think, may start to allude to with “it leaves the impression that performance in game 7s is no more valuable than any other game” then that’s where my issue is. The argument isn't made explicitly but this reads to me like implied disagreement.


At the margins, on some of the general discussion I might discuss sample sizes and how carefully we’re tracking these “late” things and how much we’re influenced by team level end result and building a blame/credit story around that and then in turn a media narrative (perhaps about a “killer instinct” or stories of “calling shots”) rather than close tracking of the individual but I don’t have a problem with the idea of increased pressure.

I agree that a player who is exceptional enough in earlier games to close out a series before it reaches game 7 is just as valuable as one that helps you win game 7

Per my post … this isn’t quite agreement. Your next sentence sort of alludes to why but continues with the equal value theme.

Winning in dominant fashion so that a star can be on the bench during the last five minutes is as valuable as one that performs well in crunch time.

I expressed a marginal preference for good early. It’s marginal but being good early means the same about the same chance of advancing in a given series* except … more chance of doing it quickly … thus saving on wear, injuries, travel for all my team and thus boosting my chances in future rounds (and in theory compounding - also boosting my chances of another quick finish). It’s marginal enough that I don’t apply it (for instance to player evaluation) in practice. But I should point out the slight distinction.
* Though – as I’ve highlighted - depending how we conceive of it manifesting itself, the “good late” might not get to play at a series level.

But eventually, players will face a situation with enhanced pressure, such as a close game in crunch time or a game 7, and that is another test of their greatness. Players who perform well under those circumstances deserve recognition for it.

Everyone who has made it to this level has already, I would venture, faced a situation with “enhanced pressure”. But okay. And all else equal you’d prefer good under pressure than not.
Is going 7 an inevitability? It’s probable to sometimes happen over a career. Games 1-4 are locked in though as things that will be played.
But sure, recognition for how they perform. I don’t see anyone disputing that. I’m just not sure it requires a multiplier or extra credit or whatever.

Players like Harden and Chris Paul may be unfairly judged by that standard considering how well they play in other games, or even in closeout games when they play well but their team loses, but it's fair to credit players who break through those barriers by performing well and winning.

It depends what you mean here. And I want to be fair here and the path for this train of thought isn’t exactly clear to me here.
One the one hand you grant team performance judgement might be unfair … “but it's fair to credit players who break through those barriers by performing well and winning”.
And it’s not clear to me the subtext of including the last two words (not entirely sure what “the barriers” are but the choice to credit the performance and then, separately, given performance already noted, team-level outcome also requires credit is where my radar is triggering to potentially sound the alarm). Or to what extent this is mitigated the prior stuff about not punishing people who incidentally lost but didn’t “deserve” to.

And for me it comes down to how well you think you can assess individuals. If someone thinks everyone on the winning team is better than everyone on the losing one or that the players on the winning team get some bonus or that winning makes the performance better (rather than better performance make winning more likely) … I’m going to disagree with them. I prefer a model where you attempt to honestly, seriously, to the best of your abilities analyze and credit the performance. And that’s difficult and not without substantial dispute.

Fwiw, I’ll just add, not "at" anyone but because I’ve seen it thrown out in the past as if some killer point that everyone missed that “winning matters” … which (depending somewhat what you mean by it) is true … and (depending what you mean by it) not contested … and misses the point. It’s imprecise. Winning is the end goal of the team in a team sport. Hopefully all players are pulling for that goal. Players contribution towards making that goal possible is what we are seeking to measure. But that doesn’t mean team level binary outcomes are good measure of individual performance.

Thanks for your reply. I'll just add that I particularly liked your last paragraph above. It is difficult to assess individual impact on a team sport like basketball, including individual impact on winning, which is the ultimate goal. I appreciate your perspective, and you've made me rethink some assumptions I've held and will hopefully think about differently going forward.

I can't keep up with many people here from a statistical/data perspective, but I love the game, and I think I understand and appreciate aspects of it more than many fans. My nature is also to look at the game from a narrative* perspective, which I do think can add value, but that can lead me astray at times. The nice thing about narratives is that they benefit from being edited, reworked, refined.

*Depending on the context, "narrative" is a word that can be fraught with negative implications. I come from an English literature and writing perspective, so in this case, narrative is a good thing, used to build a story that reveals the fundamental nature of a character or the world. I tried to think of a different word that would convey the meaning without the baggage, but I do love the word, so I kept it.

[kcktiny]

I can't keep up with many people here from a statistical/data perspective

Knowing this...

It is difficult to assess individual impact on a team sport like basketball, including individual impact on winning

What would it take for you to change your opinion on this? What would you have to see?

[Outside]

I can't keep up with many people here from a statistical/data perspective

Knowing this...

It is difficult to assess individual impact on a team sport like basketball, including individual impact on winning

What would it take for you to change your opinion on this? What would you have to see?

I'm not sure what your point is here. Just because I'm not as literate in statistics as many posters here doesn't mean I don't understand the basic concepts and limitations of, say, RAPM, impact metrics, and box stats. It also doesn't mean I don't understand the inherent problem of determining individual impact on winning in a team game.

[kcktiny]

doesn't mean I don't understand the basic concepts and limitations of, say, RAPM, impact metrics, and box stats

Who does? These are flawed, are they not?

It also doesn't mean I don't understand the inherent problem of determining individual impact on winning in a team game.

Which is why I asked.

I think I understand and appreciate aspects of it more than many fans.

In your opinion would there be perhaps some maybe video component that would help, or certain new defensive statistics, or specific sensor acquired data you think would help?

For example there is rebounding data for each player but not data for how often a player prevents an opponent from grabbing a rebound. Or stats.nba.com tracks things likes deflections, loose balls recovered, contested shots, defensive shot data, box out data, and more. None of these are used by the metrics you mentioned.

Do you have an opinion on what you would like to see in the future for helping to determine player value?

[Outside]

doesn't mean I don't understand the basic concepts and limitations of, say, RAPM, impact metrics, and box stats

Who does? These are flawed, are they not?

It also doesn't mean I don't understand the inherent problem of determining individual impact on winning in a team game.

Which is why I asked.

I think I understand and appreciate aspects of it more than many fans.

In your opinion would there be perhaps some maybe video component that would help, or certain new defensive statistics, or specific sensor acquired data you think would help?

For example there is rebounding data for each player but not data for how often a player prevents an opponent from grabbing a rebound. Or stats.nba.com tracks things likes deflections, loose balls recovered, contested shots, defensive shot data, box out data, and more. None of these are used by the metrics you mentioned.

Additional data incorporated into metrics would theoretically be helpful, but too much peripheral or supplementary data may also drown out core data. Much would depend on how the additional data is weighted. In general though, I'm inclined to think that more data would be helpful in creating more robust metrics.

Do you have an opinion on what you would like to see in the future for helping to determine player value?

We're probably in, maybe not the infancy, but the early stages of the databall era. Defense, in particular has too little individual impact data available.

Certain players point out the limitations of currently available data. How do you quantify the gravity of Curry? Of Shaq? The impact of shots not taken because of Wemby's length? The offensive genius of Jokic? How can we quantify fit or chemistry between players? Portability? The ability of players to adapt to different offensive or defensive systems? And for all these things, if you do have data, how would you apply it? APM-type data begins to get at these things but doesn't fully capture them.

Improvements will come with the combination of additional data being available, but the key component is people with expertise about both statistics and basketball. They are in the best position to determine which additional data is helpful and how to use that data once it is available. Such people are rare. Perhaps it will be pairings or teams of people, like Billy Beane and Peter Brand as portrayed in Moneyball.

But I'm skeptical that individual impact on winning will ever be completely, definitively quantifiable. Basketball is a team game, with more complex interaction compared to most team sports, and with players having a higher degree of both offensive and defensive responsibility and interaction with teammates than most sports. Parsing out individual impact from that is inherently problematic. Defining impact on winning is perhaps an order of magnitude more difficult.

Is it impossible? I don't know. With a variety of tools, we do what we can to approximate it, but as this forum shows, there are disagreements even among those who understand the nature of the available tools and their limitations and approach discussions with good faith and an open mind. Everyone has biases and blind spots.

Even with better tools, we have the problem of how to assess players from prior eras. Also across eras, considering how much the game differs over time. That may be the most difficult aspect to the discussion.

I hope that addresses your question.

[kcktiny]

Defense, in particular has too little individual impact data available.

stats.nba.com shows individual player shot defense, taken from camera data:

https://www.nba.com/stats/players/defense-dash-overall?Season=2024-25&PerMode=Totals

So for example this shows that this past season Amen Thompson in 729 FGAs faced on defense allowed just a 40.7% FG%, but Jordan Poole allowed a 52.2% FG% on 630 FGAs faced. I find this defensive data valuable.

They also break this defensive shot data down to just 2s, just 3s, and to <6' of the basket, <10' of the basket, and from >15' and out. That's a lot of defensive information.

And this is publicly available defensive data.

The impact of shots not taken because of Wemby's length?

You can look at this data and simply divide a player's FGAs faced on defense by their minutes played and multiple by some constant (say 40) to get FGAs faced on defense per 40 min for all players.

So for example last season Giannis Antetokounmpo faced just 13.8 FGA/40min but Brook Lopez faced 23.9 FGA/40min and Bobby Portis 20.1 FGA/40min. That's a huge discrepancy. The question is why? If you watch the Bucks you would know that few on offense challenge Giannis on defense on a regular basis.

How can we quantify fit or chemistry between players?

There is lineup data available too, best 5 man, 4 man, 3 man lineups. Does this help for this?

Improvements will come with the combination of additional data being available

So this is my question. What new data would you (i.e. anyone) like to see?

but as this forum shows, there are disagreements even among those who understand the nature of the available tools and their limitations

Not sure what available tools you are describing, but in relation to the plus/minus stuff I find it funny/ironic how it is often discussed as gospel here yet the above mentioned publicly available defensive data - which this plus/minus does not use - has been around for more than a decade.

[Outside]

I didn't say there wasn't any defensive data available or that the data available wasn't valuable. Modern defensive schemes are complex and variable by team. The best defenses have all players "on a string," anticipating each other's moves, doubling and collapsing according to scheme. Does the available data identify a weak defender in an otherwise strong defense, especially when the scheme is designed to hide that player's weakness? Can it identify a strong team defender in an otherwise mediocre or bad defense when the end result is the defense giving up points?

My sense is that the available data allows us to identify the best defenders but may not necessarily distinguish between them or between less than elite defenders.

You're asking me to specify what additional data is needed, but as I pointed out, I am not well-versed in what statistics are available or have the knowledge in statistics to know how best to use them or what additional data to ask for.

I benefit from discussion by knowledgeable posters here and try to occasionally add my perspective from my limited knowledge base. This all started with me responding to a post by one of those knowledgeable posters (Owly) where I attempted to add my meager perspective and show appreciation for posters like them who are central to the quality discussion here. I'm not the right person to answer your questions, but I've attempted to do so as best I can.