GOAT lists and the metaphysics of player comparisons

[Tim Lehrbach]

I'm gonna choose this moment to quietly re-enter the Player Comparisons fray after a long time away from regular participation. Such re-entry will be clumsy but, I hope, not solipsistic. I want to engage and support the work done here, but I am never quite sure how my POV fits or what I have to contribute. I want to learn from y'all, but I am never quite sure of the right questions to ask. After many deleted drafts and false starts over the years, I'm just gonna throw some text up and see if anything sticks, or at least see if it loosens me up enough to start posting again.

*******

"There is no GOAT."
"GOAT debates are useless."
"Ranking and measuring competitors with ever-increasing precision is a myopic view of sport."
etc.

These were my usual responses of old to any big questions at Player Comparisons and around the water cooler IRL. Of course, there's no bigger or more popular question than deciding the Greatest of All Time. I've long disbelieved in ranked GOAT lists because the question concerning superlative greatness will always mean different things to different individuals, groups, generations, cultures, etc. And, even supposing that there could be one best or most meaningful referent for greatness, it seems impossible (and, to my mind, right that it should be impossible) to arrive at a perfect measurement of it. Although I haven't posted much here in the last decade or so, my real-world discussions have continued to go little farther than a curmudgeonly disapproval of or bemusement at the exercise.

Lately (i.e., the last 3 years or so), I've increasingly softened my opposition and opened myself up to the idea that behind every GOAT list, there is an attempt to celebrate what one values in sports and in its sportsmen and women. While I may continue to doubt the importance of quarreling over where players rank in an absolute sense, I've begun paying more attention to the arguments people advance for why this player or that one is extraordinarily worthy (or less worthy, depending) of celebration. When people's criteria differ, they are expressing differences -- often very, very subtle differences, but interesting ones nevertheless -- of what they want to see exalted in an all-time great. When criteria more or less agree, but rankings differ, people represent differing stories in support of competing accounts over a shared value. Still other times, criteria may seem to agree, but the contests between stats and stories reveal that the speakers/writers mean different things by the quality in dispute.

Gradually, all of this has begun to seem a lot more interesting to me.

A second thing that's happened, alongside of this development, is admitting to myself that I do believe there are, at least, greater and lesser narratives, supported by data and oral histories, around top sportsmen. No, I do not believe there is a single GOAT, but the idea has cultural currency, and there is a more or less socially agreed upon pantheon of GOAT-tier players, which is worthy of sociological examination. And, on a personal level, am I really so different from anybody else when something inside my gut tells me I want to see one player elevated over another in order to reflect the values such elevation promotes? I must conclude that I am not!

Finally, I've softened on GOAT lists because I've realized that my original position -- namely, that there cannot possibly be any such official list -- is precisely what should grant me permission to engage in the exercise without excessive seriousness. Previously I had been overly concerned that ranking players according to one or more GOAT metric is potentially a harmful activity for appreciation of the sport and its place in our lives and culture. Now, I have come around to believe that it is very well possible to explore the ideas behind our comparative evaluations of players -- even using the GOAT as the vehicle -- without committing ourselves to the centrality of any one thesis about ranking and measuring competitors. In other words, ranking and measuring need not limit our view of sport altogether the way I feared before. Rather, it can be a lens we turn back on ourselves to examine why we support the players and teams we do, and in turn what this says about ourselves.

In short, there is no need to oppose an activity which reveals interesting narratives and values, which educates participants about the game and its important figures, and which is ultimately harmless no matter how imperfect the answers it yields may be.

Why post my GOAT confessional? I suppose my purposes are threefold. One, I want to reintroduce myself to the board as a participant. Hi everybody. Two, my favorite conversations here are the ones that get into the "metaphysics" of player comparisons, and I'll be delighted if my awkward posts spark any such discussion, especially if such discussion increases my awareness and ability to take part in the fun projects and threads around here! Three, I expect that there are others who have shared my skepticism about the GOAT list and other ordered rankings of players, and I wonder what their own relationship to these topics is like. Do you also struggle with questions like:

How do we deal with the under-determined nature of the questions we ask concerning relative greatness?
The uncertain answers?
The disagreements over what we value in basketball (i.e., what is greatness)?
What does a well-conceived idea about greatness look like?
How do we distinguish between better and worse answers to questions that invoke subjectivity or non-numeric factors?
Has anybody put forward something like a glossary to explain the meaning of their terms?
Where does character fit into discussions about the GOAT? Cultural import? How can we hope to fairly or accurately measure these?
Are we better off dismissing, as some do, anything which cannot be quantified, or does this sacrifice too much?
What do we take ourselves to be doing when we argue the relative merits of players? Why is this meaningful to us?
What does basketball mean to us?

I know that a lot of these questions are taken up in the projects themselves, so I'd appreciate any links to greatest hits. I'd also love to have a standalone conversation over any of them.

*******

Anyway, that mess is all I've got for now. I will attempt to bring some of my questions into PC conversations where I can, but please do not hesitate to jump in here wherever you might have thoughts to offer me. Contrasting my heretofore underdeveloped conceptions about player comparisons with the conversations I've seen around here (e.g., concerning the GOAT and other lists, criteria, and methods), I know that a lot of you have gone much deeper into these questions than I have and are further along in understanding your relationship to basketball and your answers to these kinds of questions. This, in turn, has unlocked your ability to thoughtfully engage in player comparisons in a manner that remains elusive to me. I'd love to benefit from your wisdom!

[LukaTheGOAT]

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

[Slax]

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

I love Ben Taylor's work on this topic. He does such a great job of clarifying the different ways you can think about ranking players across eras, what the rankings actually mean, how much ambiguity there is when players get compared against each other, the different approaches you can take to balancing performance and longevity when making comparisons, etc. I find all that stuff more valuable than his actual rankings.

[penbeast0]

I think the GOAT lists are more an excuse to compare great players than any kind of definitive ranking . . . not even a definitive ranking here on the PC board. They do produce some of the best debate we have here with the statistical analysts, the more holistic approaches, and the eye test guys all putting in input. So, although I was skeptical on the first one here, I got on board during it and have been very active in it since.

That said, it's just a fun board for posting and I'm glad to have you back. We can use all the thoughtful posters we can find.

[tsherkin]

I'm gonna choose this moment to quietly re-enter the Player Comparisons fray after a long time away from regular participation.

OG Tim Lehrbach! Welcome back to the fold!

[Tim Lehrbach]

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

Spoiler:

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

I love Ben Taylor's work on this topic. He does such a great job of clarifying the different ways you can think about ranking players across eras, what the rankings actually mean, how much ambiguity there is when players get compared against each other, the different approaches you can take to balancing performance and longevity when making comparisons, etc. I find all that stuff more valuable than his actual rankings.

Ben Taylor must surely be the MVP of GOAT lists. I have read some but not nearly enough of his work away from RealGM, I admit. Appreciate these links because I hadn't seen the video or podcast content before. I'll enjoy them!

Taylor is great not only because he has so thoroughly and consistently analyzed players with video and statistics and put it to a digestible measurement, but because he takes the time to acknowledge and study the foundations of player comparisons and even the rough edges which must be reckoned with.

The most interesting consequence of his work for me is that he has crystalized the oft-shared (yet, pre-ElGee, challenging-to-articulate) belief that evaluating players' GOAT candidacy (or "goodness," generally) is largely about measuring how much they contribute to winning. Championship odds or championship value-added is a terrific way to encapsulate player impact on winning games (and, for some, though less so for Taylor, the most important games). In addition to being deeply researched and meticulously calculated, the approach stands out as intuitively meaningful. We of course want to know how much a player contributes to winning. He and some others have done such a good job taking a really complicated problem (i.e., "how much does a player contribute to winning basketball games?") and yielding understandable results that the approach has become impossible to ignore and difficult to deny when taking up questions concerning relative player greatness.

My memories of basketball chatter only go back to the early 90s, but tracking the evolution of GOAT conversations since then, I can say that the question of how much a player impacts winning has always been lurking beneath arguments over rings and scoring titles and legendary moments, but any idea that basketball fans and outsiders could answer that question with extensive video analysis and player impact data was basically inconceivable. Then, the blessings of high-speed Internet, plus/minus data, and our early heroes of player analysis started delivering new methods of evaluation. You all know the story from here. What I credit Taylor and others with is making sense of the noise for the uninitiated. It's pretty wild that a layperson can comfortably use concepts like CORP in intelligible conversation. We've come so far!

Two questions persist in my mind when I read Taylor and others who take similar approaches. These are not challenges to their work or anything Taylor or anybody is responsible for answering, really, but are just lingering questions for further consideration:

1. So, we've arrived at player impact on winning being an indispensable way of measuring player goodness. Let's not take this to be true by definition. We had to do conceptual work and make commitments to get there. That is to say, going back to my reference to early 90s discussion, we didn't always have in mind a concept of measurable player impact on winning, or if we did, it was crudely using box scores (although Russell vs. Chamberlain was always the counterpoint). A concrete measurement simply wasn't available to us. Thus, my question is, what have we let go of to zero in on player impact on winning as central to the conversation about greatness? To take it even a step further, are we sure that winning games and championships is the ultimate or only purpose of the sport, such that we can reduce a player's import or success as a basketball player to his impact on winning? Not all sport is primarily about winning: take rec and youth sports, for example -- the object of the game never changes, but it doesn't typically occur to us to say that the participants are more or less valuable than one another because there is nobody (or just parents, perhaps) consuming the product with a rooting interest or an analytical eye. Further, as participants our goals are seldom to be the greatest or win the most games. But when we as spectators look at elite sport, we tend to share with Lombardi the view that winning is the only thing. Measuring and ranking the competitors according to their contributions to winning follows very naturally from this. Taylor explicitly acknowledges what his player evaluations do not account for, and I appreciate that. But when we take his analysis and apply it directly to questions like "who is the GOAT?" or "who was better?" do we leave out anything else that matters to basketball or its value to us as spectators or the players as participants? I will explore these questions further at a later time, but here are two teasers: (1) players have more motivations than winning basketball games for their participation in the sport, (2) spectators have more reasons to watch and enjoy the sport than seeing winners win. I do not simply assume that these should matter to player evaluations, but I'd like to discuss the possibilities.

2. Leadership and intangibles: where do they fit into discussions of basketball greatness? JordansBulls is roundly criticized, even mocked, around here because -- sorry, JB -- we generally take his arguments to be archaic or unable to stand up to sophisticated statistical scrutiny. But give him credit for this much: JordansBulls always provides reasons for his views. One explanation of his has drawn particular interest from me of late: his oft-repeated claims that Michael Jordan is responsible for making his teammates, especially Scottie Pippen, who they were as players and winners. There are two versions of the retort to this case. First is to say that this discounts what those players actually did on the court and gives the credit to Jordan. Second is to say that this is not admissible because it is unfalsifiable. Both are valid responses. Yet, what have we really done to look into the claim he's put forward? Couldn't Michael Jordan's effects on player development and team commitment to winning have had a major impact on the Bulls' success? And, even if the statistical or impact-based outcomes are properly credited to the players who logged the minutes, made the baskets, got the stops, etc., isn't it possible we're overlooking important inputs to the ultimate product? We know that a lot goes into what makes a player succeed: genetics, work ethic, circumstance, etc. One such factor is surely the influence of others. Yet, because we can't (or don't yet know how to) measure it, I think we have a tendency ignore it. Maybe it's something that's just beyond the science of player evaluations. That's OK, but it needs to be an acknowledged limitation to the work that we do with numbers to answers questions concerning player greatness.

That's enough for now!

[Tim Lehrbach]

I think the GOAT lists are more an excuse to compare great players than any kind of definitive ranking . . . not even a definitive ranking here on the PC board. They do produce some of the best debate we have here with the statistical analysts, the more holistic approaches, and the eye test guys all putting in input. So, although I was skeptical on the first one here, I got on board during it and have been very active in it since.

That said, it's just a fun board for posting and I'm glad to have you back. We can use all the thoughtful posters we can find.

Yeah, I don't know how it's taken me so long to come around to my "soft" view of GOAT lists, but it feels liberating! Glad to be here.

[Slax]

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

Spoiler:

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

I love Ben Taylor's work on this topic. He does such a great job of clarifying the different ways you can think about ranking players across eras, what the rankings actually mean, how much ambiguity there is when players get compared against each other, the different approaches you can take to balancing performance and longevity when making comparisons, etc. I find all that stuff more valuable than his actual rankings.

Ben Taylor must surely be the MVP of GOAT lists. I have read some but not nearly enough of his work away from RealGM, I admit. Appreciate these links because I hadn't seen the video or podcast content before. I'll enjoy them!

Taylor is great not only because he has so thoroughly and consistently analyzed players with video and statistics and put it to a digestible measurement, but because he takes the time to acknowledge and study the foundations of player comparisons and even the rough edges which must be reckoned with.

The most interesting consequence of his work for me is that he has crystalized the oft-shared (yet, pre-ElGee, challenging-to-articulate) belief that evaluating players' GOAT candidacy (or "goodness," generally) is largely about measuring how much they contribute to winning. Championship odds or championship value-added is a terrific way to encapsulate player impact on winning games (and, for some, though less so for Taylor, the most important games). In addition to being deeply researched and meticulously calculated, the approach stands out as intuitively meaningful. We of course want to know how much a player contributes to winning. He and some others have done such a good job taking a really complicated problem (i.e., "how much does a player contribute to winning basketball games?") and yielding understandable results that the approach has become impossible to ignore and difficult to deny when taking up questions concerning relative player greatness.

My memories of basketball chatter only go back to the early 90s, but tracking the evolution of GOAT conversations since then, I can say that the question of how much a player impacts winning has always been lurking beneath arguments over rings and scoring titles and legendary moments, but any idea that basketball fans and outsiders could answer that question with extensive video analysis and player impact data was basically inconceivable. Then, the blessings of high-speed Internet, plus/minus data, and our early heroes of player analysis started delivering new methods of evaluation. You all know the story from here. What I credit Taylor and others with is making sense of the noise for the uninitiated. It's pretty wild that a layperson can comfortably use concepts like CORP in intelligible conversation. We've come so far!

Two questions persist in my mind when I read Taylor and others who take similar approaches. These are not challenges to their work or anything Taylor or anybody is responsible for answering, really, but are just lingering questions for further consideration:

1. So, we've arrived at player impact on winning being an indispensable way of measuring player goodness. Let's not take this to be true by definition. We had to do conceptual work and make commitments to get there. That is to say, going back to my reference to early 90s discussion, we didn't always have in mind a concept of measurable player impact on winning, or if we did, it was crudely using box scores (although Russell vs. Chamberlain was always the counterpoint). A concrete measurement simply wasn't available to us. Thus, my question is, what have we let go of to zero in on player impact on winning as central to the conversation about greatness? To take it even a step further, are we sure that winning games and championships is the ultimate or only purpose of the sport, such that we can reduce a player's import or success as a basketball player to his impact on winning? Not all sport is primarily about winning: take rec and youth sports, for example -- the object of the game never changes, but it doesn't typically occur to us to say that the participants are more or less valuable than one another because there is nobody (or just parents, perhaps) consuming the product with a rooting interest or an analytical eye. Further, as participants our goals are seldom to be the greatest or win the most games. But when we as spectators look at elite sport, we tend to share with Lombardi the view that winning is the only thing. Measuring and ranking the competitors according to their contributions to winning follows very naturally from this. Taylor explicitly acknowledges what his player evaluations do not account for, and I appreciate that. But when we take his analysis and apply it directly to questions like "who is the GOAT?" or "who was better?" do we leave out anything else that matters to basketball or its value to us as spectators or the players as participants? I will explore these questions further at a later time, but here are two teasers: (1) players have more motivations than winning basketball games for their participation in the sport, (2) spectators have more reasons to watch and enjoy the sport than seeing winners win. I do not simply assume that these should matter to player evaluations, but I'd like to discuss the possibilities.

2. Leadership and intangibles: where do they fit into discussions of basketball greatness? JordansBulls is roundly criticized, even mocked, around here because -- sorry, JB -- we generally take his arguments to be archaic or unable to stand up to sophisticated statistical scrutiny. But give him credit for this much: JordansBulls always provides reasons for his views. One explanation of his has drawn particular interest from me of late: his oft-repeated claims that Michael Jordan is responsible for making his teammates, especially Scottie Pippen, who they were as players and winners. There are two versions of the retort to this case. First is to say that this discounts what those players actually did on the court and gives the credit to Jordan. Second is to say that this is not admissible because it is unfalsifiable. Both are valid responses. Yet, what have we really done to look into the claim he's put forward? Couldn't Michael Jordan's effects on player development and team commitment to winning have had a major impact on the Bulls' success? And, even if the statistical or impact-based outcomes are properly credited to the players who logged the minutes, made the baskets, got the stops, etc., isn't it possible we're overlooking important inputs to the ultimate product? We know that a lot goes into what makes a player succeed: genetics, work ethic, circumstance, etc. One such factor is surely the influence of others. Yet, because we can't (or don't yet know how to) measure it, I think we have a tendency ignore it. Maybe it's something that's just beyond the science of player evaluations. That's OK, but it needs to be an acknowledged limitation to the work that we do with numbers to answers questions concerning player greatness.

That's enough for now!

Great points.

On your first point, I definitely think sometimes that people underrate the importance of how entertaining a player is, cultural influence, innovation in play style, etc. Great example is Allen Iverson, whose reputation has slid a lot over the last decade as we have moved toward efficiency-based player evaluations, but whose talent and cultural impact surely still place him among the NBA greats. But this type of impact is substantially more subjective and personal and unmeasurable than "how much did he contribute to winning".

On your second point, I definitely agree, it's not just what a player brings to the court that contributes to winning. An extreme example of this is Kyrie, whose off-court behavior actually blew apart each of the three different teams. I suspect that will feed into most people's retrospective on Kyrie's greatness when his career is over, as much as his on-court talent helps teams win. Another weird example in the team-up era is that any team that LeBron was on in his prime was a free agency and trade demand magnet - does that count toward his contributions to winning or his greatness? But it's really hard to define/discover/quantify this type of value, whether positive or negative.

[Tim Lehrbach]

Great points.

On your first point, I definitely think sometimes that people underrate the importance of how entertaining a player is, cultural influence, innovation in play style, etc. Great example is Allen Iverson, whose reputation has slid a lot over the last decade as we have moved toward efficiency-based player evaluations, but whose talent and cultural impact surely still place him among the NBA greats. But this type of impact is substantially more subjective and personal and unmeasurable than "how much did he contribute to winning".

On your second point, I definitely agree, it's not just what a player brings to the court that contributes to winning. An extreme example of this is Kyrie, whose off-court behavior actually blew apart each of the three different teams. I suspect that will feed into most people's retrospective on Kyrie's greatness when his career is over, as much as his on-court talent helps teams win. Another weird example in the team-up era is that any team that LeBron was on in his prime was a free agency and trade demand magnet - does that count toward his contributions to winning or his greatness? But it's really hard to define/discover/quantify this type of value, whether positive or negative.

Good examples, illustrative of why oral/written histories and contemporary accounts are important research, especially insofar as they speak to impact beyond the hardwood.

Iverson might be one of the toughest cases in NBA history. He was such a phenomenon. Even if we could quantify that, do we assign positive or negative values to his influence? And, how much do we situate him as a man of his time (as we often do with pre-merger players) vs. giving him credit or blame for his status? Hard questions that I know have been debated in various threads here.

Kyrie may be an extreme case, but he's an important one. At this point, how can we possibly isolate his on-court production and say that's who he is as a basketball player?

LeBron, whew, is another toughie. I am generally positive on LeBron's impact on player movement for his own teams and player empowerment altogether, but there is so much to unpack with the team-up era, LeGM, etc. Think I'll leave that untouched for the moment, lest this become a LeBron debate.

[penbeast0]

Welcome back!

A view I generally believe in: https://thinkingbasketball.net/2017/12/11/the-backpicks-goat-the-40-best-careers-in-nba-history/

Some of the best players ever at their best
https://youtube.com/playlist?list=PLtzZl14BrKjSMb4IFWSy0qh_nFGiy7PoZ

Then some of his most recent podcasts where he goes through his new top 40 careers

Eps-141 to 151

https://podcasts.apple.com/us/podcast/thinking-basketball/id1428290303

Spoiler:

I love Ben Taylor's work on this topic. He does such a great job of clarifying the different ways you can think about ranking players across eras, what the rankings actually mean, how much ambiguity there is when players get compared against each other, the different approaches you can take to balancing performance and longevity when making comparisons, etc. I find all that stuff more valuable than his actual rankings.

Ben Taylor must surely be the MVP of GOAT lists. I have read some but not nearly enough of his work away from RealGM, I admit. Appreciate these links because I hadn't seen the video or podcast content before. I'll enjoy them!

Taylor is great not only because he has so thoroughly and consistently analyzed players with video and statistics and put it to a digestible measurement, but because he takes the time to acknowledge and study the foundations of player comparisons and even the rough edges which must be reckoned with.

The most interesting consequence of his work for me is that he has crystalized the oft-shared (yet, pre-ElGee, challenging-to-articulate) belief that evaluating players' GOAT candidacy (or "goodness," generally) is largely about measuring how much they contribute to winning. Championship odds or championship value-added is a terrific way to encapsulate player impact on winning games (and, for some, though less so for Taylor, the most important games). In addition to being deeply researched and meticulously calculated, the approach stands out as intuitively meaningful. We of course want to know how much a player contributes to winning. He and some others have done such a good job taking a really complicated problem (i.e., "how much does a player contribute to winning basketball games?") and yielding understandable results that the approach has become impossible to ignore and difficult to deny when taking up questions concerning relative player greatness.

My memories of basketball chatter only go back to the early 90s, but tracking the evolution of GOAT conversations since then, I can say that the question of how much a player impacts winning has always been lurking beneath arguments over rings and scoring titles and legendary moments, but any idea that basketball fans and outsiders could answer that question with extensive video analysis and player impact data was basically inconceivable. Then, the blessings of high-speed Internet, plus/minus data, and our early heroes of player analysis started delivering new methods of evaluation. You all know the story from here. What I credit Taylor and others with is making sense of the noise for the uninitiated. It's pretty wild that a layperson can comfortably use concepts like CORP in intelligible conversation. We've come so far!

Two questions persist in my mind when I read Taylor and others who take similar approaches. These are not challenges to their work or anything Taylor or anybody is responsible for answering, really, but are just lingering questions for further consideration:

1. So, we've arrived at player impact on winning being an indispensable way of measuring player goodness. Let's not take this to be true by definition. We had to do conceptual work and make commitments to get there. That is to say, going back to my reference to early 90s discussion, we didn't always have in mind a concept of measurable player impact on winning, or if we did, it was crudely using box scores (although Russell vs. Chamberlain was always the counterpoint). A concrete measurement simply wasn't available to us. Thus, my question is, what have we let go of to zero in on player impact on winning as central to the conversation about greatness? To take it even a step further, are we sure that winning games and championships is the ultimate or only purpose of the sport, such that we can reduce a player's import or success as a basketball player to his impact on winning? Not all sport is primarily about winning: take rec and youth sports, for example -- the object of the game never changes, but it doesn't typically occur to us to say that the participants are more or less valuable than one another because there is nobody (or just parents, perhaps) consuming the product with a rooting interest or an analytical eye. Further, as participants our goals are seldom to be the greatest or win the most games. But when we as spectators look at elite sport, we tend to share with Lombardi the view that winning is the only thing. Measuring and ranking the competitors according to their contributions to winning follows very naturally from this. Taylor explicitly acknowledges what his player evaluations do not account for, and I appreciate that. But when we take his analysis and apply it directly to questions like "who is the GOAT?" or "who was better?" do we leave out anything else that matters to basketball or its value to us as spectators or the players as participants? I will explore these questions further at a later time, but here are two teasers: (1) players have more motivations than winning basketball games for their participation in the sport, (2) spectators have more reasons to watch and enjoy the sport than seeing winners win. I do not simply assume that these should matter to player evaluations, but I'd like to discuss the possibilities.

2. Leadership and intangibles: where do they fit into discussions of basketball greatness? JordansBulls is roundly criticized, even mocked, around here because -- sorry, JB -- we generally take his arguments to be archaic or unable to stand up to sophisticated statistical scrutiny. But give him credit for this much: JordansBulls always provides reasons for his views. One explanation of his has drawn particular interest from me of late: his oft-repeated claims that Michael Jordan is responsible for making his teammates, especially Scottie Pippen, who they were as players and winners. There are two versions of the retort to this case. First is to say that this discounts what those players actually did on the court and gives the credit to Jordan. Second is to say that this is not admissible because it is unfalsifiable. Both are valid responses. Yet, what have we really done to look into the claim he's put forward? Couldn't Michael Jordan's effects on player development and team commitment to winning have had a major impact on the Bulls' success? And, even if the statistical or impact-based outcomes are properly credited to the players who logged the minutes, made the baskets, got the stops, etc., isn't it possible we're overlooking important inputs to the ultimate product? We know that a lot goes into what makes a player succeed: genetics, work ethic, circumstance, etc. One such factor is surely the influence of others. Yet, because we can't (or don't yet know how to) measure it, I think we have a tendency ignore it. Maybe it's something that's just beyond the science of player evaluations. That's OK, but it needs to be an acknowledged limitation to the work that we do with numbers to answers questions concerning player greatness.

That's enough for now!

Great points.

On your first point, I definitely think sometimes that people underrate the importance of how entertaining a player is, cultural influence, innovation in play style, etc. Great example is Allen Iverson, whose reputation has slid a lot over the last decade as we have moved toward efficiency-based player evaluations, but whose talent and cultural impact surely still place him among the NBA greats. But this type of impact is substantially more subjective and personal and unmeasurable than "how much did he contribute to winning".

On your second point, I definitely agree, it's not just what a player brings to the court that contributes to winning. An extreme example of this is Kyrie, whose off-court behavior actually blew apart each of the three different teams. I suspect that will feed into most people's retrospective on Kyrie's greatness when his career is over, as much as his on-court talent helps teams win. Another weird example in the team-up era is that any team that LeBron was on in his prime was a free agency and trade demand magnet - does that count toward his contributions to winning or his greatness? But it's really hard to define/discover/quantify this type of value, whether positive or negative.

We do have the "Mount Rushmore" threads that pop up every few months which focus more on impact on the sport (but not really on cultural impact -- that's one we haven't done and maybe someone can come up with a good way to clarify it for discussion).

[Tim Lehrbach]

We do have the "Mount Rushmore" threads that pop up every few months which focus more on impact on the sport (but not really on cultural impact -- that's one we haven't done and maybe someone can come up with a good way to clarify it for discussion).

In addition to impact on the game and cultural importance -- or maybe this is a generalization of and expansion on both of these -- I am interested in different ways of looking at what players most epitomize what sportsmanship means to us. This is to me an interesting question on its own merits and also one that gets at why we celebrate the players we do and what our appreciation of them means for us. If it went deep enough, I think it would potentially help me clarify my GOAT criteria for participation in next year's list.

I'd start a mini-project on this but I fear a lot of the would-be contributors are absent lately. Do you suppose there's any interest? Maybe just a single, general thread inquiring into people's thoughts on these questions would be the appropriate place to start?

[penbeast0]

You never know what will catch fire. Definitely worth throwing out a trial balloon.

[Doctor MJ]

I'm gonna choose this moment to quietly re-enter the Player Comparisons fray after a long time away from regular participation. Such re-entry will be clumsy but, I hope, not solipsistic. I want to engage and support the work done here, but I am never quite sure how my POV fits or what I have to contribute. I want to learn from y'all, but I am never quite sure of the right questions to ask. After many deleted drafts and false starts over the years, I'm just gonna throw some text up and see if anything sticks, or at least see if it loosens me up enough to start posting again.

*******

"There is no GOAT."
"GOAT debates are useless."
"Ranking and measuring competitors with ever-increasing precision is a myopic view of sport."
etc.

These were my usual responses of old to any big questions at Player Comparisons and around the water cooler IRL. Of course, there's no bigger or more popular question than deciding the Greatest of All Time. I've long disbelieved in ranked GOAT lists because the question concerning superlative greatness will always mean different things to different individuals, groups, generations, cultures, etc. And, even supposing that there could be one best or most meaningful referent for greatness, it seems impossible (and, to my mind, right that it should be impossible) to arrive at a perfect measurement of it. Although I haven't posted much here in the last decade or so, my real-world discussions have continued to go little farther than a curmudgeonly disapproval of or bemusement at the exercise.

Lately (i.e., the last 3 years or so), I've increasingly softened my opposition and opened myself up to the idea that behind every GOAT list, there is an attempt to celebrate what one values in sports and in its sportsmen and women. While I may continue to doubt the importance of quarreling over where players rank in an absolute sense, I've begun paying more attention to the arguments people advance for why this player or that one is extraordinarily worthy (or less worthy, depending) of celebration. When people's criteria differ, they are expressing differences -- often very, very subtle differences, but interesting ones nevertheless -- of what they want to see exalted in an all-time great. When criteria more or less agree, but rankings differ, people represent differing stories in support of competing accounts over a shared value. Still other times, criteria may seem to agree, but the contests between stats and stories reveal that the speakers/writers mean different things by the quality in dispute.

Gradually, all of this has begun to seem a lot more interesting to me.

A second thing that's happened, alongside of this development, is admitting to myself that I do believe there are, at least, greater and lesser narratives, supported by data and oral histories, around top sportsmen. No, I do not believe there is a single GOAT, but the idea has cultural currency, and there is a more or less socially agreed upon pantheon of GOAT-tier players, which is worthy of sociological examination. And, on a personal level, am I really so different from anybody else when something inside my gut tells me I want to see one player elevated over another in order to reflect the values such elevation promotes? I must conclude that I am not!

Finally, I've softened on GOAT lists because I've realized that my original position -- namely, that there cannot possibly be any such official list -- is precisely what should grant me permission to engage in the exercise without excessive seriousness. Previously I had been overly concerned that ranking players according to one or more GOAT metric is potentially a harmful activity for appreciation of the sport and its place in our lives and culture. Now, I have come around to believe that it is very well possible to explore the ideas behind our comparative evaluations of players -- even using the GOAT as the vehicle -- without committing ourselves to the centrality of any one thesis about ranking and measuring competitors. In other words, ranking and measuring need not limit our view of sport altogether the way I feared before. Rather, it can be a lens we turn back on ourselves to examine why we support the players and teams we do, and in turn what this says about ourselves.

In short, there is no need to oppose an activity which reveals interesting narratives and values, which educates participants about the game and its important figures, and which is ultimately harmless no matter how imperfect the answers it yields may be.

Some beautiful thought Tim. I've been taking a break and just periodically lurking for a while, trying to get some distance so I can enjoy the process here once more, and when I saw your post here I thought it was a great place to start back.

I like your last conclusion specifically, though I will say that I think that unfortunately there absolutely can be harm in the process. Putting in effort on GOAT analysis can lead to polarization, like so many other things nowadays, and polarization seems to guarantee some degree of toxicity in the contemporary social internet culture at the very least.

Why post my GOAT confessional? I suppose my purposes are threefold. One, I want to reintroduce myself to the board as a participant. Hi everybody. Two, my favorite conversations here are the ones that get into the "metaphysics" of player comparisons, and I'll be delighted if my awkward posts spark any such discussion, especially if such discussion increases my awareness and ability to take part in the fun projects and threads around here! Three, I expect that there are others who have shared my skepticism about the GOAT list and other ordered rankings of players, and I wonder what their own relationship to these topics is like. Do you also struggle with questions like:

How do we deal with the under-determined nature of the questions we ask concerning relative greatness?
The uncertain answers?
The disagreements over what we value in basketball (i.e., what is greatness)?
What does a well-conceived idea about greatness look like?
How do we distinguish between better and worse answers to questions that invoke subjectivity or non-numeric factors?
Has anybody put forward something like a glossary to explain the meaning of their terms?
Where does character fit into discussions about the GOAT? Cultural import? How can we hope to fairly or accurately measure these?
Are we better off dismissing, as some do, anything which cannot be quantified, or does this sacrifice too much?
What do we take ourselves to be doing when we argue the relative merits of players? Why is this meaningful to us?
What does basketball mean to us?

1 - Hey Tim!
2 - Some of my favorites too.
3 - Well, my relationship with it all is certainly complicated. First and foremost, I think the tendency to compare & rank is something that really works for me as a driver of ontology-building. I think there are other processes for ontology-building that are less polarization-prone, but in terms of what drives me to keep chewing on a topic, and thus keep learning within the domain with intensity for a long period of time, it's a hook that I often seem to need, and thus the negative aspects of it are things I just have to reckon with.

To your questions:

Uncertain & Under-Determined Answers


This is something I find to be a source of major divergence in approach between people, and I think it's really important to try to understand.

First off, to get it out of the way, one approach to this is to delude oneself about the domain you study and insist you know everything important about it. Simplifying the field into something you can confidently label and navigate, and then developing pride, identity, stubbornness, and aggression on the topic when challenged.

On the other end of one spectrum are the approaches that might be called "epistemically modest". The classic statement along these lines would be something like, "I only evaluate players I watched". While analysts like us would take issue with how coarse that particular phrasing is, it certainly makes sense to look to avoid drawing conclusions about stuff we don't have enough information on, and I myself do try to be epistemically modest...

but I think I'm seen by many on here as having an approach that might be characterized as ontologically aggressive. I have a tendency to stack conclusions as bricks of a tower while acknowledging that the "bricks" I'm using should not asserted with complete credence. I know I may be wrong, but I keep going anyway.

When asked why I work like I do, I tend to say a few things:

1. The idea that we truly know anything with complete credence is always a delusion. It is not possible to "only take the right steps", and if you think that's what you've done, then you're almost certainly falling prey to the same sort of missteps that you see as the danger of those more aggressive than you.

2. Not explicitly drawing conclusions is not the same thing as a clean slate. If we think about ontology building from a perspective of Bayesian inference - that is priors & posteriors - anything we have schema for, we're effectively defaulting the values of any known fields to what seems to be the most likely estimate based on more general context - aka, a prior we didn't realize we had.

A question then needs to be asked:

Is there anything noble about being wrong because you used an unconscious prior, as opposed to being wrong because we explicitly thought about what we thought was the most likely possibility in context?

I would say the answer to that question depends on the societal context, but in general, I think we'd be wise to disabuse ourselves of the notion the former wrong is any less wrong than the latter.

And so if one's goal is to optimize toward correctness, and one believes that they can do a better job in that field through conscious process than through inherited, unexamined assumptions, the way forward seems pretty clear to me.

3. But regardless of the level of aggression one used in taking epistemic steps, the most important thing to keep in mind for maintaining a relatively clean ontology is a regular circling back to inspect from first principles (aka "metaphysical analysis") with the expectation that you will need to prune some branches in your tree of knowledge that you shouldn't have let grow in the first place.

All the better if you can think about this when the information enters your brain in the first place, and you can attach explicit conscious assumptions that went in to the manufacturing of that node of know, which can make changing your mind a process with much less friction.

Criteria for Greatness

I don't think I've ever made an exhaustive list of all the different dimensions here, but what's definitely the case is that when we do GOAT lists, boundaries need to be put into place in order to create meaningful discussion. To list those that come to mind for me right now:

- Span - career, prime, n-year peak, playoff, series, season split, game, scenario
- Actual-Potential Axis - what actually happened?, what was most likely to happen?, what could have happened?
---- Sub-group: Types of Potentiality - team setting, era context, health
- Principal Lens - player, team, franchise, league, sport
- Breadth of Competitive Accomplishment - only in-game, only basketball-skill related, real-time only, unlimited
- Breadth of Non-Competitive Accomplishment - innovation, influence, aesthetic, popularity, education, virtue, symbolism, social good

I'd say in general that the RealGM PC GOAT projects tend to:
- Specify Span
- Assume NBA-lens unless specifically otherwise stated
- Assume Competitive Accomplishment only
- Leave the rest up to the predilections of the participants

I'll note though that I see you've made a thread about Sportsmen, my first thought there would be something that emphasized the Non-Competitive Accomplishment and specifically took a principal lens of "sport" rather than "league" or anything more narrow.

[Tim Lehrbach]

I'm so sorry it took me nearly a month to respond. I came charging in here like I was gonna be full-on active again, then got sidetracked by a time-consuming job search. Hope I haven't lost all momentum.

Some beautiful thought Tim. I've been taking a break and just periodically lurking for a while, trying to get some distance so I can enjoy the process here once more, and when I saw your post here I thought it was a great place to start back.

I like your last conclusion specifically, though I will say that I think that unfortunately there absolutely can be harm in the process. Putting in effort on GOAT analysis can lead to polarization, like so many other things nowadays, and polarization seems to guarantee some degree of toxicity in the contemporary social internet culture at the very least.

No doubt there can be polarization and with it harm done to one's appreciation for the sport generally or its greats in particular.

My process of growth is more along the lines of graduating from complete skepticism (i.e. "this is an impossible task which therefore can only cause damage and should not be undertaken") to one of measured openness ("this is an impossible task which nevertheless tugs at people's deeper convictions with respect to players, basketball, sporting altogether, and even culture; I should therefore be very interested in investigating how we go about answering these (impossible) questions and why they are so attractive to us"). GOAT lists and their brethren are a messy business, but if we take the time to examine those "first principles" and pause when debate is most intense to get at what matters to us personally or societally in these otherwise seemingly mundane exchanges, I have come around to believe it's a very fruitful activity indeed. If nothing else, I'd say that because this era of sport favors measuring and ranking competitors as an ultimate purpose, we would do ourselves a disservice (which, indeed, I believe I was doing to myself) by waving away the meaningfulness of these discussions.

But yes, of course, this can all go very wrong, and in fact I believe it frequently does.

1 - Hey Tim!
2 - Some of my favorites too.
3 - Well, my relationship with it all is certainly complicated. First and foremost, I think the tendency to compare & rank is something that really works for me as a driver of ontology-building. I think there are other processes for ontology-building that are less polarization-prone, but in terms of what drives me to keep chewing on a topic, and thus keep learning within the domain with intensity for a long period of time, it's a hook that I often seem to need, and thus the negative aspects of it are things I just have to reckon with.

If I've understood you here, "a driver of ontology-building," or, a way to make sense of the landscape in a detailed and meaningful way, is a good way to put this. Now, to make myself clear, I should say that I have serious misgivings about our contemporary tendency to treat measuring and ranking competitors as the very purpose of sport (this is a topic on which I've written elsewhere, and a position which I follow several historians and sociologists of sport in holding), but what I acknowledge is that, for better or worse, it simply is a primary lens through which we understand competition, especially elite sport. Thus, part of what I'm getting at above is an admission that rankings are an important opportunity to examine what we are doing when we watch and study sport.

And, to be more transparent, one of my ulterior purposes in all of this is to sneak such considerations into the "metaphysical" discussions we have about how we measure and rank competitors because what interests me is what we (individually and collectively) value in sports, and what the stories we tell about sports tell us about the competitors, ourselves as viewers, and society on the whole. More specifically, I'd like us to talk about this on two levels: (1) most basically, to identify what "the tendency to compare & rank [...] as a driver of ontology-building" says about basketball's meaning to us (i.e., "Why do we rank?") and (2) to identify how this lens affects what we do in this activity of ranking (i.e., "Why does each of us rank that way that we do?"). I think (1) is a really hard question, and I won't be disappointed if we make little progress there, while (2) is one that gets a lot of attention in "pre-ranking" discussions and in the heat of argument, but I find that (no disrespect intended to anybody) we could stand to go deeper in answering why we emphasize and weight what we do when it comes to comparing players.

To your questions:

Uncertain & Under-Determined Answers


This is something I find to be a source of major divergence in approach between people, and I think it's really important to try to understand.

First off, to get it out of the way, one approach to this is to delude oneself about the domain you study and insist you know everything important about it. Simplifying the field into something you can confidently label and navigate, and then developing pride, identity, stubbornness, and aggression on the topic when challenged.

On the other end of one spectrum are the approaches that might be called "epistemically modest". The classic statement along these lines would be something like, "I only evaluate players I watched". While analysts like us would take issue with how coarse that particular phrasing is, it certainly makes sense to look to avoid drawing conclusions about stuff we don't have enough information on, and I myself do try to be epistemically modest...

but I think I'm seen by many on here as having an approach that might be characterized as ontologically aggressive. I have a tendency to stack conclusions as bricks of a tower while acknowledging that the "bricks" I'm using should not asserted with complete credence. I know I may be wrong, but I keep going anyway.

When asked why I work like I do, I tend to say a few things:

1. The idea that we truly know anything with complete credence is always a delusion. It is not possible to "only take the right steps", and if you think that's what you've done, then you're almost certainly falling prey to the same sort of missteps that you see as the danger of those more aggressive than you.

2. Not explicitly drawing conclusions is not the same thing as a clean slate. If we think about ontology building from a perspective of Bayesian inference - that is priors & posteriors - anything we have schema for, we're effectively defaulting the values of any known fields to what seems to be the most likely estimate based on more general context - aka, a prior we didn't realize we had.

A question then needs to be asked:

Is there anything noble about being wrong because you used an unconscious prior, as opposed to being wrong because we explicitly thought about what we thought was the most likely possibility in context?

I would say the answer to that question depends on the societal context, but in general, I think we'd be wise to disabuse ourselves of the notion the former wrong is any less wrong than the latter.

And so if one's goal is to optimize toward correctness, and one believes that they can do a better job in that field through conscious process than through inherited, unexamined assumptions, the way forward seems pretty clear to me.

3. But regardless of the level of aggression one used in taking epistemic steps, the most important thing to keep in mind for maintaining a relatively clean ontology is a regular circling back to inspect from first principles (aka "metaphysical analysis") with the expectation that you will need to prune some branches in your tree of knowledge that you shouldn't have let grow in the first place.

All the better if you can think about this when the information enters your brain in the first place, and you can attach explicit conscious assumptions that went in to the manufacturing of that node of know, which can make changing your mind a process with much less friction.

Wow, you found in my tossed-out-there questions an inherent clash of characters: the "epistemically modest" vs. the "ontologically aggressive" approaches to answering underdetermined problems (e.g. "Who is the GOAT?"). There is, you must have guessed, an underlying plea for modesty in my writings on this subject.

But you also make good points about what Rumsfeld and I might agree to call "known unknowns" (using probability to make assumptions) and "unknown knowns" (unconscious priors). You are polite not to cast anybody into either camp, but I'll say that there are a whole lotta these unconscious priors underlying our opinions (here and everywhere... about everything). While I would agree that there is no special privilege to this kind of "knowledge," it gets at exactly where I'd like to turn our attention: the underlying, usually not quantitative, beliefs that inform why we think what we think about basketball. There is a scaffolding to the "tower of ontology" that helps explain why "career value," "peak season," or "championship probability added," on the one hand, and "dominance," "skillset," and "competitive fire," on the other, may be more or less meaningful terms to us. And, to hint at another item on my agenda, I think we can do better than a model of debate which proceeds basically the same whether our paradigms are aligned or altogether incommensurable. We don't have to agree on first principles, but we should at least know how to translate between one another's belief systems. The best contributors here make great efforts at this, FWIW, but I would love to see more projects predicated on such a standard. How this works, I'm not entirely sure, but it's among the things I'm thinking about in advance of the 2023 GOAT list.

[Doctor MJ]

I'm so sorry it took me nearly a month to respond. I came charging in here like I was gonna be full-on active again, then got sidetracked by a time-consuming job search. Hope I haven't lost all momentum.

Sending nothing but good vibes your way Tim.[/quote]

Some beautiful thought Tim. I've been taking a break and just periodically lurking for a while, trying to get some distance so I can enjoy the process here once more, and when I saw your post here I thought it was a great place to start back.

I like your last conclusion specifically, though I will say that I think that unfortunately there absolutely can be harm in the process. Putting in effort on GOAT analysis can lead to polarization, like so many other things nowadays, and polarization seems to guarantee some degree of toxicity in the contemporary social internet culture at the very least.

No doubt there can be polarization and with it harm done to one's appreciation for the sport generally or its greats in particular.

My process of growth is more along the lines of graduating from complete skepticism (i.e. "this is an impossible task which therefore can only cause damage and should not be undertaken") to one of measured openness ("this is an impossible task which nevertheless tugs at people's deeper convictions with respect to players, basketball, sporting altogether, and even culture; I should therefore be very interested in investigating how we go about answering these (impossible) questions and why they are so attractive to us"). GOAT lists and their brethren are a messy business, but if we take the time to examine those "first principles" and pause when debate is most intense to get at what matters to us personally or societally in these otherwise seemingly mundane exchanges, I have come around to believe it's a very fruitful activity indeed. If nothing else, I'd say that because this era of sport favors measuring and ranking competitors as an ultimate purpose, we would do ourselves a disservice (which, indeed, I believe I was doing to myself) by waving away the meaningfulness of these discussions.

But yes, of course, this can all go very wrong, and in fact I believe it frequently does.

Yup, not saying not to encourage it certainly, but we can learn always learn new things that are wrong.

1 - Hey Tim!
2 - Some of my favorites too.
3 - Well, my relationship with it all is certainly complicated. First and foremost, I think the tendency to compare & rank is something that really works for me as a driver of ontology-building. I think there are other processes for ontology-building that are less polarization-prone, but in terms of what drives me to keep chewing on a topic, and thus keep learning within the domain with intensity for a long period of time, it's a hook that I often seem to need, and thus the negative aspects of it are things I just have to reckon with.

If I've understood you here, "a driver of ontology-building," or, a way to make sense of the landscape in a detailed and meaningful way, is a good way to put this. Now, to make myself clear, I should say that I have serious misgivings about our contemporary tendency to treat measuring and ranking competitors as the very purpose of sport (this is a topic on which I've written elsewhere, and a position which I follow several historians and sociologists of sport in holding), but what I acknowledge is that, for better or worse, it simply is a primary lens through which we understand competition, especially elite sport. Thus, part of what I'm getting at above is an admission that rankings are an important opportunity to examine what we are doing when we watch and study sport.

And, to be more transparent, one of my ulterior purposes in all of this is to sneak such considerations into the "metaphysical" discussions we have about how we measure and rank competitors because what interests me is what we (individually and collectively) value in sports, and what the stories we tell about sports tell us about the competitors, ourselves as viewers, and society on the whole. More specifically, I'd like us to talk about this on two levels: (1) most basically, to identify what "the tendency to compare & rank [...] as a driver of ontology-building" says about basketball's meaning to us (i.e., "Why do we rank?") and (2) to identify how this lens affects what we do in this activity of ranking (i.e., "Why does each of us rank that way that we do?"). I think (1) is a really hard question, and I won't be disappointed if we make little progress there, while (2) is one that gets a lot of attention in "pre-ranking" discussions and in the heat of argument, but I find that (no disrespect intended to anybody) we could stand to go deeper in answering why we emphasize and weight what we do when it comes to comparing players.

1) I've thought quite a bit about the compulsion to rank, so glad to talk about it.

A ranking implies the ordering of a one-dimensional entity along an axis. This provides a seemingly definitive information gain by doing the following:

a) Focusing in a domain imbued with meaning to us - else we wouldn't bother - and thus removing any doubt as to the utility of the process.

b) Simplifying the domain to a single dimension, which makes the gain feels all the more significant (which can allow one to avoid challenging complications).

c) It lacks any real quantitative scale, which also allows us to feel like we've reached a more definitive conclusion (and also allows us to avoid mathematical complexities if we want).

So it's something that allows us to trick ourselves into a sense of ontological accomplishment by clearing the landscape of nuance like a gardener weeds his garden.

It's thus a part of the same broader category that draws people to black/white thinking, and of course, just like black/white thinking can foster identity which then arms itself against correction.

Of course, that doesn't mean that when people rank, they all just simplify as much as possible and then just decide they've arrived at the definitive right answer. It all depends on one's epistemic process how valid one's conclusions are, but the fundamental draw of the dopamine reward at the end of a successful selection is probably pretty universal.

In terms of why ranking has become more of a society-wide phenomenon recently, I think it has a lot to do with the internet and the meaning that comes from putting down your rankings in a public place where it will be seen and responded to, along with the draw of the fight that comes with disagreements.

2) In terms of why I rank as I do, well, I've been a compulsive ranker from a young age. While other kids were drawing pictures, I had notebooks which were often dominated by ranked lists.

Key to that: These ranked lists were not something I saw as permanent. I would keep list progressions which allowed me to see how an entity's given ranking changed with time, and to do meta-analysis on those shifts. So specific placements on a ranked list weren't really something I was tied to as an identity, and in fact once I got a bit older I actually kept this ranking compulsion secret from my peers so it wasn't something I was proud of, it was just something I felt a need to do.

Further, I knew about cognitive dissonance as a feeling long before I ever heard the term. I began to dig deeper and deeper in my domains of interest specifically when I felt the vertigo that came from a failed schema. I did so not to be "a thorough scientist" or anything like that, but rather to just make that bad vertigo feeling go away.

And I think others do other things to repel away from that vertigo, with the most scary to me being a capacity to create one's own reality with defense mechanisms that can successfully ward off actual reality. We all can fall into this accidentally through cognitive dissonance, but the internet has made me realize that we are capable of fabricating a reality explicitly that then actually feels more real than the truth.

Stuff like this, as a teacher, makes me want to develop a curriculum to help build the skills necessary to avoid such traps, but it's such an abstract thing that it can't realistically be understood before one could have already fallen into a trap. I find that to be quite daunting when I consider where humanity goes from here.

To your questions:

Uncertain & Under-Determined Answers


This is something I find to be a source of major divergence in approach between people, and I think it's really important to try to understand.

First off, to get it out of the way, one approach to this is to delude oneself about the domain you study and insist you know everything important about it. Simplifying the field into something you can confidently label and navigate, and then developing pride, identity, stubbornness, and aggression on the topic when challenged.

On the other end of one spectrum are the approaches that might be called "epistemically modest". The classic statement along these lines would be something like, "I only evaluate players I watched". While analysts like us would take issue with how coarse that particular phrasing is, it certainly makes sense to look to avoid drawing conclusions about stuff we don't have enough information on, and I myself do try to be epistemically modest...

but I think I'm seen by many on here as having an approach that might be characterized as ontologically aggressive. I have a tendency to stack conclusions as bricks of a tower while acknowledging that the "bricks" I'm using should not asserted with complete credence. I know I may be wrong, but I keep going anyway.

When asked why I work like I do, I tend to say a few things:

1. The idea that we truly know anything with complete credence is always a delusion. It is not possible to "only take the right steps", and if you think that's what you've done, then you're almost certainly falling prey to the same sort of missteps that you see as the danger of those more aggressive than you.

2. Not explicitly drawing conclusions is not the same thing as a clean slate. If we think about ontology building from a perspective of Bayesian inference - that is priors & posteriors - anything we have schema for, we're effectively defaulting the values of any known fields to what seems to be the most likely estimate based on more general context - aka, a prior we didn't realize we had.

A question then needs to be asked:

Is there anything noble about being wrong because you used an unconscious prior, as opposed to being wrong because we explicitly thought about what we thought was the most likely possibility in context?

I would say the answer to that question depends on the societal context, but in general, I think we'd be wise to disabuse ourselves of the notion the former wrong is any less wrong than the latter.

And so if one's goal is to optimize toward correctness, and one believes that they can do a better job in that field through conscious process than through inherited, unexamined assumptions, the way forward seems pretty clear to me.

3. But regardless of the level of aggression one used in taking epistemic steps, the most important thing to keep in mind for maintaining a relatively clean ontology is a regular circling back to inspect from first principles (aka "metaphysical analysis") with the expectation that you will need to prune some branches in your tree of knowledge that you shouldn't have let grow in the first place.

All the better if you can think about this when the information enters your brain in the first place, and you can attach explicit conscious assumptions that went in to the manufacturing of that node of know, which can make changing your mind a process with much less friction.

Wow, you found in my tossed-out-there questions an inherent clash of characters: the "epistemically modest" vs. the "ontologically aggressive" approaches to answering underdetermined problems (e.g. "Who is the GOAT?"). There is, you must have guessed, an underlying plea for modesty in my writings on this subject.

But you also make good points about what Rumsfeld and I might agree to call "known unknowns" (using probability to make assumptions) and "unknown knowns" (unconscious priors). You are polite not to cast anybody into either camp, but I'll say that there are a whole lotta these unconscious priors underlying our opinions (here and everywhere... about everything). While I would agree that there is no special privilege to this kind of "knowledge," it gets at exactly where I'd like to turn our attention: the underlying, usually not quantitative, beliefs that inform why we think what we think about basketball. There is a scaffolding to the "tower of ontology" that helps explain why "career value," "peak season," or "championship probability added," on the one hand, and "dominance," "skillset," and "competitive fire," on the other, may be more or less meaningful terms to us. And, to hint at another item on my agenda, I think we can do better than a model of debate which proceeds basically the same whether our paradigms are aligned or altogether incommensurable. We don't have to agree on first principles, but we should at least know how to translate between one another's belief systems. The best contributors here make great efforts at this, FWIW, but I would love to see more projects predicated on such a standard. How this works, I'm not entirely sure, but it's among the things I'm thinking about in advance of the 2023 GOAT list.

Indeed, your modest orientation was clear.

I like that you're bringing up Rumsfeld, but let me smash Rumsfeld with Reagan here, as Reagan talked about "The trouble with our Liberal friends is not that they're ignorant; it's just that they know so much that isn't so." Now let's just push the ridiculousness of Reagan saying that aside and recognize that he's talking about a dimension distinct from Rumsfeld's two-dimensions.

As you allude to, the first of the two un/knowns relates to conscious awareness, while I'd say the second is about factual specifics.

What Reagan is talking about is the correctness of information.

I would say it's this 3rd dimension that marries with the conscious awareness into something truly malignant. Nothing is more stifling to growth than the state of someone who "knows they know something (that isn't so)".

To your rest: Love what you're saying. I know that a lot of the younger folks take "agreeing to disagree" as something of a cop out, and understandably so, but I want to have debates where we are able to zoom in on why we disagree and either focus on those areas of disagreement, or move on. Really hate it when "bias" gets thrown out there in conversations because I think most of the time, it really misses the mark. As I say that, I must acknowledge that I've thrown the term out there too, so I'm casting the stone from a glass house.

[Tim Lehrbach]

1) I've thought quite a bit about the compulsion to rank, so glad to talk about it.

A ranking implies the ordering of a one-dimensional entity along an axis. This provides a seemingly definitive information gain by doing the following:

a) Focusing in a domain imbued with meaning to us - else we wouldn't bother - and thus removing any doubt as to the utility of the process.

b) Simplifying the domain to a single dimension, which makes the gain feels all the more significant (which can allow one to avoid challenging complications).

c) It lacks any real quantitative scale, which also allows us to feel like we've reached a more definitive conclusion (and also allows us to avoid mathematical complexities if we want).

So it's something that allows us to trick ourselves into a sense of ontological accomplishment by clearing the landscape of nuance like a gardener weeds his garden.

It's thus a part of the same broader category that draws people to black/white thinking, and of course, just like black/white thinking can foster identity which then arms itself against correction.

Of course, that doesn't mean that when people rank, they all just simplify as much as possible and then just decide they've arrived at the definitive right answer. It all depends on one's epistemic process how valid one's conclusions are, but the fundamental draw of the dopamine reward at the end of a successful selection is probably pretty universal.

In terms of why ranking has become more of a society-wide phenomenon recently, I think it has a lot to do with the internet and the meaning that comes from putting down your rankings in a public place where it will be seen and responded to, along with the draw of the fight that comes with disagreements.

Great insight! This is the sort of psychological perspective on ranking that I didn't imagine we could get at here, but I should have expected that somebody would be way ahead of me. You've given me a lot to think about.

What immediately springs to mind is to investigate how this "trick" of the mind relates to the historical context: when does this compulsion to rank begin to appear in sports, and why does it become so powerful? (I have ideas, but I'm not studied in any of this and rely on trusted secondary sources. I will come back to this after consulting them.) And then, as you point out, we get to the Internet/social media/24 hour cable news (and sports) age, where controversy is courted and dropping knowledge is a form of personal expression and identity formation.

Why should all of this happen? Well, working backwards from today, I imagine you are quite correct to identify the Internet as a major accelerator. And, recalling your psychological account, it makes sense that there would be a basic incentive to order things simply and definitively to achieve cognitive neatness and a feeling of satisfaction. I suspect there must be more ground to cover here, and you've got my wheels turning. Is this a distinctly modern story? Is it bound up in scientism? Elitism? Where does this story end, or how does it evolve from here? (As I said, I will consult some historical sources and return to these questions.)

2) In terms of why I rank as I do, well, I've been a compulsive ranker from a young age. While other kids were drawing pictures, I had notebooks which were often dominated by ranked lists.

Key to that: These ranked lists were not something I saw as permanent. I would keep list progressions which allowed me to see how an entity's given ranking changed with time, and to do meta-analysis on those shifts. So specific placements on a ranked list weren't really something I was tied to as an identity, and in fact once I got a bit older I actually kept this ranking compulsion secret from my peers so it wasn't something I was proud of, it was just something I felt a need to do.

Ha, I did this too. Songs, movies, athletes, teachers, uh... friends. I also was not proud of it. I would bet this is a common childhood or adolescent activity of future RealGMs.

Further, I knew about cognitive dissonance as a feeling long before I ever heard the term. I began to dig deeper and deeper in my domains of interest specifically when I felt the vertigo that came from a failed schema. I did so not to be "a thorough scientist" or anything like that, but rather to just make that bad vertigo feeling go away.

And I think others do other things to repel away from that vertigo, with the most scary to me being a capacity to create one's own reality with defense mechanisms that can successfully ward off actual reality. We all can fall into this accidentally through cognitive dissonance, but the internet has made me realize that we are capable of fabricating a reality explicitly that then actually feels more real than the truth.

Stuff like this, as a teacher, makes me want to develop a curriculum to help build the skills necessary to avoid such traps, but it's such an abstract thing that it can't realistically be understood before one could have already fallen into a trap. I find that to be quite daunting when I consider where humanity goes from here.

Oof, yeah. What's perhaps even scarier than the thought that we can fabricate a reality around us as a defense against cognitive dissonance is the idea that AI can fabricate our reality such that we have both the contents and the defenses programmed for us. Maybe we're saying similar things.

Indeed, your modest orientation was clear.

I like that you're bringing up Rumsfeld, but let me smash Rumsfeld with Reagan here, as Reagan talked about "The trouble with our Liberal friends is not that they're ignorant; it's just that they know so much that isn't so." Now let's just push the ridiculousness of Reagan saying that aside and recognize that he's talking about a dimension distinct from Rumsfeld's two-dimensions.

As you allude to, the first of the two un/knowns relates to conscious awareness, while I'd say the second is about factual specifics.

What Reagan is talking about is the correctness of information.

I would say it's this 3rd dimension that marries with the conscious awareness into something truly malignant. Nothing is more stifling to growth than the state of someone who "knows they know something (that isn't so)".

Points taken about the third dimension and which form of "knowing" is most malignant.

Let me just clarify what I mean about the "second" dimension and why I think it needs examination before we get to determining whether the things we know we know are really so! I am wondering whether there are unacknowledged facts or beliefs motivating the ways we choose to measure players and teams. I am further suggesting not only that the answer is surely yes but that these "unknown knowns" may be forms of social knowledge or shared values (e.g., what it means to be a winner, or to compete, etc.). We don't all share exactly the same set of this knowledge or these values, but perhaps we unconsciously recognize them in the arguments of others and react either for or against them. In other instances, we don't recognize them at all, and another's words are incorrectly taken as agreeing with our own underlying assumptions. (I think this is a particular hazard when we enter into discussion without very clear terms of the art or science. You made a similar point really well when you explored metaphor and "rim protection" in another thread.) So, when it comes to something as potentially abstract as "greatness," we especially need to get to this second layer of knowledge if we are to be comprehensible to one another. The distinction I'm drawing is that I think a prior task is to identify, without judgment, what our underlying premises, beliefs, and values are before we get to whether they are true or valid (and therefore before we can evaluate our arguments, too). My primary reason for calling these two separate steps (identify, then evaluate) rather than one (identify and evaluate) is that if we know that our assumptions may be subject to being called prejudiced or unfounded as soon as they are revealed, I'm afraid we are less likely to be forthcoming in laying them bare.

I'm (finally) getting sleepy, so I'll leave it there for now.

To your rest: Love what you're saying. I know that a lot of the younger folks take "agreeing to disagree" as something of a cop out, and understandably so, but I want to have debates where we are able to zoom in on why we disagree and either focus on those areas of disagreement, or move on. Really hate it when "bias" gets thrown out there in conversations because I think most of the time, it really misses the mark. As I say that, I must acknowledge that I've thrown the term out there too, so I'm casting the stone from a glass house.

Ahhh, I just used "bias" in the sportsmanship thread. At least it was self-directed.

[Doctor MJ]

Great insight! This is the sort of psychological perspective on ranking that I didn't imagine we could get at here, but I should have expected that somebody would be way ahead of me. You've given me a lot to think about.

What immediately springs to mind is to investigate how this "trick" of the mind relates to the historical context: when does this compulsion to rank begin to appear in sports, and why does it become so powerful? (I have ideas, but I'm not studied in any of this and rely on trusted secondary sources. I will come back to this after consulting them.) And then, as you point out, we get to the Internet/social media/24 hour cable news (and sports) age, where controversy is courted and dropping knowledge is a form of personal expression and identity formation.

Why should all of this happen? Well, working backwards from today, I imagine you are quite correct to identify the Internet as a major accelerator. And, recalling your psychological account, it makes sense that there would be a basic incentive to order things simply and definitively to achieve cognitive neatness and a feeling of satisfaction. I suspect there must be more ground to cover here, and you've got my wheels turning. Is this a distinctly modern story? Is it bound up in scientism? Elitism? Where does this story end, or how does it evolve from here? (As I said, I will consult some historical sources and return to these questions.)

Thank you for your kind words Tim!

Re: when ranking appear in sports and why so powerful?

Well, I think at the roots of it are the fact that sports are a form of competition, and thus an ordering of better/worse is inherently apart of the process. The ranking we're speaking of is just us using our minds as the field of play rather than the field itself, which comes from the phenomenon being so attractive to us that it actively lives in side our minds even when no immediate stimuli activate it.

As such, I think the fundamental "player comparison" aspect of things is basically as old as such sport have existed, and the modern ranking has exploded because our societal infrastructure allows to build more elaborate castles in the sand.

Re: bound up in scientism? Good question. I would say it's very much not scientism - as I think faith in scientists has dropped to dangerously low levels - but something often closely connected to scientific paradigm formation:

Quality and quantity of data.

With data comes "the tyranny of the quantifiable" which comes with many other things, both good and bad. Ontology construction scrapes skies but only along specific worn paths while impeding other possibly ways.

Re: Elitism? Yes, but I think more fundamentally, I'd call it identity becoming attached to knowledge (one could call it a modern strain of Gnosticism). It tends to foster elitism, but there's something deeper going on than better-than-you pride. I used the term "vertigo" for what I experienced with cognitive dissonance, and as someone who has also suffered from actual physical vertigo before, I think the analogy works very well. When the foundation of your identity is based on what you think you know, being wrong can crack the psychology ground your tread on.

Where does the story end? Huh, good question. I think it's reasonable to look at our current time period as a kind of societal digital adolescence, but the mere fact that we'll probably move to digital adulthood in the decades to come doesn't necessarily tell us a lot about what will be new in that adulthood compared to now.

To me the biggest question of all is probably, "What usurps righteous anger?" It's proving right now to have a greater gravitational pull than any other feeling in the context of these contemporary mediums, and so long as it remains dominant, we're going to go to keep get more polarizable and more prone toward destructive impulses.

It's possible that the best case scenario is for the people to just get so sick of the these negatively charged media that they go back to living their social life in their local scene, but it's hard to imagine that happening dramatically enough that it kills the most toxic online networks - in part because they and their walls probably will allow them to remain in a tidepool even as the sea falls.

2) In terms of why I rank as I do, well, I've been a compulsive ranker from a young age. While other kids were drawing pictures, I had notebooks which were often dominated by ranked lists.

Key to that: These ranked lists were not something I saw as permanent. I would keep list progressions which allowed me to see how an entity's given ranking changed with time, and to do meta-analysis on those shifts. So specific placements on a ranked list weren't really something I was tied to as an identity, and in fact once I got a bit older I actually kept this ranking compulsion secret from my peers so it wasn't something I was proud of, it was just something I felt a need to do.

Ha, I did this too. Songs, movies, athletes, teachers, uh... friends. I also was not proud of it. I would bet this is a common childhood or adolescent activity of future RealGMs.

Right on! I'd love to hear if many others on these boards were similar.

Further, I knew about cognitive dissonance as a feeling long before I ever heard the term. I began to dig deeper and deeper in my domains of interest specifically when I felt the vertigo that came from a failed schema. I did so not to be "a thorough scientist" or anything like that, but rather to just make that bad vertigo feeling go away.

And I think others do other things to repel away from that vertigo, with the most scary to me being a capacity to create one's own reality with defense mechanisms that can successfully ward off actual reality. We all can fall into this accidentally through cognitive dissonance, but the internet has made me realize that we are capable of fabricating a reality explicitly that then actually feels more real than the truth.

Stuff like this, as a teacher, makes me want to develop a curriculum to help build the skills necessary to avoid such traps, but it's such an abstract thing that it can't realistically be understood before one could have already fallen into a trap. I find that to be quite daunting when I consider where humanity goes from here.

Oof, yeah. What's perhaps even scarier than the thought that we can fabricate a reality around us as a defense against cognitive dissonance is the idea that AI can fabricate our reality such that we have both the contents and the defenses programmed for us. Maybe we're saying similar things.

Ah algorithmic media. Yup, that's a big deal, and for me as a guy with a background in AI, not something I saw coming. The idea that recommendation engines could be useful? Absolutely. That they'd have such a strong entropic force toward extremes? It makes sense in retrospect, but I didn't have the schema to see the gradient that would emerge until after the gradient roared into life.

And yes I'd say that's similar to what I was saying, but also a distinct phenomenon. I feel more optimistic about this problem going forward. I think it's something we can incrementally mitigate. It's going to continue to do damage, but being able to perceive the source of the problem gives the engineers a shot to make things better.

Indeed, your modest orientation was clear.

I like that you're bringing up Rumsfeld, but let me smash Rumsfeld with Reagan here, as Reagan talked about "The trouble with our Liberal friends is not that they're ignorant; it's just that they know so much that isn't so." Now let's just push the ridiculousness of Reagan saying that aside and recognize that he's talking about a dimension distinct from Rumsfeld's two-dimensions.

As you allude to, the first of the two un/knowns relates to conscious awareness, while I'd say the second is about factual specifics.

What Reagan is talking about is the correctness of information.

I would say it's this 3rd dimension that marries with the conscious awareness into something truly malignant. Nothing is more stifling to growth than the state of someone who "knows they know something (that isn't so)".

Points taken about the third dimension and which form of "knowing" is most malignant.

Let me just clarify what I mean about the "second" dimension and why I think it needs examination before we get to determining whether the things we know we know are really so! I am wondering whether there are unacknowledged facts or beliefs motivating the ways we choose to measure players and teams. I am further suggesting not only that the answer is surely yes but that these "unknown knowns" may be forms of social knowledge or shared values (e.g., what it means to be a winner, or to compete, etc.). We don't all share exactly the same set of this knowledge or these values, but perhaps we unconsciously recognize them in the arguments of others and react either for or against them. In other instances, we don't recognize them at all, and another's words are incorrectly taken as agreeing with our own underlying assumptions. (I think this is a particular hazard when we enter into discussion without very clear terms of the art or science. You made a similar point really well when you explored metaphor and "rim protection" in another thread.) So, when it comes to something as potentially abstract as "greatness," we especially need to get to this second layer of knowledge if we are to be comprehensible to one another. The distinction I'm drawing is that I think a prior task is to identify, without judgment, what our underlying premises, beliefs, and values are before we get to whether they are true or valid (and therefore before we can evaluate our arguments, too). My primary reason for calling these two separate steps (identify, then evaluate) rather than one (identify and evaluate) is that if we know that our assumptions may be subject to being called prejudiced or unfounded as soon as they are revealed, I'm afraid we are less likely to be forthcoming in laying them bare.

I'm (finally) getting sleepy, so I'll leave it there for now.

Great points. I might call what you're describing as a tacit metaphysical root in one's ontology. If you yourself are unable to trace the branches of your knowledge back to first principles, then beyond a certain point there becomes no reasoning with you.

v

To your rest: Love what you're saying. I know that a lot of the younger folks take "agreeing to disagree" as something of a cop out, and understandably so, but I want to have debates where we are able to zoom in on why we disagree and either focus on those areas of disagreement, or move on. Really hate it when "bias" gets thrown out there in conversations because I think most of the time, it really misses the mark. As I say that, I must acknowledge that I've thrown the term out there too, so I'm casting the stone from a glass house.

Ahhh, I just used "bias" in the sportsmanship thread. At least it was self-directed.

Hate the sin, love the sinner.

[eminence]

2) In terms of why I rank as I do, well, I've been a compulsive ranker from a young age. While other kids were drawing pictures, I had notebooks which were often dominated by ranked lists.

Key to that: These ranked lists were not something I saw as permanent. I would keep list progressions which allowed me to see how an entity's given ranking changed with time, and to do meta-analysis on those shifts. So specific placements on a ranked list weren't really something I was tied to as an identity, and in fact once I got a bit older I actually kept this ranking compulsion secret from my peers so it wasn't something I was proud of, it was just something I felt a need to do.

Ha, I did this too. Songs, movies, athletes, teachers, uh... friends. I also was not proud of it. I would bet this is a common childhood or adolescent activity of future RealGMs.

Right on! I'd love to hear if many others on these boards were similar.

Count me in, and neat, excessive list making as a shared experience.