RealGM Top 100 List -- 2011

[mysticbb]

Yeah, except that's not what I'm getting at. Court vision, genius hoops IQ, the extraordinary size for his position -- all of that can be measured. I'm talking about the verve and passion with which he devoured the game. How he made even a notorious grouch like Kareem -- another intangible, albeit a negative one -- realize that basketball is fun. How Michael Cooper marveled that he never saw him not ready to play, even at Utah on the back end of a 10-game road trip in the middle of February.

Did all that made his teammates play better basketball?

[DocHoops]

Had you ever played the game, you wouldn't ask that question...of course it did.

[mysticbb]

Great, thanks for the answer, I appreciate the help. Maybe you can help me with another question.

Do you think a better playing team is scoring more points than the opponent?

[DocHoops]

Stupid or a-hole what are you?


One or the other is the only answer.

If you really think that that attempt at baiting is even close to a valid point...I pity you. Nothing personal, I don't know you, but I won't read another character you type.

[Rapcity_11]

A question for the stat fans - if these advanced stats are so much better at determining who is better, as opposed to expert opinion and conventional wisdom, shouldn't you be able to get rich predicting the outcome of games?

I always laugh when I hear questions like this.

Advanced stats (and all stats) tell us what happened in the past, not what will necessarily happen in the future. They aren't always predictive.

Not always, but they should provide an edge that trumps conventional wisdom more often than not, right?

As predictors for team success? No.

Does Vegas use conventional wisdom?

[Rapcity_11]

^Why bother. These guys are too stubborn to admit they are wrong.

They just try to twist the argument, like they do the numbers, to fit their point of view.

Let's get back to meaningful discourse like that of Baller24, Drza ec.

I think it's more letting the numbers shape their point of view. *For most*

Too stubborn to admit being wrong? How about you do? You posted a list of supposedly immeasurable things that was in large part, wrong.

[mysticbb]

Stupid or a-hole what are you?

Likely both in your eyes. ;)

One or the other is the only answer.

Maybe you haven't understood the question. Which is unfortunately not unlikely. Thus, I will try another one: Do you think that "playing better" is related to winning basketball games?

Nothing personal, I don't know you, but I won't read another character you type.

No offense taken, but please, let me learn from your great knowledge and give me some clear answers to my questions. I really want to understand basketball.

[Fencer reregistered]

You posted a list of supposedly immeasurable things that was in large part, wrong.

Still, there are many that were correct, or would have been if rephrased.

Everything about a TEAM can be measured in points, W/L, whatever. However, important things that can not be measured by a player's stats, even +/-, include how a player does or doesn't inspire teammates to play hard, play unselfishly, or practice conscientiously. Things that can perhaps be measured by +/-, but not in a player's more itemized stats, include offensive intimidation (drawing coverage), defensive intimidation (discouraging shots, passes, drives, etc.), blockouts for other rebounders, passes that don't lead to assists, picks, or verbal leadership of the defense.

[Fencer reregistered]

What does that imply, that the only thing that matters (winning -- your own words here) can be summed up by a pair of numbers? The primary thing is that every action, or at least every action that has any impact at all on the final score, needs to have a statistical foot print. If there were something truly intangible, truly invisible to our statistical measures, then it's also invisible to the scoreboard and thus doesn't affect the outcome of the game.

Some things, however, are unprovable or unattributable (e.g., psychological effects on other teammates of one's behavior over time). I.e., one's OFF-COURT effect on the game's score can't be captured by any stat.

Further, there are many examples of cases in which one's ON-COURT impact can't be captured except by:

A. Measures based on total team (or opponent) score.
B. Subjective visual observation (whether human or automated).

[Fencer reregistered]

^Dumbest thing I ever heard.

And I literally mean that. I've never heard/seen/read anything that is some immensely wrong.

You're not paying sufficient attention to the concept of +/-.

[Fencer reregistered]

I've already covered a lot of the ones I agree with in prior posts. But I'd like to add endorsement of:

deflections
individual help defense
team defensive rotations
trash talk
pressure (applied to a ball handler)
properly/poorly executed switches
saves
momentum
fatigue
showing on pick and rolls
mismatches created
match-up changes due to foul trouble or substitution patterns
tips/back-taps
mishandled passes

Some of those can be measured by opponents' individual stats, but some can only be captured by something like +/-

[Fencer reregistered]

A question for the stat fans - if these advanced stats are so much better at determining who is better, as opposed to expert opinion and conventional wisdom, shouldn't you be able to get rich predicting the outcome of games?

I always laugh when I hear questions like this.

Advanced stats (and all stats) tell us what happened in the past, not what will necessarily happen in the future. They aren't always predictive.

Sloppy answer. As long as they're partially predictive, the challenge is valid.

[kabstah]

Some things, however, are unprovable or unattributable (e.g., psychological effects on other teammates of one's behavior over time). I.e., one's OFF-COURT effect on the game's score can't be captured by any stat.

You can expand your time table to include changes in the team's performance when a player joins/leaves a team. Would it then be hard to attribute the change to an exact quality of the player? Most definitely. But this is a top 100 list for players, not a top 100 list for basketball attributes, so it's typically sufficient to correctly (more or less) assign credit to a player without worrying about the exact distribution of his value among his constituent attributes.

[Rapcity_11]

I've already covered a lot of the ones I agree with in prior posts. But I'd like to add endorsement of:

deflections
individual help defense
team defensive rotations
trash talk
pressure (applied to a ball handler)
properly/poorly executed switches
saves
momentum
fatigue
showing on pick and rolls
mismatches created
match-up changes due to foul trouble or substitution patterns
tips/back-taps
mishandled passes

Some of those can be measured by opponents' individual stats, but some can only be captured by something like +/-

I'm confused. What are your thoughts on APM and on/off stats?

I ask because all of those things above can be captured in those statistics.

It's funny how so many people rip on stats for not measuring the immeasurable aspects of basketball such as those on the above list, yet don't understand that APM and on/off is the best we can do to measure those things.

[Rapcity_11]

A question for the stat fans - if these advanced stats are so much better at determining who is better, as opposed to expert opinion and conventional wisdom, shouldn't you be able to get rich predicting the outcome of games?

I always laugh when I hear questions like this.

Advanced stats (and all stats) tell us what happened in the past, not what will necessarily happen in the future. They aren't always predictive.

Sloppy answer. As long as they're partially predictive, the challenge is valid.

The eye test is partially predictive as well, shouldn't Laimbeer be able to get rich betting on games? Is that challenge valid?

There's also the fact that advanced statistics largely deal with individuals, yet teams are the ones we bet on and are much, much harder to break down.

[Gongxi]

The elephant in the room, of course, is that if those attributes and skills are not measured or impossible to measure (which is false, but whatever for the sake of argument), how would a person who hasn't watched the entire body of work of two players successfully compare them in those terms? Even if you had watched them, what about recency bias, or a whole host of other cognitive biases?\

If they were truly immeasurable, trying to both measure and then compare would be a fool's quest anyway, and really just an excuse to trumpet one's own favorite player- which too often is the point of the "eye testers" anyway.

[lorak]

Too often I see posters hide behind the flag of "intangibles" or "the eye test" in order to push an agenda or prop up their favorite player without providing any verifiable, corroborative evidence. Yes, stats can be manipulated, misinterpreted, cherry picked, or sometimes outright incorrect, but at least there's a transparency of method there that leaves room for critique and discussion. Falling back to "intangibles" or "things that don't show in the stats" is often just another way of saying "because I think so, NYAH!" That doesn't lead to intelligent discourse so much as a pissing contest of Nuh uh!/Uh huh!

This is perfect summary of this discussion.

[Fencer reregistered]

The eye test is partially predictive as well, shouldn't Laimbeer be able to get rich betting on games? Is that challenge valid?

Sure.

And we can be very certain that bettors do use eye tests when making their bets.

[Fencer reregistered]

Some things, however, are unprovable or unattributable (e.g., psychological effects on other teammates of one's behavior over time). I.e., one's OFF-COURT effect on the game's score can't be captured by any stat.

You can expand your time table to include changes in the team's performance when a player joins/leaves a team.

A naive approach to such an analysis would run into all sorts of noise.

And an attempt to clean it up statistically would be difficult.

[Fencer reregistered]

I've already covered a lot of the ones I agree with in prior posts. But I'd like to add endorsement of:

deflections
individual help defense
team defensive rotations
trash talk
pressure (applied to a ball handler)
properly/poorly executed switches
saves
momentum
fatigue
showing on pick and rolls
mismatches created
match-up changes due to foul trouble or substitution patterns
tips/back-taps
mishandled passes

Some of those can be measured by opponents' individual stats, but some can only be captured by something like +/-

I'm confused. What are your thoughts on APM and on/off stats?

I ask because all of those things above can be captured in those statistics.

It's funny how so many people rip on stats for not measuring the immeasurable aspects of basketball such as those on the above list, yet don't understand that APM and on/off is the best we can do to measure those things.

I diverge strongly from Doc Hoops there. I think they tell us a great deal.