JPSeraph

In the end, what are we left with? If intangibles are out, and defense is simply arguable (and thus inconclusive in cases where it's 'close'), then anything that isn't adequately quantified is simply not included?

We have McGrady and Erving roughly equal in scoring capability. I'd actually say T-Mac is deadlier, but who knows, maybe a healthy Erving in today's ruleset would catapult down the lane ad nauseam.

We have McGrady as the superior passer and ball handler(?), by a significant, but not huge gap.

And we have Erving as a better rebounder, especially on the offensive boards.

What else? We can compare steals and blocks, which Erving wins, but it counts for relatively little.

To me, logically, the more you exclude from analysis, the less general application the results will have. We're comparing players as a whole, but only including a portion of their attributes in the actual analysis.

Doesn't it seem reasonable to say that the analysis would benefit from some consideration of the portion not included? Certainly one would think if we're applying the results to comparing players as a whole and not merely "who was a better offensive player?".

That's my take anyway. I understand that you abhor the myriad pitfalls inherent in qualitative analysis, but surely if we can weed through all the chaff, there's a kernel of insight waiting.

Doing a credible qualitative analysis is just as difficult in its own right as making a credible quantitative analysis. Poorly done, either can yield silly or downright inaccurate results.

How else would Hollinger know that he had found a better overall measure than "efficiency" in PER without SOME sort of unquantified assumptions or observations about the game??

__________________
On a given day, in a given circumstance, you think you have a limit...

Minstrel

I checked one example and immediately found a counter. The player I chose was Paul Pierce. I picked him because he's been most comparable to McGrady in his generation, as far as rebounding goes among wing players.

Pierce's RbR:

Age 25: 10.6
Age 26: 9.6
Age 27: 10.6
Age 28: 10.5

I guess, in the end, we'll just need to see how McGrady's career progresses to make this comparison more fairly.

__________________
You'll never live like common people
You'll never do what common people do
You'll never fail like common people
You'll never watch your life slide out of view
And dance and drink and screw
Because there's nothing else to do.

2k

I thought Dantly won a title with the Pistons

ThaRegul8r

Nope...

__________________
THE REGULATOR
"I'm not the showstopper, I'm RUNNING the show!"

Minstrel

When did I say defense should be thrown out? I just said it's subjective.

Again, I'm all for including defense. There's just no way to conclusively rule on it, so it'll remain in the realm of opinion. Which is fine by me.

As for intangibles, to be totally frank, I think opinions on it is more noise than information. Just my take. I don't render opinions on it and I don't find opinions on it very compelling. Obviously, everyone is free to weight intangibles as they like, but I'll disagree when it's weighed in to any significant degree.

That's always a good question. The best way is if you can combine the player production into a team total and those team totals correlate well with team wins (the "objective" standard for success).

__________________
You'll never live like common people
You'll never do what common people do
You'll never fail like common people
You'll never watch your life slide out of view
And dance and drink and screw
Because there's nothing else to do.

JPSeraph

How does that which is in the "realm of opinion" (i.e. defense) factor into a player analysis which is based primarily on quantitative measures?

Suppose, for instance, that we agree to a 50/50 split between offense and defense with respect to how each category should factor into a player's overall value. If the measures for 50% of a player's value are, as you say, merely opinions, then how should they be weighted against the other 50% which is more quantitatively sound?

If they are weighed equally, it seems contradictory to your stance that quantitative measures should be the primary (or is it only?) factor in such an analysis. If they are not weighed equally, then the analysis for a player's value (which theoretically should be 50/50) is based primarily upon only 50% of the theoretical model for a player's value; such that even if we have correctly evaluated both offense (primarily quantitative) and defense (primarily qualitative), our final analysis could still be wrong due to reducing the value of any qualitative measures.

i.e. T-Mac's offense is a 9.5/10 and his defense is a 7.5/10. This should make him an 8.5/10 overall, but due to our weighting of offense as, say, twice as important as defense (since it was quantitatively measured), the final rating we assign McGrady is an 8.8/10.

This can certainly influence comparisons with similar great players or just lead us to tend to overvalue offense in general.

How do we discuss Dennis Rodman? Especially prior to his rebounding revolution.

And that's just defense. I'm deliberately biting my shirt and holding back on bringing up the argument for intangibles!

But is that what Hollinger did? I haven't read his books, but on his website I clearly recall him using arguments for why other "single number" player ratings were inferior like "any measure that has David Robinson over Michael Jordan is flawed". It could just be rhetoric, but I've always thought the drive to devise a method for rating players' overall value with a single number was at least in part due to our own sense of what results such a number "should" reveal.

I know for me personally that the joy of statistics is watching the numbers describe what I see in reality, or vice versa.

__________________
On a given day, in a given circumstance, you think you have a limit...

BadBaronRudigor

Minstrel, the original argument by you was that Erving's ABA numbers which show him as significantly better than McGrady should be ignored because upon his move to the NBA, his numbers dropped across the board. You particularly pointed to his rebounding numbers.

I suggested an alternative explanation; that it was a natural process of age and used the first three comparables on the 3rd party list of GOAT SGs (the peer group in discussion) to show that this particular drop was in fact common among this type of player.

You countered with a single example of a player not on the 3rd party peer group list who you felt was comparable to McGrady (not Erving). So, let me review the full 8 peer list from the original suggested SF Goat list.

Bird, English, Baylor, Barry. Pippen, and Worthy experienced significant drops in rebounding stats after age 26,26,27,28,29,29; very similar to Erving.

Barry, the only other player to go ABA/NBA, experienced a strong rebounding GAIN returning to the NBA at age 26 before the age dropoff caught up 2 years later.

Havlicek didn't experience a significant drop until age 31, Wilkens never experienced a rebounding drop at all.

Therefore, it seems likely that since 75% of the comparable players experienced a similar significant rebounding drop in their late 20's that THIS IS A VERY LIKELY EXPLANATION for Erving's statistical drop upon moving to Philidelphia.

ABA numbers are inflated compared to equivalent NBA numbers but the only stat survey I ever saw estimated it for late ABA at only about 10% (and attributed it to the star system that caused teams that payed big money for free agents to expect them to take on big roles . . . the same thing Cunningham attributed it to in interviews). As you go early in the ABA, numbers like Connie Hawkins's experienced greater distortion.

__________________
If you have to shoot, shoot, don't talk

Minstrel

That isn't my stance. My stance is to accept numbers over observation when there are measures available that seem to properly quantify the issue in question, removing most or all confounding factors.

There are such measures for scoring, rebounding, passing (though far from unquestionable), so I prefer them over using observation to decide who was better. Defense has no such numbers (though there are some interesting ones), and therefore, there's a lot of subjectivity involved.

But defense is clearly a significant part of the game. That's the difference between it and intangibles. We know defense is a huge part of the game and a big determining factor in who wins and loses. We don't know, at all, that intangibles are a huge part of the game and a big determining factor in who wins. Some people may theorize that it is, but with no proof. It's akin to arguing whether God exists: there's no way to prove it one way or the other.

Defense exists, though. I don't think anyone denies that.

No. Hollinger, I believe, builds measures by what he considers logical and then evaluates them by standards like repeatability: do players tend to repeat their PER performance from season to season? Doing so is evidence that it's getting at their talent level, rather than being somewhat random, because players repeat their talent level from year to year, so their PER should (by and large, with accepted statistical variance) by similar from year to year (adjusted by age).

That's far from unquestionable, as I said above. But I think his logical assumptions (like that scoring and assists are affected by team pace and rebounds are affected by how many rebounding chances there are, etc) are strong ones.

Same same.

__________________
You'll never live like common people
You'll never do what common people do
You'll never fail like common people
You'll never watch your life slide out of view
And dance and drink and screw
Because there's nothing else to do.

Minstrel