Carmelo Anthony vs Gilbert Arenas?

[tsherkin]

Eh, I'm sort of 'iffy' about this statement. I tend to give more credit to Allen Iverson than anyone else. Iverson went through the biggest metamorphosis, including improving his shot selection while picking and choosing when to be aggressive and when to back off. It was Iverson who sacrificed a lot of possessions for the good of the team.

Actually, Iverson just reverted to the playing style he had exhibited earlier in his career, emphasizing playmaking a little more than scoring because he had the help and this time, the help was compatible with his scoring zones instead of conflicting. Because Melo starts his offense under the arc and prefers to stay inside 18 feet, the entire perimeter remained open for AI to do his isolation work and he didn't clash with Melo vying for his sweet spots.

It was really Iverson, who in my opinion, doesn't get enough credit for how he's improved his play in Denver.

I wouldn't credit him for that; he should have been doing that his entire career; he's not actually THAT good a scorer without the aid of the no-handcheck rule and 'Melo's presence actually makes it more palatable for AI to take fewer shots because Anthony isn't an inefficient scorer the way AI has generally been.

Incidentally, nate's explanation has no value here; he does a good job of explaining why you might look to take 3s and to credit a 3-baller with better efficiency than is indicated but it still doesn't change the idea that if you're looking for a basket, 50% versus 35% looks a lot better when you need to make a single shot in a key moment. The payoff for a 3 is greater in the context of the game overall but when you only need 1 or 2 points, there's little sense in taking the three.

[tsherkin]

Just to expand on the whole AI/Melo dynamic...

It would have been senseless for Anthony to reduce his possessions; he's the superior scorer and AI's the better passer. AI projected as a Chris Paul-like point guard when he came into the league but he's a shot-happy fool, or was in Philly for the bulk of his career. He took the wrong path, though given his streetball history, that's hardly a surprise.

In Denver, he did what he was told and it works... it wouldn't if they tried it the other way.

Yeah, I guess credit AI for doing the thing that wasn't stupid but I wouldn't overcredit him for doing what made basic sense. That's like crediting a player for working on during the season to maintain fitness, or for improving his diet or lifting in the off-season because he's getting tossed around. That's all basic sense and yeah, you give a shrug and a nod, a word of compliment but it is expected and should hardly be awarded with great praise and adoration.

AI's presence in Denver has made things a bit easier for Anthony, sure, but Iverson is also the team's critical flaw because he's another high-usage player who works primarily out of isolation sets and the last thing that Denver needed was another isolation scorer with a weak perimeter shot because they already had issues with stalled ball movement and defenses collapsing on the paint because the Nuggets couldn't pose a significant threat from outside.

And of course, AI has a noticeable regression in efficiency when he gets to the playoffs; his career playoff FG% is almost 3% lower than his regular season FG (40.1% in the playoffs, 42.6% in the regular season, buoyed by his recent time in Denver). There is a concordant increase in 3PA that helps explain the drop but there's also only a 1.3% increase in 3P% that generally indicates that he's not doing himself any favors by shooting more threes, that's it is more a result of his inability to be as effective on account of his style of play. AI was about as bad as it gets in his first year in the playoffs with Denver and wasn't much better this year (about 6% better from the field and about 11% worse from the line, 8% worse from downtown).

Crediting AI for his body of work in Denver is not really worth the time or effort it would take to explain his massive postseason implosions. The Nuggets are 1-8 in the playoffs with AI, which is hardly any better than the 3-12 they posted in the three seasons leading up to his arrival (1-8 is a winning percentage of 0.111 versus the 0.200 they posted each year without him, matching his single-season best with Denver).

Anyway, I wouldn't be crediting 'Melo's improvement to AI so much as the presence of a significant second player. Before AI, 'Melo's best player was Andre Miller... who was very good but very much not a scorer.

[Malinhion]

Eh, I'm sort of 'iffy' about this statement. I tend to give more credit to Allen Iverson than anyone else. Iverson went through the biggest metamorphosis, including improving his shot selection while picking and choosing when to be aggressive and when to back off. It was Iverson who sacrificed a lot of possessions for the good of the team.

In 05-06 while playing for Philadelphia, Iverson had a usage rate of 35.8. In Denver, he decreased that to 27.1 in 06-07 and 26.7 in 07-08. That's basically a decrease of 1/4th of his possessions.

In 05-06, Carmelo had a usage rate of 32.0. Now, when Iverson came on-board, that actually increased in 06-07 to 33.2 (in fact, I remember many Nuggets fans posting how Anthony wasn't giving anything up and how Iverson was doing the sacrificing). To Anthony's credit, he did reduce his usage rate this season to 30.2, but that's still pretty darn high. He didn't change all that much. It was really Iverson, who in my opinion, doesn't get enough credit for how he's improved his play in Denver.

Usage rate doesn't paint the picture, my friend. Usage isn't calculated by the amount of time you have the ball in your hands or how many dribbles you take. It's a metric calculated from a box score.

USAGE RATE = {[FGA + (FT Att. x 0.44) + (Ast x 0.33) + TO] x 40 x League Pace}

So, as you can see, this thing is calculated by actually adding FGs. Let's take a situation where Iverson dribbles the ball around, Melo comes off a screen, and there's a pick-and-pop. Even though Iverson may have been holding the ball for ten times longer, Melo is getting three times the bump on his usage rate than Iverson is getting from the assist. That's not having the ball in your hands or making plays, its a simple metric that tells you one guy had the ball for three times longer.

In reality, Melo spends a lot of time getting position on the high post (as tsherkin said) or on the right block. He also runs off screens for shots. But a lot of his time is spent off-ball. You wouldn't see this flesh out in usage rates because of the way its calculated. He's assisted on 59% of his field goals. Find me another primary scorer, besides Amare, who creates such a small percentage of his own shots. This is what allows Melo to play alongside balldominant playmakers.

Yao and McGrady don't coexist because they both sacrifice some of their 30% Usage rate to play with each other. They do well because Yao is assisted on 61% of his field goals. The same is true with Carmelo, except that he's a guy who can do what Amare or Yao does from the wing. You're only sacrificing a little FG% to get more points.

Once again, this just seems a bit fishy. And instead of me trying to explain my point of view, I'll let nate33 do it, since he is able to actually make sense out of it. This is from a Gilbert debate in a thread a while back on the General NBA Board, discussing Arenas' efficiency and the pros to shooting 3's....

Again, you're missing my point. tsherkin actually covered it, but I'm not talking about point production or keeping possessions. I'm talking about the final minutes of the game, when you need a bucket to catch up, stay in it, or simply seal the deal. High percentage is better. You might be able to shoot yourself back into a game with three-pointers, but on that last possession you're better going with the higher FG% option.

That's where we'll just have to disagree. I view the term "point guard" as having a certain role -- to get his team into their offense and to put distribution of the basketball above all else. I don't think that "having the ball in his hands" a lot dictates his position. I am a firm believer that both system and role dictate position. Gilbert's role is to score, and the variation of the Princeton Offense that Eddie Jordan runs is predicated on one-on-one play and isolation sets, making set positions irrelevant.

And I am obliged to agree with miller; Arenas isn't a point guard. He's too much like Oscar Robertson in that he dominates the ball but he isn't a distributor, he's a scorer first who also distributes.

I have acknowledged a million times that Gil is either a guard or a combo guard. Perhaps I was not clear enough here. My point was just that he has the ball in his hands enough to rack up the assists, if he was going to get them. That's just not his playing style. I realize that he's not a point guard because he's not a facilitator. But that's just a description of his style, not a reason for his average assist totals.

Carmelo Anthony's standard deviation for his scoring was 7.2 this past season (07-08); in 06-07, Arenas' was 10.7. This seems to bear out the notion that Arenas fluctuates a lot more readily than does Melo in terms of his scoring output. Melo was therefore generally around 19 to 33 points most nights, with anything outside of that range existing as a significant departure from his average performance.

I see we're still using my standard deviation metric for seeing how consistent a player is. But remember, while the point production is easier to look at in terms of how it fleshes out, you really have to compare the % of deviation from the actual PPG. I forget what this number is called. Variance?

Either way, I think that if the main asset you're getting from a guy is his scoring, its imperative for that guy to be consistent. This is where FG% bears out--it has a strong correlation to point deviation.

[tsherkin]

I have acknowledged a million times that Gil is either a guard or a combo guard. Perhaps I was not clear enough here. My point was just that he has the ball in his hands enough to rack up the assists, if he was going to get them. That's just not his playing style. I realize that he's not a point guard because he's not a facilitator. But that's just a description of his style, not a reason for his average assist totals.

Actually, it is. Gilbert is the first option in Washington, which means he's relied upon to produce volume offense, which limits the number of playmaking opportunities he can get. Yes, some of the limitations come from his penchant for volume 3PT shooting and that's been acknowledged, he has flaws and among them is shot selection (perhaps most notably, at that).

But he's still an extremely valuable playmaker because he's one of the comparably few players who produce high-volume offense and still post 5+ apg.

He's not a finisher like Wade, so his opportunities for the penetratre-and-pitch (so valuable to Wade and a prime Stephon Marbury) are lower than you might otherwise thing. He produces assists at a rate that's roughly comparable to Kobe, Lebron and T-Mac and yet is impugned for not producing assists at the rate of a pure point guard, which is entirely unfair.

He plays a role very much like that of McGrady (who's assists are up now that his points are down) and Kobe and produces in a very much similar fashion.

You're unfairly criticizing him based on a standard that doesn't apply to him and one you do not apply to other players of comparable styles. Perhaps it is height bias, because he's only 6'3 and those guys are 6'6 - 6'9?

I see we're still using my standard deviation metric for seeing how consistent a player is. But remember, while the point production is easier to look at in terms of how it fleshes out, you really have to compare the % of deviation from the actual PPG. I forget what this number is called. Variance?

Either way, I think that if the main asset you're getting from a guy is his scoring, its imperative for that guy to be consistent. This is where FG% bears out--it has a strong correlation to point deviation.

Well, I could use the CoV, which would make Carmelo's deviation 28% and Arenas' 41.2%, would that make it clearer?

The Coefficient of Variation is just the standard deviation divided by the mean and since it's best used in the ratio scale, it should be a decent measure here. Just the StDevs doesn't give you a base scale but if you see that Gil is deviating roughly 13% more from his mean, then you might get a better idea of just how significant the difference in their consistency is. This is also why I included their scoring spread, how often they scored in each of those ranges. Melo peaks lower but doesn't have the same bottom end that Arenas does; he doesn't score as low as Gil tends to on his off-nights, while sacrificing some of the top-end offensive explosiveness Gil displays.

Anyway, the point I was making was actually in support of Anthony; I think he's definitely a more consistent scorer than Arenas and I think the reasoning is clear; he hits his field goals more regularly and since they are higher-percentage shots, they're more reliable which leads to that efficiency and consistency. Arenas relies on high-payoff, low-reliability shots from outside the arc on a much more significant basis, lowering his consistency.

[Malinhion]

We aren't actually arguing about Gil's assist totals. Both of us think they are average, and both of agree that he is a combo guard or merely a "princeton" guard, not a point guard. I'll never expect PG assist totals out of Gil. Its not his game. He brings other PG-like qualities, most notably his ability to push the tempo.

Reread my first post:

"People who prefer Gilbert will argue in favor of his strengths. He is arguably the best in the league at pushing the tempo. Look how much slower the Wizards played this year without him on the court. While not being a superior playmaker, he is somewhat of a better passer and will get his assists."
"He also keeps his turnovers down for how often he's got the ball in his hands."
"The only argument you used to be able to make here was that while Arenas racked an average number of assists, Melo was a below-average rebounder."

I, personally, will take consistency over explosiveness for a scorer. What really sells me is this...

The drop I'm going to get going from Carmelo to a Tayshaun Prince, Lamar Odom, or Caron Butler is larger than it would be if I was going from Arenas to an Allen Iverson, Dwyane Wade, or Joe Johnson.

[tsherkin]

We aren't actually arguing about Gil's assist totals. Both of us think they are average, and both of agree that he is a combo guard or merely a "princeton" guard, not a point guard. I'll never expect PG assist totals out of Gil. Its not his game. He brings other PG-like qualities, most notably his ability to push the tempo.

OK, I'm with you so far.

I, personally, will take consistency over explosiveness for a scorer. What really sells me is this...

It's something to consider but I suggest to you that it is not so obvious a thought as you'd think if the rest of your team is decent. See, with the Wizards, because they have some other offensive forces, if Arenas doesn't pop off, then they can still win. With Denver, they have pretty much two scorers and if both of them aren't decent, then they don't win, more often than not. So I think in that case, consistency is more valuable because the roster isn't balanced but I still think it's situational.

The drop I'm going to get going from Carmelo to a Tayshaun Prince, Lamar Odom, or Caron Butler is larger than it would be if I was going from Arenas to an Allen Iverson, Dwyane Wade, or Joe Johnson.

There is a gigantic gap between Arenas and Wade. Between Arenas and AI? I don't think there's that much of a gap at all, I guess, except that Arenas is a better streak shooter. Between Arenas and JJ? There's a defensive gap but I think in principle, Arenas is a lot more dangerous than JJ and I think the gap there is at least the equal of Carmelo to the first group (Tay, Odom, etc).

[Malinhion]

It's something to consider but I suggest to you that it is not so obvious a thought as you'd think if the rest of your team is decent. See, with the Wizards, because they have some other offensive forces, if Arenas doesn't pop off, then they can still win. With Denver, they have pretty much two scorers and if both of them aren't decent, then they don't win, more often than not. So I think in that case, consistency is more valuable because the roster isn't balanced but I still think it's situational.

I charge you with the task of finding me a team who has enough consistent scoring options that the production they get from their #1 option is an afterthought. The Wizards have an argument for this, but its not like they’ve got the playoff success to prove that it works. And that’s where it matters, because you need to know that your star can pop off in a critical playoff game when the rest of the guys are struggling.

I realize this isn’t a great point to make against Melo in particular, who has had playoff struggles in terms of his stats and team success. However, it still stands to make my point in general.

I don’t care if my primary playmaker is an inconsistent scorer, as long as he’s not my lead scorer. I can deal with Nash putting up 42 points when you take away Amare and his other options. But if that’s the main thing I’m counting on a guy for, I want to know what I’m getting every night. It makes it much easier to fit into a structured system so you can make appropriate adjustments for your team without the guesswork of which player will be “hot.”

There is a gigantic gap between Arenas and Wade. Between Arenas and AI? I don't think there's that much of a gap at all, I guess, except that Arenas is a better streak shooter. Between Arenas and JJ? There's a defensive gap but I think in principle, Arenas is a lot more dangerous than JJ and I think the gap there is at least the equal of Carmelo to the first group (Tay, Odom, etc).

Right, that’s my point. There are a lot more combo guards at/above/below Gilbert’s level. If I really wanted, I could trade for one. However, if I’ve already got Arenas on my team I can’t find a Melo without selling the farm. If I’ve got Melo, I can package Andre Miller with a big guy, an expiring contract, and a pick, and pick up an MVP-calibre combo guard.

The league is much more talent-stacked on the perimeter. Having an inside scorer like Melo is invaluable because of the drop off to the next best guy. If you’ve done reading on VoRP (Value over Replacement Player) that is the type of thing I’m getting at. Except from an organizational perspective, not how much production you lose while the star is on the bench.

[Ballings7]

Carmelo

While neither are wide-range players outside of scoring, between the two, Carmelo's better because of his rebounding, and has improved his creating for others noticably over the years. Carmelo's also the better scorer between the two because of his size. Also has been more durable.

Comparision aside, Gilbert needs to become a more balanced offensive player, and utilize what he can do defensively on a consistent basis, outside of going for steals. Picking his spots when to really commit, and be at least adequate the rest of the time.

Similar goes for Carmelo defensively.

[tsherkin]

I charge you with the task of finding me a team who has enough consistent scoring options that the production they get from their #1 option is an afterthought. The Wizards have an argument for this, but its not like they’ve got the playoff success to prove that it works. And that’s where it matters, because you need to know that your star can pop off in a critical playoff game when the rest of the guys are struggling.

Washington's problem is their inefficacy on the glass, their poor defense and that fact that the vast majority of their offense is perimeter-oriented as opposed to post-oriented. They're flawed in those regards and that is why they do not succeed in the postseason.

You have a sound point, I suppose, and I'm really just playing Devil's Advocate while I wait and see if Carmelo continues to develop as a playmaker or will continue to kill ball movement with stall isos the way he still does. He holds the ball too long and that's something that really bugs me about him.

I don’t care if my primary playmaker is an inconsistent scorer, as long as he’s not my lead scorer. I can deal with Nash putting up 42 points when you take away Amare and his other options. But if that’s the main thing I’m counting on a guy for, I want to know what I’m getting every night. It makes it much easier to fit into a structured system so you can make appropriate adjustments for your team without the guesswork of which player will be “hot.”

Who on the Wizards should Arenas go to more often to produce the higher volume of assists? Who should he trust to take the shots he has to give up? Jamison is just as much of an inconsistent perimeter shooter and gets his post buckets opportunistically rather than primarily through isos through a line of deployment. I guess they might look at Butler but in a full season without Arenas, he wasn't even the team's leading scorer... Jamison was. Butler's a fine young scorer but he's not a guy from whom the team would benefit overmuch by increasing his shot volume. The points need to be scored and the problems for the team aren't primarily on offense (aside from the lack of a post option and maybe the underutilization of Brendan Haywood in that role) but what else would you have Arenas do?

Yeah, he's a bit inconsistent but his scoring efficiency remains solid on a nightly basis, and those FTAs keep coming at a higher rate, usually enough to off-set the shooting streaks. You'll notice that while Arenas had a few games with lower scoring outputs than 'Melo, he had a lot more days where he scored more explosively and that was in slightly fewer games.

Check this out:

Over 77 games, 'Melo scored over his average 36 times.

Over 74 games, Arenas (in 06-07) scored over his higher average 39 times.

Over 77 games, 'Melo scored UNDER his average 41 times.

Over 74 games, Arenas scored under his average 35 times.

Now who's more consistent?

'Melo scored above his average 46.8% of the time; Arenas, 52.7%.

'Melo scored under his average 53.2% of the time, 6.4% more often than he scored above it; Arenas scored under his average 47.3% of the time, 5.4% of the time.

So 'Melo scored over his average less and under his average more.

Carmelo scored 25-27 ppg (basically +/- 1 from his average) in 17 of his 77 games, or 22.1%. Arenas scored 27-29 points 8 times, or 10.8%.

So yeah, Anthony was more consistent about scoring his average, by 11.3%...

But Arenas scored about his average (which was higher than 'Melo's) 5.9% of the time. If you want to be picky about it, Arenas scored higher than 'Melo's scoring average 45 times.

How much the value of Carmelo's consistency there? If Arenas is consistently scoring above 'Melo's average in raw numbers and in terms of proportion, what value is there to 'Melo's consistency?

Sure, Arenas pitted out pretty badly 5 times but he only scored at or under 15 points the same number of times as did 'Melo in a sample 3 games shorter than 'Melo's. That's hardly indicative of any major problem with Arenas' game when he's hitting the higher peaks so much more often, don't you think?

Right, that’s my point. There are a lot more combo guards at/above/below Gilbert’s level. If I really wanted, I could trade for one. However, if I’ve already got Arenas on my team I can’t find a Melo without selling the farm. If I’ve got Melo, I can package Andre Miller with a big guy, an expiring contract, and a pick, and pick up an MVP-calibre combo guard.

AI isn't an MVP-caliber combo guard and didn't even deserve the MVP he won in 01, so let's not do that dance.

If you’ve done reading on VoRP (Value over Replacement Player) that is the type of thing I’m getting at. Except from an organizational perspective, not how much production you lose while the star is on the bench.

I'm semi-familiar with VORP, yes. Frankly, I disagree with the assessment on account of the reasoning behind Iverson's availability and you overrating his status as an "MVP-caliber combo guard."

I mean, Philly didn't get great return value for AI on that deal regardless of how it panned out for them and those deals aren't common occurrences. It would NOT be just as easy to have a 'Melo-type player and add someone of AI's or Arenas' caliber. More to the point, Arenas is better than Iverson, so the comparison doesn't work anyway.

[Malinhion]

Well, disagreed. I would rather have Iverson than Arenas. I'm not saying he deserved his 2001 MVP because of his atrocious efficiency and the fact that there were better candidates, but the same argument could be made for Nash. Fact is, you can't win an MVP without at least being an MVP-calibre talent. That's what gets you in the top-5 of voting.

Iverson is an MVP-calibre combo guard. So is Wade. Arenas, to me, is not.

True, Iverson was available by pure circumstance. But if Melo was under the same circumstance I think it would take a lot more to get him. Also, since there are more top-flight combo guards in the NBA, they are more likely to be available more frequently, whereas you can only get a Melo-like SF through the draft or by a once-in-a-decade trade.

[tsherkin]

Well, disagreed. I would rather have Iverson than Arenas. I'm not saying he deserved his 2001 MVP because of his atrocious efficiency and the fact that there were better candidates, but the same argument could be made for Nash. Fact is, you can't win an MVP without at least being an MVP-calibre talent. That's what gets you in the top-5 of voting.

Iverson is an MVP-calibre combo guard. So is Wade. Arenas, to me, is not.

I'm gonna go with "not touching that with a 10-foot pole" because it'll derail the thread, but suffice it to say, I disagree with you vehemently. Iverson is not MVP-caliber and never has been. He's been overrated his entire career.

True, Iverson was available by pure circumstance. But if Melo was under the same circumstance I think it would take a lot more to get him. Also, since there are more top-flight combo guards in the NBA, they are more likely to be available more frequently, whereas you can only get a Melo-like SF through the draft or by a once-in-a-decade trade.

Well, if he gets moved as is being rumored, we shall see, shan't we?

[Malinhion]

We shall, we shall.

I'm willing to take up that Iverson argument any time you're feeling bored. :P

[tsherkin]

Come bug me on MSN, it'll take up too much space on the server otherwise, lol.