John Wall Appreciation Thread - Part III

[stevemcqueen1]

Serious question: are there more than 87,000 Wizards fans?

[stevemcqueen1]

Actually... there are probably more John Wall fans than Wizards fans. And most of his votes are probably people voting multiple times.

[payitforward]

The Bulls with their all-wrong roster are 17-18. Please explain why they're better than the Kings.

Since you bring up the Bulls, I think they're a good example to make my point. The Bulls are essentially an All-Star WS48 roster. They should be one of the best teams in the league. Going by last year's WP48 numbers, Rondo is the 4th best PG in the league, Wade is the 21st best starting SG in the league, Butler is the 5th best SF in the league (and 8th best overall player in the league), Taj Gibson is the 8th best PF in the league (6th best if you only count starters), and Robin Lopez ranks 20th among starting centers. That's 2 All-NBA players and a possible All Star, plus a relatively ordinary shooting guard and center. How could they not be a 55-win team?

There's quite a disconnect here, Nate. & it indicates you haven't taken the time to figure out what it is WP48 actually claims to do. One thing it doesn't claim to do is support statements of this kind: "That's 2 All-NBA players and a possible All Star, plus a relatively ordinary shooting guard and center. How could they not be a 55-win team?"

In fact, this year, to date, WP48 suggests the Bulls should have won 18 games. They've won 17 games. That's pretty close, wouldn't you say? This year, to date, WP48 suggests the Wizards should have won 15.75 games. We've won 16 games -- again, that's pretty close.

If you want to question WP48, i.e. to point out where there might be flaws, the obvious thing to say is: "Yes, that's quite accurate at the team level, but I don't believe that the way WP48 apportions responsibility for those wins among players on the team is accurate."

One could have a discussion about that question, and it could well be a useful one. Basically, WP48 says that the lion's share of our wins comes from the play of Wall, Porter, Gortat & Beal. Both because they have been the best players and because they've played the most minutes. Perhaps you disagree?

[payitforward]

One more thing...

...But therein lies the weakness of WP48 as a prediction tool. It relies heavily on rebounds but it doesn't account for the difficulty of getting each additional marginal rebound in a team context.

1. What makes you think WP48 is "a prediction tool?"
2. What makes you think WP48 "relies heavily on rebounds?"
3. What makes you think WP48 "doesn't account for the difficulty of getting each additional marginal rebound in a team context?"

Look forward to hearing your answers to those questions.

[payitforward]

Nate -- the above is a little snarky. Why? Because I get tired of having to defend something in which I do not have the slightest interest. Something whose value is obvious, and whose limitations are obvious as well.

[pineappleheadindc]

It's a travesty that John Wall is #7 in voting in East guards. It's like Zaza Pachulia having more votes than Kawhi Leonard.

I hate mass voting. I'm becoming convinced that the majority of people are just plain stupid.

[Meliorus]

It's a travesty that John Wall is #7 in voting in East guards. It's like Zaza Pachulia having more votes than Kawhi Leonard.

I hate mass voting. I'm becoming convinced that the majority of people are just plain stupid.

It's not even bad. The current voting is just fans. When the media/player voting kicks in, he'll be much closer. Also, he has to win to get more recognition. We were the 2 seed when he got voted as a starter. I think Kemba has a right to be more upset given the way Charlotte is playing.

[CobraCommander]

Read on Twitter

"7th place!? Yo Maverick. *7th* place!?

"I'm going to use this as fuel. It's tough: we're not on national TV, so the fans don't get to see me play. All the guards ahead of me are great players; no disrespect. But they also play in bigger markets. Never mind the fact that their teams all have better records than ours. It's disrespectful. But it's good, though. I'm gonna keep playing my game, working hard on defense, and trying to lead my team. WallWay. #5Deep.

"Isaiah's having a great year. He's making 6.587 million a year pre-tax. No signature shoe; he's got some rotation player-level deal with Nike. He's doing his thing. I'm not watching his money, tho'. He signed his deal when he signed it.

"Lowry and DeMar; they got a whole country riding for them. I'm proud of them boys. Kyle's making 12 million. That's 15,881,160 in loonies. Adidas gave him some player edition shoes, but they're not paying him Harden money, so I'm cool with that. Just 'cause you stitch your initials on the tongue of a shoe don't make it a signature. Kristen Ledlow could rock Lowry's shoes on Inside Stuff and you'd think they're hers.

"I don't know what DeMar makes; he never plays point guard so it doesn't matter.

"D Rose and D Wade are like my little brothers, feel me? They get that veteran respect. Rose has worked hard. 21.3 million, and he's earned it. I thought I'd get his slot as the number one guy at Three Stripes, but then he moved to New York, and they signed Harden. But hey; I'm glad he's eating. I heard a rumor that he signed a $500 million deal with Armani Exchange back in 2014; now he's as washed as Armani Exchange is. LOL. Nah, I'm just playing. Brothers mess with each other. Rose is my man hundred grand. Hundred grand pre-tax, that is (DC takes too much of my check as it is).

"Wade's making that Li-Ning money -- he makes $15 million on off-court deals alone. He gets microfiber cooling wipes from Mission Athlete Care. He's basically making my on-court salary off-court. And his wife is making that Being Mary Jane money, too -- not a bad come up from Star Jones. No jealousy here, though. Wade put in that work. He's a legend. I mean, sure, he won his rings with LeBron and Shaq and cashed in with a non-name sneaker company, but I don't hand out asterisks. Nothing but love for Wade.

Eff Kyrie."

He ain't counting they money tho! John gonna get voted in by coaches and media. Fans dont love us DC...and Beal not even on the list while avg 22 points a game...but Lin is...and he been hurt more than Beal...FANS SHOULD FIND OUT WHO IN THE GAME WHEN THE GAME HAPPENS...no more voting for fans...because yall (we) crazy

[CobraCommander]

That's as may be, & I don't know how one defines "elite point guard," but I'm quite comfortable saying that the following guys are having better seasons than John Wall, and most of them do so every year: Chris Paul, James Harden, George Hill (before being injured), Russell Westbrook, Kyle Lowry, Steph Curry, Mike Conley, & Jeff Teague. Not a stretch to put Ricky Rubio on that list as well.

If you want to include guys who play fewer minutes -- but still play a lot -- you could also add Patrick Beverly & Patty Mills.

I'm sure there'll be plenty of push-back and lots of justification narratives; that's what fans do. Doesn't change anything.

Im simply here to remind PayItForward that he said...R.I.C.K.Y. Rubio was having a better year than Wall. Rickyyy Rubio...you embarrass yourself. If you watched the game you know you owe the board an apology...now do as your name says...Pay It Forward and go vote for Wall (and Beal) for the allstar game.

[AFM]

[nate33]

1. What makes you think WP48 is "a prediction tool?"

Well, if it fails to predict how a player will fare in a different setting, then it's not very useful. It's easy to backtest a metric that starts with wins and then assumes the players responsible for those wins are good. But which players are good, and how important are they? When I see how WP48 fails so horribly in the case of Chicago (where multiple players from different teams were assembled), then I conclude that it's not as useful as you think it is.

2. What makes you think WP48 "relies heavily on rebounds?"

It values an individual marginal rebound as being as important as a steal or a turnover. That seems intuitively wrong to me. A steal is a clear event where a possession is taken from the other team and given to your team, usually with the additional advantage of a high percentage fast break. A defensive rebound missed by an individual player has a significant chance of being grabbed by a teammate, so it doesn't result in a lost possession. When I see Jeremy Lamb, a 24% 3P shooter with a -1.8 on/off differential playing 19 minutes per game for a .500 team wildly outranking Demarr Derozan and Bradley Beal, it seems highly questionable. And then when I see Thabo Sefolosha outranking Stephen Curry and John Wall, I know it fails the smell test.

3. What makes you think WP48 "doesn't account for the difficulty of getting each additional marginal rebound in a team context?"

I explained this in a post above. And it's not just rebounds. It's defense too. It basically takes any good defensive team and assumes the team is good defensively because each individual player is pretty good or great defensively, when it fact it may just be a good defensive anchor carrying them. When the weak individual defenders who are carried by good teammates end up getting traded to different teams, their weak defense is exposed and they suddenly plummet in the WP48 metric.

For goodness sake, the top 3 shooting guards in the WP48 metric are Thabo Sefolosha, Jeremy Lamb and Joe Ingles. Do you really want to stand behind this?

[payitforward]

That's as may be, & I don't know how one defines "elite point guard," but I'm quite comfortable saying that the following guys are having better seasons than John Wall, and most of them do so every year: Chris Paul, James Harden, George Hill (before being injured), Russell Westbrook, Kyle Lowry, Steph Curry, Mike Conley, & Jeff Teague. Not a stretch to put Ricky Rubio on that list as well.

If you want to include guys who play fewer minutes -- but still play a lot -- you could also add Patrick Beverly & Patty Mills.

I'm sure there'll be plenty of push-back and lots of justification narratives; that's what fans do. Doesn't change anything.

Im simply here to remind PayItForward that he said...R.I.C.K.Y. Rubio was having a better year than Wall. Rickyyy Rubio...you embarrass yourself. If you watched the game you know you owe the board an apology...now do as your name says...Pay It Forward and go vote for Wall (and Beal) for the allstar game.

But, no, I didn't say Ricky Rubio was having a better year than Wall.

Since you give me the opportunity, let me say that as well as John Wall was playing 10 games ago -- at which point he was having a career year -- he is playing even better since then.

On the season, & keeping it to guys who've played 1000+ minutes so far, I'd say the only PGs having better years than John are Chris Paul, James Harden, Kyle Lowry, Russell Westbrook & Steph Curry. If you want to count the Greek Freak as a PG, he'd have to be on the list as well.

If you don't like that list, Commander, nothing to be done for you, sorry.

[payitforward]

1. What makes you think WP48 is "a prediction tool?"

Well, if it fails to predict how a player will fare in a different setting, then it's not very useful. It's easy to backtest a metric that starts with wins and then assumes the players responsible for those wins are good. But which players are good, and how important are they? When I see how WP48 fails so horribly in the case of Chicago (where multiple players from different teams were assembled), then I conclude that it's not as useful as you think it is.

Have you ever taken a course in Statistics, Nate? All these tools are statistical, which means that their predictions are increasingly accurate the larger the group over which the prediction ranges.

In fact, that's the only way you measure accuracy of a statistical analysis, isn't it? Just as, if I gave you an example of a Mexican family that did better and better generation to generation you wouldn't accept that as disproving (or even evidence against) the larger statistical fact that Mexican families overall haven't shown that improvement, so too the fact that Rajon Rondo is having a bad year after having a good one last year doesn't establish what you suggest above. Not even if it's true of Lopez as well.

2. What makes you think WP48 "relies heavily on rebounds?"

It values an individual marginal rebound as being as important as a steal or a turnover. That seems intuitively wrong to me. A steal is a clear event where a possession is taken from the other team and given to your team, usually with the additional advantage of a high percentage fast break. A defensive rebound missed by an individual player has a significant chance of being grabbed by a teammate, so it doesn't result in a lost possession. When I see Jeremy Lamb, a 24% 3P shooter with a -1.8 on/off differential playing 19 minutes per game for a .500 team wildly outranking Demarr Derozan and Bradley Beal, it seems highly questionable. And then when I see Thabo Sefolosha outranking Stephen Curry and John Wall, I know it fails the smell test.

A. WP48 doesn't "value an individual marginal rebound as being as important as a steal or a turnover." It does value an offensive rebound that way, however. Let me ask this... do you understand how someone like Dave Berri (the Phd sports economist who developed WP48) would get to a "value" for any kind of event in a game? Or how anyone develops & tests any statistical analytical tool? Or, I could just ask if you are familiar with software like SAS?

B. Do you know who the highest-WP48 2 guard in the league is this year? Danuel House. Of course, he's only played 1 minute. Just as statistical predictions are more accurate over larger numbers than small numbers, so too they are more significant for an individual case over a large sample than a small sample. Jeremy Lamb's only played 400+ minutes this year, but somehow he's managed to get 10+ rebounds per 40 minutes (edit: make that almost 11 boards/40 minutes). It's an outlier number; he's never done anything like that in the past (tho he is a pretty good rebounder for a guard). I imagine it'll come back to earth by the end of the year. For the moment it pushes up his WP48. It's a fact without any meaning whatever. See my point above about an outlier family.

C. "fails the smell test" ?? -- did you see my list above of the top PGs this year? Chris Paul, James Harden, Kyle Lowry, Russell Westbrook, Steph Curry & John Wall. You have a problem with that list? Or, does the fact that Danuel House ranks #1 in WP48 among SGs give you a problem? A yes in either case seals the deal that you don't understand how statistical tools work.

3. What makes you think WP48 "doesn't account for the difficulty of getting each additional marginal rebound in a team context?"

I explained this in a post above. And it's not just rebounds. It's defense too. It basically takes any good defensive team and assumes the team is good defensively because each individual player is pretty good or great defensively, when it fact it may just be a good defensive anchor carrying them. When the weak individual defenders who are carried by good teammates end up getting traded to different teams, their weak defense is exposed and they suddenly plummet in the WP48 metric.

Sigh. That's not a response, is it? How would a social scientist "account for the difficulty of getting each additional marginal rebound in a team context?" Or, in general, account for any such problem or issue? By running regressions in the stat software. If you are really suggesting that this academic economist can't run regressions, or somehow doesn't, what's the point of this discussion?

For goodness sake, the top 3 shooting guards in the WP48 metric are Thabo Sefolosha, Jeremy Lamb and Joe Ingles. Do you really want to stand behind this?

You forgot Danuel House!

Again, you miss the point, i.e. you insist on not understanding how statistical tools work. So, lets look at shooting guards who've played 1000+ minutes (so that we have the most reliable sample we can come up with). Top 5 in WP48 are: Jimmy Butler (who does play some minutes at the 2 but really would be better thought of as a 3), Avery Bradley, Kentavious Caldwell-Pope, Bradley Beal & DeMar DeRozan.

K C-P & Beal are the outliers on that list, compared to previous years. But if you look at their numbers, you can see they are both way improved over earlier years. As to Sefolosha, btw, he is a terrific player. You don't have to be a big-time scorer to be a terrific player.

As to your point about defense, you invented it out of whole cloth. Overall, as they move from team to team, players' WP48 doesn't change much. Now, players are people not machines, so there's some variation year to year. And there's also an arc of development - peak - decline. But regressing out the latter, there's not much change in a guy's WP48 as he moves from team to team.

Once again, like any fact of its kind, that's truer over a large number than a small number, and of course it's also truer over larger samples than small. A guy plays 500 minutes one year, instead of 2500 minutes, his WP48 has a higher likelihood of high variance from his norm than if he plays his usual 2500 minutes.

WP48 isn't magic; it isn't "the truth"; it can't predict everything. It's easy to come up with a case like the Bulls where, at least so far, guys who've had more good seasons than not are having "not" seasons at the same time. Only a fool would deny that this can happen. Or deny... really anything at the level of singular events in the real world. But that's not an argument against a statistical tool. Sorry.

As I've said a bazillion times now (sorry, anyone but nate who bothers to read this), WP48 is only better than anything else of its kind. That's all. Not a religion, not establishing any metaphysical facts about the "goodness" of a player.

[payitforward]

Nate -- if you really want to continue this, I'll honor that by being responsive. But I don't think we should do it in this forum, ok? We'd have to do via PMs. Otherwise, the discussion is hijacking too much space.

[nate33]

For goodness sake, the top 3 shooting guards in the WP48 metric are Thabo Sefolosha, Jeremy Lamb and Joe Ingles. Do you really want to stand behind this?

You forgot Danuel House!

Again, you miss the point, i.e. you insist on not understanding how statistical tools work. So, lets look at shooting guards who've played 1000+ minutes (so that we have the most reliable sample we can come up with). Top 5 in WP48 are: Jimmy Butler (who does play some minutes at the 2 but really would be better thought of as a 3), Avery Bradley, Kentavious Caldwell-Pope, Bradley Beal & DeMar DeRozan.

Now you're just being disingenuous because you know my point is compelling. I saw how you slipped in a 1000 minute cutoff to eliminate all the horrible selections made by WP48. Why not a 700 minute cutoff so Sefolosha, Ingles, Tony Allen, Seth Curry, and Belinelli are included? Most of those guys are starters or play heavy minutes off the bench.

By arbitrarily picking a 1000 minute cutoff, you are reducing the possible choices of shooting guards to just 11, the 11 shooting guards trusted most by their coaches to play heavy minutes. Of course when you reduce the available players to just the 11 most utilized SG's in the league, you're going to end up with mostly good players. Heck, if you just rank the SG's solely by minutes played per game, and throw out anyone playing for a team with a .400 record or worse, you get this ranking:

DeRozan
Bradley
McCollum
Thompson
Beal
Caldwell-Pope

That's as good of a ranking as WP48. Maybe even better.

But I agree that there's no point in discussing this further. Neither of us are budging from our viewpoint.

[payitforward]

I don't have a viewpoint, nate, I grew up watching Dragnet: "just the facts, ma'am."

But, yeah, from that I'm not budging so lets drop it.

[CobraCommander]

That's as may be, & I don't know how one defines "elite point guard," but I'm quite comfortable saying that the following guys are having better seasons than John Wall, and most of them do so every year: Chris Paul, James Harden, George Hill (before being injured), Russell Westbrook, Kyle Lowry, Steph Curry, Mike Conley, & Jeff Teague. Not a stretch to put Ricky Rubio on that list as well.

If you want to include guys who play fewer minutes -- but still play a lot -- you could also add Patrick Beverly & Patty Mills.

I'm sure there'll be plenty of push-back and lots of justification narratives; that's what fans do. Doesn't change anything.

Im simply here to remind PayItForward that he said...R.I.C.K.Y. Rubio was having a better year than Wall. Rickyyy Rubio...you embarrass yourself. If you watched the game you know you owe the board an apology...now do as your name says...Pay It Forward and go vote for Wall (and Beal) for the allstar game.

But, no, I didn't say Ricky Rubio was having a better year than Wall.

Since you give me the opportunity, let me say that as well as John Wall was playing 10 games ago -- at which point he was having a career year -- he is playing even better since then.

On the season, & keeping it to guys who've played 1000+ minutes so far, I'd say the only PGs having better years than John are Chris Paul, James Harden, Kyle Lowry, Russell Westbrook & Steph Curry. If you want to count the Greek Freak as a PG, he'd have to be on the list as well.

If you don't like that list, Commander, nothing to be done for you, sorry.

Hmm not buying it. Kyle Lowry will have to see John and if Beal is healthy we will see how those guards do against our guard. Paul and Steph are both having down years...while JWs stock is rising...steadily. Did you go to the AllStar vote and Pay it forward...and vote for JW and Beal?

[Meliorus]

With Lowry, when a 30 year old suddenly jumps to an incredible TS% of 64.2, it's more of an aberration than a sustainable thing. He'll fall back down to Earth.

[payitforward]

Im simply here to remind PayItForward that he said...R.I.C.K.Y. Rubio was having a better year than Wall. Rickyyy Rubio...you embarrass yourself. If you watched the game you know you owe the board an apology...now do as your name says...Pay It Forward and go vote for Wall (and Beal) for the allstar game.

But, no, I didn't say Ricky Rubio was having a better year than Wall.

Since you give me the opportunity, let me say that as well as John Wall was playing 10 games ago -- at which point he was having a career year -- he is playing even better since then.

On the season, & keeping it to guys who've played 1000+ minutes so far, I'd say the only PGs having better years than John are Chris Paul, James Harden, Kyle Lowry, Russell Westbrook & Steph Curry. If you want to count the Greek Freak as a PG, he'd have to be on the list as well.

If you don't like that list, Commander, nothing to be done for you, sorry.

Hmm not buying it. Kyle Lowry will have to see John and if Beal is healthy we will see how those guards do against our guard. Paul and Steph are both having down years...while JWs stock is rising...steadily. Did you go to the AllStar vote and Pay it forward...and vote for JW and Beal?

Well, you are certainly consistent no question about that. Consistently wrong (not to mention insulting, but twerps like you aren't worth responding to on that issue).

The idea that Chris Paul is having "a down year" is particularly ludicrous! This is a guy who will likely go down in history as the best point guard ever to play the game, and he is having literally the best year of his career. Shows how much you know, huh?

Curry is down it's true -- he's better than John, obviously, but not playing at the level of the last couple of years. Who knows? He could get worse and worse until he's awful, and that would prove how great John Wall is, right?

What I don't understand is why it isn't enough for John Wall to be having a career year, for him to be the terrific player he is. I don't understand why that isn't good enough, why it's necessary to ding some other player to make John look good. Because, you know, it doesn't. And he doesn't need it.

As to whom I vote for in the All Star game, what business is that of yours? I don't need to prove my fan creds to you. I'm sure I've been an NBA fan longer than you have, and I'd be surprised if you've been a Bullets/Wizards fan as long as I have for that matter.

Tell you what, since you like telling me what my handle says I should be doing, why don't you change yours from CobraCommander to Corporal Garter Snake -- just to get your name a little closer to your place in the scheme of things.

Oh gosh... I guess I did respond to the insulting part. Go eat a mouse.

[payitforward]

That's all meant tongue in cheek, CabroCermmonda. Stay just as you are & never change.