#19 Highest Peak of All Time (Ewing '90 wins)

[Doctor MJ]

Dirk Nowitzki '11 has been enshrined. We move on.

[Doctor MJ]

Vote: West '68

For the same reasons I've said before.

[Dr Positivity]

Edit - Changing my vote to 1990 Ewing

Vote 2003 Tmac

He's a more physically imposing talent than West to me. Enough to make a slight difference

[thizznation]

Does anyone have a good explanation for '03 TMac's poor RAPM?

[C-izMe]

Vote: 03 TMac

Can't go with the guy who missed 30 games yet.

[thizznation]

You can't assume that '68 West is always going to miss 30 games. The 30 missed games tell us that the sample size is going to be smaller when trying to determine how high the peak in examination really is. I'm pretty sure the '68 West voters are not voting him under the assumption he is going to miss a large chunk of the season.

[Doctor MJ]

the one thing that really confuses me is why you - one of the biggest proponents of the plus minus stats - didn't mention high error earlier ? in particular when you were using 4.5 game sample to determine Malone's value without Stockton ? it didn't seem to bother you then so why change your approach ? is it possible that supremely unbiased ElGee lost his way ?

furthermore West's in/out is a RS stat. Lakers weren't playing like 8 SRS team in the playoffs. yet West's RS stat is being used as an argument here as if RS had relevance... but then you say missing 40% of RS games is irrelevant. so I ask - which is it ? you can't have it both ways. either RS is important and you have to punish West greatly for missing its major part, or you go with PS results only, using in/out as your context info, not as a main argument. this in/out is the main reason why West 68 is getting voted in before his other seasons.

I think that's the Elgee's double standards about which mysticcbb is talking recently and I agree with him and you.

Okay so guys, when I see stuff like this in concerns me. This is an awful lot of talking about a poster, and not about the actual basketball.

You're all certainly entitled to your opinions about other posters, but when you agree to be a part of one of these projects, you're agreeing to operate with a level of politeness above what you'd get in a normal RealGM thread.

Why? Because these projects require large amounts of commitment, and if it gets too negative, we lose good people. TrueLAfan retired from RealGM twice (the second for good it seems), both happened after stints in the thick of projects which he felt were too negative.

So let's move on folks. All of us - myself included - get heated about these things, and it's worth it for all of us to take a step back and remember this is supposed to be fun.

~Doc

[Doctor MJ]

Can't go with the guy who missed 30 games yet.

Champion another of West's years if you believe in West but not in his '68 season.

[kasino]

67 Rick Barry

[Doctor MJ]

Does anyone have a good explanation for '03 TMac's poor RAPM?

Oh wow. I hadn't notice that before. I don't know what to make of it to be honest.

Here's one thing I'll say though:

Engelmann uses prior seasons as part of his algorithm. This makes a lot of sense if you want to get a sense of who truly is the most valuable player right now, but so far as I can tell, it means that a player who takes a leap in a given year is going to get underrated by the stat. Hence you'll see players have a tendency to really put up huge numbers in their second year at a super high level.

We don't see this with McGrady in '04 of course...but then he didn't maintain his level from '03. So simply put, given those facts, and the fact that McGrady was relatively high up on the list for Rosenbaum's APM for '03, I think McGrady's plainly getting underrated by the stat.

However, that number is REALLY low. I don't feel like what I've said above can alone explain it.

[Doctor MJ]

67 Rick Barry

Huh, no reasoning?

[An Unbiased Fan]

Can someone explain how 2003 Tmac is better than 2009 Dwight?

[thizznation]

Doc, I was thinking something similar. Since they didn't have as much previous data in '03, the results could be not as accurate when compared to that of some of the more recent years that have been using almost a decade of back-data.


A lesser version of the '11 Dirk that just got voted in, '03 Dirk, scored 4.0 on RAPM on the same year as T-Mac's 1.4. I know RAPM isn't the end all to be all and has it's kinks but I thought it should be brought up.

[Lightning25]

Can someone explain how 2003 Tmac is better than 2009 Dwight?

He's better than 2009 Dwight, but not 2011 Dwight imo.

My vote is the same as last time '68 West, best two-way player remaining.

Tracy Mcgrady shouldn't be in so early. I don't even think Tmac was ever even better than current Kevin Durant.

[Doctor MJ]

Incidentally, I decided to do some analysis of Engelmann's 1-year RAPM outlier performers by normalizing with standard deviation, and since McGrady's RAPM got brought up here, it seems rather relevant.

Preface: mystic mentioned that there's more that ideally should be done to make these measurements better, and he's right. This is what I had time to do a few nights back, but if someone has better stuff, that'd be awesome.

Okay, listed by rank, the performances 3.5+ StDev or better:

1. '10 LeBron +4.35
2. '04 Garnett +4.26
3. '09 LeBron +3.95
4. '03 Garnett +3.89
5. '07 Duncan +3.75
6. '11 Nowitzki +3.74
7. '10 Wade +3.73
8. '12 Nowitzki +3.67

When you look at the results with values of 2.5+, a "Big 10" emerges from the RAPM era. Here's a chart, listing the player, and his number of years at +4, +3.5, +3, and +2.5 respectively. The 10 listed are the only players with at least 2 +2.5 years.

Code: Select all

Player   +4 +3.5 +3 +2.5
Bryant    0   0   0   2
Duncan    0   1   3   5
Garnett   1   2   4   7
Ginobili  0   0   2   6
James     1   2   3   6
Nash      0   0   0   4
Nowitzki  0   2   2   4
O'Neal    0   0   1   3
Paul      0   0   0   2
Wade      0   1   1   3

McGrady, now that I check him, only just barely hits +1.5 once in his entire career. It's kind of crazy.

Of note: The Houston years don't seem that odd to me. There's where he's hovering around in the 1-1.5 range which you might find a bit low, but obviously he was only a contender here based on his Orlando work. But those Orlando years look downright mediocre, and that's weird enough I don't know if I really have confidence in them.

[bastillon]

if we're honestly doing this "what if he's on another team" I can't believe why you would take 68 West over 66/70. the argument that ElGee is mainly propping up is that RS doesn't matter on title contenders because they'll get to the playoffs anyway and then what you're actually doing when healthy in the postseason matters more. but where's the evidence that championship caliber teams can withstand 30 games without its best player ? why would I make a false assumption they'll just get to the playoffs easily ? there were many championship teams that struggled badly without their best player. 58-69 Celtics were championship caliber team every year and they were 10-18 without Russell. Rockets 92-96 were 7-27 without Olajuwon and they won two titles in that time span. Mavs were like what, 1-6 without Dirk ? 05-07 Suns were 4-13 without Nash. there have been plenty of title caliber teams who would could very easily miss the playoffs altogether if they played in a strong year with their best player missing that many games.

for that reason I can't fathom how possibly someone would take slightly better impact in the RS over 66/70 West playing 20-25 more games. seems pretty insane to me to be honest.

[bastillon]

some posts that were left without response in the last thread:

West's overrated 68 postseason:

What's 68 West argument over 11 Dirk in the playoffs? West beat 2 teams with an average of 36 wins in the western conference finals before losing a 3.87 SRS team in the finals. It's not like West stats improved a lot in the playoffs. If West had a true 8 SRS+ team like Elgee is claiming, why did they lose in the finals with a SRS half of that? Wouldn't that be a huge blackmark against West that his team lost to a team half as good?

West's defensive and clutch reputation without any evidence to back it up:

Doc MJ, imagine we wouldn't have any kind of pbp data now and someone in 40 years would tell you that Bryant was an awesome clutch player and great defender. What would be your reaction? Well, you are doing the same with West right now, proclaiming him as being incredible clutch and a great defender based on hearsay. The story of the witnesses and the media might not be the accurate one. Like Bryant, West might have been just more often in such close game situations, making it more likely to have some clutch success without saying anything about his efficiency during those minutes. You should keep that in mind as well.

That's great point. Even now, when we basically are able to watch every game, this kind of bias is strong, even among media members. So imagine what happened during 60s when journalist usually watched players only when they were playing in media members city. Maybe finals were more available in national TV during 60s? That would explain West's clutch reputation, who obviously often played in the finals and that created his image as mega clutch player, because nobody (except Celtics players - but that was Russell's team, and BTW, Sam Jones also have great clutch reputation) during 60s played as often in these clutch situations which were seen in national media.

lack of playing time. I would safely assume a guy who plays much fewer minutes has a lower impact than his other version who plays a lot more.

Another mark against 68 West is that his MPG was the lowest it was since his rookie year. So in 66 he played 68% more regular season minutes than 68 and 62% more in 70 compared to 68. I doubt West's impact in those minutes is enough to overcome 68% more minutes of a slightly worse West.

Of course his playoffs in 68 was very overrated. The Lakers didn't beat a team with a positive SRS in that postseason. In fact the average SRS of their opponents in the Western conference playoffs was -2.39 SRS and 36 wins or the equivalent of last season's Golden State Warriors/TWolves. The Lakers beat a 29 win team in that playoff. Thats the equivalent of sending the Toronto Raptors to the playoffs.

my take on why 70 West over 68:

why West 66 or 68 over 70 ? historical analysis tells us he became much more of a lead guard that year, ended up bringing the ball up the floor and just played with the ball a lot more. statistical analysis shows how that impacted his boxscore stats - more ppg, much more apg. so what do we have for 66/68 ? I have a hard time believing that a player like West would have an objectively worse impact playing more without the ball, the very point of wings being better than bigs on offense is their ability to control the offense simply because they play with it more often, instead of relying on entry passes into the post. I can't stress enough how dubious that makes me about selecting 66/68. do you honestly expect West to play better if you give him less opportunities to produce ? it's not just this narrative that makes me so confident in 70 over 60s peaks, his offensive production also rose to new levels because of his passing.

66: 34.2 ppg 5.6 apg 6.3 rpg @ 58.1% TS
68: 30.8 ppg 5.5 apg 5.4 rpg @ 59.6% TS
70: 31.2 ppg 8.4 apg 3.7 rpg @ 55% TS

you could see a pattern... 70s West avged many more assists in the playoffs than 60s West. I like late peaks. I'm giving the benefit of the doubt to the more experienced player playing in a role that suits him better. in general I want my best perimeter player to handle the ball.

as for his defense, I'd like to follow up on drza's point... why would I expect ANY superstar guard to have more impact than Dirk on defense ? I'm not just talking about Jerry West, Wade and Kobe. superstar guards are by definition offensive anchors and their responsibilities are so large that they can't exert much energy on defense. as a result you pretty much don't see superstar guards to have defensive impact (talking about in/out type of stats). I mean freakin Michael Jordan left the team in 94 and Bulls defense improved. I've also recently created a thread outright questioning Frazier/Payton defense... no responses. I just don't expect a guard to make a large impact defensively when he's exerting so much energy on offense. even Manu Ginobili who drza brought up as an example was a low minute player who went all out in his (relatively) limited time on the floor. I just don't see him maintaining quite the same defensive intensity if he was playing 40 mpg. I think everyone agrees with me on this. with Dirk measuring better than the best defensive guards of the last generation, why would we expect West to beat him on defense ?

one more thing, I can't believe someone brought up West's clutch scoring as a point against 2011 Dirk. I'm going to go out on a limb and say this: nobody in history approached Dirk's crunch time dominance from that postseason. nobody. not Bird, not Jordan, not Kobe, not anybody.

@ ElGee, you said RS measures "goodness" with quite big sample size... how about the fact that West 70 31/7.5/4.6 @ 57% TS vs 26/6/6 @ 59% TS in 68 ?

all in all I see 70 West as a better playoff performer, who was much more durable, didn't stat-padd against poor playoff competition, gave you the same skillset but was playing more with the ball (I will argue West playing off ball doesn't give you quite as much impact). all this seems like 68 West is being voted in based on in/out stats because he doesn't really have anything over his far superior 70 version.

[Doctor MJ]

if we're honestly doing this "what if he's on another team" I can't believe why you would take 68 West over 66/70. the argument that ElGee is mainly propping up is that RS doesn't matter on title contenders because they'll get to the playoffs anyway and then what you're actually doing when healthy in the postseason matters more. but where's the evidence that championship caliber teams can withstand 30 games without its best player ? why would I make a false assumption they'll just get to the playoffs easily ? there were many championship teams that struggled badly without their best player. 58-69 Celtics were championship caliber team every year and they were 10-18 without Russell. Rockets 92-96 were 7-27 without Olajuwon and they won two titles in that time span. Mavs were like what, 1-6 without Dirk ? 05-07 Suns were 4-13 without Nash. there have been plenty of title caliber teams who would could very easily miss the playoffs altogether if they played in a strong year with their best player missing that many games.

for that reason I can't fathom how possibly someone would take slightly better impact in the RS over 66/70 West playing 20-25 more games. seems pretty insane to me to be honest.

I know you're not replying to me specifically, but my own personal feeling (which I phrase this way because I know many won't agree) is this:

If a regular season injury actually didn't get in the way of the team's post-season opportunities, why exactly should I care about it?

I understand that anyone focused on essentially tallying all the good things a player did throughout the year would care, and I respect that.

But for me, what I'm typically interested in is just the most impressive player. Injuries absolutely can get in the way of a player's ability to impress, even if he's truly healthy when he does play, but if there are no real consequences in a given situation, it just seems so pedantic to fixate on them.

I'll also say that if you believe the missed time allowed the player in question to, in effect, try harder in the games he did play, I can certainly understand that as an argument to make against that players candidacy too, but I don't think anyone's making that argument.

The arguments I'm typically seeing made are fishy because they are extremely passionate, and they tend to fluctuate in their attack. Oscar got in several threads ago, and so to me the obvious reason why West drops is because of the disagreement about the '68 season, and that's really quite understandable. But in the last thread, I started seeing such harsh attacks about West, such disrespect to contemporary opinion without any analysis supporting a different conclusion, and I start wondering what the heck is happening. Where was that vitriol when Oscar was getting voted in?

[thizznation]

some posts that were left without response in the last thread:

West's overrated 68 postseason:

What's 68 West argument over 11 Dirk in the playoffs? West beat 2 teams with an average of 36 wins in the western conference finals before losing a 3.87 SRS team in the finals. It's not like West stats improved a lot in the playoffs. If West had a true 8 SRS+ team like Elgee is claiming, why did they lose in the finals with a SRS half of that? Wouldn't that be a huge blackmark against West that his team lost to a team half as good?

What, that he lost to Bill Russel's Celtcs in the finals and now it turns into an overrated post-season?

West's defensive and clutch reputation without any evidence to back it up

With all due respect this seems like borderline trolling. This isn't even worth a response.

[bastillon]

Dude, you're the one getting so miffed when people use an APM other than RAPM, and using multi-year instead of single year. I look at Engelmann's single years, measure it in terms of standard deviations, and this is what I get:

'03 - 2.3 stdev
'04 - 2.5 stdev
'05 - 2.4 stdev
'06 - 2.4 stdev
'07 - 2.6 stdev
'08 - 2.5 stdev
'09 - 1.7 stdev
'10 - 1.5 stdev
'11 - 3.9 stdev
'12 - 3.7 stdev

Don't tell me that's not weird. Something changed dramatically in '11. Either it was Dirk becoming a fundamentally better player, or it was the team becoming a great fit for him, or we're seeing something weirder than that.

What do you think it is?

oh, since you're here, why didn't you apply the same standard to Jerry West ? wasn't his impact higher in 68 because team around him improved in terms of fit ?