Everyone LOVES LOVES to suck on the teet that is Win Shares/PER, WHATEVER.

Here's a quick example of 1 current guy vs. a prominent past player, and how, despite almost IDENTICAL stats in a particular year, the Win Shares & Win shares/48 are WAYYYYYYYYYYYYYY off.

98 Jordan ~vs.~ 13 Melo

http://www.basketball-reference.com/play-index/pcm_finder.cgi?request=1&sum=0&p1=jordami01&y1=1998&p2=anthoca01&y2=2013

^ look at the numbers CLOSELY.

Some context: Jordan played in more games that year, but because of their almost IDENTICAL raw AND advanced stats, their win shares/per 48 should've ALSO been damn near identical.

I don't know how the hell to list things on here, but to quickly break some things down for you guys:

Melo scored the exact same points as MJ, on better TS% AND EFG% AND 3PT% AND FT%, with more rebounds, same blocks, and less than 1 steal and 1 assist shy of MJ.

AND YET - even when you account for MJ playing more games that season, their Win Shares/48 are VERY DIFFERENT: .184 to .238.

Jordan won the MVP that year on essentially the SAME kind of numbers that Melo put up last year, while Melo only finished 3rd. Now I'm not suggesting Melo should've finished higher (hell no he shouldn't have), but this should just go to show how RETARDED Win Share and Win Shares/48 really are.

Not that it even holds weight, but their PER's were actually identical.

And I do NOT wanna hear about "defensive differences" between the 2 that year. Jordan was WAY past his prime both athletically and defensively. He was a sieve at times, and average for most of the year. I've watched a ton of his playoff games that season, and he almost NEVER guarded the best or even 2nd best player available. He was literally HID (much like we do with Melo) b/c he could no longer handle defensive duties like Harper or Pippen could.

Basically my point is, independent of the obvious win differences between the 2 teams (and I'm excluding playoff performances here), during the regular season, Melo was every bit as good as MJ's washed up ass in 1998.

And yet MJ wins his weakest MVP, putting up the lowest # of assists in his career, playing mediocre defense. Just goes to show you how advanced stats can be utterly meaningless.

"Basically my point is, independent of the obvious win differences between the 2 teams (and I'm excluding playoff performances here), during the regular season, Melo was every bit as good as MJ's washed up ass in 1998."

Melo cant even sniff the drawers of a 34 year old MJ....melo's efficiency is strictly tied to the three ball in which he took over 300 more than MJ in far less games....thats the only thing that boosted his TS%

Jordan STILL was the more efficient player all across the board and that was with Melo playing closer to the basket as the 4.

Jordan at 34>>>>Melo at 28.

Bonn1997 wrote:All the respected advanced stats (and the eyeball test once you open your eyes to more than flashy volume scoring) confirm this actually. It's not just win shares.
Note that .184 and .230 are not that far apart and Melo had a good regular season. You're talking about the difference from being maybe top 10 vs. top in the league. You seem to be downplaying a large difference in assists and steals and exaggerating a small-to-medium difference in WS, but that's not as important as a few other issues. First, validation of these stats has to be done through statistical analyses predicting wins, not lay eyeball perception of the stats. We really care (or should at least) about how closely connected the stats are to winning, not to amateur perception of the players. Second, note that the stats are adjusted against the NBA average for the current season. The same number of points or assists or shooting efficiency is not necessarily equal when they're separated by a few decades.

First off, I don't appreciate your tone suggesting I'm just looking at "flashy volume scoring". That crap doesn't even make sense. When Melo was overall MORE efficient than MJ that year. MJ took MORE shots to score THE SAME points. Are you clueless as to what "flashy volume scoring" actually means?

Second cut the "amateur perception" BS. I'm not "amateur" and for all I know YOU ARE.

Third, all you did was prove my point. DWS aren't extrapolated the right way, which OBVIOUSLY boosted MJ's advanced stats since the 98 Bulls were clearly a better defensive team than the Knicks last year. (re: see Boozer consistently being among top DWS holders despite being a terrible defender).

On top of it, if you have to continuously adjust WS by "decade", then it shouldn't even be a stat you use to compare players in separate generations, and yet that's EXACTLY what people do.

And that's a PRETTY big difference in WS/48. ESPECIALLY considering their individual stats are almost identical. You say I "overrated" 0.9 steals and 0.9 assists? Lol WHAT? Are you serious guy? There is NO significant statistical difference in 0.9 assists more. The steals, perhaps, but again, it looks like maybe YOU'RE underrating the 1.1 rebound difference that Melo held that year.

Also there IS NO league-wide baseline for Win Shares as there is for PER, so you're wrong about that, and your post seemed to be a complex rambling of statements that have no actual context or meaning.

My point is simple - WS/48 are unreliable as hell. The eyeball test is what I prefer, but it seems like Win Shares' formula is so complicated, and ill-fitted for the NBA, where synergy and court dynamics play a big role.

It also doesn't even take into account defense in any impactful way. DWS is a HORRIBLE stat and that absolutely SKEWS the results.

If Melo had 98 Pippen and Rodman and Kukoc and Harper along with Phil Jackson, I'm pretty sure he could easily duplicate what they did in the regular season atleast.

hungerstrike wrote:

Bonn1997 wrote:All the respected advanced stats (and the eyeball test once you open your eyes to more than flashy volume scoring) confirm this actually. It's not just win shares.
Note that .184 and .230 are not that far apart and Melo had a good regular season. You're talking about the difference from being maybe top 10 vs. top in the league. You seem to be downplaying a large difference in assists and steals and exaggerating a small-to-medium difference in WS, but that's not as important as a few other issues. First, validation of these stats has to be done through statistical analyses predicting wins, not lay eyeball perception of the stats. We really care (or should at least) about how closely connected the stats are to winning, not to amateur perception of the players. Second, note that the stats are adjusted against the NBA average for the current season. The same number of points or assists or shooting efficiency is not necessarily equal when they're separated by a few decades.

First off, I don't appreciate your tone suggesting I'm just looking at "flashy volume scoring". That crap doesn't even make sense. When Melo was overall MORE efficient than MJ that year. MJ took MORE shots to score THE SAME points. Are you clueless as to what "flashy volume scoring" actually means?

Second cut the "amateur perception" BS. I'm not "amateur" and for all I know YOU ARE.

Third, all you did was prove my point. DWS aren't extrapolated the right way, which OBVIOUSLY boosted MJ's advanced stats since the 98 Bulls were clearly a better defensive team than the Knicks last year. (re: see Boozer consistently being among top DWS holders despite being a terrible defender).

On top of it, if you have to continuously adjust WS by "decade", then it shouldn't even be a stat you use to compare players in separate generations, and yet that's EXACTLY what people do.

And that's a PRETTY big difference in WS/48. ESPECIALLY considering their individual stats are almost identical. You say I "overrated" 0.9 steals and 0.9 assists? Lol WHAT? Are you serious guy? There is NO significant statistical difference in 0.9 assists more. The steals, perhaps, but again, it looks like maybe YOU'RE underrating the 1.1 rebound difference that Melo held that year.

Also there IS NO league-wide baseline for Win Shares as there is for PER, so you're wrong about that, and your post seemed to be a complex rambling of statements that have no actual context or meaning.

My point is simple - WS/48 are unreliable as hell. The eyeball test is what I prefer, but it seems like Win Shares' formula is so complicated, and ill-fitted for the NBA, where synergy and court dynamics play a big role.

It also doesn't even take into account defense in any impactful way. DWS is a HORRIBLE stat and that absolutely SKEWS the results.

If Melo had 98 Pippen and Rodman and Kukoc and Harper along with Phil Jackson, I'm pretty sure he could easily duplicate what they did in the regular season atleast.

Sorry, for the confusion. You're new here and you probably don't realize that I've used the phrase "flashy volume scoring" many times and it was not in reference to you. Same with the amateur comment. You're making many of the comments I hear from people newer to the advanced stats but you're right I don't know anything about your training. You're right that the DWS are less well-validated than the OWS are, but there was a meaningful difference (about 3.0 WS) between Melo's and Jordan's OWS that year. Also, the WS are not adjusted by decade. Rather the raw stats differ in value by year. In low scoring seasons (and the 90s were lower scoring), the same total points would have more value, for example. 0.9 assists more is about 2 more PPG assuming some were 3 pointers. How many games do teams lose by 2 points? You'd have to do a more complex adjustment if the players didn't have similar usage rates but that's not relevant here. Actually, Melo had 0.9 fewer assists despite a slightly higher usage rate. (Also, you have to look at the usage to assists ratio, rather than just assists. Someone who touches the ball a lot will get more assists just "by accident" but may still be hurting the team by touching the ball so much and not moving the ball well enough.) BTW, I never said there was a league-wide baseline. I don't know what you mean by that phrase. Finally, you're still exaggerating a fairly small difference in their WS48 numbers.

RonRon wrote:We can't even compare Melo with

Lebron, Durant, or even Paul George or Harden

Suddenly a UK member created of 12/4/13 wants to compare him with Jordan?
Really?
SMH

No, you can compare him with Harden.
Lebron is out of reach, Durant is better at what Melo is good, PG can defend. Harden is not that far imho.

Bonn1997 wrote:However, a weakness of WS that you didn't mention is that it doesn't adjust the stats for the player's position. For example, 5.8 Rbs a game from an SG (about 2 more than you'd expect from the position) will help your team much more than 6.9 from a PF (about 2 fewer than you'd expect). Some of the other position adjustments would help Melo in this regard but they'd be out-weighed by the gigantic difference in rbs.

why should I have to mention it? youre just grasping for straws here man, to undercut an argument that already held water and that's that win shares is so flawed I have no idea why people even use it.

win shares doesn't discriminate based on position, even if it should. I see no asterisks next to players win shares numbers.

the whole thing is ridiculous anyway.

hungerstrike wrote:

Bonn1997 wrote:Sorry, for the confusion. You're new here and you probably don't realize that I've used the phrase "flashy volume scoring" many times and it was not in reference to you. Same with the amateur comment. You're making many of the comments I hear from people newer to the advanced stats but you're right I don't know anything about your training. You're right that the DWS are less well-validated than the OWS are, but there was a meaningful difference (about 3.0 WS) between Melo's and Jordan's OWS that year. Also, the WS are not adjusted by decade. Rather the raw stats differ in value by year. In low scoring seasons (and the 90s were lower scoring), the same total points would have more value, for example. 0.9 assists more is about 2 more PPG assuming some were 3 pointers. How many games do teams lose by 2 points? You'd have to do a more complex adjustment if the players didn't have similar usage rates but that's not relevant here. Actually, Melo had 0.9 fewer assists despite a slightly higher usage rate. (Also, you have to look at the usage to assists ratio, rather than just assists. Someone who touches the ball a lot will get more assists just "by accident" but may still be hurting the team by touching the ball so much and not moving the ball well enough.) BTW, I never said there was a league-wide baseline. I don't know what you mean by that phrase. Finally, you're still exaggerating a fairly small difference in their WS48 numbers.

stop the sarcasm buddy boy. I didn't say "flashy volume scoring" had anything to do with me, and stop fake insulting my intelligence in regards to Win Shares when it's pretty obvious YOU'RE no expert either. and I'm not "new" in that I've stalked this site for awhile before deciding to sign up b/c there's literally no traffic but atleast the overall IQ seems to be a bit higher than other sites.

the fact that you typed all that out trying to justify your bare-bones understanding of a baseball applied sabermetrics by making excuses about how you'd have to apply league-wide differences PROVES MY POINT FOR ME ANYWAY.

win shares is a ****ty metric to use b/c of all the possible complications you already tried to mention.

you kinda tailed off there at the end, but the point still holds true - when you have two guys, who put almost identical stats, in similar minutes, on similar shots, in similar roles, minus a few MINOR differences here and there, and yet the WS/48 aren't close ENOUGH (and believe me if you know anything about Win Shares I'm not exaggerating jack, it's still a discrepancy that doesn't deserve to be there in the first place).

Am I the only one thinking this kind of post is out of place?
Mate, having an issue with win shares is one thing, being overly aggressive in a forum is another.

I for one won't spend anymore time replying to this post, until you calmed down a bit.

raven wrote:
Am I the only one thinking this kind of post is out of place?
Mate, having an issue with win shares is one thing, being overly aggressive in a forum is another.

I for one won't spend anymore time replying to this post, until you calmed down a bit.

How the hell is out of place when I didn't insult him, call him names, or use profanity.

Get outta here with that weaksauce. I've seen you guys have PLENTY of "strong opinions" on here and no one bats a damn eyelash, but when a "new poster" comes on, it's so threatening. No ones being overly aggressive AT ALL. I think maybe you're very sensitive, or don't realize that I was defending my point, which I have a right to...

And frankly I could care LESS if you "spend time" here or not. Matter fact, you should leave. B/c all you've done is stalk me in 2 threads now just to try and get under my skin and troll me randomly.

Are you experiencing "new member of the family" jealousy or something? lol

martin wrote:hungerstrike, I think you are bit out of line here and should heed some of the other posters. IMHO you are being a bit aggressive and it won't lead to good discussion. Thanks.

Whatever. I don't think I'm being that "out of line" at all when there's like 3 different guys attacking my views, and not exactly in the most polite way either.

And I certainly hope that's not some veiled "threat" that you'd "suspend" me or something, b/c frankly I could care less. I was hoping some people would appreciate my intention was to say something positive about Melo in spite of how people crap on him all the time. Sorry if people couldn't or didn't want to comprehend that. I was just trying to say something interesting is all. My bad...

I mean one guy sarcastically fake shakes his head, which is pretty damn condescending. Guy couldn't even comprehend which YEAR I was referring to. THAT doesn't make for "good discussion" either.

But I'll tell ya what - I'll just leave this thread. I think some of the guys on here are either really obstinate to ANY new view, or are just cranky as hell and can't stand a positive view on things.

Maybe I should be like that guy TKF and make a thread showing how Melo sucks worse than Mark Blount. I bet I'd get kisses for that.