mrKnickShot wrote:

ChuckBuck wrote:

mrKnickShot wrote:

Bonn1997 wrote:

mrKnickShot wrote:

Bonn1997 wrote:

Anji wrote:What's his total Bases per AB's???

Much lower than Gallo's

Negative bashing.

That was a joke. You took that seriously???? You know there's no such thing as total bases and ABs in basketball, right?

I did not know that. Weird wild stuff!

And, Hilarious!

You two really need to go out on a Veggiefishing date!

Thats mean ChuckBuck! Even I did not go there.

Sorry, just trying to break up the exchanges like a ring ref.

mrKnickShot wrote:In Melo's worst year, his VI (Versatility Index - By Hollinger) is 9.3. Tied for 18th in the league with Kobe Bryant and Chris Paul. He is also 3rd among SF's behind only Lebron and KD.

WOW!! I'm impressed!

All that said, Melo actually is having a good season and is a good player - not great but good. His FG% is down but everything else is up from his career numbers.

jrodmc wrote:

Bonn1997 wrote:

mrKnickShot wrote:In Melo's worst year, his VI (Versatility Index - By Hollinger) is 9.3. Tied for 18th in the league with Kobe Bryant and Chris Paul. He is also 3rd among SF's behind only Lebron and KD.

WOW!! I'm impressed!

That is irrelevant to this thread but anyway...the Hollinger stats are garbage. As far as I am aware, he gives variables whatever weights he feels are reasonable. Win shares and wins produced are statistically validated approaches - variables are given their weights in regression equations based on how accurately the weighting predicts actual wins. Basically, you start with the outcome you're interested in (that is, wins) and see how strongly each variable (e.g., scoring, scoring efficiency, rebounds, etc.) correlates with the outcome. The strength of the correlation indicates how much weight the variable should have. It's a better approach than just guessing the weight.

All that said, Melo actually is having a good season and is a good player - not great but good. His FG% is down but everything else is up from his career numbers.

Still, any reasonable person would still rather be watching Gallo on the IR. A true superstar, with the much taller, more attractive variable correlation. Plus he's not near as fat as Melo.

IrishKnickFan wrote:

jrodmc wrote:

Bonn1997 wrote:

mrKnickShot wrote:In Melo's worst year, his VI (Versatility Index - By Hollinger) is 9.3. Tied for 18th in the league with Kobe Bryant and Chris Paul. He is also 3rd among SF's behind only Lebron and KD.

WOW!! I'm impressed!

That is irrelevant to this thread but anyway...the Hollinger stats are garbage. As far as I am aware, he gives variables whatever weights he feels are reasonable. Win shares and wins produced are statistically validated approaches - variables are given their weights in regression equations based on how accurately the weighting predicts actual wins. Basically, you start with the outcome you're interested in (that is, wins) and see how strongly each variable (e.g., scoring, scoring efficiency, rebounds, etc.) correlates with the outcome. The strength of the correlation indicates how much weight the variable should have. It's a better approach than just guessing the weight.

All that said, Melo actually is having a good season and is a good player - not great but good. His FG% is down but everything else is up from his career numbers.

Still, any reasonable person would still rather be watching Gallo on the IR. A true superstar, with the much taller, more attractive variable correlation. Plus he's not near as fat as Melo.

We could have had both of these guys which would have been great since melo has proven he could work the 4. Just melo could have waited for Free agency

Yes, and Gallo's proven the well known idiom that (Pat Riley notwithstanding) it truly takes an eyetalian to look truly good in those Armani suits. Melo could play the 4. Gallo could man the IR instead of Stat. Perfectly brilliant scenario!

Bonn1997 wrote:

mrKnickShot wrote:In Melo's worst year, his VI (Versatility Index - By Hollinger) is 9.3. Tied for 18th in the league with Kobe Bryant and Chris Paul. He is also 3rd among SF's behind only Lebron and KD.

WOW!! I'm impressed!

That is irrelevant to this thread but anyway...the Hollinger stats are garbage. As far as I am aware, he gives variables whatever weights he feels are reasonable. Win shares and wins produced are statistically validated approaches - variables are given their weights in regression equations based on how accurately the weighting predicts actual wins. Basically, you start with the outcome you're interested in (that is, wins) and see how strongly each variable (e.g., scoring, scoring efficiency, rebounds, etc.) correlates with the outcome. The strength of the correlation indicates how much weight the variable should have. It's a better approach than just guessing the weight.

All that said, Melo actually is having a good season and is a good player - not great but good. His FG% is down but everything else is up from his career numbers.

People always forget that stats are only good for showing what happened in the past. Way too much weight is given to predictions/projections drawn from inadequate data samples and it is the touted like its the end all be all in a discussion. I agree with you in that Hollinger's stuff is mostly garbage. He has no explanations for the constants he uses or weights given. I looked up win shares and didn't see anywhere that 3 point buckets are treated like home runs in baseball. I believe someone posted that as a reason among others to discard win shares.

I always look for stats that are essentially simple so I am still not completely sold on Win shares but the premise seems to be solid. But the important thing is to remember that past performance is only a broad indication of future possibility and Correlation does NOT equal Causation. What it does not mean is that a player with 8 years of solid performance is better in the present moment than a rookie with no past data. Though some people seem to draw such conclusions routinely.

It does have some pure entertainment value though. You posted Melo's stats for this season without even expressing an opinion, and it was immediately labelled negative bashing. Hilarious.

Anji wrote:" I looked up win shares and didn't see anywhere that 3 point buckets are treated like home runs in baseball."

This poster never said that. I said that baseball has a value system that you can assign weights more easily to for an equation like W/S to convey exactly what plays and players are and can contribute to "winning" you games. Because 4bases/homerun being the top and an out being the bottom.

Steve Novack was being dubbed the best of our three forwards, while coming in last in any stat that didn't involved 3point shooting. One aspect of his game completely distorted his W/S, as if he was achieving the a play that held the highest weight of the equation. There is no Home Run in basketball, and that was my point.

What part of the formula led you to conclude that only his 3 point shooting " distorted" the win shares in his favor? Can you back this up? I don't need another baseball analogy, just tell me why you think his 3 pt shooting is being given undue preference.

loweyecue wrote:

Bonn1997 wrote:

mrKnickShot wrote:In Melo's worst year, his VI (Versatility Index - By Hollinger) is 9.3. Tied for 18th in the league with Kobe Bryant and Chris Paul. He is also 3rd among SF's behind only Lebron and KD.

WOW!! I'm impressed!

That is irrelevant to this thread but anyway...the Hollinger stats are garbage. As far as I am aware, he gives variables whatever weights he feels are reasonable. Win shares and wins produced are statistically validated approaches - variables are given their weights in regression equations based on how accurately the weighting predicts actual wins. Basically, you start with the outcome you're interested in (that is, wins) and see how strongly each variable (e.g., scoring, scoring efficiency, rebounds, etc.) correlates with the outcome. The strength of the correlation indicates how much weight the variable should have. It's a better approach than just guessing the weight.

All that said, Melo actually is having a good season and is a good player - not great but good. His FG% is down but everything else is up from his career numbers.

People always forget that stats are only good for showing what happened in the past. Way too much weight is given to predictions/projections drawn from inadequate data samples and it is the touted like its the end all be all in a discussion. I agree with you in that Hollinger's stuff is mostly garbage. He has no explanations for the constants he uses or weights given. I looked up win shares and didn't see anywhere that 3 point buckets are treated like home runs in baseball. I believe someone posted that as a reason among others to discard win shares.

I always look for stats that are essentially simple so I am still not completely sold on Win shares but the premise seems to be solid. But the important thing is to remember that past performance is only a broad indication of future possibility and Correlation does NOT equal Causation. What it does not mean is that a player with 8 years of solid performance is better in the present moment than a rookie with no past data. Though some people seem to draw such conclusions routinely.

It does have some pure entertainment value though. You posted Melo's stats for this season without even expressing an opinion, and it was immediately labelled negative bashing. Hilarious.