[dmo21Participant]

 So I have done this from time to time throughout the years. A simple rundown of it is that it is a positional ranking based on this current year’s stats that were put into my formula, and that’s it. No biassed opinions or anything here. If you want to know more or see the last rankings, click here: http://www.nbadraft.net/forum/positional-rankings-so-far. If you do have an opinion, please comment, but read the stuff in the link before as it is explained more in there. So here they are:

PG:

1. Chris Paul (141.6)
2. Stephen Curry (109.2)
3. Kyle Lowry (102.0)
4. Isiaah Thomas (98.5)
5. Damian Lillard (95.0)
6. Ty Lawson (94.9)
7. Goran Dragic (94.4)
8. Mike Conley (89.8)
9. Kyrie Irving (88.6)
10. John Wall (88.6)

SG:

1. James Harden (84.9)
2. Wesley Matthews (78.9)
3. Dwayne Wade (74.2)
4. Manu Ginobili (73.6)
5. Arron Afflalo (72.3)
6. Kyle Korver (69.7)
7. Eric Bledsoe (67.9)- figured I would call him a SG as I put Dragic as a PG
8. Kevin Martin (67.0)
9. Lance Stephenson (64.5)
10. Monta Ellis (63.8)

SF:

1. Kevin Durant (126.2)- surprised he took over Lebron
2. LeBron James (124.7)
3. Paul George (90.5)
4. Carmelo Anthony (81.0)
5. Chandler Parsons (73.8)
6. Nicolas Batum (68.4)
7. Andre Iguodala (65.7)
8. Martel Webster (63.9)
9. Kawhi Leonard (62.9)
10. Leol Deng (59.9)

PF:

1. Kevin Love (110.9)- LeBron would be first with 122.5
2. Blake Griffin (92.4)
3. Anthony Davis (89.5)
4. Dirk Nowitzki (89.5)
5. David Lee (84.8)
6. LaMarcus Aldridge (81.6)
7. Serge Ibaka (73.5)
8. Paul Millsap (72.9)
9. Tim Duncan (72.2)
10. Kenneth Faried (71.4)

C:

1. DeMarcus Cousins (93.3)- A.Davis would be first with 96.5
2. DeAndre Jordan (92.4)
3. Dwight Howard (92.0)
4. Andre Drummond (90.5)
5. Nikola Pekovic (84.7)
6. Andrew Bogut (83.6)
7. Chris Bosh (83.6)
8. Jordan Hill (81.9)
9. Robin Lopez (80.7)
10. Roy Hibbert (76.5)

I left out some guys due to injury (if they haven’t logged enough games or aren’t returning for the rest of the year).

 

 

 

[Toronto16Participant]

 This is misleading, because nobody would take Kyle Korver over DeMar DeRozan.

[dmo21Participant]

 This is just based off this year so far, not looking into the future or the past. The key difference between these two is efficiency and shooting. Korver can shoot threes and is very efficient while Demar is the opposite. They have similar rebounds, assists, blocks, and steals. Demar does put up around ten more points, but how many more shots does he have to take to get them? 

If there was 1 game to be played, I don’t think Demar would be the unamious decision.

[mikeyvthedonParticipant]

Sure, Kyle Korver is absolutely better and more efficient shooter. His TS% is MUCH better than DeMar’s, and he takes far fewer shots. However, is there not something that comes into play with this that might work in DeMar’s favor? Such as, who is better at actually getting these shots off, or getting to the basket? As far as Offensive Rating is concerned, I think that this category can indeed be distorted based on the team one plays on or the role they are used for. 

Kyle’s role is to take the open shot whenever he gets a chance. He is one of the best shooters in the league and this year is playing 35 minutes per game. He is averaging 12.8 ppg per 36 minutes, DeRozan is at 20.2 per 36 (and is playing more than that, however this is just a per minute example). DeMar is not scoring more efficiently than Kyle, though I do believe the reason he is shooting more is based on his ability to create these opportunities, not only because he is being a hog. Of course a team would want Korver to shoot as much as possible, just think he does not create those opportunities at the level DeRozan does.

Also agree that if there is 1 game played DeRozan would not be a unanimous decision, it would definitely depend on the position. What I do believe is, if you were starting a team from scratch and looking for a starting 2 guard, you would more than likely take DeMar. He may shoot only 80% to Kyle’s 90% from the line, but he gets to the line about 7 times more often! Korver being more efficient is one thing, however that does not necessarily mean the Raptors would be better with him in place of DeMar. Think in the end, there is a reason DeMar is getting more shots and while his efficiency shooting leaves something to be desired, I think there is a reason that he does indeed shoot more and score more that may not be factored into your formula.

[mikeyvthedonParticipant]

Sure, Kyle Korver is absolutely better and more efficient shooter. His TS% is MUCH better than DeMar’s, and he takes far fewer shots. However, is there not something that comes into play with this that might work in DeMar’s favor? Such as, who is better at actually getting these shots off, or getting to the basket? As far as Offensive Rating is concerned, I think that this category can indeed be distorted based on the team one plays on or the role they are used for. 

Kyle’s role is to take the open shot whenever he gets a chance. He is one of the best shooters in the league and this year is playing 35 minutes per game. He is averaging 12.8 ppg per 36 minutes, DeRozan is at 20.2 per 36 (and is playing more than that, however this is just a per minute example). DeMar is not scoring more efficiently than Kyle, though I do believe the reason he is shooting more is based on his ability to create these opportunities, not only because he is being a hog. Of course a team would want Korver to shoot as much as possible, just think he does not create those opportunities at the level DeRozan does.

Also agree that if there is 1 game played DeRozan would not be a unanimous decision, it would definitely depend on the position. What I do believe is, if you were starting a team from scratch and looking for a starting 2 guard, you would more than likely take DeMar. He may shoot only 80% to Kyle’s 90% from the line, but he gets to the line about 7 times more often! Korver being more efficient is one thing, however that does not necessarily mean the Raptors would be better with him in place of DeMar. Think in the end, there is a reason DeMar is getting more shots and while his efficiency shooting leaves something to be desired, I think there is a reason that he does indeed shoot more and score more that may not be factored into your formula.

[dmo21Participant]

 First off I don’t agree with all of my ratings (hence why it is unbiased), but I do try to stick up and explain for the results. I personally would choose DeRozan (I am a die hard Raptors fan, so that might be biassed haha), especially for the future, but these rankings aren’t about the future. They are about the 30-40 games these guys have played this year so far, and that’s it. Doesn’t matter what you did last year or what you might be able to do in the future.

To get to your points, there are categories that do favor DeRozan, I was just explaining the ones that favored Korver before as that was the argument. DeRozan, being the 1st or 2nd option for offense (Lowry would be the other) and creating his own shots comes out in the USG% (how much a player is used). For DeRozan, 27 of his total 63 points in my rating come from his USG%, while Korver’s usage is only 13.8, giving DeRozan about 13 more points than Korver in that one category.

The offensive and defensive rating stats were hard for me to value (this is the first season I started using advanced stats in my calculation, I used to use per games) as they are very dependent. So I tried to let those effect the final ranking minimally (but it still may be to high, still working to get this more full proof). The ratings aren’t a null stat and do mean something though. Having Korver on your team definitely makes your offense run much smoother, but would probably make your defense worse. If you swapped DeRozan and Korver to the other team, I don’t think their offensive ratings change that drastically because, like I said, having a guy like Korver on your team makes your offense run so much better.

Lastly, I’m a big no-believer on per 36 stats because they are usually unrealistic. A guy who plays 10 minutes a game cannot produce the same rate at 36 minutes. There is no fatigue factor in there or the fact that the other team might change the way they play if that said player scores 10 points in 10 minutes. 

I’ll make a comment on your other post as well. I like to keep my discussions organized. Thanks for your constructive criticism though as it is my goal to make this the best it can be.

  

 

 

[dmo21Participant]

 First off I don’t agree with all of my ratings (hence why it is unbiased), but I do try to stick up and explain for the results. I personally would choose DeRozan (I am a die hard Raptors fan, so that might be biassed haha), especially for the future, but these rankings aren’t about the future. They are about the 30-40 games these guys have played this year so far, and that’s it. Doesn’t matter what you did last year or what you might be able to do in the future.

To get to your points, there are categories that do favor DeRozan, I was just explaining the ones that favored Korver before as that was the argument. DeRozan, being the 1st or 2nd option for offense (Lowry would be the other) and creating his own shots comes out in the USG% (how much a player is used). For DeRozan, 27 of his total 63 points in my rating come from his USG%, while Korver’s usage is only 13.8, giving DeRozan about 13 more points than Korver in that one category.

The offensive and defensive rating stats were hard for me to value (this is the first season I started using advanced stats in my calculation, I used to use per games) as they are very dependent. So I tried to let those effect the final ranking minimally (but it still may be to high, still working to get this more full proof). The ratings aren’t a null stat and do mean something though. Having Korver on your team definitely makes your offense run much smoother, but would probably make your defense worse. If you swapped DeRozan and Korver to the other team, I don’t think their offensive ratings change that drastically because, like I said, having a guy like Korver on your team makes your offense run so much better.

Lastly, I’m a big no-believer on per 36 stats because they are usually unrealistic. A guy who plays 10 minutes a game cannot produce the same rate at 36 minutes. There is no fatigue factor in there or the fact that the other team might change the way they play if that said player scores 10 points in 10 minutes. 

I’ll make a comment on your other post as well. I like to keep my discussions organized. Thanks for your constructive criticism though as it is my goal to make this the best it can be.

  

 

 

[dmo21Participant]

 This is just based off this year so far, not looking into the future or the past. The key difference between these two is efficiency and shooting. Korver can shoot threes and is very efficient while Demar is the opposite. They have similar rebounds, assists, blocks, and steals. Demar does put up around ten more points, but how many more shots does he have to take to get them? 

If there was 1 game to be played, I don’t think Demar would be the unamious decision.

[Toronto16Participant]

 This is misleading, because nobody would take Kyle Korver over DeMar DeRozan.

[Lebron’s HairlineParticipant]

Looks alot like an Espn insider list, they love odd rankings

[Lebron’s HairlineParticipant]

Looks alot like an Espn insider list, they love odd rankings

[ChoppyParticipant]

 Was wondering why players values differ based on the positions you arbitrarily put them in. For example, if you consider LeBron a SF, his value is 124.7, yet if you consider him a PF, his value is now 122.5. Same goes for Anthony Davis. Just curious as to how your formula works.

[dmo21Participant]

 The formula is weighted differently for different positions dependent on how important the stat is for that position. For example, blocks and rebounds are weighted more for centres and power forwards than they are for point guards. The assists still count toward the final result, they just don’t mean as much. This comes from what you’d want your typical position to do (rather have a rebounding center than an assisting one). 

This also makes the numbers for the different positions incomparable.

 

[dmo21Participant]

 The formula is weighted differently for different positions dependent on how important the stat is for that position. For example, blocks and rebounds are weighted more for centres and power forwards than they are for point guards. The assists still count toward the final result, they just don’t mean as much. This comes from what you’d want your typical position to do (rather have a rebounding center than an assisting one). 

This also makes the numbers for the different positions incomparable.

 

[ChoppyParticipant]

 Was wondering why players values differ based on the positions you arbitrarily put them in. For example, if you consider LeBron a SF, his value is 124.7, yet if you consider him a PF, his value is now 122.5. Same goes for Anthony Davis. Just curious as to how your formula works.

[Scrooge McDuckParticipant]

Lance Stephenson at #9 is an anomaly.

[Scrooge McDuckParticipant]

Lance Stephenson at #9 is an anomaly.

[rileymcshea3Participant]

 Shows how dominant the PG position is in today’s game, especially compared to SGs. The top 10 PGs all have a higher rating then anyone on your SGs list.

[rileymcshea3Participant]

 Shows how dominant the PG position is in today’s game, especially compared to SGs. The top 10 PGs all have a higher rating then anyone on your SGs list.

[mikeyvthedonParticipant]

As he said, every positions rankings are weighted differently. It looks like the weighting for PG’s naturally gives them higher scores in comparison to how SG’s are evaluated. If it is being weighted in terms of position, sort of hard to use it as an evaluation on player ranking in general. So, while PG definitely seems to be a deep position, looks like PG’s are prone to getting high scores in his position ranking grid. At least appears to be that way.

[dmo21Participant]

 You’re definitely correct on that. Each position is ranked separetly based on skill set and what you usually want out of a player at that position.

 It is hard to rank players at different positions with using stats, the eye test, or anything. It’s hard to gauge if Dwight Howard’s blocks and rebounds mean more than Chris Paul’s scoring and assists. It’s much easier to rank at each position because you expect different things from each position. The best rebounding PG isn’t always your best PG, but your best rebounding C or PF will definitely be one of the top players at their position. That’s why I do it per position.

[dmo21Participant]

 You’re definitely correct on that. Each position is ranked separetly based on skill set and what you usually want out of a player at that position.

 It is hard to rank players at different positions with using stats, the eye test, or anything. It’s hard to gauge if Dwight Howard’s blocks and rebounds mean more than Chris Paul’s scoring and assists. It’s much easier to rank at each position because you expect different things from each position. The best rebounding PG isn’t always your best PG, but your best rebounding C or PF will definitely be one of the top players at their position. That’s why I do it per position.

[mikeyvthedonParticipant]

As he said, every positions rankings are weighted differently. It looks like the weighting for PG’s naturally gives them higher scores in comparison to how SG’s are evaluated. If it is being weighted in terms of position, sort of hard to use it as an evaluation on player ranking in general. So, while PG definitely seems to be a deep position, looks like PG’s are prone to getting high scores in his position ranking grid. At least appears to be that way.

[mikeyvthedonParticipant]

Just would really like to see the formula. I commend you for coming up with this system, just do not know if it truly is unbiased as I do not know what goes into it. Is it based on totals or averages? Does it take pace adjustment into account? Do blocks count for PG’s and 3-PT% for centers? Are the categories weighted on importance? Do actual minutes played also take part, or is it meant to be a per minute thing like PER?

Would like to know what goes into it, because while the ratings might be trying to take bias out of the picture, certain categories could definitely tip the scales in certain players favor. You claimed you would post it, so am interested to see what you factor in for the rankings of each position. Just because it is statistically based does not mean it is completely without bias, even PER definitely seems to favor offensively minded players, while defense is almost entirely based on steals and blocks, categories not every great defender rack up.

I like the concept, just want to know what is behind it, because I think even you would agree, if you made a list of top players at a given position, it would look different than your rankings. I looked up the formula for PER a while ago, it is confusing as all get out. However, it is only meant to be a gauge of per minute productivity when a player is on the court. It is not meant to be a factor that all players should be ranked by. I am absolutely guessing you do not mean for these rankings to be concrete lists of the best players in the league at their position, just wonder what they are supposed to represent? Players having the best season? Or is it a list you go by to eliminate external bias?

[mikeyvthedonParticipant]

Just would really like to see the formula. I commend you for coming up with this system, just do not know if it truly is unbiased as I do not know what goes into it. Is it based on totals or averages? Does it take pace adjustment into account? Do blocks count for PG’s and 3-PT% for centers? Are the categories weighted on importance? Do actual minutes played also take part, or is it meant to be a per minute thing like PER?

Would like to know what goes into it, because while the ratings might be trying to take bias out of the picture, certain categories could definitely tip the scales in certain players favor. You claimed you would post it, so am interested to see what you factor in for the rankings of each position. Just because it is statistically based does not mean it is completely without bias, even PER definitely seems to favor offensively minded players, while defense is almost entirely based on steals and blocks, categories not every great defender rack up.

I like the concept, just want to know what is behind it, because I think even you would agree, if you made a list of top players at a given position, it would look different than your rankings. I looked up the formula for PER a while ago, it is confusing as all get out. However, it is only meant to be a gauge of per minute productivity when a player is on the court. It is not meant to be a factor that all players should be ranked by. I am absolutely guessing you do not mean for these rankings to be concrete lists of the best players in the league at their position, just wonder what they are supposed to represent? Players having the best season? Or is it a list you go by to eliminate external bias?

[dmo21Participant]

Finally somebody asks for the formulas haha. I used to use per game and hard stats, like PPG and MP, but have recentley changed to advanced statistics as they do a better job at illustrating what a player does on the court. I’m not going to explain the stats here and just use their abbreviated names, but if you don’t know what they mean or want to know where they come from just go to http://www.basketball-reference.com. Here they are (hope you like math and algebra haha):

 PG:

 (((TRB%*0.5)+(AST%*2)+(STL%*2)+(BLK%*0.5)+(TOV%*(-2))+(USG%)+(ORtg*0.25)+(DRtg*0.25))+(WS*10))*TS%

 SG:

 ((TRB%*0.5)+(AST%*1)+(STL%*2)+(BLK%*0.5)+(TOV%*(-1.5))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

 SF:

 ((TRB%*1)+(AST%*1)+(STL%*1)+(BLK%*1)+(TOV%*(-1.5))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

 PF:

 ((TRB%*1.5)+(AST%*0.5)+(STL%*0.5)+(BLK%*1.5)+(TOV%*(-1))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

 C:

 ((TRB%*2)+(AST%*0.5)+(STL%*0.5)+(BLK%*2)+(TOV%*(-1))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

  Hopefully this didn’t come out as a mess because I was having trouble with the format.

Enjoy.

 

[mikeyvthedonParticipant]

Could definitely use this formula, can do simple math and algebra. Though have you seen the formula for PER? That is confusing as hell, lol. Understand what you are trying to do, just think the guy in the post below this had some good points. I think offensive and defensive rating can kind of skew things a bit, depending on the pace a team plays at.

Also agree about USG% and trying to get actually player scoring involved. Leaving ones scoring ability completely to USG% and ORating probably does not really serve the purpose of determining a players actual scoring production or ability. I know by using the %’s rather than the actual statistical production, you are trying to even the playing field, just think this kind of rewards situational players in the process. Which might lead to the reason that some of them appear so highly in the rankings, as opposed to players with larger roles/more responsibility.

Even as John Hollinger says about PER, it is more used as a guide than as a ranking system. As the only defensive components applied in it are blocks and steals, some defensive minded players rank quite low. While you bring in defensive rating, those can be significantly altered by the team one plays for. I think that individual ORtg and DRtg are not necessarily the best ways to determine value in each category as they tend to favor role players in general, especially ORtg in many cases.

If you do not want to use Per 36 or Per 40, why not just points per minute? You can just go through peoples Per 48 minute averages, divide the average by 48 and you have a viable per minute total. It still keeps everyone on the same platform, though it does not use a finite amount of time played. Just an observation, I appreciate that you made this formula, just not sure if it cannot be improved a tad. Their will be detractors from any formula used, just feel that while this may be unbiased, it still could use a bit of a tweak to better rank where these guys actually are in the league.

[mikeyvthedonParticipant]

Could definitely use this formula, can do simple math and algebra. Though have you seen the formula for PER? That is confusing as hell, lol. Understand what you are trying to do, just think the guy in the post below this had some good points. I think offensive and defensive rating can kind of skew things a bit, depending on the pace a team plays at.

Also agree about USG% and trying to get actually player scoring involved. Leaving ones scoring ability completely to USG% and ORating probably does not really serve the purpose of determining a players actual scoring production or ability. I know by using the %’s rather than the actual statistical production, you are trying to even the playing field, just think this kind of rewards situational players in the process. Which might lead to the reason that some of them appear so highly in the rankings, as opposed to players with larger roles/more responsibility.

Even as John Hollinger says about PER, it is more used as a guide than as a ranking system. As the only defensive components applied in it are blocks and steals, some defensive minded players rank quite low. While you bring in defensive rating, those can be significantly altered by the team one plays for. I think that individual ORtg and DRtg are not necessarily the best ways to determine value in each category as they tend to favor role players in general, especially ORtg in many cases.

If you do not want to use Per 36 or Per 40, why not just points per minute? You can just go through peoples Per 48 minute averages, divide the average by 48 and you have a viable per minute total. It still keeps everyone on the same platform, though it does not use a finite amount of time played. Just an observation, I appreciate that you made this formula, just not sure if it cannot be improved a tad. Their will be detractors from any formula used, just feel that while this may be unbiased, it still could use a bit of a tweak to better rank where these guys actually are in the league.

[dmo21Participant]

Finally somebody asks for the formulas haha. I used to use per game and hard stats, like PPG and MP, but have recentley changed to advanced statistics as they do a better job at illustrating what a player does on the court. I’m not going to explain the stats here and just use their abbreviated names, but if you don’t know what they mean or want to know where they come from just go to http://www.basketball-reference.com. Here they are (hope you like math and algebra haha):

 PG:

 (((TRB%*0.5)+(AST%*2)+(STL%*2)+(BLK%*0.5)+(TOV%*(-2))+(USG%)+(ORtg*0.25)+(DRtg*0.25))+(WS*10))*TS%

 SG:

 ((TRB%*0.5)+(AST%*1)+(STL%*2)+(BLK%*0.5)+(TOV%*(-1.5))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

 SF:

 ((TRB%*1)+(AST%*1)+(STL%*1)+(BLK%*1)+(TOV%*(-1.5))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

 PF:

 ((TRB%*1.5)+(AST%*0.5)+(STL%*0.5)+(BLK%*1.5)+(TOV%*(-1))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

 C:

 ((TRB%*2)+(AST%*0.5)+(STL%*0.5)+(BLK%*2)+(TOV%*(-1))+(USG%)+(ORtg*0.25)+(DRtg*0.25)+(WS*10))*TS%

  Hopefully this didn’t come out as a mess because I was having trouble with the format.

Enjoy.

 

[Jester87Participant]

What about using the net rating between team offensive rating when the player is on and when is off and the same for defense instead of pure Offense rating and defense rating (both of them being pretty meaningless for individual evaluation)? That gives a more realistic value of how much that player gives to his team with his offense/defense. I mean, a guy like Boozer has 99.6 DRtg because the Bulls are an elite defensive team and the can somehow hide him, but the Bulls have a 95.0 DRtg when he’s off, so he’s a minus 4.6 on defense. While a guy like Asik last year had a 101.3 DRtg because there were too many bad defenders playing for the Rockets, but his team defensive rating was 107.0 when he was out (so he was such a huge value for his team on D). 

I also wouldn’t put USG in the equation, it’s a tricky stat because a low USG rate means a role player who doesn’t do anything creative on offense, but a the same time a high value isn’t good either, especially with bad TO% or TS% numbers it means you’re doing more than you can and you’re basically an irresponsible player. Including USG in the formula rewards would give a higher overall value to players chucking shots and using most of their team’s possessions.

I would also try to include scoring ability somehow, offensive rating is a good measure of one player’s ability to help his team on offense, but that could be a result of his passing ability, of his outside shooting opening the lanes, or of his ability to move off the ball. I think scoring ability should be in the equation, otherwise skills like pass and shooting efficiency would be overvalued (because they all affect offensive rating but there are also single stats for them such as Ast% and TS%). Obviousy points per game is not a good indicatore, because guys with higher playing times and/or usage would have an unjustified advantage, so would guys playing for fast pace teams. It’s pretty much the same for points per 36/40 minutes, but points per 36/40 minutes pace adjusted would be a good compromise. Oh, and of course WS/48 are probably better than WS.

By the way, mine are just suggestions, I really appreciate you job and your effort anyways.

 

[Jester87Participant]

What about using the net rating between team offensive rating when the player is on and when is off and the same for defense instead of pure Offense rating and defense rating (both of them being pretty meaningless for individual evaluation)? That gives a more realistic value of how much that player gives to his team with his offense/defense. I mean, a guy like Boozer has 99.6 DRtg because the Bulls are an elite defensive team and the can somehow hide him, but the Bulls have a 95.0 DRtg when he’s off, so he’s a minus 4.6 on defense. While a guy like Asik last year had a 101.3 DRtg because there were too many bad defenders playing for the Rockets, but his team defensive rating was 107.0 when he was out (so he was such a huge value for his team on D). 

I also wouldn’t put USG in the equation, it’s a tricky stat because a low USG rate means a role player who doesn’t do anything creative on offense, but a the same time a high value isn’t good either, especially with bad TO% or TS% numbers it means you’re doing more than you can and you’re basically an irresponsible player. Including USG in the formula rewards would give a higher overall value to players chucking shots and using most of their team’s possessions.

I would also try to include scoring ability somehow, offensive rating is a good measure of one player’s ability to help his team on offense, but that could be a result of his passing ability, of his outside shooting opening the lanes, or of his ability to move off the ball. I think scoring ability should be in the equation, otherwise skills like pass and shooting efficiency would be overvalued (because they all affect offensive rating but there are also single stats for them such as Ast% and TS%). Obviousy points per game is not a good indicatore, because guys with higher playing times and/or usage would have an unjustified advantage, so would guys playing for fast pace teams. It’s pretty much the same for points per 36/40 minutes, but points per 36/40 minutes pace adjusted would be a good compromise. Oh, and of course WS/48 are probably better than WS.

By the way, mine are just suggestions, I really appreciate you job and your effort anyways.

 

[dmo21Participant]

 A huge thanks for your ideas. Like I said this was my first time using advanced stats and that it is a work in progess, and I think your ideas will definitely help out. 

I’ll use that offensive/defensive rating idea you suggested as that way it says more about the player rather than the team. There’s always that player that gets over rated because of his team (like your example, Boozer, and also like Korver who I have discussed before). 

I really like using the USG% stat as it shows how important a that player is to their team. For me, I would rate a guy higher if he put up 20-5 being the 1st option on his team (someone like Derzon) than a guy who puts up the same stats being a 3rd option on their team (someone like Bosh) as I figure it’s harder to be a better player when the other team keys on you (DeRozan) rather than you playing off the better players on your team (Bosh). USG% may boost chuckers for that stat alone, but having TO% and TS% in there will offset that as chuckers will usually have higher turnovers and lower shooting percentages. 

I was trying to find a good scoring stat but I couldn’t find any. Assists, rebouns, steal and all the other stats have advanced stats for them but points doesn’t. Plus I didn’t look around that hard.

As I’ve stated before, I don’t like using per36 or per40 stats as there is too much variability in it.

[Jester87Participant]

Yep, I know, per36 or per 40 minutes stats have lot of noise but when they’re pace adjusted they’re a little better. Also you can exclude players who haven’t played at least (say) 15-16 mpg and at least say 2/3 of the games, so that you don’t have garbage time heroes messing up the stats. 15-16 minutes usually means solid rotation player, so in my opinion while it’s still not perfect, you can have at least a decent indicator. Another hypotesis would be points per possessions, that is pretty much the scoring equivalent of AST% and REB%, but that’s far from perfect because it gives such a great advantage to low usage players and 3 point shooters. If one could find a way to weigh PPP with usage that would be a good compromise.

Again appreciation for what you’re doing and for your willingness to accept suggestions.

[Jester87Participant]

Yep, I know, per36 or per 40 minutes stats have lot of noise but when they’re pace adjusted they’re a little better. Also you can exclude players who haven’t played at least (say) 15-16 mpg and at least say 2/3 of the games, so that you don’t have garbage time heroes messing up the stats. 15-16 minutes usually means solid rotation player, so in my opinion while it’s still not perfect, you can have at least a decent indicator. Another hypotesis would be points per possessions, that is pretty much the scoring equivalent of AST% and REB%, but that’s far from perfect because it gives such a great advantage to low usage players and 3 point shooters. If one could find a way to weigh PPP with usage that would be a good compromise.

Again appreciation for what you’re doing and for your willingness to accept suggestions.

[dmo21Participant]

 A huge thanks for your ideas. Like I said this was my first time using advanced stats and that it is a work in progess, and I think your ideas will definitely help out. 

I’ll use that offensive/defensive rating idea you suggested as that way it says more about the player rather than the team. There’s always that player that gets over rated because of his team (like your example, Boozer, and also like Korver who I have discussed before). 

I really like using the USG% stat as it shows how important a that player is to their team. For me, I would rate a guy higher if he put up 20-5 being the 1st option on his team (someone like Derzon) than a guy who puts up the same stats being a 3rd option on their team (someone like Bosh) as I figure it’s harder to be a better player when the other team keys on you (DeRozan) rather than you playing off the better players on your team (Bosh). USG% may boost chuckers for that stat alone, but having TO% and TS% in there will offset that as chuckers will usually have higher turnovers and lower shooting percentages. 

I was trying to find a good scoring stat but I couldn’t find any. Assists, rebouns, steal and all the other stats have advanced stats for them but points doesn’t. Plus I didn’t look around that hard.

As I’ve stated before, I don’t like using per36 or per40 stats as there is too much variability in it.

[Afactor4Participant]

 Just out of curiousity, what would Duncan be ranked as a Center?

[Afactor4Participant]

 Just out of curiousity, what would Duncan be ranked as a Center?

[Author]

Posts