Player's 13 (P13- a new advanced stat for player comparison)

[PetroNet]

1) How many guys (can) score 25 ppg? FIVE guys in the entire league averaged over 25 in 2012-13. Only NINE guys over 20 ppg. Eight of those nine guys were on Playoff teams. Two of the nine on the champion Heat. I don't overvalue PPG. If anything, BUCKETS -- like Bill Russell said -- is undervalued. But I like the ring of Player's 13.

obviously points are going to be heavily weighted in an metric. but you overweight them to the point where it obscures every other category. obviously scoring 25 ppg is great... but it shouldnt make you autmatically better then someone who scores 22 ppg regardless of how well you shoot it. if one guy score 25 ppg on 42% shooting he shouldnt be ranked higher then a similar guy who scores 23 on 55% shooting. in your metric, this is extremely possible and even likely for comporable players

2) Of course I don't care about Turnovers. Great players with the ball in their hands most lead the league in Turnovers, whereas not-so-good players tend to have lower Turnovers because the ball ISN'T in their hands as much. I realize this is a difficult concept for you to grasp. With you logic, I should punish Steve Nash, who led the league in TO's TWICE?!? What about Jason Kidd??? Magic Johnson? Charles Barkley??? (*crickets*)

No one is saying that Turnovers need to be weighted high. but to completely ignore them is simply ignorant. obviously guys with the ball more are going to turn it over more. That doesnt mean that the degree to which you turn it over needs to be ignored. Harden and Paul both have the ball in their hands a ton. Harden turns it over a ton more. should that not matter?

3) If football is a game of inches, basketball is the game of fractions. Yes, every .09 counts. But I'm not going to punish Kobe Bryant for never shooting over 47%. I'll take Kobe's 30 points on 44% shooting over Paul Millsap's 17 points on 53% shooting. Same goes for Allen Iverson, Ray Allen, Kevin Martin, or any other volume scorer.

47 vs 53 is one thing. but your weight on FG percentage is so low that shooting 55% vs. 43% doesnt even matter because you put such a small weight on it. scoring is counted 25 times. FG% 3 times. thats insanely disproportional. Kobe scoring 30 points on 44% shooting would be worth more then if another player scored 27 points on 54% shooting. thats simply ridiculous

Again, I invite you to run the numbers and compare players. See if it fits your eye test.

PS - Why is everyone so down on Monta Ellis 10-11, which is that stat I posted for him? WE ALL KNOW -- or maybe 'we' don't -- that GSW's offense was high speed and Monta played over 40 mpg. I've written it 10 times, but it obviously bears repeating -- P13 is part of a conversation. If you're talking about Monta Ellis in the GSW years, you'll probably mention the style of play and his minutes.

But, then again, some people... don't have a clue... about that type of stuff.

the only person here without a clue is you. i already gave you an example where a player who was better at everything with an extremely large advantage in shooting percentages rated lower because he scored 0.5 less points per game.

Stephan curry being below monte ellis is terrible conidering curry was better in just about every possible aspect of basketball.

[PetroNet]

Still waiting for your response on my example of a guy being better at everything having a worse rating because of a 0.5 points per game difference

[ChiLA Bullkers]

Still waiting for your response on my example of a guy being better at everything having a worse rating because of a 0.5 points per game difference

Still waiting for you to post two players in REAL LIFE where this is applicable...

[PetroNet]

Still waiting for your response on my example of a guy being better at everything having a worse rating because of a 0.5 points per game difference

Still waiting for you to post two players in REAL LIFE where this is applicable...

How is that not real life? is it not possible for those stat line to be achieved.

if your stat is only applicable to the past, its not a really good stat. your, in fact, is awful and by your own admission was built by taking a ranking of players and finding a stat that would keep said rankings in place. its completely arbitrary and ignores several major aspects of basketball.

you need to really stop with the charade.

but even only going by statline from last year, there is no way stephan curry can be ranked below monte ellis while shooting significantly higher from the floor, from 3, and from the 3 point, assiting more, rebounding more, and scoring more.

[thxfrthmmrs]

By your calculation, Wilt Chamberlain's 1962 season should be the best ever, even if he doesn't have Blocks and Steals to go with that, because linear weight you put on the numbers.

The biggest flaw I see is your calculations are not pace adjusted, and it's not' compared relatively to the rest of the league.

The factors you used are also pretty arbitrary, and you don't consider the scarcity of the stats. Even the best blockers in the league gets about 3 blocks a game, yet a block is only worth a factor of 5, while the best scorers score about 30 points, it gets a factor of 13. By your calculation, a player who scores 15 points is as valuable as a player who blocks 4 shots, when there are plenty of guys who can score 15 points, but very few can block 4 shots, that's when the value of the stats comes in.

[therealbig3]

You've made it clear that it's not the only stat you use, and yet in the "rank the starters for the Nets" thread, you rated Deron Williams 5th out of the 5 starters, and said something along the lines of "Wow, a lot of guys like D-Will. To me, he's the most overrated player in the league. Didn't you guys see my P13 stat, which proves he's the most overrated player in the league?"

Sounds a lot like you're HEAVILY weighing this stat in your player evaluation.

Oh I don't need the stat to tell you Deron Williams is the worst starter on the Nets...

Ok. Go for it.

Tell me how Deron Williams is the worst starter on the Nets. This should be fun.

[Doctor MJ]

So, please let me know your thoughts.

Well let's see.

I applaud you for getting into thinking like this. I wish more would.

Have you thought about why you're adding in the "13" step? By definition, you rounding the number makes the stat less precise. It's a bad move. Think about how you could apply a mathematical tranformation that could achieve what you want without any loss of precison.

Have you thought about how Bill Russell himself would have done by your stat? I'm going to guess that Wilt would have thrashed him. If I thought Wilt was a far superior player to Russell, I don't think I'd be making stats inspired by Russell. As is, I think that most of the time Russell was playing better than Wilt.

In general, if we're really evaluating the stat on the big boy level, I think a stat like this is mostly late for the party. Weighting a bunch of factors has been done for a very long time and we reached a point a while ago where the weights were determined by methods with more basis than guesstimations. Your stat would be useful if we were not yet in the age of "advanced stats", but that's not where we've been for a long time.

And I apologize if people are rude to you here, but this is pretty much how it goes whenever anyone jumps in giving any indication that they think they've determined a great way to do things but really they are behind the curve. The feeling of people is naturally going to be that you need to spend more time understanding all the incredible statistical work that's out there before you come up with your own stats. This isn't entirely true because what you're doing is good practice, but when share what you're doing publicly you've got to show awareness where your work is in the context of the stats that people in that community are using.

Also, we have a Stat Analysis board here which is where this thread should be. I'll move it for you.

[ChiLA Bullkers]

So, please let me know your thoughts.

Well let's see.

I applaud you for getting into thinking like this. I wish more would.

Have you thought about why you're adding in the "13" step? By definition, you rounding the number makes the stat less precise. It's a bad move. Think about how you could apply a mathematical tranformation that could achieve what you want without any loss of precison.

Have you thought about how Bill Russell himself would have done by your stat? I'm going to guess that Wilt would have thrashed him. If I thought Wilt was a far superior player to Russell, I don't think I'd be making stats inspired by Russell. As is, I think that most of the time Russell was playing better than Wilt.

In general, if we're really evaluating the stat on the big boy level, I think a stat like this is mostly late for the party. Weighting a bunch of factors has been done for a very long time and we reached a point a while ago where the weights were determined by methods with more basis than guesstimations. Your stat would be useful if we were not yet in the age of "advanced stats", but that's not where we've been for a long time.

And I apologize if people are rude to you here, but this is pretty much how it goes whenever anyone jumps in giving any indication that they think they've determined a great way to do things but really they are behind the curve. The feeling of people is naturally going to be that you need to spend more time understanding all the incredible statistical work that's out there before you come up with your own stats. This isn't entirely true because what you're doing is good practice, but when share what you're doing publicly you've got to show awareness where your work is in the context of the stats that people in that community are using.

Also, we have a Stat Analysis board here which is where this thread should be. I'll move it for you.

Thanks, Doc MJ.

It's always good to be able to defend one's side. And I realized I'd run into some people who seem just mad, and that anger would be turned towards my post. That's okay. Says more about them then it does me. And I invited it by saying, "let me know your thoughts."

Interesting what you wrote about Russell vs. Wilt. Since they didn't record blocks or steals until 73-74, I couldn't really accurately run Russell. Of course Wilt's stats were off the charts. But knowing that Russell was a Red system guy, knowing his titles, knowing his style of play, even if the stat favored Wilt, of course I'd always go with Russell.

[acrossthecourt]

It's not advanced and we have more than enough player metrics.

It's just a convoluted linear weights system.

You also don't adjust for pace. Of course if you play on a fast team you can pick up more box score stats. But the notable problem is, do you think the box score tells you everything you need to know about a player?

You obviously have a tough time reading and/or comprehending what I explicitly wrote along with introducing the stat, which is, and I'll repeat:

"Obviously this isn't the ONLY stat I use in discussions/arguments for best player ever or best player season-to-season or even player's impact on their team.

But the stat does show LeBron is #1, Durant #2, and everyone who we all agree is Top 10 or so are up there. There aren't any big shockers."

Yes, the stat doesn't account for pace. But if you know the Game, you can add that into the discussion. P13 isn't the end all be all. It's PART -- AGAIN -- of the conversation.

And I challenge you to name that ONE stat that best represents a player's overall impact or greatness...

Please. Let me know.

Because I just came up with one.

No you didn't.

Do you really think the box score tells everything?

[ChiLA Bullkers]

It's not advanced and we have more than enough player metrics.

It's just a convoluted linear weights system.

You also don't adjust for pace. Of course if you play on a fast team you can pick up more box score stats. But the notable problem is, do you think the box score tells you everything you need to know about a player?

You obviously have a tough time reading and/or comprehending what I explicitly wrote along with introducing the stat, which is, and I'll repeat:

"Obviously this isn't the ONLY stat I use in discussions/arguments for best player ever or best player season-to-season or even player's impact on their team.

But the stat does show LeBron is #1, Durant #2, and everyone who we all agree is Top 10 or so are up there. There aren't any big shockers."

Yes, the stat doesn't account for pace. But if you know the Game, you can add that into the discussion. P13 isn't the end all be all. It's PART -- AGAIN -- of the conversation.

And I challenge you to name that ONE stat that best represents a player's overall impact or greatness...

Please. Let me know.

Because I just came up with one.

No you didn't.

Do you really think the box score tells everything?

I repeat: "It's PART -- AGAIN -- of the conversation."

[azuresou1]

It makes me sad that there's so much conversation on this "advanced stat" but I can't get people to review my shot at VORP.

[giordunk]

Most of the conversation is ripping on how this isn't an advanced stat.

Here's my opinion.

I don't mind your 'P13', but it's ultimate no more advanced a stat than adding put a player's points, rebounds, assists, blocks, and steals then calling it a stat. P13 is just a stat. There is absolutely nothing 'advanced' with it.

I think to break into 'Advanced stat' realm, you really need to consider a lot more variables. For example, having only 3 point percentage as part of your 'P13' while not taking into account how many attempts a player takes. If you want to be considered an 'advanced stat' you need to be able to compensate and realize all the different tiny details. Things like pace adjusted numbers, how to weigh attempts with records, or effective field goal or true shooting percentage.

But when there exists like 10-20 other different statistics that actually consider pace, TS%, +/-, points allowed, and all that ****, your statistic is hardly advanced at all.

[Doctor MJ]

Thanks, Doc MJ.

It's always good to be able to defend one's side. And I realized I'd run into some people who seem just mad, and that anger would be turned towards my post. That's okay. Says more about them then it does me. And I invited it by saying, "let me know your thoughts."

Interesting what you wrote about Russell vs. Wilt. Since they didn't record blocks or steals until 73-74, I couldn't really accurately run Russell. Of course Wilt's stats were off the charts. But knowing that Russell was a Red system guy, knowing his titles, knowing his style of play, even if the stat favored Wilt, of course I'd always go with Russell.

Glad you took my comments well.

Re Russ & Wilt, I'd absolutely expect that blocks & steals wouldn't give Russell any new big advantage. From what I can tell, Wilt at his most block-happy might have out-blocked Russell. Of course, the goal isn't the block, it's causing the disruption of the other team without disrupting your own team. When it came to that nuance, Russell was always lightyears ahead of anyone. I'd venture to say only really old Duncan & Garnett really can see the game like Russell did among bigs, and of course at this point there physical skills are nowhere near their prime.