K. Pelton: Lakers non believers in advanced stats

[ALL HAIL]

It is possible to construct a thriving team that meshes well while also drafting well without the use of analytics as a guiding force.

[EArl]

Statistics are a very important tool in every day life, but its biggest problem is trying to implement empirical data onto humans. The biggest flaw in that, is that humans are very unpredictable. Think of this similarly to the stock market and how economics work. That will only take you so far if you don't include the human variable which is impossible to put a number on. you can make predictions in how humans will act and be right most of the time, but they can also react completely different. Its a big reason as to why economics, psychology, sociology etc are social sciences and not hard sciences. In real life situations the so called eye test which is another word for having experience will help over advances statistics.

Think about the Lakers. People say that Kobe isn't a great player by providing statistics, but look at the success that he has given us. He is a walking contradiction.

TLDR version: Advanced Stats are effective, but you can't rely on them completely.

[Jedi32]

@ earl I wish I had that standing ovation gif for you. .. that was very well said.

[TyCobb]

My beef with advanced stats is how it's ruining the overall viewing pleasure of the sport.

Baseball right now is currently suffering from this. The three most commons outcomes for hitters these days are walks, strikeouts, and homeruns. Why? Because advanced stats told front office guys that these are the most productive players.

Basketball will slowly turn into a 3 point and free throw contest.

[Kilroy]

Advanced stats are used in some form or other in all walks of life. There's no problem with the data in and of itself...

What's a problem is when some fools watch Moneyball and miss the point entirely, thinking that in order to be successful they HAVE to put Advanced Stats out in front of the Org in some way... Like, when in doubt, advanced stats out... But check with Advanced Stats first before you even think about doing anything.

They are a decision making tool... Not a decision maker.

Should the Lakers use advanced stats? Of course, and both Jim and Mitch said they do. Is it imperative they market themselves as an 'advanced stats org?' Of course not.

[dipstick]

Using advanced stats only go so far. And it probably works better in some sports more than others. While it may work to an extent for basketball, I think everyone only uses it as a guide. You still need to put up your foundations which are your stars. From there, statistics may be useful in filling out the roster in deciding which player and which skills any team is lacking in. In my opinion, it won't work that well in looking for guys with the x-factor that your team will be heavily relying upon.

[aaron_gray]

My beef with advanced stats is how it's ruining the overall viewing pleasure of the sport.

Baseball right now is currently suffering from this. The three most commons outcomes for hitters these days are walks, strikeouts, and homeruns. Why? Because advanced stats told front office guys that these are the most productive players.

Basketball will slowly turn into a 3 point and free throw contest.

That's not how advanced stats work. If that was true, than Harden wouldn't be allowed to iso and he would instead have to play more team ball. Instead, he still gets to iso like every other play because his isos are more efficient than most post ups (which frankly makes me wonder why teams don't just double him in the half court when he gets the ball and live with a trev/beverly 3).

Advanced stats are just stats. They're not even the most interesting part of analytics. The most interesting ones have to do with probability.

http://grantland.com/features/why-teams ... hip-go-it/

This isn't quite the article I'm looking for, but it has to do with the 5% theory and it talks about how contenders are built. A top 10 player, a good value free agent signing etc. In another article, one of the things was an unexpected breakout of a young player. If I'm not mistaken, each of these things contribute to the creation of a contender, and what you try to do is create the best odds of something like that happening. That might mean loading up on young players like a tryout, signing the cheapest free agents, not going after middle class veterans to chase superstars, etc. At least that's the part I find most interesting about analytics. The stuff of nba.com, basketball-reference and 82games, those are just stats. Loading up on favorable odds, now that requires some analysis and some serious balls.

[EArl]

My beef with advanced stats is how it's ruining the overall viewing pleasure of the sport.

Baseball right now is currently suffering from this. The three most commons outcomes for hitters these days are walks, strikeouts, and homeruns. Why? Because advanced stats told front office guys that these are the most productive players.

Basketball will slowly turn into a 3 point and free throw contest.

That's not how advanced stats work. If that was true, than Harden wouldn't be allowed to iso and he would instead have to play more team ball. Instead, he still gets to iso like every other play because his isos are more efficient than most post ups (which frankly makes me wonder why teams don't just double him in the half court when he gets the ball and live with a trev/beverly 3).

Advanced stats are just stats. They're not even the most interesting part of analytics. The most interesting ones have to do with probability.

http://grantland.com/features/why-teams ... hip-go-it/

This isn't quite the article I'm looking for, but it has to do with the 5% theory and it talks about how contenders are built. A top 10 player, a good value free agent signing etc. In another article, one of the things was an unexpected breakout of a young player. If I'm not mistaken, each of these things contribute to the creation of a contender, and what you try to do is create the best odds of something like that happening. That might mean loading up on young players like a tryout, signing the cheapest free agents, not going after middle class veterans to chase superstars, etc. At least that's the part I find most interesting about analytics. The stuff of nba.com, basketball-reference and 82games, those are just stats. Loading up on favorable odds, now that requires some analysis and some serious balls.

From watching basketball throughout the years there is one thing that I have noticed in winning teams: dumb luck and that you can't account for. That Robert Horry shot against the Kings. When Phil played the bench for a while against Portland (could have easily gone the other way). That lob missed by Courtney Lee against Orlando and those 3s that Fisher nailed against Nelson. Those clutch shots that Metta hit against Celtics could have gone the other way. You really do need a little bit of luck to win a ring.

[aaron_gray]

My beef with advanced stats is how it's ruining the overall viewing pleasure of the sport.

Baseball right now is currently suffering from this. The three most commons outcomes for hitters these days are walks, strikeouts, and homeruns. Why? Because advanced stats told front office guys that these are the most productive players.

Basketball will slowly turn into a 3 point and free throw contest.

That's not how advanced stats work. If that was true, than Harden wouldn't be allowed to iso and he would instead have to play more team ball. Instead, he still gets to iso like every other play because his isos are more efficient than most post ups (which frankly makes me wonder why teams don't just double him in the half court when he gets the ball and live with a trev/beverly 3).

Advanced stats are just stats. They're not even the most interesting part of analytics. The most interesting ones have to do with probability.

http://grantland.com/features/why-teams ... hip-go-it/

This isn't quite the article I'm looking for, but it has to do with the 5% theory and it talks about how contenders are built. A top 10 player, a good value free agent signing etc. In another article, one of the things was an unexpected breakout of a young player. If I'm not mistaken, each of these things contribute to the creation of a contender, and what you try to do is create the best odds of something like that happening. That might mean loading up on young players like a tryout, signing the cheapest free agents, not going after middle class veterans to chase superstars, etc. At least that's the part I find most interesting about analytics. The stuff of nba.com, basketball-reference and 82games, those are just stats. Loading up on favorable odds, now that requires some analysis and some serious balls.

From watching basketball throughout the years there is one thing that I have noticed in winning teams: dumb luck and that you can't account for. That Robert Horry shot against the Kings. When Phil played the bench for a while against Portland (could have easily gone the other way). That lob missed by Courtney Lee against Orlando and those 3s that Fisher nailed against Nelson. Those clutch shots that Metta hit against Celtics could have gone the other way. You really do need a little bit of luck to win a ring.

Yeah of course, and some teams are in position to be lucky but never get the luck they need to go all the way. That's fine. Sometimes, that's the best you can do, but the vast majority will get a shot sooner or later. The problem is, windows close so fast now with shorter contracts. The hard part is creating a sustainable team with a long window of contention.

[TyCobb]

My beef with advanced stats is how it's ruining the overall viewing pleasure of the sport.

Baseball right now is currently suffering from this. The three most commons outcomes for hitters these days are walks, strikeouts, and homeruns. Why? Because advanced stats told front office guys that these are the most productive players.

Basketball will slowly turn into a 3 point and free throw contest.

That's not how advanced stats work. If that was true, than Harden wouldn't be allowed to iso and he would instead have to play more team ball. Instead, he still gets to iso like every other play because his isos are more efficient than most post ups (which frankly makes me wonder why teams don't just double him in the half court when he gets the ball and live with a trev/beverly 3).

Advanced stats are just stats. They're not even the most interesting part of analytics. The most interesting ones have to do with probability.

http://grantland.com/features/why-teams ... hip-go-it/

This isn't quite the article I'm looking for, but it has to do with the 5% theory and it talks about how contenders are built. A top 10 player, a good value free agent signing etc. In another article, one of the things was an unexpected breakout of a young player. If I'm not mistaken, each of these things contribute to the creation of a contender, and what you try to do is create the best odds of something like that happening. That might mean loading up on young players like a tryout, signing the cheapest free agents, not going after middle class veterans to chase superstars, etc. At least that's the part I find most interesting about analytics. The stuff of nba.com, basketball-reference and 82games, those are just stats. Loading up on favorable odds, now that requires some analysis and some serious balls.

How many Harden iso's end in a 3 pointer or trip to the FT line? My guess is thats what makes it such an efficient play.

[aaron_gray]

My beef with advanced stats is how it's ruining the overall viewing pleasure of the sport.

Baseball right now is currently suffering from this. The three most commons outcomes for hitters these days are walks, strikeouts, and homeruns. Why? Because advanced stats told front office guys that these are the most productive players.

Basketball will slowly turn into a 3 point and free throw contest.

That's not how advanced stats work. If that was true, than Harden wouldn't be allowed to iso and he would instead have to play more team ball. Instead, he still gets to iso like every other play because his isos are more efficient than most post ups (which frankly makes me wonder why teams don't just double him in the half court when he gets the ball and live with a trev/beverly 3).

Advanced stats are just stats. They're not even the most interesting part of analytics. The most interesting ones have to do with probability.

http://grantland.com/features/why-teams ... hip-go-it/

This isn't quite the article I'm looking for, but it has to do with the 5% theory and it talks about how contenders are built. A top 10 player, a good value free agent signing etc. In another article, one of the things was an unexpected breakout of a young player. If I'm not mistaken, each of these things contribute to the creation of a contender, and what you try to do is create the best odds of something like that happening. That might mean loading up on young players like a tryout, signing the cheapest free agents, not going after middle class veterans to chase superstars, etc. At least that's the part I find most interesting about analytics. The stuff of nba.com, basketball-reference and 82games, those are just stats. Loading up on favorable odds, now that requires some analysis and some serious balls.

How many Harden iso's end in a 3 pointer or trip to the FT line? My guess is thats what makes it such an efficient play.

He does shoot a lot of 3s and he is arguably the best in the league at getting to the line. But you can't complain about lack of variation due to the development of statistics.

[TyCobb]

He does shoot a lot of 3s and he is arguably the best in the league at getting to the line. But you can't complain about lack of variation due to the development of statistics.

Yeah I can. lol Like I said, I'm witnessing it right now as a fan of baseball. When you have a league made up of specific skill set, because advanced analytics told you they were the most productive types, the (you're gonna like this) probability that it becomes the most common output will be recognizable to even the casual fan.

Because of this, in baseball, they are messing around with pitch clocks and the like because the game is 'too slow' now, nevermind the fact that the players who are held in higher regard watch pitches all day instead of swinging the bat early in the count (the guys who put the ball in play majority of the time).

[aaron_gray]

Yeah I can. lol Like I said, I'm witnessing it right now as a fan of baseball. When you have a league made up of specific skill set, because advanced analytics told you they were the most productive types, the (you're gonna like this) probability that it becomes the most common output will be recognizable to even the casual fan.

Because of this, in baseball, they are messing around with pitch clocks and the like because the game is 'too slow' now, nevermind the fact that the players who are held in higher regard watch pitches all day instead of swinging the bat early in the count (the guys who put the ball in play majority of the time).

I'm not familiar with how baseball works, but in the nba, different players have different skill sets. What these stats do is encourage players to maximize their skill set by playing to their strengths. The mid-range game is still the dominant skill for a lot of good wing players such as Rudy Gay, Gerald Henderson and DeMar DeRealgm. While LMA has messed around with a 3, his mid-range game is still what makes him arguably a top 15 player. As long as colleges don't pump out identical robots in the future, than the different physical attributes and upbringings of players will lead to guys playing differently, and the stats available will encourage them to do so.

[aaron_gray]

I get that teams will start shooting more 3s and try to get to the line more, but I think that the extinction of the mid-range game is a bit outlandish.

Can you also elaborate on why you hate 3s in the first place? Free throws, yeah I find them pretty boring as well (unless it's Dwight shooting them, in which case there's real drama). But have shooters lined up around the arc is terrifying, especially when the shooters belong to the opposing team. IMO during a close game, it becomes a lot more thrilling when there's shooters on the floor rather than when a bunch of brick layers are in.

[TyCobb]

I just watched Kobe for 18 years... the least amount of shot resistance does nothing for me.

[Slava]

SportsVu cameras produce some much data that I doubt even 90% of the teams that have that info found out how to use them yet. You are basically getting the coordinates of 10 players on the floor every second for 48 minutes at a time. That's one of those cases where technology creates more information than solutions.

[Slava]

Thanks to Grantland, here's an excellent research paper discussing defensive metrics using SportsVu data: http://www.sloansportsconference.com/wp ... oints2.pdf

[Kilroy]

He does shoot a lot of 3s and he is arguably the best in the league at getting to the line. But you can't complain about lack of variation due to the development of statistics.

Yeah I can. lol Like I said, I'm witnessing it right now as a fan of baseball. When you have a league made up of specific skill set, because advanced analytics told you they were the most productive types, the (you're gonna like this) probability that it becomes the most common output will be recognizable to even the casual fan.

Because of this, in baseball, they are messing around with pitch clocks and the like because the game is 'too slow' now, nevermind the fact that the players who are held in higher regard watch pitches all day instead of swinging the bat early in the count (the guys who put the ball in play majority of the time).

Then you have a transcendent player come along and blow the whole 'plan' out of the water... What if like you indicated, Shaq arrived in the league today? It would take a decade for a lot of the 3pt or bust teams to recover. He would eat like nobody we've ever seen in today's league. Like 40ppg against what passes as a C today.

The bad part is because he's regarded as an 'inefficient' player, maybe he doesn't get the support he needs to develop into the monster he was, in today's world.

Baseball's still recovering from the 'steroid era'... But eventually, there are going to be some specific talent in baseball that's going to make everyone have to start rethinking what they're doing today.

The same is true for Basketball, only more so... Because the NBA is a much faster league. Things happen, come and go, much faster.

It literally would only take one Shaq, or Kareem, or Magic, to turn the league on it's ear again.

The thing about analytics or more specifically, analysts is, there's always a stat to 'prove' they saw it coming all along. They're never really wrong.

[Michael Lucky]

It was well documented over the last few years that the front office was late to the party but had started looking more into advanced stats. What i think is misleading from the article is the justification for the classification when it uses the coach as a reason for it. Last time i checked the coach wasn't the one drafting and signing players.

[Slava]

It was well documented over the last few years that the front office was late to the party but had started looking more into advanced stats. What i think is misleading from the article is the justification for the classification when it uses the coach as a reason for it. Last time i checked the coach wasn't the one drafting and signing players.

I think dock iterated this in the last page, the use of analytics is not just limited to draft and free agency but teams like the Spurs & Rockets use it to genuinely build their basketball strategy. That's where the coach comes into play. What the Lakers have is a very flawed and frankly amateurish model that's lagging way behind the league wide trajectory.