538 Article: "The Hidden Value of the NBA Steal"

[Doctor MJ]

Just noting that in response to the criticism, the writer (Morris) is responding with a series of FOUR articles, of which he's posted 3...none of which has addressed my criticism. Here's hoping the last one gets to it.

I have to say, I like him responding...but the spread over 4 articles makes me chuckle. Clearly he's used to having many articles to say his point as that's what he did with his awesome Rodman analysis, but damn, if you're going to split things up like that, then you had better respond to every major argument thoroughly. So, if he gets to his 4th and final article and doesn't respond to my point (whether directly to me or to someone else), the guy is just going to cement his reputation in as another one of these guys who simply doesn't know how to enter into a new field and evaluate the lay of the land at all.

And that's the reputation 538 is now getting, and what it's what Berri did before which then tainted Gladwell.

Here's a quote from Krugman's op-ed on NYT:

http://krugman.blogs.nytimes.com/2014/0 ... blogs&_r=0

Unfortunately, Silver seems to have taken the wrong lesson from his election-forecasting success. In that case, he pitted his statistical approach against campaign-narrative pundits, who turned out to know approximately nothing. What he seems to have concluded is that there are no experts anywhere, that a smart data analyst can and should ignore all that.

But not all fields are like that — in fact, even political analysis isn’t like that, if you talk to political scientists instead of political reporters. So, for example, before glancing at some correlation and asserting causation, you really should talk to the researchers.

This is absolutely the issue here.

What's so bizarre to me here is that I wouldn't expect Silver to make the same mistakes Gladwell did. Gladwell's not a guy who can make up his own stats so he ended up relying on a guy with a stat, and the issue is that he didn't know how to choose.

Silver here hires a guy in Morris who isn't making his own stats...and yet seems so cocky in his conclusions. I don't understand how that works. If you can't make a stat better than the people have already made, then obviously the discipline you're working in is pretty mature and any article you write on something as prominent as 538 needs to be from someone either deep into the mature community or working with such guys.

What's also bizarre is that Silver hired a deep community guy in Neil Paine in addition to Morris, but Paine's primarily writing on baseball right now. If that's Paine's passion that's fine, but we're left with a situation where 538's basketball guy probably isn't their best basketball guy. That is frustrating.

I'll quite here, and see what else Morris writes. I don't want to be to reactionary, but damn, this is exactly the type of thing you fear will happen when your hobby gets a mainstream analytic voice presented to the world.

[mysticbb]

Here's a quote from Krugman's op-ed on NYT:

http://krugman.blogs.nytimes.com/2014/0 ... blogs&_r=0

Unfortunately, Silver seems to have taken the wrong lesson from his election-forecasting success. In that case, he pitted his statistical approach against campaign-narrative pundits, who turned out to know approximately nothing. What he seems to have concluded is that there are no experts anywhere, that a smart data analyst can and should ignore all that.

But not all fields are like that — in fact, even political analysis isn’t like that, if you talk to political scientists instead of political reporters. So, for example, before glancing at some correlation and asserting causation, you really should talk to the researchers.

This is absolutely the issue here.

That is in fact true. I can only say something about the basketball stuff and the climate stuff they posted, and I have to say that there isn't much quality involved.

For basketball:

I just ran a multiple regression on player RAPM data by using possession based and normalized boxscore values, and I found that STL% is not significant at the player level on offense, and found a +0.16 as the coefficient on defense. Meaning, one standard deviation higher STL% gives about 0.16 standard deviation higher defensive RAPM value. So, the standard deviation for STL% is about 0.7 (it is the weighted standard deviation). What worries me about that article is the fact that the method is only vaguely described, but what I sense from that is that Morris is just using a smaller sample (which may be biased, or likely is biased) as well as having a huge effect of the frequency of the respective events in his results. To test the latter hypothesis, I looked at the standard deviation on the team level per game for each entry and found for this season (as well as the found coefficients by Morris in brackets):
Rebounds: 2.1 (1.7)
Assists: 1.6 (2.2)
Steals: 0.8 (9.1)
Blk: 0.7 (6.1)
Turnover: 1 (5.4)

The comparison here should give an answer. If I find no correlation, I have to conclude that my hypothesis is wrong. But if I find a correlation, I have further evidence that the frequency of the events is actually giving here a misleading result. And well, every person should see that the higher the standard deviation is the lower the coefficient becomes. R is -0.88. Or in other words: About 77% of his findings can be explained by the frequency of the events!
So, there might be some inherent value of steals in comparison to other entries (I found that to be significant for the defense, not significant as said before on offense), but his results as well as the standard deviation of each event on a per game basis actually implies: That a steal seems to be more valuable than a rebound, because a steals happens less often than a rebound.

Btw, your assessment of the Rodman stuff he wrote seems also more effected by the pure effort part by Morris rather than the quality of the research itself. When you read the article series you will find a pretty common theme: Morris is actually trying to prove his preconception. He is not seriously questioning his findings at all, he is not making an effort to check for mere coincendences (that Rodman missed more road than home games has an effect, that Rodman missed games when a backup also wasn't avaialable, that Rodman missed games when the opponents were in average slightly weaker), part of the effect he found is simply based on a biased sample he used. And I also suspect that from the remaining 23% for the steals effect article a big part can be explained by a biased sample. And that's where Morris lacks the necessary skill or honesty in order to provide useful research (and where similarities to Berri are seen from my POV).


Regarding climate science:

Pielke Jr lacks the necessary skills in the underlying physics. In order to fully understand climate science it is necessary to understand things like non-linear dynamics (especially fluid dynamics), you need to understand radiation physics; so, it really helps to have a greater understanding of theoretical meteorology in order to understand the influences of additional energy (temperature is just the measurement of the average speed of particles, so just an representation of kinetic energy) in the atmosphere. And then you have the ocean, cryosphere, land mass and biomass as the other components of the climate system. You can't just simply ignore those things based on weak correlation analyses, especially when you are basing that of on biased samples (like Pielke's studies regaring landfall hurricanes, yeah, they don't happen very often, therefore finding a statistical relationship needs a HUGE amount of data which isn't there yet).
I supsect that the most people simply don't appreciate the difference in knowlegde regarding the underlying physics. There is a Dunning-Kruger effect at work, where a lot of people without that knowledge actual believe they can say something about it, because they have experienced different kinds of weather as well as have seen not accurate forecasts based on the NWP models. Way too often I ran into people trying to argue that those kind of things would "prove" that climate models are wrong (e.g.: they can't predict the weather, thus they can't predict the "weather" aka climate 100 years from now or something like that); completely ignoring the underlying fact that weather prediction is actually a initial value problem while climate is actually a bounderary value problem, thus a really important difference which can't be ignored at all.

What's also bizarre is that Silver hired a deep community guy in Neil Paine in addition to Morris, but Paine's primarily writing on baseball right now. If that's Paine's passion that's fine, but we're left with a situation where 538's basketball guy probably isn't their best basketball guy. That is frustrating.

Paine said on APBR that he isn't allowed to write about basketball at that point yet, but that this will change in the future. So, he left out whether that is related to his depature on bbr or based on his possible work as a consultant for a NBA team. Having Paine instead of Morris should greatly increase the quality for that topic.

[Doctor MJ]

Btw, your assessment of the Rodman stuff he wrote seems also more effected by the pure effort part by Morris rather than the quality of the research itself. When you read the article series you will find a pretty common theme: Morris is actually trying to prove his preconception. He is not seriously questioning his findings at all, he is not making an effort to check for mere coincendences

You're right. When I read the Rodman articles I wasn't reading it from a perspective evaluating his analytic bonafides, I was just reading it from the information it gave me, which meant I was basically giving him a thumb's up for the data mining effort he put in.

Wow, all the more embarrassing that 538 was sold on him.

[bondom34]

I feel almost like it's a spinoff site that wants/wanted to be a "smarter" Grantland with politics instead of pop culture, but in the end it just missed the mark. I don't really plan on reading sports content from them as the NBA stuff has been so subpar compared to Lowe or Mahoney from SI. It's just nothing with any real insight.

[bondom34]

FYI, acrossthecourt blogged on it too:
http://ascreamingcomesacrossthecourt.blogspot.com/

[mysticbb]

I don't really plan on reading sports content from them as the NBA stuff has been so subpar compared to Lowe or Mahoney from SI. It's just nothing with any real insight.

Wait until Paine is allowed to write about basketball. I suspect the content gets better then.

FYI, acrossthecourt blogged on it too:
http://ascreamingcomesacrossthecourt.blogspot.com/

So, at the end he shows a similar thing like I did here, and concluding essentially the same. Good!

[Dr Positivity]

I thought the "Four Strikes and You're Out" article about catcher's calls changing/framing, was pretty well done and the quality I expected of 538

Can't say I've liked anything else from them

[Chicago76]

Late to this, but a couple of observations that I find particularly troubling:

Statistical nuances aside, the biggest issue I have with the author of the piece isn't that he does not reside within or at least reach out to the deeper community, it's that the questions related to steal value aren't that deep to begin with. Any outsider with decent knowledge of the game should be able to pick up on these things.

Four positive individual factors and one negative factor that leads to steals: reach (length), lateral quickness, hand quickness, intelligence/anticipation, and the negative of the gamble. This isn't esoteric knowledge. A couple of external effects: playing with good defensive teammates may help another player get steals due to pressure elsewhere (the credit problem), and to some extent, steals are position-based (perimeter guys are more likely to get them at high rates. The position-based element is related to three factors: 1) perimeter guys defend players who more often receive the ball from long passes 2) they defend players who shoulder more of the ball handling responsibility, and 3) they have greater luxury of taking a risk 25 feet from the basket that an interior player 8 feet from the basket because there is more space/time for defensive recovery. This is not rocket science.

I find it unbelievable that the author was taken aback when people talked about the unmeasured downside of a steal attempt. A reasonable observer who understands the importance of defensive rotation and position should pick up on this. Hell, a six year old who is taught to always stay between his man and the basket should understand this.

The mere suggestion that steals come at a cost via gambling should not be a revelation. Non-rigorous statistical analysis should certainly raise the issue. Of the top 20 STL% players this season (min 1000 minutes to date), the player's on-court team opponent eFG% is lower than the players off-court eFG% only 25% of the time. This may be a product of better players having higher STL% and lineups of better players going against better players being more offensive friendly, ie, incremental offensive ability among better players outpaces incremental defensive ability. But it might not. Certainly something that should have been considered going in.

Then there is the idea that players who exhibit traits that lead to steals may simply be better at things not completely picked up by traditional box score stats. Players with better lateral and hand quickness, anticipation, and length will generally be better shot creators (for themselves or others), ball handlers, and general facilitators. Things that might not fully be expressed by shot volume, efficiency, or assists. The author casually mentioned these issues, but generally ignored the intuitive side whenever it got in the way of a good story.

Overall, I think the problem with 538 to date is that they are trying to do too much too soon with their audience.

Being a baseball guy, Silver should abhor counting stats. RBI, hits, stolen bases tell you nothing really without success rates. Triples are really scarce, but is that important? PPG and STLs are no different. The author should know better, or at the very least should have been more actively critiqued by someone who could ask him these types of questions before publication. Silver should know better. The writers should understand the basic concept of opportunity cost. In baseball, the basic form of opportunity cost is runner advancement vs. outs. In basketball and other sports with greater player interaction, this is obviously more difficult to discern. You just can't casually mention this and then move in a piece.

The organization needs to come to grips with the fact that not every piece they write will be a revelation. The bar has been set higher than they believe in various fields. To make every piece more groundbreaking, they're going to need writers that they frankly can't afford. You can't pay someone $50,000 (or whatever) per year, ask them to move to high cost NYC, and get the volume of quality the organization needs to pull that off. The minds they need to attract are either making more money in a day gig or they're making approximately the same money in academia where they get to really understand their areas of expertise without the pressure of producing puff pieces every week or so for publication. The field is too competitive for any one of those authors to become the next Silver, Gladwell, Lowe, or Simmons and carve out a long-term sustainable existence and they know it. They're a select group that are doing it purely for the love of doing it. The problem is, there aren't enough of those people willing to make the economic sacrifice.

Basically, the model doesn't work. They need better editorial review. They need to build up their graphics and production team. They don't need the full in-house writing team though. They need a few in-house writers who are good at explaining relatively simple (but not groundbreaking things) to a wide audience. Mona Chalabi is good at this. She compiles some interesting and somewhat revealing sound bites, puts a nice graphic presentation on them, and publishes at a decent clip. She's not telling me things that I don't generally already know, but she gives me something every day that I don't have time to dig for myself if I want a clearer picture. For the deeper dives, they need to resort to freelancers, each of whom is responsible for producing 3-4 strong pieces a year in their given field. A lot of people aren't going to give up their more lucrative day job for a full time position, but they will produce high quality content for a relative pittance while keeping that day job, provided they take satisfaction from this as a serious hobby and love the idea of their name's association with a nationally branded media venture.

I'm hoping the formula gets better, but it's going to take some time. I suppose we'll know by this time next year.

[Doctor MJ]

I feel almost like it's a spinoff site that wants/wanted to be a "smarter" Grantland with politics instead of pop culture, but in the end it just missed the mark. I don't really plan on reading sports content from them as the NBA stuff has been so subpar compared to Lowe or Mahoney from SI. It's just nothing with any real insight.

A "smarter" Grantland is kinda what I was hoping for too. Grantland to me has been such a phenomenal success, and I didn't see it coming. Simmons is a fun little guy, but I had assumed he'd set the bar for the website that that's not that high. Instead they immediately started hiring guys seemingly with the rule "Find people from Simmons' interests who understand those interests much better than he does". It's been great in general, then there's been the emergence of Zach Lowe. Over at SI Lowe was on the shortlist of guys I'd refer people to for "serious basketball bloggers", but that's it. At Grantland he's become by far the best basketball writer I've ever seen.

I have a lot of respect for Silver, so I guess I figured that if Simmons could do it, so could Silver. It might be harder to find guys who can run circles around Silver because Silver's so sharp, but as long as found the best out there, that's be good enough. Instead, it's just gimmicks.

[Doctor MJ]

I find it unbelievable that the author was taken aback when people talked about the unmeasured downside of a steal attempt. A reasonable observer who understands the importance of defensive rotation and position should pick up on this. Hell, a six year old who is taught to always stay between his man and the basket should understand this.

The mere suggestion that steals come at a cost via gambling should not be a revelation.

Good point. I mean hell, whenever the word "steal" is used in any context I'd argue it'd done specifically to hammer it that it IS a gamble that might get you burned. In baseball for example the value of the stolen base has been much analyzed in terms of the cost of the failed attempts. All one has to do is think through what the word means and one understands the issue even if you've never played before.

And of course if you have played, you know what it's like to gamble and get burned and shouldn't need convincing.

[Chicago76]

I find it unbelievable that the author was taken aback when people talked about the unmeasured downside of a steal attempt. A reasonable observer who understands the importance of defensive rotation and position should pick up on this. Hell, a six year old who is taught to always stay between his man and the basket should understand this.

The mere suggestion that steals come at a cost via gambling should not be a revelation.

Good point. I mean hell, whenever the word "steal" is used in any context I'd argue it'd done specifically to hammer it that it IS a gamble that might get you burned. In baseball for example the value of the stolen base has been much analyzed in terms of the cost of the failed attempts. All one has to do is think through what the word means and one understands the issue even if you've never played before.

And of course if you have played, you know what it's like to gamble and get burned and shouldn't need convincing.

The most frustrating thing to me is that he could have developed a pretty interesting article on Rubio had he actually understood the tradeoff. Rubio's on-off when Love is always on the court:

MIN defensively: steal rate 3 pts better on, overall opp TOV rate jumps from abt 14 to 20%. That's pretty ridiculous. All else equal, that's a 6.5 pt/100 improvement when Rubio is on the court. All else isn't equal though. They allow teams to score more efficiently when they don't turn the ball over (the gambling component). On non-TOV possessions, they Ortg is something like 132 when Rubio is on, 125 when he's off. Net, they're still 2 pts better when he's on. I'm not saying he's a +2 defender (because presumably a lot of his non-Love minutes are vs. subpar Os, and the team O is still bad). It's not a stretch to say he's average to slightly above avg defensively. I don't think he's a +1.5 defender the way many forms of RAPM say he is. A high risk strategy isn't something most teams should imitate, but for MIN's purpose, it gives them enough to make the team's D not as bad as it otherwise would. They're slightly below averagish with Rubio on and Love off.

Offensively: they're much better with Rubio on (7 pts better). Figure possessions off a steal are 0.5 to 0.6 pts more successful than other possessions, and the extra steals only explains maybe 25% of that 7 pt improvement. The rest is that he's simply a great facilitator with excellent vision. Probably better than his ast totals would dictate considering he is a bit turnover prone and can't shoot to save his life.

This is a blunt tool and summary, but throw some fancy graphs in there with some comparable elite point guard comparisons and maybe a couple of Lowe-esque gifs to show the gambling aspect and how it places a ceiling on the D to flesh this out and I think someone could make some interesting points.

[bondom34]

I find it unbelievable that the author was taken aback when people talked about the unmeasured downside of a steal attempt. A reasonable observer who understands the importance of defensive rotation and position should pick up on this. Hell, a six year old who is taught to always stay between his man and the basket should understand this.

The mere suggestion that steals come at a cost via gambling should not be a revelation.

Good point. I mean hell, whenever the word "steal" is used in any context I'd argue it'd done specifically to hammer it that it IS a gamble that might get you burned. In baseball for example the value of the stolen base has been much analyzed in terms of the cost of the failed attempts. All one has to do is think through what the word means and one understands the issue even if you've never played before.

And of course if you have played, you know what it's like to gamble and get burned and shouldn't need convincing.

The most frustrating thing to me is that he could have developed a pretty interesting article on Rubio had he actually understood the tradeoff. Rubio's on-off when Love is always on the court:

MIN defensively: steal rate 3 pts better on, overall opp TOV rate jumps from abt 14 to 20%. That's pretty ridiculous. All else equal, that's a 6.5 pt/100 improvement when Rubio is on the court. All else isn't equal though. They allow teams to score more efficiently when they don't turn the ball over (the gambling component). On non-TOV possessions, they Ortg is something like 132 when Rubio is on, 125 when he's off. Net, they're still 2 pts better when he's on. I'm not saying he's a +2 defender (because presumably a lot of his non-Love minutes are vs. subpar Os, and the team O is still bad). It's not a stretch to say he's average to slightly above avg defensively. I don't think he's a +1.5 defender the way many forms of RAPM say he is. A high risk strategy isn't something most teams should imitate, but for MIN's purpose, it gives them enough to make the team's D not as bad as it otherwise would. They're slightly below averagish with Rubio on and Love off.

Offensively: they're much better with Rubio on (7 pts better). Figure possessions off a steal are 0.5 to 0.6 pts more successful than other possessions, and the extra steals only explains maybe 25% of that 7 pt improvement. The rest is that he's simply a great facilitator with excellent vision. Probably better than his ast totals would dictate considering he is a bit turnover prone and can't shoot to save his life.

This is a blunt tool and summary, but throw some fancy graphs in there with some comparable elite point guard comparisons and maybe a couple of Lowe-esque gifs to show the gambling aspect and how it places a ceiling on the D to flesh this out and I think someone could make some interesting points.

Great analysis, and really shows what could have been done. Another article I just read on the 538 topic, comparing RAPM and steals, but breaking it down by player position. Still shows you can't use steals as the conclusive stat, but they are somewhat helpful for guards in particular (which would basically be what's expected).

http://georgetownsportsanalysis.wordpre ... -position/

[Chicago76]

Great analysis, and really shows what could have been done. Another article I just read on the 538 topic, comparing RAPM and steals, but breaking it down by player position. Still shows you can't use steals as the conclusive stat, but they are somewhat helpful for guards in particular (which would basically be what's expected).

http://georgetownsportsanalysis.wordpre ... -position/

Thanks. That study basically confirms my observations. If you can get a good defensive SF (like a Leonard or George type), your D has a lot of flexibility in terms of switching, opportunistically overplaying, and general strategy. High steal guys that seem to move the needle for teams defensively tend to be SFs, and the coefficients and r-squared tend to bear that out.

Personally, if I were Silver, I'd tell my basketball guys to run more in depth statistical analysis on somewhat divisive players like a Rubio to better identify what they're both good and bad at and express those numbers in terms of pts per 36 (normal starter range minutes at normal pace) to convey this to a broader audience. You can be fairly robust from a statistical standpoint while translating this to the public. Joe Six Pack might be interested that Rubio's defensive style might generate points on the offensive end and defensive stops worth X points, but some of that is given back when this aggression leaves his teammates out to dry in a way that a Chris Paul type doesn't.

Rather than getting so completely technical and bogged down in methodology, it needs to be more grounded. Take what Lowe does and shift the numerical side about half to two thirds of the way between where he is now and where 538 is currently trying to go. Cut his strategy/analysis section by about half to compensate.

[acrossthecourt]

I don't really plan on reading sports content from them as the NBA stuff has been so subpar compared to Lowe or Mahoney from SI. It's just nothing with any real insight.

Wait until Paine is allowed to write about basketball. I suspect the content gets better then.

FYI, acrossthecourt blogged on it too:
http://ascreamingcomesacrossthecourt.blogspot.com/

So, at the end he shows a similar thing like I did here, and concluding essentially the same. Good!

Thanks. I just saw what you did.

By the way, a few months ago he was tweeting about the value of steals and Rubio, and I started arguing with him. I was just saying that while there's a correlation between steals and great players, it's not a consistent one because some guys gamble too much (Cousins) and some guys can get steals without gambling (Chris Paul.) He strongly disagreed. I guess he's been working on that content for a while.


I hope they try to present some of the basics of basketball stats to a wider audience. They can at least do that. Show more of the world the power of point differential for teams, how possessions are defined and why it matters, a summary of one topic (like a history of the +/- movement would be interesting), etc. They don't have to invent a new magic metric.

[colts18]

...

[mysticbb]

Mystic, you are really a smart guy

Yeah, probably, and the interesting thing about being smart is that someone realizes when the very basic error of your comment is actually mentioned in the very part of my post you just quoted.

Also, something being "complex" doesn't mean "completely unpredictable on all levels". Throwing a dice is also complex with a lot of "complex variables" going in (the various molecules of the dice itself, the position of the hand relative to the ground/table, the wind, surface structure of the ground/table, applied acceleration to the dice, etc. pp.) and the result of that throw depends on those initial conditions. But after enough throws I can predict with great certainty that the average value will be around 3.5, and that the distribution will be close to uniform while each number was thrown about 1/6th of the time (assuming the dice is not loaded, or for more than 1 dice I may use the binomial distribution function). So, despite the fact that I can't accurately predict the outcome of one throw, I can say a lot about the average and the distribution by just understanding the underlying physics and math of the issue. A very similar thing happens with weather and climate predictions. The NWP model results are depending on the initial values much more than the climate models, and while the NWP is off after a few days (the best models can actually predict the weather for certain regions for about 4 weeks in advance with pretty high accuracy for temperature and rainfall; at about 0.8 to 0.85), but I can say a lot more about the climate which is to weather in essence like the average and the distribution in that dice throwing experiment.

We can also use basketball games as an example. If we want to predict the exact result (exact points scored for each team) of a specific game, we have to deal with a myriad of complex variables going into that, thus the error will be rather big. If we predict just the scoring margin, the error will already be smaller. And then, when we just predict the average scoring margin for a team over a specific "healthy" timespan, the error will even get smaller. At the end, the bigger the sample becomes, the smaller the error for predictions get. So, and then we can use the average scoring margin to predict wins by applying the pythagorean expectation. Thus, without having the ability to predict the outcome of a single game with great certainty, we can still say a lot about the average performance as well as about the most likely distribution. And then we may use that average performance level of a specific team to get a better prediction for a given game ... Sure, it is not 100% accurate, because we don't know whether a specific player will get injured or will have a bad day or whatever influence you want to imagine, but within the contrains of the layed out scenario of a "healthy timespan" we can say a lot about the outcome of a 7-game series for example. Especially in baskeball were each game gives us about 190 throws of loaded dices ... And that's essentially what the results of climate models present: the most likely result (there is always a range of possible results anyway) in a given scenario. Change the scenario and the result will probably/likely change.

[colts18]

We can also use basketball games as an example. If we want to predict the exact result (exact points scored for each team) of a specific game, we have to deal with a myriad of complex variables going into that, thus the error will be rather big. If we predict just the scoring margin, the error will already be smaller. And then, when we just predict the average scoring margin for a team over a specific "healthy" timespan, the error will even get smaller. At the end, the bigger the sample becomes, the smaller the error for predictions get. So, and then we can use the average scoring margin to predict wins by applying the pythagorean expectation. Thus, without having the ability to predict the outcome of a single game with great certainty, we can still say a lot about the average performance as well as about the most likely distribution. And then we may use that average performance level of a specific team to get a better prediction for a given game ... Sure, it is not 100% accurate, because we don't know whether a specific player will get injured or will have a bad day or whatever influence you want to imagine, but within the contrains of the layed out scenario of a "healthy timespan" we can say a lot about the outcome of a 7-game series for example. Especially in baskeball were each game gives us about 190 throws of loaded dices ... And that's essentially what the results of climate models present: the most likely result (there is always a range of possible results anyway) in a given scenario. Change the scenario and the result will probably/likely change.

It's easy to predict the NBA next year if you use information from this year, but the 100 year model is like trying to predict the NBA 5-10 years down the road. Mystic are you willing to go on the record and make any climate predictions for 100 years down the line?

[mysticbb]

It's easy to predict the NBA next year if you use information from this year, but the 100 year model is like trying to predict the NBA 5-10 years down the road.

No, it isn't. The climate system has a specific response time to internal and external forces and is far more stable over the course of the years than the NBA is. In fact, the climate is considered to be the average of about 30 years. So, in order to predict the state of the climate system in 100 years from now and get an equivalent for NBA predictions, it is like predicting the outcome of the playoffs by looking at the past season and then take the first half of the current season into account as well. Sure, if an important player gets injured, the scenario changes, and if some bigger volcanoes erupt in a short time span or a bigger asteroid/meteor is hitting the Earth (which is rather unlikely over the next 100 years), the scenario will also change, but don't make absurd statement like yours.
And even my comparison isn't actually the best way to describe that. Much better analogy would be trying to predict the average and the distribution of throwing a fair dice like 10,000 times.

Mystic are you willing to go on the record and make any climate predictions for 100 years down the line?

Predicting the development of the climate system is something very different from predicting the weather at a specific location at a specific point of time. You probably believe that climate models would actually do the latter, otherwise your comments wouldn't make any sense at all.

I actually think that you really underestimate the overall knowledge about that topic, and specifically what kind of knowledge I have in that field. And given your responses you way overrate your ability to make an judgement about the scientific value and accuracy of climate models. Please, don't feel offended and don't act on that Dunning-Kruger effect. Just to put something into perspective here: my knowledge in climate physics exceeds my knowledge about basketball analytics easily. ;)

[Chris_SoCal]

I think I understand the origional authors point. Points Rebounds and Assists tend to be team or system stats and stats such as blocks or steals are based much more on an indivuals talent or as mentioned more irreplaceable.

If you play amazing defense you can probably hold a team to 40% shooting if you play virtualy no defense the team will shoot around 50%. If you get a steal you hold the team to 0%. Yes sometimes you get burned but teams score so often anways, they may still have scored regardless.

Statisticly speaking amost everything averages out in the course of games. There isn't a huge differnce between the best offenses and worst. If you want to win games you must make game changing plays. Steals and blocks are obvious ways to do this. Often just a couple of these plays seem to make a BIG difference at key moments.

As to scoring being a team stat.... consider the scoring champ KD he averages around 30 points a game on around 50% shoot. Amazing numbers. However if KD missed a game would you expect OKC to score 30 less points at a much lower fg%? Nope OKC would probably be able to compensate for about 90% of his effectivness. His "game changing" plays they would not be able to compensate for.

Yes steals are undervalued/underappreciated. However are they worth 9 points? What a strange way to look at it. First you do everything you can to devalue scoring buckets then you list how many buckeks a steal is worth to pump it up. While I get what he is trying to say... it is a very awkward way to put it.

[Chris_SoCal]

Thinking about this some more... while steals are undervalued even the best ball theifs only average around 2 steals per game. So while they make a big differnce in the course of games I don't think it is particularly usefull when describing an indivual players effectiveness. Or put differntly there isn't a big enough differnce between the best ball theifs and the worst to matter.