Around the NBA part 2

[Red Vines]

Clippers are so deep it's not even funny and Grant Hill is coming back, unreal.

[Capn'O]

Your overconfidence is pathetic. You don't even understand that you made a terrible bet, just from a probability theory aspect.

Your overconfidence, on the other hand, is AWESOME!!!!! Congratulations on having AWESOME!!!!! overconfidence.

I'm hardly overconfident. I have probability on my side and it's a shame that you can't understand why.

You absolutely are. You're overconfident in your assumptions that basketball is a coin flip and that his crappy first month is statistically relevant. Otherwise you wouldn't have so gleefully taken the bet.

In fact, you are so overconfident that you came in crowing after one game like you won the bet when somebody as statistically savvy as yourself ought to know we're playing game of averages.

In the end, we'll be haggling over assists.

[fdr2012]

You absolutely are. You're overconfident in your assumptions that basketball is a coin flip and that his crappy first month is statistically relevant. Otherwise you wouldn't have so gleefully taken the bet.

In fact, you are so overconfident that you came in crowing after one game like you won the bet when somebody as statistically savvy as yourself ought to know we're playing game of averages.

In the end, we'll be haggling over assists.

You cannot predict future basketball statistics, just like you cannot predict the stock market return over the next year.

The first rule of any data analysis is that you NEVER throw away data. Saying that Lin's first month is irrelevant is based on nothing and simply demonstrates your bias. Lin just played a game that showed you that month was VERY relevant.

We are playing a game of averages. We've already seen 36 coin flips and 46 still remain. I'll take my chances that the first 36 flips weren't biased.

Nevermind. You're not going to understand anything I say anyway. You'll just end up blaming Mchale or Harden or the "system" or "injuries" when you lose the bet. I can't really blame you here. Guys like Bill Bradley on the other hand, know what I'm talking about, but they're still going to be blinded by their homerish glasses.

[ibraheim718]

CeePeeThree looking like a statue out there.

[Capn'O]

You absolutely are. You're overconfident in your assumptions that basketball is a coin flip and that his crappy first month is statistically relevant. Otherwise you wouldn't have so gleefully taken the bet.

In fact, you are so overconfident that you came in crowing after one game like you won the bet when somebody as statistically savvy as yourself ought to know we're playing game of averages.

In the end, we'll be haggling over assists.

You cannot predict future basketball statistics, just like you cannot predict the stock market return over the next year.

The first rule of any data analysis is that you NEVER throw away data. Saying that Lin's first month is irrelevant is based on nothing and simply demonstrates your bias. Lin just played a game that showed you that month was VERY relevant.

We are playing a game of averages. We've already seen 36 coin flips and 46 still remain. I'll take my chances that the first 36 flips weren't biased.

Nevermind. You're not going to understand anything I say anyway. You'll just end up blaming Mchale or Harden or the "system" or "injuries" when you lose the bet. I can't really blame you here. Guys like Bill Bradley on the other hand, know what I'm talking about, but they're still going to be blinded by their homerish glasses.

Enjoy stats 102 this semester.

I, on the other hand, will enjoy watching basketball and keep my fingers crossed for the assists fairy to show up. Because that's what it's coming down to.

[fdr2012]

Enjoy stats 102 this semester.

I, on the other hand, will enjoy watching basketball and keep my fingers crossed for the assists fairy to show up. Because that's what it's coming down to.

No it won't. We made a bet on three different stats. It's not more likely to "come down" to assists than FG% and PPG. You're making completely baseless statements that do nothing but demonstrate how clueless you were when you proposed the bet.

[Bravery]

Clippers are so deep it's not even funny and Grant Hill is coming back, unreal.

I've watched the Clippers all season long (mostly to see CP3/Griffin and how Crawford is doing in that system)

I don't know if they can overcome Durant and OKC's playoff/finals experience, but the Clips are incredibly deep from top to bottom. They all just kinda fell into their roles early and that's what happens when you have a superior PG out there handling all of the traffic. They also have a mix match of vets and youth needed to go far. Hill, Billups, Turiaf, Butler, CP3 are all battle tested guys and it's especially important when you have the ears of young players, Griffin, Jordan and Bledsoe who need leadership and someone to mold them.

They are one of the few teams that have what I would like to call the holy trinity: They have the go-to superstars. They have a balanced auxiliary. And they have their bench. They possess the highest margin of victory in the league at around 11 pts I think - and they do it with their chemistry and a top 5 defensive efficiency rating.

One final note - Although they will not admit this openly, I also believe the constant talk surrounding the Lakers and their star acquisitions during the off-season also contributed to their instant chemistry as well. They are simply hungrier by proxy and thus it fueled their motivation... and when you have a lot of intelligence and pure athletically-gifted talent on that team, that can become very dangerous.

[Capn'O]

Enjoy stats 102 this semester.

I, on the other hand, will enjoy watching basketball and keep my fingers crossed for the assists fairy to show up. Because that's what it's coming down to.

No it won't. We made a bet on three different stats. It's not more likely to "come down" to assists than FG% and PPG. You're making completely baseless statements that do nothing but demonstrate how clueless you were when you proposed the bet.

Look, maybe you missed it hanging out by the tracks flippin' coins with Hollinger... but I have provided plenty of basis for these predictions since before the season started, essentially predicting the arc of his production to date. I would extrapolate more but you've been kind of a jackass since we made this bet so I don't really feel like it.

[fdr2012]

Look, maybe you missed it hanging out by the tracks flippin' coins with Hollinger... but I have provided plenty of basis for these predictions since before the season started, essentially predicting the arc of his production to date. I would extrapolate more but you've been kind of a jackass since we made this bet so I don't really feel like it.

Sorry, nothing personal, but you Lin homers are terribly annoying. He's a 12/6 guy. No more, no less. I don't know what about his game warrants so much discussion as if he is some unique talent, a potential allstar (real one) or a future HOF. Guys with Lin's talent are a dime a dozen in the league (and outside of the league).

Lin is nothing special. There is no "arc" to his production. He had a very bad November, the a good December and he may very well fall off again in January. There is no trend in the data. What you see is what you get.

[Capn'O]

Look, maybe you missed it hanging out by the tracks flippin' coins with Hollinger... but I have provided plenty of basis for these predictions since before the season started, essentially predicting the arc of his production to date. I would extrapolate more but you've been kind of a jackass since we made this bet so I don't really feel like it.

Sorry, nothing personal, but you Lin homers are terribly annoying. He's a 12/6 guy. No more, no less. I don't know what about his game warrants so much discussion as if he is some unique talent, a potential allstar (real one) or a future HOF. Guys with Lin's talent are a dime a dozen in the league (and outside of the league).

Lin is nothing special. There is no "arc" to his production. He had a very bad November, the a good December and he may very well fall off again in January. There is no trend in the data. What you see is what you get.

Well, I think Bill Bradley has lumped me with the skeptics recently so bias is often in the eye of the beholder, innit.

Anyway, I'm well aware that from a raw probability perspective you're looking at 3 categories, each with a probability of 50% going up or down based on this year's statistics as a discrete data set (slightly more in your favor if you consider that neutral counts and goes to you). 1/2*1/2*1/2 = 1/8 and that's where you're getting your numbers from about my odds of winning. Lin's stats aren't data from a textbook about an unknown commodity though. When I make a basketball prediction, past performance is certainly the starting point from where I make that prediction. However, even at that - what is your discrete data set? Is it this year only? Does it include last year (pre-injury where his numbers were better)? Games where he's played "significant minutes"? Per 36? What is the cutoff to create a discrete data set?

To use the stock market example - yes nobody can really predict what will happen with a stock on any given day. However, if you've done your due diligence and know a little bit about the company you're investing in, where they're starting from, and what strengths they might have (e.g. what guys like Warren Buffet do) you can have some degree of confidence in your long term investments (particularly if you make several or many based on due diligence). You know what kind of company you're investing in.

With Lin, I knew three things about him going into the year and predicted a slow start compared with last year's #s. These were:

1. He's coming back from a significant knee injury
2. He's being integrated as the PG in a new system with all new main pieces
3. He's a young player and younger guys need a little more time to learn and have less leeway implementing their own methods

You're saying the above factors had no influence on how he started the season and that this year's cumulative stats are the only basis for predicting future stats. I think that's a crock. He looked stiff and confused for about a month before he and his teammates figured out where to be. Playing with confidence/full physical ability/a distinct role I think the subsequent fluctuations we're seeing now are more indicative of where his season means will fall. I misspoke calling his earlier numbers outliers - Bringing in basketball observations, I think they're different data altogether. "Lin with injury" vs. "Lin at full strength", we could call it.

"Lin at full strength" seems to round out at about 14/6 on 45% shooting. Did last year. Has since November.

The assists are still a concern for me. Harden does a fair amount of the facilitating and I'm not seeing a whole lot of room for improvement on those numbers. Another concern is that other teams will figure out how to stop the Rockets as N'Awleens did tonight and Lin's role will change again. We shall see.

[GettinitDone]

There are so many Supermans playing for Lakers... Shaq, Dwight, and Earl Clark (22pts and 10 rebounds vs Spurs)... Clarksanity!!!

[fdr2012]

Well, I think Bill Bradley has lumped me with the skeptics recently so bias is often in the eye of the beholder, innit.

So apparently there is a scale of insanity, where the uber-insane call the less insane a "skeptic". Good to know.

Anyway, I'm well aware that from a raw probability perspective you're looking at 3 categories, each with a probability of 50% going up or down based on this year's statistics as a discrete data set (slightly more in your favor if you consider that neutral counts and goes to you). 1/2*1/2*1/2 = 1/8 and that's where you're getting your numbers from about my odds of winning.

That's just a naive and simple way of doing it. In reality FG% and points are (positively) correlated, so your odds are slightly higher.

Lin's stats aren't data from a textbook about an unknown commodity though. When I make a basketball prediction, past performance is certainly the starting point from where I make that prediction. However, even at that - what is your discrete data set? Is it this year only? Does it include last year (pre-injury where his numbers were better)? Games where he's played "significant minutes"? Per 36? What is the cutoff to create a discrete data set?

I'm going to take the entire dataset (including last season). The issue with Lin is that he's much more effective when he's used as the primary scorer and when he has spacing. After melo came back last season, his stats took a nose dive. Obviously, I'm not looking at old games where he played less than 20 minutes.

My best guess of his "true" ability in a secondary role is around 12-13 PPG on 40% shooting.

To use the stock market example - yes nobody can really predict what will happen with a stock on any given day. However, if you've done your due diligence and know a little bit about the company you're investing in, where they're starting from, and what strengths they might have (e.g. what guys like Warren Buffet do) you can have some degree of confidence in your long term investments (particularly if you make several or many based on due diligence). You know what kind of company you're investing in.

False - Warren Buffet cannot pick firms/stocks better than a random number generator (or a monkey). Warren Buffet is successful because he invests only in low-risk stocks (i.e. low cash-flow volatility) and he has very low financing costs (via his insurance companies). Anybody who has access to money at -2% like Buffest does can do the same thing.

With Lin, I knew three things about him going into the year and predicted a slow start compared with last year's #s. These were:

1. He's coming back from a significant knee injury
2. He's being integrated as the PG in a new system with all new main pieces
3. He's a young player and younger guys need a little more time to learn and have less leeway implementing their own methods

You're saying the above factors had no influence on how he started the season and that this year's cumulative stats are the only basis for predicting future stats. I think that's a crock. He looked stiff and confused for about a month before he and his teammates figured out where to be. Playing with confidence/full physical ability/a distinct role I think the subsequent fluctuations we're seeing now are more indicative of where his season means will fall. I misspoke calling his earlier numbers outliers - Bringing in basketball observations, I think they're different data altogether. "Lin with injury" vs. "Lin at full strength", we could call it.

"Lin at full strength" seems to round out at about 14/6 on 45% shooting. Did last year. Has since November.

I saw no evidence of Lin being injured in November. He simply sucked. IMO, your points are nothing more than excuses to justify the data after the fact. I see Lin on this Rockets team as a 12/6 guy on 40% shooting. If Harden goes out, he can do much better.

The assists are still a concern for me. Harden does a fair amount of the facilitating and I'm not seeing a whole lot of room for improvement on those numbers. Another concern is that other teams will figure out how to stop the Rockets as N'Awleens did tonight and Lin's role will change again. We shall see.

Teams constantly adjust and players constantly get hot/cold. That's why you have mean reversion in player stats. Lin will revert back to his mean and as the #3 option on that Huston team, I don't think that's more than 12/6 on 40% shooting.

P.S - Regardless of everything, you still made a terrible bet. Completely irrational.

[Rasho Brezec]

right now with felton out it wouldn't be horrible to have lin coming off the bench. we could use someone to attack the basket. see, i'm fair. i'd also take gallo over copeland. don't tell you know who though.

If the team signed Kenyon would you take Gallinari over Stodemire?

what stoudemire are we getting ? it's hard to answer that.

The real one, not the mythical one.

[Falstaffxx]

If the team signed Kenyon would you take Gallinari over Stodemire?

what stoudemire are we getting ? it's hard to answer that.

If we're thinking about them both as starters, then I'd slot Gallinari into this lineup over Stoudemire, at the 3 with Anthony at the 4. Stoudemire could be a good backup 4 for this team though. I guess I'd take Gallinari over Stoudemire. He's a better defender, he draws fouls at a rate similar to Stoudemire, has more range, and can drive with the ball.

[Falstaffxx]

Of course, the guy we presumably could have had instead of Stoudemire is Lee. At the time, I thought it was a great move, but now I feel dumb.

[RHODEY]

The Clippers' lone Achilles's heel is that they look rather ordinary in a half court set. We saw this last night for stretches in the 3rd qtr. versus Dallas. Blake Griffin does not impress me when he is not dunking and CP3 is not really a go to stud, he's a play making stud.

IMO, they are a deeper versus of those Nets teams that made the finals... .

When it comes down to it all the depth in the world(which they have) wont makeup for these shortcomings in the later rounds of the playoffs.

[god shammgod]

Marc Stein ‏@ESPNSteinLine

Only four West teams over last 15 seasons, according to Elias, made it to playoffs at 15-20 (or worse) through 35 games. Lakers now 15-20

that's a pretty damning stat. i still think they have a chance though.

[Rasho Brezec]

I don't:

Arash Markazi ‏@ArashMarkazi

According to AccuScore the Lakers have a 48.3% chance of making the playoffs, finishing with an average of 42 wins after 10,000 simulations.

Arash Markazi ‏@ArashMarkazi

The last five teams to make the postseason as an 8 seed in the West had a combined win percentage of .584, which is about 48 wins.

Arash Markazi ‏@ArashMarkazi

The Lakers would have to go 33-14 the rest of the season to finish 48-34.

[god shammgod]

I don't:

Arash Markazi ‏@ArashMarkazi

According to AccuScore the Lakers have a 48.3% chance of making the playoffs, finishing with an average of 42 wins after 10,000 simulations.

Arash Markazi ‏@ArashMarkazi

The last five teams to make the postseason as an 8 seed in the West had a combined win percentage of .584, which is about 48 wins.

Arash Markazi ‏@ArashMarkazi

The Lakers would have to go 33-14 the rest of the season to finish 48-34.

they won't reach that record but i'm not sure portland will either. they don't have a great chance that's for sure.

[Thugger HBC]

No way of knowing how many games they HAVE to win, but they need to start winning some games real soon.