Behold, the idiocy that is John Hollinger

[FINtastic]

I do think it's funny that he continues to gloss over the strength of schedule being a factor in the MOV.

[kchan]

LOL. First ever post. I made 4 accounts over the last year ever since Jason Kidd came into the lineup and none of them gets activated. LOL. I just tried my new OviMail here and it worked!

Now for this particular thread, I'll quote Vin Diesel:

"It doesnt matter if you win by an inch or a mile, winning is winning."

[Usually Lurkin]

Originally Posted by kchan

LOL. First ever post. I made 4 accounts over the last year ever since Jason Kidd came into the lineup and none of them gets activated. LOL. I just tried my new OviMail here and it worked!

Now for this particular thread, I'll quote Vin Diesel:

"It doesnt matter if you win by an inch or a mile, winning is winning."

Your first post is a Vin Diesel quote? Haven't you been lurking long enough now to know that he hates the mavs and quoting him will get you automatically banned?

[kchan]

Its not about that he hates the mavs, but about how the quote applies to Hollinger's thread on winning margins. Besides, he only said that in a movie. He did not make that line up himself. It was in The Fast and The Furious.

[Dirkadirkastan]

C'mon UL, stop messing with people's minds.

[nowhereman]

It's ridiculous and a little classless how riled up Mavs fans are getting about this. Last time Mavs fans got so up in arms about Hollinger's stats, we were swept in the first round. If we hate his stats so much, it's kinda funny to think that the Mavs have reinforced his model over the last few years. I'm not a hater, I'm just saying... maybe we should chill out and see what the team actually accomplishes before we burn him alive.

Many sports statisticians will tell you that MOV is a very good predictive stat, and wins and losses best serve as a retrodictive stat. Hollinger's bottom line reflects what I feel intuitively - taking the whole season, and especially the last 25% (15-16) games the mavs have played, their overall output reflects a team that probably would be eliminated in the first round. In 10 or 15 games, when the mavs move up in the rankings as the pre-trade games are taken out of the 25% part of the equation, the rankings will povide an objective model that shows PREDICTIVELY how the mavs will do in the playoffs - i'd expect it to say we are a 2nd round or WCF team. The model isn't built to effectively reflect skewed trades. This might be because very skewed trades are an aberration and weren't worth accounting for in the system.

If you take our numbers exclusively after the trade under the model, we look like a WCF team. We're slightly behind the Lakers and slightly ahead of the Jazz/Nuggets/Suns. And this is information based on our strongest 10 game stretch of the season. Isn't that an accurate, reasonable statement of how an objective viewer would assess this team?

After the Pau trade, the model had the Lakers considerably lower than where they actually stood and Laker fans were similarly up in arms. The model will play out and we will look better. Major trades aren't built into the system and that's a flaw in the system, but MOV IS a strong predictive stat and on a whole the system is at least a worthwhile model worth considering, along with other information available at sites like basketball-value.com, hoopdata.com, 82games.com, knickerblogger.net, and other such sites.

[nowhereman]

Originally Posted by nowhereman

Hollinger's bottom line reflects what I feel intuitively - taking the whole season, and especially the last 25% (15-16) games the mavs have played, their overall output reflects a team that probably would be eliminated in the first round.

and here, lets be fair. We didn't even play like a playoff team ever since Dirk's teeth went into Landry's elbow. We got absolutely hammered by Toronto, Denver, and LA. We barely eked out wins against Golden State. We lost at home to Minny. It was really really bad. Sure, there are reasons (Damp's absence, Hung-over players..) but can you REALLY expect anyone to create a model that cares about anything other than a team's output? Our output was terrible for at least a third of the season so far.

[DubOverdose]

I think it is hilarious that Miami lost ground in the rankings after beating the Lakers. They weren't supposed to win, so how is it negative that they did?

[nowhereman]

Originally Posted by DubOverdose

I think it is hilarious that Miami lost ground in the rankings after beating the Lakers. They weren't supposed to win, so how is it negative that they did?

Probably because the win replaced a more dominant win in their last 25% of games.

E.G. our ranking went down considerably after the NY game wasn't considered a "recent" game for us anymore. Replacing a 50 point win with a 3 point win will greatly undercut your score. Similarly, in the next couple of weeks, even after a loss I expect our ranking to go up because the ass kicking Denver handed us will be taken out of that subset of the equation.

[kg_veteran]

More "magical clutch powers" on display tonight!

[bernardos70]

Do you believe in magic?
http://popup.lala.com/popup/504684637832576022

[Usually Lurkin]

I just realized: John Hollinger's stuipid basketball nickname would be . . . J-Ho!!! I think he's full of post-trade bitter, and wants to stick it to us.

[chumdawg]

Originally Posted by kg_veteran

More "magical clutch powers" on display tonight!

Let me know where you're coming from here. An eight-point margin is stout. Not clutch. Stout.

And certainly not anywhere near "magical."

[Usually Lurkin]

Originally Posted by chumdawg

Let me know where you're coming from here. An eight-point margin is stout. Not clutch. Stout.

And certainly not anywhere near "magical."

so says J-Ho's magic formulas. But people that watch the game know better!

[jthig32]

I enjoyed some of these notes from DB.com on the subject:

John Hollinger 2006-2007
"The Spurs' problem isn't age or a lack of fire or any of the other ideas trotted out in recent weeks. It's that they're 5-10 in games decided by five points or less, while the Mavs are 12-2. That's the main reason the teams are 8½ games apart in the standings, not any difference in the quality of their play. "

John Hollinger 2007-2008
Utah is 26-18, but has the scoring margin of a team that should be 30-14. The Jazz haven't been fortunate in close games (5-8 in games decided by five points or fewer) but have been in nearly every other contest. Utah's only two blowout losses have come in the infamous 101-77 loss to the Lakers in which they scored only six points in the fourth quarter, and a 96-83 loss to Atlanta in which they pulled the starters when trailing by 27 in the middle of the third quarter. As a result, the standings don't reflect Utah's true strength. Most notably, the Jazz trail Dallas by three games in the standings despite a much better scoring margin.


John Hollinger 2008-2009
Of course, luck can swing both ways, and in this case it did. Dallas went 18-5 in games decided by five points or less, enabling the Mavs to win 50 games despite having the scoring margin of a 47.6-win team. The boost from winning so many close games more than offset the bad luck from their opponents' free throw success.

John Hollinger 2009-2010
Actually they're 16-5 in those [decided by 5 points or less] games, but let's proceed -- I have this debate every year with fans from a different team that happens to be fortunate in close games. They all think their players have magical clutch powers, but they don't. There is NO correlation -- none -- in that stat from year to year, even with teams that keep their personnel remarkably stable

Mav's record in games decided by 5 points or less over the last 10 years

2009-2010 16-5 (after 61 games)
2008-2009 17-5
2007-2008 8-10
2006-2007 19-4
2005-2006 15-7
2004-2005 12-9
2003-2004 9-9
2002-2003 9-9
2001-2002 9-7
2000-2001 10-9

So after Steve Nash left and Dirk became effectively the main man here, the Mavs are 87-40 (about 69%) in games decided by 5 points or less.

Someone else broke down the winning percentage of the other elite teams over recent years, and no one was above 58% in close games. Maybe Hollinger is right in that most teams that win a high number of close games on year regress back to the mean in the subsequent years. But I'm not sure how many years the Mavs (Dirk and Terry specifically) have to do this before he admits it's more than just luck with this team.

[FINtastic]

Originally Posted by jthig32

I enjoyed some of these notes from DB.com on the subject:





Someone else broke down the winning percentage of the other elite teams over recent years, and no one was above 58% in close games. Maybe Hollinger is right in that most teams that win a high number of close games on year regress back to the mean in the subsequent years. But I'm not sure how many years the Mavs (Dirk and Terry specifically) have to do this before he admits it's more than just luck with this team.

Excellent info there. It is interesting the numbers you get when you actually test out Hollinger's assumption on the Mavericks. If we assume that the results of these games really are coin flips, then the probability of the Mavericks winning at least 87 out of 127 close games is roughly 0.0008%.

I wonder when ol' Hollinger is going to reject the null hypothesis here.

[chumdawg]

Originally Posted by jthig32

But I'm not sure how many years the Mavs (Dirk and Terry specifically) have to do this before he admits it's more than just luck with this team.

At the risk of running myself up the flagpole, it is indeed just luck. There's not a team anywhere in the NBA--Jordan and Kobe included--who has demonstrated an uncanny ability to win games that are up in the air with seconds remaining. Not a single team.

The dirty secret that is not often mentioned by Mavs fans is that the team has a god-awful predilection toward making games closer than they need to be. That stat about the Mavs winning fifteen in a row, or whatever, by one point? That's nice and all. But half those games should have been five- or six-point wins.

The Mavs have a terribly annoying predilection toward giving up a basket immediately, and without even a hint of defense, when they are up two possessions with less than a minute to play. Why do they do this?

The point is that that the Mavs make games closer than they need to be. They don't run away from teams very often. They exist in that statistical grey area that Hollinger and the others harvest.

You blame Hollinger in this thread, but the entity you should blame is the Mavs.

[FINtastic]

Originally Posted by chumdawg

At the risk of running myself up the flagpole, it is indeed just luck. There's not a team anywhere in the NBA--Jordan and Kobe included--who has demonstrated an uncanny ability to win games that are up in the air with seconds remaining. Not a single team.

The dirty secret that is not often mentioned by Mavs fans is that the team has a god-awful predilection toward making games closer than they need to be. That stat about the Mavs winning fifteen in a row, or whatever, by one point? That's nice and all. But half those games should have been five- or six-point wins.

The Mavs have a terribly annoying predilection toward giving up a basket immediately, and without even a hint of defense, when they are up two possessions with less than a minute to play. Why do they do this?

The point is that that the Mavs make games closer than they need to be. They don't run away from teams very often. They exist in that statistical grey area that Hollinger and the others harvest.

You blame Hollinger in this thread, but the entity you should blame is the Mavs.

Even if we grant that your argument is true, it still means Hollinger's analysis is off.

[jthig32]

Originally Posted by chumdawg

At the risk of running myself up the flagpole, it is indeed just luck. There's not a team anywhere in the NBA--Jordan and Kobe included--who has demonstrated an uncanny ability to win games that are up in the air with seconds remaining. Not a single team.

The dirty secret that is not often mentioned by Mavs fans is that the team has a god-awful predilection toward making games closer than they need to be. That stat about the Mavs winning fifteen in a row, or whatever, by one point? That's nice and all. But half those games should have been five- or six-point wins.

The Mavs have a terribly annoying predilection toward giving up a basket immediately, and without even a hint of defense, when they are up two possessions with less than a minute to play. Why do they do this?

The point is that that the Mavs make games closer than they need to be. They don't run away from teams very often. They exist in that statistical grey area that Hollinger and the others harvest.

You blame Hollinger in this thread, but the entity you should blame is the Mavs.

Up in the air with seconds remaining is not the discussion at hand and not Hollinger's point. The Mavs consistently winning a significantly higher percentage of five point MOF wins over the past five years than anyone in basketball is not luck.

You can say that the Mavs make games closer than they need to be, and maybe they do. But how a game ends up being decided by five points is not relevant to the discussion. Hollinger states that ANY game decided by a low MOV is a coin flip, regardless of how it got there. And in the case of the Mavs, he's wrong, and they've proven it over and over again.

[G-Man]

Originally Posted by nowhereman

It's ridiculous and a little classless how riled up Mavs fans are getting about this...
If you take our numbers exclusively after the trade under the model, we look like a WCF team. We're slightly behind the Lakers and slightly ahead of the Jazz/Nuggets/Suns. And this is information based on our strongest 10 game stretch of the season. Isn't that an accurate, reasonable statement of how an objective viewer would assess this team?

And it is what Holliger would have written in this article, if he was objective. He would have written something like " Wow, suddenly my model has the Mavs looking as good as any team in the West outside of the Lakers". Instead, he keeps saying things like "I'm still highly skeptical. Scoring margin is a better predictor of future success than win-loss record; Dallas is 8th of the 8 likely West playoff teams in that metric. The trade makes them better, but those numbers haven't improved significantly since the deal, and of course they're not the only team that got players at or near the deadline."

And this is my real beef with Hollinger's recent article. His own model says the Mavs after the trade are a much better team than they were before the trade. Yet he persists in saying they are not. I believe this is because his system showed Butler and Haywood to be statistically similar to Howard and Damp, so they would only add 1 win. The 10 game win streak has blown that prediction to bits, yet JH refuses to admit it.

What we see in the Mavs are 2 new guys who's intangibles are not well measured by his system. But rather than admitting that his own system shows the impact of the trade is much better than he had forecast, he downplays the post trade stats. That is because it is only by rolling in the full season stats that he can continue to claim he was right that the trade makes no difference.

I don't think its classless to point this out. And its way more fun than ragging on the refs or something that is classless. Because, in this case we get to hoist the stat genius on his own numbers, and see how he cherry picks them.

You are right that in 3 weeks, the last 25 stats he uses will look better, because it will be all post trade. That will let him adjust his predictions and say his model works. Perhaps it does. But right now, he can only defend it by cherry picking stats, and its fine to call him on it.

[chumdawg]

Originally Posted by jthig32

Up in the air with seconds remaining is not the discussion at hand and not Hollinger's point. The Mavs consistently winning a significantly higher percentage of five point MOF wins over the past five years than anyone in basketball is not luck.

You can say that the Mavs make games closer than they need to be, and maybe they do. But how a game ends up being decided by five points is not relevant to the discussion. Hollinger states that ANY game decided by a low MOV is a coin flip, regardless of how it got there. And in the case of the Mavs, he's wrong, and they've proven it over and over again.

It is entirely relevant to the discussion, and I am shocked that a reasonable person such as yourself would begin to think otherwise.

[nowhereman]

Originally Posted by G-Man

And it is what Holliger would have written in this article, if he was objective. He would have written something like " Wow, suddenly my model has the Mavs looking as good as any team in the West outside of the Lakers". Instead, he keeps saying things like "I'm still highly skeptical. Scoring margin is a better predictor of future success than win-loss record; Dallas is 8th of the 8 likely West playoff teams in that metric. The trade makes them better, but those numbers haven't improved significantly since the deal, and of course they're not the only team that got players at or near the deadline."

And this is my real beef with Hollinger's recent article. His own model says the Mavs after the trade are a much better team than they were before the trade. Yet he persists in saying they are not. I believe this is because his system showed Butler and Haywood to be statistically similar to Howard and Damp, so they would only add 1 win. The 10 game win streak has blown that prediction to bits, yet JH refuses to admit it.

What we see in the Mavs are 2 new guys who's intangibles are not well measured by his system. But rather than admitting that his own system shows the impact of the trade is much better than he had forecast, he downplays the post trade stats. That is because it is only by rolling in the full season stats that he can continue to claim he was right that the trade makes no difference.

I don't think its classless to point this out. And its way more fun than ragging on the refs or something that is classless. Because, in this case we get to hoist the stat genius on his own numbers, and see how he cherry picks them.

You are right that in 3 weeks, the last 25 stats he uses will look better, because it will be all post trade. That will let him adjust his predictions and say his model works. Perhaps it does. But right now, he can only defend it by cherry picking stats, and its fine to call him on it.

I wouldn't chalk any of this to an anti-mav bias. I'd chalk this up to an editor imposed deadline and espn's constant desire to stir up trouble and get as many hits as possible. This might be academic dishonesty, but it's a way better read this way. It's also not good for business to openly acknowledge a flaw in the system. If the Mavs move up his rankings, he'll give them their due. Until then he will find reasons to justify their low position in a way that doesn't undermine the model.

[bobbyfg7]

I'm still confused on the luck part. How is it luck if you have a good free throw shooting team? Isn't that what happens in the final seconds of close game? The losing team will foul to stop the clock, right?

Is it luck that the Mavs are usually the team on top in those final seconds? Is it luck that the Mavs were able to make their shots and keep the other team from making theirs? Is it luck that the Mavs have the experience to know what it takes to win in those final seconds?

To me, calling anything luck is taking away from the skill of the players to make their shots, pick up their defense to another level, and do what ever it takes to get the W in the final seconds of a close game.

[Dirkadirkastan]

Originally Posted by chumdawg

It is entirely relevant to the discussion, and I am shocked that a reasonable person such as yourself would begin to think otherwise.

I'm not quite sure what you two are squabbling about here. Both of you seem to be arguing that the Mavs' close wins are not coin flips. I guess Chum kind of puts it in an "almost choking" context while Thig has a "super clutch" outlook, but in the end, how are your claims different from each other with respect to Hollinger's "coin flip" claim?

[kg_veteran]

Originally Posted by chumdawg

Let me know where you're coming from here. An eight-point margin is stout. Not clutch. Stout.

And certainly not anywhere near "magical."

It was a close game down the stretch. Didn't the Kings get to within 2 with a shot to tie pretty late in the game? Oddly enough, because it's an 8 point win (thanks to some late FTs), Hollinger probably considers it a quality win instead of "freakish luck", which sort of makes my point for me.

[kg_veteran]

Hollinger picks an arbitrary cut-off point (5 points or less) for his analysis. The point of my wisecrack is that: 1) this game won't even show up on the grid for him, when it should (as a close game that was won in the final minutes); and 2) the Mavs won yet another close game, and it sure didn't look like luck.

BTW, I agree with Chum. The Mavs have a bad habit of letting opponents hang around rather than putting them away. That could eventually come back to bite them, because when you play with fire you do eventually get burned. But, I would argue that when they get into those late-game situations, it is NOT complete luck that allows them to prevail most of the time.

[Usually Lurkin]

Wow. It's good to know that we lost the Finals because of a series of unfortunate coinflips, and not because of bad coaching decisions, or mental collapses, or biased referees.

[SeriousSummer]

I think there is a reason why the Mavs are an outlier to Hollinger's statistical analysis, and I also expect that very soon the Mavs unusual performances in close games will come back in line with other teams--and it isn't because the Mavs are either super-clutch or chokers. It's the unusual way the team is constructed.

The Mavs have, until recently, been a very poor defensive team at the perimeter, but a very, very good free throw shooting team. The strategy at the end of close games with the Mavs leading has really been to avoid fouling and then when the other team fouls the Mavs to make the free throws.

So the Mavs go into the last minute up 3 points. The other team drives and scores because the perimeter defense is bad, the interior defense is no more than fair, and the Mavs above all want to avoid the foul. Then the inbounds pass goes to Dirk (or Jet or Kidd) and he makes the foul shots and the sequence repeats until the game is over and Dallas squeaks out a win.

Other good teams have better perimeter defense and don't shoot free throws as well. That means their strategy is higher risk and higher reward. They can't just let the other team score and make free throws. So they defend; sometimes foul; sometimes miss free shots and lose.

In any event, that strategy is more likely to lead to bigger wins or to losses. The Mavs are like a good chess player squeezing a pawn advantage into an end game win. You can win constantly that way if you're good enough, but it's a cautious, boring strategy.

I think the Mavs will be changing their strategy a bit now. With Marion and Butler, the perimeter defense is much improved. That should mean more end game stops and a few more wins by a larger margin.

[Murphy3]

In all honesty, you simply cannot look at the final score and decide what is and is not an actual close game without reviewing the final few minutes of every game throughout the season. What could have been a relatively easy double digit win could turn into an easy 5 point win if a team hits a couple of meaningless jumpers to close out a game. How many times have we seen the Mavs with a nice margin in the closing minutes let it get down to 5-6 points or so?

You just cannot pick an arbitrary margin of victory number and base all of your statistical reasoning off if it without running a huge risk of being off base. We have all seen a football game that's going down to the wire turn into a double digit win for one team in a matter of seconds. We see that in all sports.

Example 1 - A football team is down 4 win a minute to play with a third and goal from the 5. The QB throws a pick, and the CB runs it back for a TD. The margin of victory ends up being 11 points.

Example 2 - A football team has been down by double digits all game long..They're down 12 but scores a TD as time expires to make it a 5 point game.

Which game was closer? Well, the final score was closer in example 2. But, the game was never in doubt.

Which game was closer in doubt? Example 1.

Common sense... It has to play a factor. Obviously, I know that each and every one of you could have come up with the same example.. but I'm just trying to illustrate how common sense is left out of Hollinger's equation.

His statistics are example of a couple of things:

1. Stubbornness. Anyone with an ounce of common sense understands that his statistics cannot be full proof. There simply is no way for such a statistic to be remotely full proof. But, he has dug down so deeply in the trenches that he refuses to admit such. To admit that his statistical analysis is not full proof would be to undermine himself in his opinion. So, he cannot and will not do so at any cost. He's just another 2 game swing David.

2. His statistics also show an example of people being unable to properly use statistics. I am not saying that Hollinger is not capable. Personally, I believe that it's his stubbornness that does not allow him to admit the obvious. But, there are many people that view his statistics and take them as gospel. Many complex statistical metricls are seriously flawed. Just take a look at most defensive statistics in baseball. They simply cannot account for everything. Hollinger's statistics simply do not account for a team hitting a couple of late jumpers to make a game that was not close into a 5 point game as time dwindles down.

Common sense. That is what his statistical analysis lacks. I know that it is frustrating to some of you.. But just know that it is a statistical analysis that is based upon flawed logic..so therefore, his statistical analysis is very, very likely to be flawed. That is not to say that there isn't some merit to his analysis at times. I am not saying that his margin of victory stat is incapable of being correct. However, due to the flawed makeup of the statistic, yes, it is extremely likely to be incorrect..

[G-Man]

To be fair, I love Hollingers numbers and his analysis, even when I disagree with it. I'm just complaining now about how he acts when his stats fail to match reality. Instead of just saying "stats have their limits," he says things like "winning close games is just luck." Or winning 10 in a row incluing 5 play-off contenders doesn't indicate the post trade Mav team is better than the un-improved version that lost something like 7 of 11 going into the break.

[Murphy3]

Because it's obviously luck when you're up by 7 and a team hits their final shot to bring it down to a 5 point lead and reaching the magical close game cutoff line.

[Dirkadirkastan]

The final margin doesn't always appear dominant, but this team has already gone on many impressive runs. Now I'm not seeing what everyone else is describing. The problem isn't that they generally falter late (though this does happen on occasion); it's that all too often they come out flat early in the game. So a big run that would normally give a team a comfortable lead merely puts the Mavs back in the game. Just look at the current win streak.

Suns: Mavs fell behind by 8 early, yet took a 19-point lead in the third quarter and ultimately won by 10.
Magic: Mavs fell behind by 11 early, and again in the third, but went on a 19-0 run that turned the game around and won by 10.
Heat: Mavs fell behind by 8 at halftime, but still won by 6.
Pacers: Mavs really dominated this game despite sloppy play. Took a 22-point lead, then all the starters checked out and the Pacers scored a bunch of garbage baskets for an unimpressive final margin of 9.
Lakers: Mavs fell behind by 9 a couple minutes before halftime, then led by 5 shortly after the intermission. Probably one of the biggest runs you can have against that team. Final margin of 5.
Hawks: This one's all over the place. Mavs took an early 14-point lead, then fell behind by 15 in the 4th for a negative 29-point swing. Then the monster duo of Kidd and Dirk led the charge to an 8-point overtime win (positive 23-point swing).
Hornets: Took a huge 25-point lead by playing what was probably the most beautiful basketball of the win streak. Then they got sloppy in the second half, letting children on the other team cut the lead to 4, but the Mavs still won by 8.
Bobcats: The Mavs were down again early, this time by 10 at the half. But there was a positive 18-point swing in the second half, and the Mavs won by 5.
Wolves: This game sucked. Fell behind 7 early and took a 13-point lead at some point before the half, but who cares? They're the Timberwolves.
Kings: Mavs got frustrated and couldn't buy a call, and fell behind by 10 as Carlisle was ejected. Still, the Mavs took a 15-point lead in the third quarter, and won by 8.

So the Mavs are showing an ability to go on huge runs every game. But they're usually falling behind early and sometimes slipping late. I'm not sure if it's faulty lineups due to the current lack of depth or a drop in intensity or what, but there are lulls that are hurting the margin.

[dirno2000]

As has been pointed out the five point barrier is arbitrary. I thought it would be interesting to go back and see if the Mavs have consistently exceeded their expected win total based on the Pythagorean method (I used 16.5 as the exponent,not sure what Hollinger uses).

As we look at the past 9 years there's not the clear pattern of outperforming MOV that we see with the five point cutoff. From 2003 to 2005 we actually won fewer games than we should have. Still we've exceeded our expected win total in three of the past four years. That should be enough to at least consider the possibility that there's something going on here.


Expected Wins--- Actual Wins--- Difference
2001-2002--- 54--- 57 (+3)
2002-2003- --64--- 60- (-4)
2003-2004--- 55--- 52- (-3)
2004-2005--- 59--- 58- (-1)
2005-2006--- 60--- 60- (0)
2006-2007--- 63- --67-(+ 4)
2007-2008--- 56--- 51- (-5)
2008-2009--- 47--- 50- (+3)
2009-2010--- 37- --42- (+5)

[chumdawg]

Originally Posted by kg_veteran

It was a close game down the stretch. Didn't the Kings get to within 2 with a shot to tie pretty late in the game? Oddly enough, because it's an 8 point win (thanks to some late FTs), Hollinger probably considers it a quality win instead of "freakish luck", which sort of makes my point for me.

Excellent point that you make here, and that Murph echoes. If you really want to look at performance in "clutch" or "close game" situations, you should start your analysis at some point in the game--like with a minute or two left, probably--when there is STILL some game to be played.

[Usually Lurkin]

The Mavs will start climbing quickly up his statistical rankings.
It'll be interesting to see what he says then

[bernardos70]

He'll default to this:

Originally Posted by Bookit

I don't want to completely dismiss the Mavs' prospects because they could land home-court advantage for two playoff rounds, and I'd like to see them play a few more games with their new acquisitions. But despite the recent win streak, my outlook on Dallas really hasn't changed much. Until further notice, it's a pretty good team … but one that's unlikely to topple any of the three fantastic ones it may have to get past in the Western Conference playoffs.

In other words: I want to cover my butt just in case I am wrong and the Mavs are better than I thought. I reserve the right to change my mind.

[kg_veteran]

Originally Posted by chumdawg

Excellent point that you make here, and that Murph echoes. If you really want to look at performance in "clutch" or "close game" situations, you should start your analysis at some point in the game--like with a minute or two left, probably--when there is STILL some game to be played.

Right. As Murph pointed out, you have to know the context of each particular game (and filter the data accordingly) in order to make any sort of cut-off point relevant. I like your idea of looking at performance in close game situations. It would take a lot more work for a stat geek to find those games and analyze them, but if you could find the games where the score is within x points with 1-2 minutes left, then you could better analyze whether a team really performed well in late game situations.

Of course, that's not what Hollinger is trying to analyze. He's trying to analyze overall team strength, and part of his analysis includes the faulty assumption that teams that win close games are simply lucky.

[chumdawg]

Originally Posted by kg_veteran

Of course, that's not what Hollinger is trying to analyze. He's trying to analyze overall team strength, and part of his analysis includes the faulty assumption that teams that win close games are simply lucky.

In fairness to Hollinger, he's trying to find a way to look deeper than the standings. If W's and L's are all you are concerned about, you can find it in your morning paper.

His rankings, then, are more usefully seen as relative indicators from team to team. If two teams have the same W-L record, the team that more often kicks the crap out of the opponent may be the stronger team.

The more I think about it, the more I think it may not be worth much, for the sole reason that the entire NBA season, as long as it is, is still a small sample size (relative to the statistical models based on pythags). And more particularly, a playoff series definitely represents an extraordinarily small sample size. What difference does it make, in other words.

Another way of stating it is: If the Mavs have a knack for winning close games, how much does that help them in a playoff series?

[xrobx]

well the fake show just got beat today for their 3rd loss in a row. we're only 3 games out of the #1 seed now. IS IT POSSIBLE?

[jthig32]

Originally Posted by chumdawg

It is entirely relevant to the discussion, and I am shocked that a reasonable person such as yourself would begin to think otherwise.

I'm not saying it's not a relevant talking point in general. It is. But it's not relevant to the question of whether Hollinger is correct in his assessment of the Mavs ability to win close games.

The reason it's not relevant is this: Regardless of why the Mavs win more than their fair share of close game, be it because they're super duper clutch, or because they simply fart away big leads all the time, either option still invalidates Hollinger's analysis. If the Mavs are consistently bad at holding leads and putting bad teams away, but still find a pay to pull out more close games than the average team, then it still nullifies Hollinger's discounting of the Mavs due to their inferior MOV.

As to the rest of the discussion, I wholeheartedly agree with the idea that the final score in no way tells you the overall competitiveness of the game itself. It's the major flaw in using MOV.

Interestingly enough, a poster at DB.com did a historical analysis of the last decade's playoff results, and found that MOV was no better than winning percentage at predicting who would win a given playoff series or how deep a team would go in the playoffs.