ClutchFans Game Thread: Rockets @ Wizards 12/9/2006

Ugh...Tracy's back, again...I know, I know, but this doesn't make me feel good at all...Yao is still the best...

 

There may be a finite set of variables that have significance over a large sample, but you don't know WHICH ONE. And out of those finite set, you don't know HOW MUCH EACH ONE CONTRIBUTE. You can't isolate them.

It is fickle for you to say one way or another when you don't really know the inherent effects. If it is as simple as you say, "beating the market" would be a reality instead of a theoretical possibility.

You'd notice that nowhere did I say you can't use statistics. In fact, I clearly stated in my last debate with you, that statistics CAN be a good starting point, if used properly. The problem is that you are not using it properly.

If for example, the lights didn't go out, or if TMac didn't get injured. I'd buy your theory of "Yao only played well because it was against inferior competition" a lot more than I do now.

You've made no attempts to isolate the independent variables, which by the way, simply can't be done in this case. Yet you drew a conclusion regardless.

And by the way, I think Yao is a good play. But that is just my opinion.

What are you talking about? You are debating how Yao did when Haywood/Thomas went out. That is, BY DEFINITION, an event study.

Likewise, you talked about how you think the team will do after Gay/Swift was traded for Battier. That is also, by definition, an event study.

Situations that occurred in a game that might have had an impact on the results? That is also, an event study.

I wouldn't keep mentioning event studies if you don't keep using them.

Let's talk about risk. Risk is the deviation from expectations. You have an expected value of what Yao is expected to score against those two. The expected value may very well be higher for Lang/Booth than Thomas/Haywood.

But how Yao performed ISN'T measured by the expected value for those two set of players. It is measured by how he actually did compared to his expected value against Lang/Booth.

So with your conclusion, is the expected value for Lang/Booth 23 points a quarter? That is just, wow...

And you didn't use statistics, that is somewhat true, I'm actually wondering what you did use exactly.

Based on observations and anecdotal evidence, there are at least a few who challenged your assertions, saying that Lang/Booth did better defensively than Haywood/Thomas. One of them said Yao faced double teams from Lang/Booth which he did earlier, you dismissed that as "weak double teams."

So OK, Haywood and Thomas are better than weak double teams? By how much?

No. But you do need to back up your assertions.

Like I said, I'm not challenging your opinion that it hurts the Wizards. I'm challenging your opinion that it hurts the Wizards by 23 points.

Not according to what you just did buddy.

No. An analysis doesn't have to be about "Yao can do this against blah blah." It could be casual observations about the things Yao did right, or wrong, both of which are much easier to pick up, because you are not predicting a trend, but just making an observation.

You could look at how we ran our plays. And if we did them well, nobody is preventing you from using the qualifier, "mind you, this is against the Wizards."

Once again, that doesn't predict a trend.

No. Like I said, what I do disagree is that it's 23 points easy.

Mind you, it wasn't a direct quote. But you didn't say Yao is expected to play better against Lang/Booth? You did that even in this very post.

That is fine.

 

You absolutely can isolate variables. It's done all the time. You construct a model, and you use it to make predictions. If those predictions are generally accurate, then you have a useful model. You can't "know" with certainty how much each variable contributes, if that's what you think I'm saying.

And nice job misquoting me. I never said it was "simple".

Here's the problem, and there's no point for me to read further. That is not my "theory". I suggest you take some time to consider what I actually said, instead of debating against whatever strawman you constructed in your mind.

 

-edit-

 

I think I read somewhere Yao called the Wizard centers "great player".

 

Yeah, that's why post-game quotes, which are many a time full of "political correctness" and hyperboles, can hardly be qualified as expert opinions and be used to boost one's argument.

 

durvasa and MFW make my brain hurt.

 

Um no, you can't. The Wizards game, my personal opinion. There were 3 significant variables. Of course, that is just my guess. So let's test them. They are:

1. The lights went out, giving Yao a breather
2. TMac got hurt
3. Haywood/Thomas got hurt

Now let's test them. In order to test the effect of the lights going out, you have to find a game where TMac didn't get hurt, Haywood/Thomas weren't hurt, we are playing the Wizards in the 4th quarter on December 9, 2006, with the game starting at 7, in Washington, reffed by Joe DeRosa, Marc Davis and Sean Wright, where Thibodeau got a tech (wow, I must have missed that during the game), Battier going 3-4, Alston... "pant, pant." Then you measure how Yao did.

You are telling me you can control the variables? The first thing that you can't repeat right away is the date and time of the game.

That's insignificant you say? OK, just isolated it to TMac not getting hurt and Haywood/Thomas not getting hurt. How would Yao do with the lights before the 4th quarter. How many times have the lights gone out in a game, period? Not many in Yao's career. And even if you miraculously find a large enough sample size, you can't isolate it from TMac and Haywood/Thomas being out.

Oh, of course, you would have to then do it by having the lights not go out, TMac being hurt but Haywood/Thomas aren't, keeping everything else constant. Then do it for Haywood/Thomas being out, the lights and TMac not being out.

After that you combine them, which in its very nature, is flawed, because there may be correlation between the variables. So you say, OK, what's the effect of the lights being out and TMac being out, but not Haywood/Thomas...

You've gotta be kiddin' me when you say you can isolate the variables.

I didn't misquote you. I said "it was as simple as that." I just put the quote around the wrong place.

Strawman? Wow, that's rich. So now isolating the variables in an event study is now strawman? Good to know.

I don't mean to nitpick, but I can't help but notice you attributed Yao's good game to inferior competition while completely not addressing the fact that he played without Yao and with a couple minutes breather.

 

Make that "he played without TMac."

 

I skimmed the last 2 pages of this thread to see why it hadn't died and now I've got a headache.

Can't we all agree on the following:

1 That Lang and Booth were totally outmatched and didn't belong on the floor with Yao? They barely belong in the NBA. If your title is 3rd string center or 4th string center on the Wizards, you are verrrrry low on the NBA food chain. They are both a subspecies of NBA player. Booth had played a total of 17 minutes the entire season before Saturday. Lang had played 23, 20 of them in a blowout loss. Both Booth and Lang had played in only 3 games prior to Saturday.

2 That while Yao would not have scored 23 points against Hayes and Thomas, he still demonstrated an awesome ability to take over a game very late? That his attitude adjustment after Tracy left was beautiful to behold?

3 The 15 minute break obviously helped Yao because I don't remember ever seeing him expend so much energy in the 4th quarter of any game, much less a back to back.

What I'm saying is Yao was awesome but let's add a touch of realism. If Yao ever scores 23 4th quarter points against a legit center, that would be another level above what happened Saturday. He's the best center in the NBA and just because a smidgen of realism is added to the discussion of the Wizards game, that doesn't make someone a YOH.

Then again, I didn't read the whole debate because it was too painful. Some of you should know that debating stats with durvasa is like debating Sam Cassell with tinman: You cannot win.

 

Actually, we dont know what Yao could have done to Thomas/Haywood in the 4th. In the 4th, early at least, Yao was dominating Thomas early before he got hurt.

 

We don't know for sure. Maybe he would have scored 40. But it's a safe assumption Yao wouldn't have scored 23. It's safe to assume neither Thomas nor Haywood would have committed 4 fouls in 4 minutes. You get the point.

If Yao got hurt also, who knows, maybe Luther would have scored 23 points. It's a safe assumption that wouldn't have happened.

 

Wow,those are really good debates between durvasa and MFW.

Just, too much obbessed with the details and forgetting about the one fact: THIS IS A FIVE-MAN's GAME. Yao is great, but he couldn't have done all that without a well-executed game plan and teammate's great efforts.

For the post-game analysis, I would say, that's focus on the break down tactis as this link shows:
http://bbs.hoopchina.com/htm_data/30/0612/145888.html

Let's focus on how well those tactis excuted, how we can use this kind of game plan against the Lakers since T-Mac is not gonna play next time which is kind of same situation here?

See, I enjoyed THIS kind of after game anaylsis more.

 

You mention three factors which, in your opinion, had an impact. I agree. But I'm not going to pretend like it's possible to quantify the degree to which they had an impact just based on the information in this game. It's a pointless exercise.

I never was said you can draw strong statistical conclusions from the data in a single game. Nor have I even attempted to do this. I don't know what else I can say to get this across to you.

Attributing words and ideas to me that I've never expressed in this thread and arguing against those --- that's the strawman.

Why should I address everything that happens in a game? I made a simple observation and posted it. Am I obligated to go through every single factor? Did I begin my post with "These are all the reasons why Yao played well in the fourth quarter"?

 

Thank you. I agree with everything in your post.

 

http://www.nba.com/games/20061209/HOUWAS/playbyplay.html

Yao had two free throws early in the fourth against Lang. Etan Thomas came in, drew a shooting foul from Yao on one end, and then got called for a shooting foul on the other (the same play in which he hurt his ankle stepping on a teammates foot). Then Thomas left the game.

But you're right, we don't know what Yao could have done against Thomas/Haywood in the 4th.

 

But you might forget one thing that TMAC was hurt, so that the 3rd or 4th string centers and forwards could totally collape on Yao Ming, which MAYBE, just MAYBE offset the benefit Yao got due to the injury of their 1st/2nd string centers. By no means, Yao can constantly score 23 points in every quarter, therefore it was a spectacular performance; otherwise, it would be norm, not worth mentioning at all. But to claim it's something he's supposed to do facing 3rd/4th string centers and double/tripple teams is absurd.

I just don't understand the backhanded compliment at all. It was a good win, and it was a good 4th quarter performance, it wouldn't kill someone to compliment without a "but". But that's all fine, everyone is entitled to his/her opinion. But to get cute and play with half-truth, and partial stats to "prove" how that backhanded compliment was indeed sincere praise and decent objective analysis, that's an insult of public intellegence. If one wants to go about statistics, study, and critical reasoning, instead of being honest with it - personal observation, then, that someone has to be ready for some solid and bulletproof scientific exam.

In short, any attribute in any event WILL affect the outcome of that event. To attribute the result SOLELY to one main factor, while intentionally ignoring another major factor equally important, if not more, has nothing to do with objectivity or study, but spinning on own term.

There is not big enough sample size to draw conclusion that 3rd/4th string centers of the Wizards working extremly hard and double tripple teaming the only left opponent offensive weapon, will have far inferiorer result than their 1st/2nd string centers of the same team, who could NOT afford to focus their full defense on one player alone. However, we have an established conclusion, that Rockets played poorly without TMac by last year's result. If one opens up a discussion claiming it's all fact based, stats based, study oriented and objective conclusion, while ignoring the biggest, already proven most important factor of Rockets win - both of their star players staying healthy, that someone has to be ready to take some challenge of his STUDY.

Getting defensive, and claiming vulnerable of being labelled YOH or something isn't an honest way to conduct such study.

 

Who's spinning here?

This is nonsense. I'm not obligated to rigorously prove that Yao wouldn't have played as well against Thomas/Haywood in the fourth. I'm offering an opinion, which I think is reasonable and fairly obvious. Not a quantitative study.

Anytime someone makes a comment you don't agree with, you expect them to prove it statistically? What's your problem?

 

I am spinning, so don't take me seriously. However, I call my observation subjective. I wouldn't claim my one-liner and half-truth-telling observation as OBJECTIVE ANALYSIS.

You showed clearly that you are aware of the SIGNIFICANCE of those variables, and also talked about abstracting and constructing models. However, comparing to Haywood's injury, which hasn't been proven to have a big impact on Wizards' team result over the years, TMAC's injury has been proven to be the most important negative contributor to the result from any team he played for. Am I far-fetched to call you intentionally ignoring TMac's injury, while emphasizing Haywood's departure? Otherwise, how do you justify that obvious error in event study and abstraction, from a diligent stats researcher and thorough analyzer like yourself, Sir?

 

Firsrtly, I believe you're the one who introduced the word "objective" into this thread, not me.

But yes, I strive to be objective when possible (I think most people here do). That doesn't mean I need to exhaustively discuss every aspect of the game, to your satisfaction.

If you took the time to read what that was a response to, you'd understand I was making a general comment and not talking specifically about a single NBA game. Further, I never claimed to be carrying out a statistical study in this game -- that's a story you and MFW have made up on your own for whatever reason. Such a study is theoretically possible with enough information, but I'm not going to bother attempting it. If you care so much about it, why don't you do it?

Anyways, I'll take your advice and not take anything you say hereafter seriously.