So do you delusional steven curry bandwagoners still think he's mvp

who said it was a joke just because he used defensive stats? my point was you can't justify the assertion that curry is an elite defensive player just because his defensive numbers are good, because everyone knows defensive stats are not 100% reliable, but at the same time it is still better than eye test so should be given more weight, just not 100%. If the guy who made the video gave more weight to the eye test, then it is more problematic.

 

You criticized the video for using stats and said they paint a wrong picture. Then in the next post you suddenly said "stats are the way to go" when evaluating players and that they paint an accurate picture after 82 games.

So, what is it now?

 

you're creating a false dichotomy where there isn't one and making it sound like I made a very simple black and white argument. I said you can't solely rely on defensive stats to judge any player defensive ability because it's not 100% reliable. But at the same time, it is still better than the eye test so should be given more weight, perhaps something like 6:4 or 7:3. What I was saying is if coach nick gave 90~100% weight to the stats, that's wrong. But if he gave 90~100% weight to the eye test, that would be even more wrong since at least you can draw general conclusions by analyzing stats, whereas eye test results can be highly subjective and selective.

 

You didn't say any of that or nuance anything in your original posts, better say it that way from the beginning.

 

you drew false implications from my statement where all I said was 1. can't judge solely based on the stats 2. same goes for the eye test, and maybe this is even worse. Besides it's a moot debate, because neither of curry's defensive stats nor most people's eye test results suggest he is an elite defender. It's somewhere around above average range, and basically that title of the video was nothing more than a clickbait.

 

Yes, Curry is not an elite defender, but James is not an ELITE shooter either.

 

You never said "solely" or anything in that regard, but let's not argue about unnecessary crap.
Know that the title was weird and agree with above average.

 

imo, stats and eye tests are equally effective, that is if you are a careful observer. Why? Because they are the same thing, you are trying to describe how a player plays. Stats are compiled by human observers. The advantage here is that you won't have to watch 82 games to get the whole package - that is because someone else has watched 82 games for you and recorded their observation into numbers. Something like +/- is raw and straightforward, while some others are up to interpretations, like you can technically name 10 different kind of assists.

Analytics then take it one step further to make sense of the raw numbers by weighting what's more important versus what's less important, for ex, someone thinks free throw is efficient, then a formula would give it a higher coefficient.

Likewise, an eye test guy could value the impact of heart, say an observer might think heart outweighs consistency, so this person may have his or her own ideas on what's important, a mental equivalent of coefficients in stats.

They are all basically the same thing. Do you trust someone else's take on data collection and their generalization of them? Or do you trust the consistency of your own eye? If you are observant and actually watch the games, you'd always be the better judge versus those who don't. Stats would be good additions because while you might be watching for a certain qualities in players, you wouldn't want to count number of times something is done, so you read stats compiled by someone else who did.

It's the same game!

Stats may not record an action in full details, and it would always lack distinction of the action's quality, but it is useful because you can see a trend over large amount of similar actions over 82 games. An eye test is better for evaluating an action in as much depth as you're capable.

 

Klay Thompson still better than Harden. Carrying the Warriors through the post season. Meanwhile Harden's defense is so bad he can't even defend himself from a photographer without committing a foul.

 

While both are important, IMO the eye test is more valuable.

For instance, we've all seen the 8 minute youtube video of Harden's defensive lowlights. On one of the plays, Doug McDermott was in the corner being guarded by Harden. He caught Harden ball-watching, went backdoor, received a pass, and missed a wide open layup.

On that possession, the eye test says that Harden played terrible defense. What do the stats say?

 

That's not how stats work. You have to define what you are measuring. That play most likely isn't in whatever definition of defensive stats that the person is measuring.

 

Harden is still better than klay, because you're drawing general conclusions based on a way too small sample size (6 games w/o Curry) and the defense he faced was pretty terrible, any player can go off on Houston and Portland's sh**ty defense. Besides,

2015- 2016 Playoffs

WS Draymond 2.2 > Klay 1.8

WS/48 Draymond .231 > Klay .197

BPM Draymond 8.7 > Klay 4.0

VORP Draymond 1.2 > Klay 0.6

Ortg Draymond 118 > Klay 117

Drtg Draymond 98 > Klay 106


Draymond Green carried the Warriors in the playoffs, not Klay. That's a fact.

 

You don't need stats to see this, anyone watching the games can tell you that their entire offense and defense flowed through Green.

As far as stats vs eye-test, I think the best system needs both.

You use stats to give you ideas on what to look for, or for indicators/red flags.

Then you use the eye-test to back that up.

It's the reason the Warriors still use guys like Jerry West who said it himself, you can't use analytics to quantify heart (which a lot of defense is)

 

I think you are conflating two issues here. Yes, a human observer watches games and records points, rebounds, assists, etc. for the each player. But that's pretty standardized procedure and is subject to much less interpretation than say figuring out whether this player is better than the other in general. The issue at hand is the latter, and for that eye-test is the last thing you want to use simply because there is no way of proving yourself right or wrong to others.

Many people confuse coming to their own personal conclusions with proving yourself right to others. Eye-test is only useful for the former, because you're basically convincing yourself that what you've seen with your own eyes is correct. Of course, we all know that's not the case based on years of psychological research (confirmation bias, focusing effect, etc.)

So if you think Player A is better than Player B largely based on the eye-test, it's fruitless to engage with other people in a debate, because you have no sufficient evidence to convince others whether your opinion is true or wrong. It's called falsifiability, which is the bedrock of any sound argument and if your theory is unfalsifiable, then it's not a legitimate theory in the first place.

All you can do is state your personal opinion and then move on. Don't couch it as an objective fact and don't challenge others to change their mind. You are unqualified to do that if you are an eye-test advocate.

Now if you are in the stats camp, your argument can be falsified. It's very simple, find out what's wrong with a particular stat or its interpretation and go from there. Stats are tailor-made for this kind of discussion, because it usually contains generalized information (thus much less selective than the eye-test) and objective in nature. Not objectively true, but objectively falsifiable.

Who is the greatest scorer in the history of NBA? If you are one of the eye-test advocates, you have to go through all the past footages of every single player and then make a mental note whenever some player 'seems' to have scored more than others. Funny thing is, once you start recording their points, you're basically half-way into the stats camp. Notice how bizarre that process sounds like? I'm not arguing that stats should be given 100% weight and there is no room for casual observation, but it should be unquestionably clear as to which of the two is a more reliable source of information (to everyone, not just yourself) whenever people engage in this kind of discussion.

 

link to podcast

Around 17:15

“You can’t look in the heart of someone–you can take every test…

“One of the best players in the league, one of the very best players in the league is James Harden. I really admire the way he plays. And he’s one of those players who reminds me kind of an old-fashioned player. But he’s so clever.

“But my best guess is if we gave these guys all kinds of tests before they were drafted, and he was a high draft pick, I don’t know how he would he’d rate on that scale.”

 

yes i agree with you, except that I believe it's not a 50:50 thing and stats should be given more weight, if we are having a debate. read my post above for the detailed explanation.

 

I'm convinced. Draymond Green is better than Harden.

 

Although I disagree, you can make that argument for this season. Draymond finished higher in the MVP ranking this year, so at least that's not a stretch, unlike Klay.

 

To the broader point of stats vs eye-test in team building, for most organizations, stats should be given more weight.

But I'd even go as far as to say given how championship teams tend to be more than the sum of their stats, that the best eye-testers (who actually use a very logical/statistical approach as you say) will beat the best stat geeks at basketball team building.

There's a imo large human component to basketball. Stats would have told you not to draft Draymond Green or Kawhi Leonard, both who couldn't touch 25% from 3, and to a lesser extent Steph Curry. The human component -- their drive to be great, leadership, habits IE "heart" is totally unquantifiable...

Stats tells you pairing Harden and Howard would make a great team. Turns out they both have fragile egos and different motivations.

Analytics has a huge place in basketball. It definitely tells you how to optimize on floor strategy a la Morey telling Dwight he doesn't deserve post touches --- side note, the real question is, is it a net positive via Dwight's improved effort and defense to give him inefficient post touches.

Again to the broader point, I love analytics, Morey, et al. But I hate how people espouse analytics like thats it, there's nothing else to it. As in Harden's defensive metrics are average, he must be an average defender. When anyone who watches the Rockets will tell you that his defensive lapses are demoralizing to his teams, and every time he forgets to play defense it could actually have a bigger effect than simply giving up one basket. Stats also cant quantify Dwight maybe thinking if Harden doesn't play D, why should I?

Stats doesn't tell you as Jerry West stated, Harden has no heart and maybe building a team around a guy with no heart is kind of a mistake.

 

Based on what Dwight has said in this post-season, Harden being affected by Dwight's lackadaisical play is a more appropriate assumption, since Dwight himself admitted he got disinterested in the game, not because of Harden, but due to lack of offensive possessions he was getting. They both played subpar defense, but only one player publicly admitted he didn't put in 100% effort for very selfish reason, so it's only logical to think that Dwight negatively affected Harden and his teammates, not the other way around.

You're getting paid millions of dollars to just play basketball. I would say that's one of the best jobs in the world. If Dwight refuses to give maximum effort because he feels like he's not getting enough touches, then that should reflect poorly on him, not other players or Morey.

You still didn't address how eye-tests can be a reliable source of information, especially in a debate setting like this. You're not going to convince anyone with your eye-test results. What if your eyes are deceiving you? everyone's subject to confirmation bias, if you have a preconceived notions about Harden's poor defense, then that's all you gonna see. That's why stats are important, because it can counter whatever personal biases you have. Stats are just superior to eye-tests in terms of its accuracy, general applicability, and objectivity.