Barkley: Analytics are @#$%

[Rerisen]

Rip, Billups and Rasheed were all nearly equal to Harden in offensive ability,

None were even close, nor have touched his offensive numbers.

Rockets are 9th in defense, better than us, and Harden is giving much better effort on that end this year.

The problem with Harden, much like Rose had in 2011, is you can be ranked somewhere between 4th and 10th among the best players in the NBA, but around 80% of the titles are won by the teams with a top 3 guy.

Just like Chuck was 2nd-5th best in 1993 but unfortunately MJ was #1.

[te887848]

For my money, Detroit had 3-4 players better than Harden and perhaps Barkley. Defense is that important.

I'd take Derrick over Harden and Barkley. I like his defense more when it matters most and I doubt Harden gets so many FTs in the playoffs. It's not a landslide or anything but the narrative that Harden has arrived as some sort of true superstar cause of his numbers is a foul one to me.

Detroit didn't have a single player in the same universe as Harden or Barkley. Rose used to be a superstar and whether you'd take a healthy/pre-injury version of him over Harden/Barkley is perfectly debatable, but since the injury he hasn't proven to be any better than Brandon Knight, never mind Harden or Barkley.

[Gmoney3]

Rip, Billups and Rasheed were all nearly equal to Harden in offensive ability,

None were even close, nor have touched his offensive numbers.

Rockets are 9th in defense, better than us, and Harden is giving much better effort on that end this year.

The problem with Harden, much like Rose had in 2011, is you can be ranked somewhere between 4th and 10th among the best players in the NBA, but around 80% of the titles are won by the teams with a top 3 guy.

Just like Chuck was 2nd-5th best in 1993 but unfortunately MJ was #1.

Exactly. In fact most the titles go to the best player in the NBA.

Since 1991 when Jordan's reign started the best player in the NBA won the title arguably every year from 91-98 (8 titles), again from 00-02 (3 titles), then 09, 10, 12, 13 (4 titles).

That is 15 of the last 24 finals have been won by a guy who was at least arguably the best player in the game.

Other years such as 99, 03, 05, 06, 07, 11 had top 5 players in Duncan, Wade, and Nowitzki.

The only teams that have won titles without a top 5 player were the 04 Pistons, 08 Celtics, 14 Spurs.

[Rerisen]

The only teams that have won titles without a top 5 player were the 04 Pistons, 08 Celtics, 14 Spurs.

Garnett was definitely top 5 in my book in 08.

His defense was still at DPOY levels and he was still an efficient volume scorer on offense too.

Even raw production he was 4th in the league in PER.

[Gmoney3]

^ I was thinking of putting him there but I figured it was close.

[dice]

The only teams that have won titles without a top 5 player were the 04 Pistons, 08 Celtics, 14 Spurs.

Garnett was definitely top 5 in my book in 08.

ben wallace was #4 in 2004 xRAPM as well, for what that's worth (his defensive # was off-the-charts)

[mccluskey]

Today, someone tweeted that analytics have become an erroneous substitute for true understanding of a sport. It makes sense given how fantasy basketball and football have become so popular among fans. There are a lot more box score analysts today and frankly, anyone could write a mildly informed profile on a player without actually watching tape.

I understand both sides, but I will always lean eye test first, and then statistics in any sport except baseball.

100% agreed with this. Sports are big mainstream entertainment these days, and you have a lot of people with no background or serious interest in studying them who still want to appear knowledgeable and be part of the conversation without having to really understand the game.

[Rerisen]

Today, someone tweeted that analytics have become an erroneous substitute for true understanding of a sport. It makes sense given how fantasy basketball and football have become so popular among fans. There are a lot more box score analysts today and frankly, anyone could write a mildly informed profile on a player without actually watching tape.

I understand both sides, but I will always lean eye test first, and then statistics in any sport except baseball.

100% agreed with this. Sports are big mainstream entertainment these days, and you have a lot of people with no background or serious interest in studying them who still want to appear knowledgeable and be part of the conversation without having to really understand the game.

The missed irony in the tweet and its hubris, is how many people have a 'true understanding of the sport'. It suggests an arrogant, "I have nothing more to learn," attitude.

Of course 9 out of 10 will claim they do and then when you ask for their opinions on a player's skill at a particular thing, you will get 3 or 4 different answers among them on how good the guy is at it, based on all their subjective eye-tests. Just read the Arron Afflalo thread for instance, people can't even agree what the guy is good at.

Numbers like TS% or +/- never give you different answers based on the same data. They aren't a bible for exact player worth, you must be able to interpret some context, but they are at minimum an objective foundation to use to help you paint the picture of a player more accurately, based on actual facts of what happened.

Not some guy seeing a couple Kobe game winners replayed over and over on sportscenter, never showing his far more often misses, and believing he's 'uber clutch'. What % of the NBA fanbase believed that myth for years? Their 'eye-test' believed it because that is all their eyes ever were shown.

The root of analytics, is of course analysis or to analyze something, and all that means is to try to understand it better.

Where the NBA has come from the 40s till now you could say is on account of 'analytics' in a generic sense, even if past advancements of understanding the game - Tom Thibodeau scribbling his own player rating system in a notebook - were done with pencil and paper instead of computers. It's nothing more than progress, the march of civilization!

[coldfish]

I think that analytics are overused in two ways:
- Small sample sizes. When you have a small amount of data, one flyer either good or bad can drastically alter the data and make it unreliable.
- Small differences. For much the same reason. If one guy has a +2 and another is at +4, there is really no statistical difference between those guys. One small stretch of play could switch the numbers around.

As an example of where advanced stats are useful, people just looking at box scores could have come to the conclusion that Josh Smith in Detroit was fine. All the advanced stats showed that he was killing the team. Quite frankly though, the eye test did also. I have rarely seen the eye test and a full boat of advanced stats disagree.

Advanced stats are basically good for dismissing people who don't know squat about basketball and dismiss subjective eye test points. Of course, those people are tough to talk to in general. It always surprises me how many players and ex players fall into that group though. That's why you should never hire one directly as a coach or GM. They need some time away from the game to get used to watching it as an outsider.

[chitowndish]

Really loved that analytics video but I really don't think they determined that minute restrictions are BS in fact I think the discussion highlighted the opposite. In fact Colangelo credited that for the vaunted Phoenix medical staff. They do study the analytics, determine the best course of action for a particular player and then dictate it to the coach. The idea being that the coach isn't the expert here and shouldn't control that aspect. The exact thing that Thibs is chafing at. But it sounds like the idea of having health experts dictate limits for players is something that's starting to happen across the league. So in a way it sounds like this whole Jen Swanson, Thibs FO battle isn't something limited to our team, it's a change in the way that sports medicine and analytics are being applied to the game across the board. You have the old schoolers like Thibs who don't like it but it's starting to be adopted across the league because of the success Phoenix enjoyed using it. Like the example of McDyess where the guy was supposedly "healthy" and the Phoenix medical staff refused to let him play for 4 weeks because they knew better. An old school coach like say Thibs (and Stan) isn't going to like that, being told he can't play a player that seems to be completely healthy, it isn't going to help him win games immediately (and really didn't help win games in that case) but their results speak for themselves over the long run.

What I got from that whole discussion is the struggle between Thibs and Jen is likely happening across the league and it's probably a battle that guys like Thibs are going to eventually lose because they are flat out not experts in that field. It isn't about "Yoga Instructors" telling Thibs how to coach, it's more about coach's pretending they are experts on wear and tear of the human body when they don't have nearly as much expertise in that area as they think they do. Pushing through it, toughing it out, being a warrior thinking is becoming outdated and the league is starting to evolve beyond it. Really interesting video though I really enjoyed it. Tons of it was applicable to this Bulls team.

[Mech Engineer]

Gar mentioned the other day how the Bulls use analytics more in the medical recovery/rehab stuff than actual basketball. He claimed that they are a more traditional scouting team and they look at a few of these numbers(like real plus/minus etc..).

But, Jen or some assistant of hers must be a medical analytics proponent in the Bulls organization. That might be the friction between Thibs and Jen/Gar because I think Thibs understands the basketball analytics part and wouldn't be against it.

[chitowndish]

Yeah I think Stan highlighted Thibs point of view perfectly, they don't trust the science behind it and it's never been done in the past. MJ played huge minutes etc. With the basketball analytics Thibs can probably vet the results against his understanding of the game and have a grasp of where it's coming from and why it works but the medical stuff they just have no frame of reference to and it often runs counter to winning in a game and what's been done in the past. Like Stan said if you take your best guy off the court you have a worse chance of winning which is the coaches job. It's really about an investment into the future, and you can't see the benefit game to game which is where the coach operates.

Guys Like Fred Tedeschi probably come from that old school sports school of thought of if you are healthy you go and if not you don't type of thing (Which has historically been their job) where Jen is probably coming from more of this new school train of thought of analytics and managing wear and tear over time (which people haven't even concerned themselves with historically). Like that really is two different roles and frame of mind. Fred has to determine if players can play and get them back to playing in the short term (injury recovery) where Jen has to try and manage long term affects and health (injury prevention). So I would think that's why Thibs supports Fred so much is they come from that old school style of thinking where this whole Jen thing is completely new and encroaches on a coach's traditional turf (which a role like Fred's really doesn't). Looks like the league has been copy catting what Phoenix did, and you are seeing the push back from the old guard types.

[ChosunX]

Charles Barkley is the real MVP of basketball discourse for kickstarting this thread.