Byron: Analytics Will Play A "Much" Bigger Role This Coming Season

That's all I'm saying. Of course, I'd build my team around Kobe and his skillset. He's a transcendent talent.

 

That's fine. There's a notable difference between Jordan shooting a mid-range shot and Jordan Hill from last year, obviously. No doubt. If Russell and Clarkson can be that efficient from mid-range then more power to them; take the shot. Nevertheless, the way teams would generally guard Clarkson would be to go under the screen inviting the 3-point shot in some instances. That's where Clarkson needs and will improve, IMO.

Generally, though, I agree w/you.

 

I think myopic might be the better word, and it better describes the idea that decontextualized expected values should dictate strategy.

for the smartest guys in the room, the analytics folks really have no understanding of the limitations of their data and/or analyses, particularly the notions of causal direction and predictive precision.

and I'm pretty sure it was nobody's Plan A to have Jordan hill bomb away from the perimeter. when you're an awful team, you end up in awful positions, and players don't/can't play to their strengths. at least we didn't have analytics savvy mike dantoni here to encourage hill to take a few steps back before hoisting...

 

No, Mike D'Antoni had Pau hoisting from the 3 point line. LOL! It's all about that floor spacing. . There's no reason to overthink this from my perspective. I take the simplest example; Like @therealdeal stated w/the example he used w/Kobe, I'll take that information and try to use it to my competitive advantage. It's also in the Battier video that I posted...he explained how analytics helps him.

It's not a bad thing. As for Jordan Hill, you have it misunderstood; he was bombing away because we were 'tanking.' Although, the schemes were awful since the beginning of the season, and I'm thinking the analytics would have helped Byron switch what he was doing from the beginning of the season...

 

Right but that assumes two things - that Hill actually heeds the advice and that the offense is structured to give Hill looks elsewhere. Hill took a lot of mid-range shots because he got the ball there and was open. We didn't run plays for Hill, and we probably wouldn't still. Also - players have to change their habits. Hill likes the mid-range shot, apparently. You can show him shot charts (which I don't even consider analytics) but he has to make the change. A lot of what people deem analytics is stuff teams have been doing for decades....they just used to call it film and scouting. I don't think Scott and his team have any shortage of doing that. Where they may fall short is the advance stats category, which as I explained, have just as many flaws as traditional stats.

 

Gotcha. Watch the Battier video and see how it could be used for a player if you have the time; it's only about a 3:30 video.

 

The NBA's focus is beginning to shift to analytics and talent development. What are your thoughts on the use of advanced stats and scouting in the NBA?

"I think it's essential. That's the way the game is growing. The game changes and either you change with it or you get left behind. We didn't have all these stats before. In my opinion I don't think people necessarily turned their nose up at it, but it just wasn't available. You had a box score and basic averages, but with analytics, things like baseball's sabermetrics, and fantasy sports, it's amazing what you can break down.

How you get those stats and relate them, or translate them to in-game situations remains to be seen. But it really helps me in preparation. Seeing what different lineups are good together, efficiency defensively and offensively, it's just amazing how deep you can go into this stuff. I'm all about it, I think it's really cool. I know some people are a little resistant to it -- and to each their own -- but I think it's here to stay. It's not going anywhere, and I think it's a big help to the game.

These are tools. As a coach you're a leader, but you're really teaching. Why would you reject anything that would make you a better teacher? That's how I look at it. Why would you turn anything down that can help you teach these young men, and if you're a teacher, you have to broaden your own horizons. You have to keep getting better. You can't just sit in your chair like some old university professor for 80 years and not change. You have to roll with the flow. This is a fluid, evolving game and its important to stay up with it."

http://www.silverscreenandroll.com/2015/8/7/9116041/la-lakers-head-coach-casey-owens-dfenders?utm_campaign=silverscreenandroll&utm_content=article%3Atop&utm_medium=social&utm_source=twitter#

Lakers New D-League Coach Casey Ownes and his thoughts on analytics...

 

That's good news. This is key for me...."How you get those stats and relate them, or translate them to in-game situations remains to be seen"

It's not a new problem, either. There has been film for decades. Coaches saw tons of stuff on film - how do you take the good and the bad and replicate it to a T in a real game situation.

My answer is that it's not 100% possible, but it's still extremely valuable to have that info.

 

Can we bring this dude up yet?

 

Yeah, Bertka has been apparently doing it for decades, perhaps not to the same amount of detail, but still...

 

Does Byron have a "list"? I wonder if Casey just made it?

"These are tools. As a coach you're a leader, but you're really teaching. Why would you reject anything that would make you a better teacher? That's how I look at it. Why would you turn anything down that can help you teach these young men, and if you're a teacher, you have to broaden your own horizons. You have to keep getting better. You can't just sit in your chair like some old university professor for 80 years and not change. You have to roll with the flow. This is a fluid, evolving game and its important to stay up with it."

 

Not sure I follow this. The "analytics guys in the room" are providing filtered and projected data.... nothing else. What a coach believes is important and useful is where the "rubber hits the road" in this.

As a number cruncher and a long time follower and coach of youth basketball I see very little value in the so called advanced analytics in their current form. There is minor player placement and situational things that can be extrapolated however the use of predictive analytics to create a positive net effect in a team sport where the dynamics of a data set are never the same is beyond what you can do with math IMHO.

As I do my work I'm acutely aware of the limitations. I see them in real time in fact. You try and design data queries that eliminate as many of the limitations as possible however there is no way to get beyond a certain degree of probability.

Basketball is A team sport that creates an ever changing set of data points with nearly infinite variables. Basketball statistics, even in their advanced state, are all individual in nature. The interactions of the dynamic game situations cannot be modeled with even an average probability.

Matching personnel up with a need can be modeled to a certain extent. That is a lot more individual in nature than making coaching decisions based on analytics however.

 

not sure why you don't follow me; you're saying similar things. my issue is that the analytics acolytes (including the dfenders coach, btw) don't seem to understand what they're doing (and not doing) any more than the Byron scotts of the world. btw, as a university professor, I think it's sort of funny that someone touting advanced statistics would then turn and impugn career academics as not understanding the world.

anyway, part of effective information usage is knowing what to ignore, or perhaps more precisely--how to weight various indicators.

it behooves stat guys to insist they're uncovering truths that the layperson can't understand (or the ignoramuses refuse to)--these are guys who had no jobs in sports a decade ago. there's still no Oakland A's in basketball, for reasons you stated. does that mean you should take the "Burn the witch" attitude that scott seemed to espouse until he was browbeaten by the press, fans, and finally (and obviously) his bosses? no. but it also means that at the heart of his skepticism is some truth.

back to my original point: the fundamental misunderstanding of the midrange shot's value has to do with a very simple mistake in reasoning (as opposed to the other myriad issues with trying to model complex data patterns with what is really very few relevant data points in nba basketball). the theory seems to be that avoiding the midrange shot will make you better. Houston espoused this theory full-on and rode it to...a mediocre offense. this is because it's more likely the causal direction is reversed: good offenses create other kinds of shots. but what's a good offense? it can take various forms, and dantoni's suns were just one of them.

am I trying to say that we should all shoot 18 footers all day? of course not. but simply suggesting that if you stop doing that, you'll be better is frightfully reductionist. this kind of thinking is rampant in the coverage of analytics, and it makes one suspect whether it's rampant in the offices of the practitioners as well. of course, we can't really know that.

in sum, if Byron scott (or any coach) hears "hey, stop having hill shoot the midrange shot. you're welcome, idiot", I understand why he doesn't want to talk to that person. again, we don't know that this is how it goes, but it's as plausible as scott giving them surprise wedgies whilst drawing up wing isos for hill.

 

You called BS on 1) the fact that there are analytics guys who don't have a clue about basketball and you also said 2) Byron is a rock head. I disagree with both.

I'm not saying analytics guys don't know any basketball (nor that Byron is the sharpest in the tool shed)....but as others have said, just pointing to a stat and saying "we should do less of this with player x because he's ineffective doing that" is pretty useless. I think they absolutely need a translator, every team does. If Byron has a vision for offense and defense - analytics should complement it, not replace it.

The Lakers seem to have hired a guy that can say "when we run this set, we're more effective when this player does x and this player does y. It also puts us in better position for transition defense". Not just "hey we're really bad in these categories, and the Warriors won a chip because they aren't". Maybe they found a guy who knows his basketball x's and o's and can translate analytical data into sets within our offense. It sounds like that wasn't happening before according to Byron.

 

not to turn this into another scott thread, but nobody could win with what he had last year. I'm not sure the best coaches in the league could win 40 with the roster they have this year. and that's assuming health, which hasn't been there. analytics, jerry sloan, whatever...we need to keep expectations realistic. the lakers don't have 1st, 2nd, or even 3rd tier star right now (*maybe* kobe could be 3rd tier if he could stay on the floor). that's tough sledding.