Expectations, and the obliteration of

You mentioned that one of the expectations/goals was to stay competitive. Well, how does being ahead after three quarters in almost 65% of your games sound? We also didn't have a game where we trailed at the end of all four quarters the entire time Durant was on his 20 point-per-game streak. The Denver game put an end to that.

I imagine a lot of people know this site http://sonicscentral.com/apbrmetrics/viewforum.php?f=1
is out there but if you want more number-crunch reading or discussion there is a lot available.

A previous post is in moderation and will provide some additional links and could make what I said after it make more sense.

If there is a Thunder player who is doing something wrong all the time then maybe a Thunder fan needs to challenge him to online play and school him on that til he changes.

Barry probably had a pretty good memory. He might be prepping to become a GM someday though.

Most players won't read long reports ... but they will play videogames. So you use the NBA game with the "DNA" of the actual player tendency stats and get them to learn that way.

Ok he is the editor of it.

I may be misremembering on that journal founding part but he at least has been associated with it some since.

Jax for a deeper look at hoopnumbers new version of Adjusted +/- this write up is available

http://www.hoopnumbers.com/allAnalysisView?analysis=RAPM&discussion=True

Haven't found the Block paper in its entirety but it might show up in a journal eventually.

The price of anarchy paper is here
http://www.bepress.com/jqas/vol6/iss1/3

The Thunder's consultant helped found this journal.

The rate of injury is pretty steady across players except it is elevated for those under 21 or heavy.

This is old but it has injury incidence by age and that probably hasn't changed much

http://tinyurl.com/yhfuqll

That's "Crow:" not "Crowl" (shades of Get Smart)

Reading about Pryzbilla injurying himself while injured reminds me how the incredible health of the team may have contributed to how it's surpassed expectations. I wonder if its as much a function of age as it is of luck? Collison, obviously the oldest rotation player, may be the only one who has missed consecutive games because of injury.

CrowL Any studies out there on age relating to injury?

@Heysloth
There was never a chance durant was gonna leave after next season when his rookie contract was over, worst case he doesnt sign and is a restricted free agent, he then accepts the qualifying offer for 1 year and then becomes an unrestricted free agent. So that has him here through 2011-2012. The qualifying offer would be only 7MM about half a max offer, so more then likely he would sign a deal that offers more money/years then the qualifying offer either from us or another team when he is RFA and then we can match. he could pull a lebron and only sign a 3 year max level deal and become a free agent earlier, but that still puts him here for 4 more seasons.

I told a few buddies that I thought we could win 42 games and miss the playoffs by a few games, and make it next year. I was basically told I know nothing about sports and there is no way the Thunder make the playoffs on the next two years, and I should have fun warching Durant because he is gonna leave as soon as his contract is up. I was basing most of that on out .40 win percentage and I am so glad I underestimated how much improvement everyone would make, especially Westbrook. Durant is amazing but I still had my doubts that Jet 0 could put it together so fast, he has been amazing lately

david :
The big issue is that most players do not have Shane Battier’s intelligence and willingness to utilize all information so it is not clear how specific the coaches instructions can be.

I had only thought along those lines as it relates to the Thunder players, but I can see on a macro scale how this would be an issue.

I know with our guys, they're young and still gaining experience. It's asking too much of Ibaka to get his defensive rotations correct, so you tell him to do 4 things and 4 things only. I'm sure that is also true for each member of the team.

It would be nice to say, as a staff, that "we go over on high ball screens by default, because the stats tell us that is the best play for most matchups". Then, when you get to the playoffs, you start going under on those high ball screens daring Kidd to shoot that long jump shot, because the stats tell you that is your best bet.

I bet a guy like Nick Collison is at a Battier-level intelligence right now. To see him anticipate a charge is a thing of beauty. He will even anticipate that Ibaka's man will be in great position for a rebound, because he knows Ibaka will leave him for the block shot attempt, and Nick will change his approach accordingly. Not that you would track that type of activity with a stat, I'm just merely pointing out Nick's on-court intelligence and awareness.

I'm sure the line between basketball IQ/intuition and advanced scouting and proprietary stats would be really, really, really thin/blurred.

Completely agree that it would be great to integrate stats more completely. However, one thing that was clear from the conference was that even if you have a coach who is willing to adapt his schemes to reflect stats (for instanced minimizing corner 3s or pushing players toward low percentage shots) it is not clear how much information you want to give to players given that basketball is an instinctive game. This view was echoed by multiple GM's and coaches.

The big issue is that most players do not have Shane Battier's intelligence and willingness to utilize all information so it is not clear how specific the coaches instructions can be. Apparently many players in the league routinely forget what play was called even if the call happened 10 seconds previous. That said, Brent Barry mentioned that when he played for the Rockets, there was a binder of info about guys tendencies and how to best attack them as well as what the best sequence of plays was to run against different sets of players. I didn't get the sense though that he spent a lot of time looking at binder. And even if he did, how much game specific info could he process quickly enough to be effective. The problem with the NBA is that there is very little practice time during most weeks (unlike the NFL) so this type info is difficult to internalize. Now in the playoffs or division rivals, that's a different story

@david

I read your previous comments, and am jealous that you were able to attend. I work with stats some, though the majority of my job is more related to data analysis. The one thing in my job that I don't get to do is use data analysis to drive decisions, which would be interesting, and really the main interest I have in proprietary stats.

In a perfect world, advanced stats, or proprietary stats, should be driving basketball theory. In a way they already do. Battier wants to force Kobe to his left, for a 15 foot jump shot. Why? Because the advanced stats say that shot is Kobe's least efficient. But the stats could go further than a one-on-one matchup.

For instance, a scenario that fascinates me is the high ball screen. This simple 4 person play can determine 3 different offensive plays; a drive to the hoop, a pick and roll, or a pick and pop. If each offensive option had an expected value associated with it, then that makes the defense easier to choose. Depending on the matchup a team could know to go under the screen, switch or go over the screen, all based on risk management, or least expected value.

@Jax Raging Bile Duct

As I mentioned in a earlier thread, I was at the MIT Sloan conference and the Thunder did send a representative - Basketball Operations assistant Jesse Weinstein-Gould.

http://www.linkedin.com/pub/jesse-weinstein-gould/4/538/459

I doubt he would tell you how the Thunder use advance stats specifically, but perhaps he would talk about it more generally.

Regardless, Dean Oliver, Denver's stat's guru mentioned that the Thunder were one of 7/8 teams which had fully integrated advanced stats into all portions of their decision making processes.

Everyone (all the stat people) seem to agree that eventually we are going to have info on where every player is on the court for every game - similar to what they have in baseball and in the English Premier league, which has the potential to revolutionize defensive statistics.

The big problem with basketball stats (relative to baseball) is that most of the relevant things cannot be found in a box score. Teams pay people to watch every game and code up their own box score measuring things like deflections which conventional box scores miss. Unfortunately, since each team invests resources to do this, this data is never gonna be publicly available unless the NBA starts collecting more data for all of the teams

I think I was writing that I expected like 35 wins also. My bad.

I called 40. People laughed. Who's laughing now? Evidence: http://bit.ly/9cFUiq

Back to that expectations thing, I have often thought about that Kevin Durant quote from his bball camp last summer. I was actually in there with my daughter(who got her picture in the DOkie w/ Mr. Durant as well, I might brag...) at camp that day, and the look on his face was serious as the proverbial heart-attack when he said, "absolutely the playoffs are a possibility, that is what we expect, and anything else is a failure."

Seems a motivated superstar trumps expert analysis!

Durants best chance of having a 50/40/90 year would be if we had a dominant post presence that demands a double team, then Kd would get more open looks and increase his FG and 3pt %

@justin
only 7 players have ever done 50/40/90 and played more then 500 mins. Dirk,Bird,nash, mark price, jose calderon, reggie miller, steve kerr.

Nash and Bird are the only ones with multiple seasons.

KDs current pace will have him be the only player to ever have 47/37/87 and shoot more then 1200 FGA, 650FTA, 250 3PA

@Mark!

I should also mention that while KD only needs 27.0 PPG to average out to at least 29 PPG on the season, to hit those %'s, he'll need to average something like 34.6 PPG to close out the year. That's a lot.

Going back to the "hey, look at how the Thunder are blowing pre-season expectations out of the water" thing, I just thought that we only need 3 wins the rest of the season to guarantee a .500+ finish. I know plenty of people who would have paid serious money at the thought of a .500 finish this season. And yet the Thunder show no signs of slowing down, so we should be well and truly above that. Makes me very happy. :D

I should mention my % numbers were based on the assumption that KD maintains the same number of attempts per game as he has during the first 62. I wouldn't put my money on it happening, but he really is egregiously close.

Dirk did it once, and he needed a perfect storm type of year to do it. He only shot over 50% once in his career, 40%3PT twice, and 90%FT twice, and they just happened to line up in 06-07. Otherwise, his %'s look an awful lot like KD's. So I guess a similar type of thing could happen to Durant.

In moden history (i.e. excluding Dolph Schayes), Reggie Miller holds the record for most FTA in a season while shooting 90%+ from the line, he shot exactly 600FTA in the 90-91 season. And Miller could be argued as the best FT shooter of all time.

It's actually a bit of a catch-22. For Durant to shoot 50% from the field he'll have to get to the rim more often and rely less on the mid range shot.. which would lead to more FTA. Gives you respect for what Bird was doing. 50/40/90 is incredibly tough to do as a first option over the course of an entire year.

@justin
thats a great stat, being the best FT shooter for over 700 attempts all time, is a good accomplishment.

It's really tough to shoot 90% from the line when you go to the line as often as Kevin Durant does.

As it stands, he's on pace for 800FTA. If he keeps his current %, he'd have the highest FT% of anyone shooting 700FTA or more in a season in NBA history (barely beating Rick Barry). Larry Bird was shooting 400-500 FTA a season..

@Mark!
A couple of those numbers are very doable, I think. It's the percentages that I think make-or-break it. Though if KD wants to average a double-double from here on out the rest of the way, I have absolutely no problem with that.

That being said, I could totally see KD getting to "Basketball Jesus"-land within the next few years.

@Mully Mulls
if we did it right we could trade picks and end up with multiple top 5 picks, by then, say trade for 2 picks each year until 2018-19 and 20, against teams that are hopefully bad, and get a lot of good lottery picks

@Royce Young

Yeah, the tight-lipped reputation the Thunder has about it's business extends to their stat team.

I would love to develop a formula similar to the one used in determining the value of a blocked shot. The thesis is centered around expected value, which is something I think you could extend to other defensive principals, such as screen switching, or going under the high ball screen.

So close to the Basketball Jesus. For Durant to hit Simmons' prediction, he would need to average over the last 20 games...

27.0 PPG
9.4 RPG
57.3 FG%
47.0 3P%
94.8 FT%

Probably not going to happen, but I wouldn't put it outside the realm of possibility. It's absolutely ridiculous I can convince myself to think so. I love Durant.

@Jax Raging Bile Duct
Get after it.

I've been trying to lock down more on what the Thunder does in terms of advanced stats, but haven't really been able to. I find it pretty fascinating as well.

@Royce Young

I would like to chat it up with a few of our resident pro-stats commenters sometime about the TrueHoop posts over the weekend. I find them fascinating. Would really like to get my hands on more of the math than was presented in those articles, so if anyone here knows of some, let me know.

I hate to derail an article too far to talk about what I want to talk about, so maybe a quick article on it would help us out?

Maybe I'm the only one.

Best stat from last nights game. If the Thunder plays .500 ball the rest of the season then Houston (currently 9th and holding the tie breaker against OKC) would have to go 17-3 to make it over us.

Eeek!

@Royce Young
I can't wait! When I read that I was like... "So THAAAATS why he loves us so much...." I really can't believe that we have another 5-6 years of developing with these guys before they hit their prime... is it ridiculous to ask if we can just trade our draft picks out until 2018?

@Mully Mulls
Yep, I've got a column coming on that soon.

Wow! It's eerie how spot on The Sports Guy is. I bought his book and I think that we have learned "The Secret." If you don't know what I'm talking about, buy his book and enjoy being enlightened for the rest of your life... it's that good!

On another note... "Great Scott!" I DID predict a 42-40 season, and I can't wait to say that I was wrong sometime next week.

@Sas

Absolutely.

Warriors have had a ton of injuries too..

@f5alcon
I'm guessing the overestimation on the Clips part might have something to do with someone's broken patella.

Hey, if Shawn Marion is the Matrix, could Durant be the Nexus?

this blog predicted 45 wins http://www.hoopangle.com/thunder/2009/08/05/2009-10-season-schedule-preview-predictions/

@Royce
I like the "Next is Now!" slogan. If you think about KD being ESPN's NEXT athlete, it's particularly applicable.

Clark Matthews :
That’s probably the first time I’ve ever been wrong about anything. Ever.

Same here. It's a weird feeling.

cool i was hoping you would do an article like this, the experts were way off in most cases, only LA in the western conference looks to be right, they underestimated nuggets and mavs, overestimated clippers and warriors

Someone on the realgm site predicted 41-41 I think. I haven't been able to find any predictions by a professional sports writer that had the Thunder in the playoffs.

@Clark Matthews
And happy to be so?