Dean Oliver is a genius

To get FG% up they either need to

1) get more inside points by getting a postup guy, or increasing and improving the effectiveness of the motion or by creating more putback baskets.

2) get more good percentage 3 point shots.

or 3) run more

A postup guy with better / more active 3 point shooters and more motion (with the possible exception of the rebound hound) could yield strong reinforcing results.

More quick 3s, fewer turnovers. And more offensive rebounds than from 2 pt attempts because more misses > more rebound opportunities. And that can lead to inside scoring.

The postup guy is a hard thing to find these days. All the more reason to amp up the 3 ball game.

If you can increase the pick n roll yield or get more cuts for scores go for it. They don't seem that able at it though.

Winning FG% has the highest correlation with winning.

Thunder may win the contest of own FG% vs opponent FG% against bad or average teams but it is unlikely against good teams.

Being 22nd on eFG% right now is an issue when looking at roster construction and playing time allocation.

Thunder 23rd on assists. Again a roster construction issue. Very little assists from all the bigs and below average from the wings. The PGs can't do the whole job.

Several major teams seem to have a strategy based on high 3pt makes and high offensive rebounds- Houston, Portland and the Lakers are top 10 on both. I like that strategy.

@Jax Raging Bile Duct
I think the reason you do a binary on the OREB Rate is because winning is a binary value as well. I think stats aren't legitimate if you use a continuous variable as the independent and a binary value as the dependent.

The best measure would probably look at the correlation between OREB Rate and point differential, then computing the r-squared of each variable to determine how much each variable contributes to winning. You could probably do a multiple regression analysis from there to get the interactions from the various factors and predict a point differential based on a set of inputs.

I'm going to see if I can do something like this but I only have the data to do it for last season.

@Jax Raging Bile Duct

I would like that same kind of chart, only broken down by OREB%. I think a small advantage in OREB% is probably close to meaningless and it skews the statistics up there; when we're experiencing a 10%+ difference in OREB% I think the correlaiton to winning or losing should be stronger than what's represented up there. The vast majority of samples in Dean Oliver's data probably represent negligible advantages..

To be clear, I would like to see a chart like this:

Category __________________ Win %
---------------------------------
W Offensive rbds. _________ .461
+2 Offensive rbds. _______ .???
+4 Offensive rbds. _______ .???
+6 Offensive rbds. _______ .???
+8 Offensive rbds. _______ .???
+10 Offensive rbds. ______ .???
>10 Offensive rbds. ______ .???

I think this all illustrates the 15 losses we have, or at least a lot of them, we shot worse and didnt make up for it with better offensive rebounding or less turnovers, most of our close games could have been one if we improved slightly on offesive rebounding margin or turnover margin

This is fun stuff. I don't think that it can be argued that differences in FGA between two teams is a major factor in winning. FGA within a game context are entirely connected to rebounding and turnovers. If you win those battles, it goes a long long way. Offensive rebounding and forcing turnovers, especially, seem to be almost the same in value to a team. Both give the ball back to your team, tend to lead to much higher percentage shots, and have momentum effects on the opposition. It's great that we (in general) get about the same offensive rebounds as the opposition, but that's only one half of the FGA issue (which I think is the more dominant connection to make). We are 7th from last in turnover rate. So, even if we grab an equal percentage (not number) of offensive rebounds, we are almost invariably handicapped by giving the ball away somewhere else. In fact, opponents average 1.5 more field goals than the Thunder every game.

Obviously, a great defense can take away that 1.5 shot advantage, but only if we have an adept offense to hurt an opponent when they don't shoot well. What I'm seeing, essentially, is that the Thunder tend to fail at the limiting factors of those things which they do well. They crash the offensive boards but can't get them on the defensive end, thus a wash in rebounding. They hold the opponent to a lower percentage than themselves, but they the opposition more possessions. They shoot slightly more free throws than the opposition, but actually foul a bit more as well. We are 4 games above .500, but a lot of our negative stats wash out our positive ones.

Ultimately, we can look at a lot of things that the Thunder do both well and poorly. Until we start winning on both sides of coins (high offensive AND defensive rebounds, low opponent FG% AND low team turnovers) we will be just good enough to win with better talent, but just bad enough to lose to inferior talent

@justin

Excellent point. Game to game It would shake out even I guess right?

To follow up Bryan's point, it seems far more important to close out on the perimeter, hopefully lowering opp FG% than it is to be in position for the rebound.

I would like to see an average margin of offensive rebounding difference and how that affects the win/loss %. I would imagine that there is a special number of offensive boards, a median score maybe, where if your offensive rebound advantage was, say a net +5, you win, but less than that and you lose.

Oliver's chart treats the category like a binary value, and I don't think it is.

@J.G.
Look at Oliver's stats. The second weskest correlation to wins is offensive rebounding%. So if we're giving up a high percentage of offensive rebounds, it's still not that big a deal. Defensive rebounds are highly correlated to win rates only because, as a total stat, they are correlated to better defensive FG% (more opponent missed shots = more rebounds). Defensive rebounding RATE is not that well correlated with winning. So don't be too worried about the defensive rebounding rate (at least according to the data).

But then again, what do I know. I still refuse to save hundreds of dollars by switching to gieco.

And whenever the Thunder get sluggish closing out on shooters, that's when they start to fall apart.

Although, if you are not closing out on the perimeter, you are more likely to be in position for the rebound if they miss. Could have something to do with the numbers as well since the Thunder consistently contest every shot.... which puts even more pressure on their interior players to get the boards.

f5alcon :
our Offensive rebound rate is 0.11 so we get and give up Orebs at about the same rate

Right, so for the Thunder, wins depend (even more so than most) upon shooting percentages. Thankfully, they hold teams to low percentages, which helps to compensate for inconsistent shooting.

our Offensive rebound rate is 0.11 so we get and give up Orebs at about the same rate

Joe :@J.G.I’m not sure I understand your question J.G., but for references sake, we are seventh best at offensive rebound rate–at getting the most of our available offensive rebound opportunities as expressed by percentage at 28.2%.
But we are 7th worst at defensive rebound rate (or, not letting the other team get the offensive rebound) at 71.9%.
Some of it is style of play, some of it is personnel.

That's exactly what I was saying, Joe. To me the most telling stat is not the Thunder's ability to grab offensive rebounds, but their inability to stop the other team from grabbing offensive rebounds, which is another way of saying that they're AWFUL in defensive rebounding rates (just forgot of the wording of the stat).

Style of play and personnel for sure (shot selection I would definitely highlight under style of play as being "young"), but in crunch time and into the playoffs, that is the stat that hurts the most and Oliver's statistics back it up. That's, to me, what has to be addressed.

anecdotal

Joe :@J.G
.
I’m not sure I understand your question J.G., but for references sake, we are seventh best at offensive rebound rate–at getting the most of our available offensive rebound opportunities as expressed by percentage at 28.2%.
But we are 7th worst at defensive rebound rate (or, not letting the other team get the offensive rebound) at 71.9%.
Some of it is style of play, some of it is personnel.

Yeah, my guess is that it's because the Thunder's guards are aggressive on the offensive glass - especially Westbrook. Durant is also better than most SF at gathering up a put-back.

@J.G.
I'm not sure I understand your question J.G., but for references sake, we are seventh best at offensive rebound rate--at getting the most of our available offensive rebound opportunities as expressed by percentage at 28.2%.

But we are 7th worst at defensive rebound rate (or, not letting the other team get the offensive rebound) at 71.9%.

Some of it is style of play, some of it is personnel.

@Joe

This kind of proves the point though. The Suns lead the league in FGA, but they actually gave up an average of over 3 FGA more to the other team because they turned the ball over more and were a poor offensive rebounding team, maybe? The Lakers, meanwhile, had 2 FGA more than their opponents because they turned the ball over less and secured offensive rebounds. The Warriors also allowed more FGA.

Being a fast team will lead to having more FGA over the course of the season but the effect that it has on FGA in a given game doesn't seem like it'd be significant to me..

@f5alcon

I think DJ might be one of those players that plays well vs scrubs and can't bring the same game vs regulars. His defense before being shipped out was horrible!! and unless he magically found his game (also says in the article he has a post game) I bet when he gets called up he plays about the same.

@justin
I think you're right. I wasn't thinking it through properly when I originally made that point. Speed/pace shouldnt have an impact on net FGA's. That should just be dictated by TOs, ORebs, and (to a lesser extent) FTAs.

Obviously, a faster team will be more FGA's over the season, but the net per game between the two teams shouldn't be affected.

@dork

Yeah i meant the thunder girls, and i dont care of what they are wearing to the party.

In the article it says Dj whites defense has been good

justin :@Joe

This makes no sense to me maybe I’m stupid. So what if D’antoni’s team takes a shot in 7 seconds and the other team takes one in 20 seconds, they still can’t get another shot attempt until they regain possession of the ball, correct?
In a vaccuum, turnovers and everything else being equal, how fast you take a shot doesn’t seem to me like it’d affect your FGA relative to the other team in that particular game. It’s like, if you’re playing chess and you take 5 seconds to act and the other guy takes 5 minutes, you still get the same number of moves.

Well strictly speaking you are absolutely correct. If you use the chess analogy, one move at a time. I was thinking more in terms of the whole game concept since the game doesn't happen in a vacuum. The Suns for example would likely get an offensive rebound and immediately put another shot up, whereas the Bobcats would pull it out after the Oreb and work for a good shot. That, plus turns and all that multiply over a season.

For example, last year the top five teams in FGA's were the Suns, Lakers, Warriors, Pacers and Knicks; not coincidentally the 5 fastest teams were the Warriors, Knicks, Pacers, Suns and the Nuggets, followed by the Lakers.

I see your point though. You aren't wrong.

Here's my issue, since the Thunder are giving up a horrific amount of offensive rebounds to the other team, isn't that a better indicator of their defensive rebounding ability than raw #'s in defensive rebounds alone?

And since defensive rebounds are the third biggest indicator of winning percentages, shouldn't that be a huge concern?

Isn't there a statistic that shows defensive rebounds snagged out of total potential defensive rebounds? Or in other words, defensive rebounds they've grabbed versus how many they've lost to the other team. Isn't that the best measuring stick?

Bryan :
I wish the book’s author had thrown shot attempts in the mix. I would like to see how that might change the analysis.
But nice piece, Joe.

Well Oliver used two sets of data. One was from a company called STATS, where the games from 92-94 were used. In that data raw FGA's came in at 45.5%, which is just up from last, just above 3fga's. Oliver took the idea and ran his own numbers, which are the ones I put in the table, from a more current time frame and changed some of the categories. FGA wasn't one of the ones he used. But his data and the old STAT data are very similar so it is fair to say that FGA's don't correlate strongly with wins.

f5alcon :@dork
lol, maybe they need to come to our watch party

You ARE talking about the Thunder girls and not just the bikini's right? Cuz I don't put it past Royce to show up in one o.O

justin :@Vega
DJ White = Future HOF.

...in the D league. Until he learns defense he has no business being in a thunder uniform imho.

@Joe

This makes no sense to me maybe I'm stupid. So what if D'antoni's team takes a shot in 7 seconds and the other team takes one in 20 seconds, they still can't get another shot attempt until they regain possession of the ball, correct?

In a vaccuum, turnovers and everything else being equal, how fast you take a shot doesn't seem to me like it'd affect your FGA relative to the other team in that particular game. It's like, if you're playing chess and you take 5 seconds to act and the other guy takes 5 minutes, you still get the same number of moves.

justin :
@Mark!
Being a volume scoring team doesn’t give you more attempts than the other team in a given game.. possessions are still relatively equal.

Well yes and no, it depends.By being a "volume scoring team" you mean fast, it can. If, like under D'Antoni, he preached getting a shot off in 7 seconds or less and the other teams usually use more of the shot clock, the faster team would get more shots. A faster team doesn't get more possessions, but it CAN theoretically get more shots.

@Vega
"Defensively, White is extremely vocal, makes all the rotations, and definitely still brings the defensive intensity he showed as a senior at Indiana"

"He also showed prowess in the post"

@za
I wonder what the number of added possessions was, thats what we need to look at, and maybe the conversion % or the 2nd chance points difference.

@Vega

DJ White = Future HOF.

Mark! :@justin

I doubt FGA’s have a very strong correlation to wins, but instead of looking at an offensive rebound as leading to an additional attempt, you could arguably look at it as -removing- a FGA from the FG% calc, thus increasing the efficiency of the other team.

Agreed. And if the offensive rebounds come in the same possession, as when we allow three or four shots on the same possession, the effect is skewed because several rebounds only hurt you once. Obviously, that's not ideal if you need the ball back at the end of the game though.

Draftexpress has an article on the D-League Showcase, and the main focus of the article is the "Dominant DJ White". Their words, not mine.

http://www.draftexpress.com/blog/D-League-Blog/#D-League-Showcase-Day-Two-3368

Off topic but David Stern has suspended Gilbert Arenas indefinintely without pay. Calls him "unfit" to play in the NBA now.

This would be incredible of Arenas' contract were voided.

@Mark!

Being a volume scoring team doesn't give you more attempts than the other team in a given game.. possessions are still relatively equal.

Small difference in offensive rebounding % are probably useless.. I'd be interested in knowing the average difference in offensive rebounding % in the games Dean Oliver used for that data and how, say, a 10%+ difference in offensive rebounding % affects wins and losses. I'd imagine the effect is much stronger than what's shown here..

Also just another note, when you use % the differences can be so minute. Look at the 'ties' for FG%, OREB%, or A:TO ratio, there are much fewer. Some of the wins and losses for those statistics are probably very small values and effectively a 'tie'.

Heh, dunno why it escaped me that the table is looking at net. Still, I doubt there would be a strong correlation to wins if your team won on attempts. I think all the teams that have been built around volume scoring (last Golden State this year) would muddy up the waters in the win/loss column.

FGA probably only matter if there is a significant number more, if one team teakes 5 more attempts it probably isnt game changing but if a team takes 10 or 15 more it probably is.

@Mark!

Agreed.. I just meant our offensive rebounding has been so poor other teams are getting huge FGA advantage over us. Very difficult to overcome unless they shoot like crap (Bulls).

I wish the book's author had thrown shot attempts in the mix. I would like to see how that might change the analysis.

But nice piece, Joe.

@Mark!
15 more FGA is potentially 45 more points, even if its only them making 33% and 2 pointers thats still 10 more points then they would have scored, FGAs do matter, we would have blown out chicago if it were not for their offensive rebs

Let me amend that by saying I'm sure NET FGA's do have a strong correlation, but FGA's on their own (as the other stats in the table) probably do not.

@justin

I doubt FGA's have a very strong correlation to wins, but instead of looking at an offensive rebound as leading to an additional attempt, you could arguably look at it as -removing- a FGA from the FG% calc, thus increasing the efficiency of the other team.

im just waiting for somebody to make the correlation between scoring the most points and winning games,lol

I would think there is definitely a lot of correlation between several of these stats (FTs made/FTs att, etc), but I would also assume the author knew what he was doing and took this into account when explaining what was most useful.

This stuff interests me, so I may have to take a look at this book.

@dork

lol, maybe they need to come to our watch party

@f5alcon

I support this as long as the stats are in really really fine print.

While offensive rebounding % is not as big a factor for winning games as shooting %, if you're giving up 15+ offensive rebounds in a game I think it's a bit more crippling than that table would indicate. Our last two opponents have had 99 FGA in regulation, 15+ more shot attempts.