Friday, February 09, 2007

Digging Deeper: Blazer Lineup Statistics

Thanks to the good people over at 82games.com, we can look at a variety of non-traditional NBA statistics. Today, we’ll focus on Portland’s most effective 5-man lineups; who’s played the most, who’s been most effective, etc.

Portland has four sets of five-man lineups that have played more than any others. They are as follows:

1. Jack-Roy-Udoka-Randolph-Pryziblla. (304 total minutes have been played with this lineup)
2. Jack-Udoka-Webster-Randolph-Pryzbilla. (138 minutes)
3. Jack-Roy-Udoka-Randolph-Aldridge. (83 minutes)
4. Jack-Udoka-Webster-Randolph-Aldridge. (77 minutes)

Following those four, you then have 9 more lineups that have played between 42-49 minutes. Before we dive too deep, let’s ask a question: has the number of lineups that Porltand has used been abnormally large? Let’s use 40 minutes as a cutoff point. Portland has used 13 total lineups that total 40 minutes or more together. How does that compare to other Western Conference teams?

Let’s look at the top four teams in the West:

Dallas: 8 lineups
Phoenix: 13 lineups
San Antonio: 12 lineups
Lakers: 10 lineups
(average = 10.75)

And now, the bottom four teams (excluding Portland):

Memphis: 10 lineups
Seattle: 11 lineups
Sacramento: 10 lineups
New Orleans: 12 lineups
(average = 10.75)

What does that mean? Absolultey nothing. Which was the point. Good teams don't necessarily play a fewer amount of lineups together, despite what many hack sportswriters would like you to think. So, while Portland may be a touch on the high side, they certainly aren’t grossly so. And considering that they’ve suffered more injuries than most teams to-date this season, that’s to be expected.

Now that we’ve got that out of the way, and we know Portland’s top-5 most played lineups, which are the most effective? For one thing, let’s define “effective”. We can look at the top-5 lineups by total overall +/-, but that’s partially a counting stat in that two equally effective lineups will have different overall +/- depending on total time played. Hence, I propose we measure effectiveness by “win%”. Meaning, how frequently did that lineup outscore their opponent during the segments they were on the floor?

Here are Portland’s top 5-man lineups, their W-L and %, followed by total minutes played and rank. We’re only considering lineups that have played a total of 40 minutes or more together, so the rank is out of 13. (Bear in mind that Portland is playing at a .400 winning % as a team right now, so that’s the mid-point.)

1. Rodriguez-Dixon-Webster-LaFrentz-Magloire, 5-2, 71%, 42 minutes, #13
2. Jack-Udoka-Webster-Randolph-Pryzbilla, 7-3, 70%, 138 min, #2
3. Jack-Udoka-Webster-Randolph-Magloire, 4-2, 67%, 49 min, #6
4. (tie) Jack-Dixon-Udoka-Randolph-Magloire, 6-6, 45 min, #10
Jack-Dixon-Webster-Outlaw-Magloire, 6-6, 50%, 43 min, #11
5. Jack-Roy-Udoka-Randolph-Aldridge, 5-7, 42%, 83 min, #3
6. Jack-Roy-Udoka-Randolph-Pryzbilla, 8-12, 40%, 304 min, #1

I extended it out to #6 so you could see where the top lineup ranked.

Interesting. We should keep in mind one key thing: the starters, or #’s 2, 5 and 6 in the above rankings, always play against the other team’s best players, so they’re starting out at an immediate handicap. With that in mind, several items jump out:

-the lineup that Nate has used is playing right on par with the team’s win%. Not that that should be a shock to anyone, but worth noting: as those guys go, so goes the team.
-the small sample size of the top-rated lineup must be taken into consideration. No doubt they’d regress back a bit with more playing time. That lineup has played fewer than 3% of the season’s minutes.
-I’m shocked to see Martell Webster both a part of two of the top 4 lineups played and a part of 3 4 of the top 5 most effective lineups.
-LaMarcus Aldridge is not a part of a single lineup that has a 50% or better win%. Surprising.
-Nate’s getting pretty close with his second-most-used lineup also being the second-most-effective.
-Brandon Roy is nowhere to be seen; not in a single +50 lineup. One might argue “but he plays with the starters, and hence plays against the starters.” Then how do you explain Zach Randolp and Ime Udoka’s success?

In case you’re wondering, here is the rank of the individual players for how frequently the team has “won” its segments of the floor when that player plays. What follows is the rank, player name, % of minutes played, and Win%. We’re only going to pay attention to players who have participated in 15% or more of the team’s minutes.

1. Rodriguez, 19%, .500
2. Webster, 43%, .490
3. Pryzbilla, 25%, .429
4. Randolph, 71%, .408
5. Dixon, 47%, .404
6. Aldridge, 31%, .395
7. Outlaw, 34%, ..382
8. Udoka, 59%, .341
9. Magloire, 37%, .340
10. Jack, 67%, .333
11. Roy, 41%, .310

Wow. More interesting items. If you were to look strictly at these numbers, with no other factors considered, you’d come to the conclusion that Udoka, Jack, and Roy play far too much, and that Rodriguez, Webster, Pryzbilla, Dixon and Aldridge should be playing much more. I’m not saying this is right, I’m just saying that’s what these particular numbers are indicating.

I’ll take it one step further, and say if you were to look at these +/- numbers, then your starting lineup should be: Rodriguez, Dixon, Webster, Randlph, Pryzbilla, with your first guys off the benc as Aldridge, Outlaw and Udoka.

Fascinating stuff. Again, we aren’t saying this is right, but chew on this for a couple of days – I’ll be back next week with some further analysis.

Enjoy for now.

8 comments:

Anonymous said...

HOLY CRAP THAT IS SOME FASCINATING STUFF, MODERATOR. AND WHEN WILL WE LEARN YOUR NAME??

AMAZING THAT ROY DOESN'T HAVE BETTER +/- NUMBERS IT WILL BE INTERESTING TO SEE HOW HE ENDS UP BY THE END OF THE YEAR.

Anonymous said...

It is remarkable how small are the total minutes of the most used lineups for most teams. Detroit last season was the exception with the highest% of total minutes in the main 5 man lineup or the few main derivations of it.

Starting lineups matter far less than people think given these results. Rotation management is very important and usually needs more careful study.

Anonymous said...

I rеad this post cοmpletely сοncerning
the diffегence of mοst uρ-to-date аnd previous technolоgiеs, it's remarkable article. garcinia cambogia - garcinia cambogia - garcinia cambogia garcinia cambogia - gnc garcinia cambogia weight loss - garcinia cambogia

Anonymous said...

Ιt's a pity you don't hаve a donаte button!
I'd certainly donate to this brilliant blog! I suppose for now i'll settle for bookmarking and adding your RSS
feеԁ to my Google аccount. I lοoκ forwarԁ to brand
new upԁateѕ and will talk about this sitе with mу
Faсеbook group. Talk sοon!


Μy web-site: car rental

Anonymous said...

Hеllо mates, how is the whοle thing,
and what you desiгe to ѕaу on thе tοpic of this post, in my
view itѕ actuаlly amazіng designеd for me.


Also visit my web site: coffee pure cleanse walmart

Anonymous said...

Usually I do not гead article on blogs, but I ωould liκe tο say thаt thiѕ wrіte-up
vегу pгeѕsured me to
try аnd dο ѕo! Your writing
taste haѕ beеn аmazеd me. Thanks, quіte nice
aгticle.

Have a loοk at my ρage :: online quote form

Anonymous said...

Prеtty nice ρost. I just stumbled
upοn your blog anԁ wished tο ѕаy that
I've really enjoyed surfing around your weblog posts. In any case I'll be ѕubscribing іn yоuг rss fеed аnd I'm hoping you write once more very soon!

Check out my web blog :: sportsbet ()

Anonymous said...

Hi, I woulԁ like to subscгibе fоr this webpagе to
οbtaіn most recent updates, ѕo ωherе
can і do it ρleаse hеlp out.


Also ѵisit mу ωeb blog raspberry ketones