December 6, 2009

Musings on the Pacers, Nuggets, and Blazers

Filed under: Uncategorized — wwinston @ 4:46 pm

Let’s try and answer the following three interesting questions:

  1. Why are the Pacers playing so badly?
  2. Should Melo have been Western Conference Player of the Month?
  3. How badly will Greg Oden’s tragic injury hurt the Blazers?

Why are the Pacers so Bad?

In the last 2 weeks the Pacers have a 1-7 record. By looking at how the Pacers play with different players on the court it is easy to determine the major causes of the problem. After adjusting for who they play with and against we find that during the last two weeks Hibbert, Murphy and Hansborough have  played more than 20 points worse than average (per 48 minutes). I estimate that during their playing time these players have cost the Pacers nearly 30 points per game over the last 2 weeks. So what to do? Either play these guys less minutes or figure out what they are doing wrong.

Melo should not have been Player of the Month!

   Melo is averaging over 30 points a game, so he must be having a great season. Not so fast! After adjusting for who Melo plays with and against we find that per minute played Nene and Andersen have contributed more to the Nuggets success than Melo. In what follows we tell you how well the Nuggets have played (per 48 minutes) for various frontcourt combos.

  • Martin Andersen 36 min +8
  • Melo and Andersen 25 min +6
  • Melo Martin Andersen 148 min +9
  • Melo Martin and Nene 410 min +5
  • Melo Nene and Andersen 60 min +19
  • Nene and Andersen 96 min +20
  • Nene and Melo 50 min -17 pts
  • Nene and Martin 12 min +21
  • Melo Martin 13 min -45

I think these numbers make it clear that Nene and Andersen really key the Nuggets success. In particular with Andersen and Nene in (Martin and Melo) out the Nuggets play 20 points better than average.

How will the Blazers Fare without Oden?

   Again we are so sorry to see Greg Oden suffer another season ending knee injury. The Blazers should be ok without him, however. The key to the Blazers success are the minutes with Blake, Miller and Roy on the court. Whether Oden is in or out the Blazers play 21 points better than average with these three guys in. The rest of the time the Blazers play only one point better than average.


  1. First of all, let me thank you for maintaining this great blog and putting so many interesting and relevant statistical tidbits in (relatively) layman’s terms. I think you’re making a lot of people smarter, better fans.

    I’m really interested in the notion of what adjusted +/- says about teammates’ performance together, like what you’re talking about with the Blazers. For example, I would think that the success of Blake+Miller+Roy is mostly due to the fact that it’s an unconventional 3-guard combo that serves as a “change up” to a more straight-forward line-up and may, therefore, catch opponents relatively off-guard in the limited time it is used. However, if the Blazers played those three guys together more often (which, I suppose, is what you’d advise them to do), opponents would focus much more on exploiting them defensively, or alter their own on-court personnel to better match up with them, which would probably reduce the 3-guard line-up’s effectiveness.

    I’d love to hear your thoughts.

    Comment by HoopsFan — December 7, 2009 @ 12:08 am

  2. ouch..someone needs to show brandon roy this.
    something was up when they ended their losing funk with those 3 in the starting lineup and broke it up when ‘franchise’ was not having any of it

    Comment by Ray — December 7, 2009 @ 4:34 am

  3. Thanks for kind words. Yo uare right of course, that playing these guys more would make defenses more prepared for them. But these 3 should surely be on court at end of game. I do not know if they are or not.

    Comment by wwinston — December 7, 2009 @ 9:07 am

  4. I had a similar thought to HoopsFan when I first read this. This is borderline impossible I imagine, but can adjusted +/- control for the types of player a unit is against? What I’m trying to get at is, it’s very likely McMillan would use their 3 guard set vs. opposing teams that had a moderately small Guard/SF combo (like a Curry/Ellis/not huge SF combo) where he could get away with Roy defending SFs and Blake on SGs, but probably did not give them any playing time against a giant back court and small forward (something like a Tyreke Evans/Kevin Martin/X combo). The Warriors combo could certainly be more productive in the long run, but it’s likely the lineup the Blazers combo would be more effective against. Certainly there is some selection bias as to when a coach uses a lineup or player that could be in play with +/- numbers.

    Comment by Ray — December 7, 2009 @ 5:36 pm

  5. we could check whether this combo does better when other team is big or small. I just do not have time right now. good idea.

    Comment by wwinston — December 7, 2009 @ 8:10 pm

  6. Wayne,

    1. Fundamentally, I’m in agreement with the basic premise of your work. :-)

    2. However, if you’re going to advise others that a Blazers’ line-up of Blake+Miller+Roy has had success this season [i.e. 20 points above average, according to you] then it’s imperative to complete the rest of equation properly.

    By this, I mean:

    Q1. Which sets of 2 teammates have been partnered with this three-some?

    Q2. What sets of 5 opponent players have this three-some [and its concomitant five-some] been matched-up against when they’ve appeared together this season for the Blazers?

    At its core, basketball is a game of highly specific player match-ups and, IMO, it renders otherwise worthwhile information like yours relatively inconsequential if/when it appears without also including the details of A1. and A2.

    Looking forward to you providing this specific information so that I may further analyze the goings-on when the combination of Blake+Miller+Roy has been used together this season for Portland.


    Comment by khandor — December 9, 2009 @ 6:36 pm

  7. Doesn’t the success of the Nene/Andersen combination run counter to your previous theory that a team shouldn’t have 2 non-shooters on the court at the same time to be successful? Or is one of these guys a much better shooter than I’m giving him credit for?

    Comment by Malcolm — December 9, 2009 @ 11:04 pm

  8. I would agree. No theory will work 100% of the time.

    Comment by wwinston — December 10, 2009 @ 12:37 am

  9. Here are the lineups: Like 1st lineup played 18.58 pts better than average.
    POR 18.58 ( 21.12 6.34 -14.78 48.43 1.90 10.49) 134.09 minutes 54 appearances 1 A 13.60 38 $
    Aldridge Blake Miller Oden Roy 27.09 years 8841_POR_2009

    POR 13.30 ( 18.59 0.72 -17.87 50.29 1.80 10.76) 64.42 minutes 35 appearances 3 A 7.45 10 $
    Aldridge Blake Miller Przybilla Roy 28.75 years
    POR 29.41 ( 28.27 18.23 -10.04 71.11 2.59 9.35) 10.32 minutes 8 appearances 6 C 27.92 6 $
    Blake Miller Oden Outlaw Roy 27.26 years 9864_POR_2009

    68.78 ( 24.83 4.54 -20.28 91.99 2.76 9.79) 6.80 minutes 2 appearances 15 C 56.47 8 $
    Blake Miller Oden Roy Webster 26.82 years 25224_POR_2009

    Comment by wwinston — December 10, 2009 @ 12:44 am

  10. Wayne,

    Thanks for providing the front-court tandems that the Blazers have used, thus far, in conjunction with the Miller+Blake+Roy combination.

    If I’m reading your values correctly, it would seem to indicate the following information:

    [please correct if these are wrong on my part]

    PAIR 1
    Aldridge+Oden, 134.09 minutes, 54 appearances, +18.58

    PAIR 2
    Aldridge+Przybilla, 64.42 minutes, 35 appearances, +13.30

    PAIR 3
    Outlaw+Oden, 10.32 minutes, 8 appearances 6, +29.41

    PAIR 4
    Webster+Oden, 6.80 minutes, 2 appearances, +68.78

    If you could also provide this same information for the opponent players who were on the court against each of these separate PAIRS [i.e. vs PAIR1; vs PAIR 2; vs PAIR 3; and, vs PAIR 4], it would go a long way towards being able to do a further analysis of what was actually happening when the Blazers elected to play with this specific threesome at their PG, OG and SF positions.

    Thanks, in advance.

    Comment by khandor — December 17, 2009 @ 4:13 pm

  11. Wayne,

    I should also add the following …

    Since fulfilling my request would entail the detailing of upwards of 99 separate line-up appearances for the opponent’s different 5-man units used against the Blazers when Portland has had Miller, Blake and Roy on the floor together, I would certainly understand if it’s not possible for you to provide that information in this setting.

    If it isn’t, then, perhaps, you could simply point me to a different location where this specific information might be found.

    Thanks, in advance.

    Comment by khandor — December 17, 2009 @ 4:23 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

Powered by WordPress