February 8, 2010

Can the Celtics Regroup?

Filed under: Uncategorized — wwinston @ 8:51 am

Since Christmas the Boston Celtics have played like an average NBA team (.65 points better than average). This will not satisfy fans of the sports’ most storied franchise. Is there any hope for the Celtics to recapture their 2007-2008 level of greatness?

     The first point to make is that KG in his current level of health is not having a big impact. Since Christmas with Pierce, Rondo, Ray Allen, and KG on the court the Celtics have played 6 points better than average. When Pierce, Rondo and Ray are on court with KG out the Celtics have played around the same: 5 points better than average. I am not a doctor, but KG 2010 is not playing at his past level. His Adjusted +/- this year is -3. During his last 5 seasons his Adjusted +/- averaged out to +11. Unless KG can regain his previous abilities, it will be tough for the Celtics to make the conference finals.

  Still, there are ways for the Celtics to improve. Let’s partition all Celtic minutes played since Christmas based on which of Pierce, Rondo, and Ray Allen are on the court and see how things went:

  • All 3 in : 5 points better than average.
  • Pierce and Rondo in, Ray out- 20 points better than average.
  • Pierce and Allen in, Rondo out- 9 points better than average
  • Rondo and Allen in Pierce out-2 points worse than average
  • At most one of Rondo, Pierce and  Allen in-10 points worse than average.

From these numbers it looks like when at least two of Pierce, Rondo, and Ray Allen are in the Celtics are ok. Having all 3 in does not make them that much better. Therefore, it looks like the Celtics would improve if they would rest Pierce, Rondo, and Ray Allen one at a time more often and lessen the time two or more of them are out.

For example, Pierce can be rested with

BOS      10.85 (    7.74   -0.11   -7.85    3.01    0.62    6.52)  110.14 minutes   50 appearances     3 A     14.38       33 $
T. Allen      R. Allen      Garnett       Perkins       Rondo         29.01 years       1299

This lineup plays 11 points better than average and rests Pierce.

To rest Ray Allen I would recommend

BOS      30.36 (   -0.19   -5.02   -4.83  -18.94   -0.47    4.48)   22.60 minutes   15 appearances     6 C     29.73       14 $
T. Allen      Garnett       Perkins       Pierce        Rondo         28.56 years       1809_BOS_2010
This lineup has played 30 points better than average.

To rest Rondo it appears that the quartet of Ray Allen, Wallace, House and Davis has played well recently (+58 rating in 15 minutes).

Also, the return of Marquis Daniels should create some lineups that can better rest  Rondo or Ray Allen,

In summary, things can go better even if KG’s level of play does not improve. But without a return to KG’s past level of greatness, it looks like no Championship in Boston this season.

By the way, Ray Allen is mentioned in trade talk. His Adjusted +/- is +10 this season (despite a decline in box score stats), so trading Ray Allen would really reduce the Celtics level of performance during the rest of the season.


  1. [...] salary cap Telegram    Celtics fall to Magic Celtics finally at full strength Mathletics   Can the Celtics Regroup? NTCF    Rasheed Wallace Sucks WEEI    The Three-Pointer: The Celtics Forced To Face Reality [...]

    Pingback by Today’s Celtics Links 2/8 « Flceltsfan’s Weblog — February 8, 2010 @ 12:07 pm

  2. My immediate reaction to the data that shows Allen at +10 individually for all minutes and yet with Pierce and Rondo in, Ray out at 20 points better than average and then Rondo and Allen in Pierce out at 2 points worse than average is to bring Ray Allen off the bench or just used with with the bench more and with House as much as possible.

    Of this depends on how Tony Allen plays as a starter and with these particular starters. It has hardly been used in 2 seasons. But it was a smashing success in 2007-8- his most used lineup (139 minutes) for a +26 according to basketballvalue. It has worked just as well in the 22 minutes they’ve used it so far this season.

    Comment by Crow — February 8, 2010 @ 5:50 pm

  3. Actually the starting lineup isn’t the problem. It is fine on average for the season, not immediately sure exactly where it has been the last 10-20 games though.

    Maybe starting Tony Allen would shake things some and work but you don’t have to do that.

    You could just avoid Rondo and Allen in, Pierce out.

    My suggestion to start Allen reduced the time of Rondo and Allen in together but the starters are with Pierce so reducing that shouldn’t have been a focus.

    You avoid Rondo and Allen in, Pierce out by subbing Tony Allen for Ray Allen and not Pierce as much as possible.

    Comment by Crow — February 8, 2010 @ 10:15 pm

  4. Some of my comments are contradictory.

    Actually if have Rondo and R Allen in, Tony Allen for Pierce looks fine, as you show, if you have the other starters in there too for that 5 man lineup.

    It comes down to whether there is a strong enough basis to conclude about 5 man lineups or trios and how much to weight the information from both levels.

    It is tricky.

    Comment by Crow — February 8, 2010 @ 10:22 pm

  5. You could tweak this a little further in general and certainly some more for specific matchups and based on play any given night but this 6 lineup set looks pretty decent to me as a starting point for further refinement:

    Rondo- R Allen- Pierce- Garnet- Perkins 12 minutes
    Rondo- T Allen- R Allen – Garnett- Perkins 8 minutes
    Rondo- T Allen- Pierce- Garnett- Perkins 8 minutes
    Rondo- T Allen- Pierce- Davis- Wallace 6 minutes
    House- T Allen, Tony – R Allen- Davis- Wallace 8 minutes
    House- R Allen, Ray – Daniels- Williams- Wallace 6 minutes

    Player minutes

    rondo 28
    house 20
    r allen 34
    tallen 30
    daniels 6
    pierce 26
    garnett 28
    davis 14
    williams 6
    wallace 20
    perkins 28
    Total 240

    Comment by Crow — February 10, 2010 @ 5:03 pm

  6. Doc believes in lineup concentration … for the starting lineup but then still uses about 220 others. Why stop with lineup concentration essentially after the first one or three lineups? I wouldn’t. A six lineup set is unrealistic for 100% of minutes but 12 lineups ought to be enough for 80% of minutes I’d think if you have a well-designed roster and appropriate play the best combos thinking. Celtics top 12 lineups have been used about 50% of the time this season. Injuries are a valid and significant excuse but usually is not enough of an explanation. Atlanta. OKC, Houston and Laker top 12s are close to 2/3rds of all minutes. They are the most concentrated. Even they could go farther. I’d guess it would be better if they did but of course I can’t say for sure how appropriate the micro-use lineups are in a given moment.

    Comment by Crow — February 12, 2010 @ 4:25 pm

  7. Atlanta and Memphis are the only teams who have 3 players on the court together over half the game on average. Boston has some combos above 40%. In the regular season I can understand spreading your stars around and maybe learning about other options too. The playoffs are different in this regard.

    Comment by Crow — February 13, 2010 @ 1:09 am

  8. I totally agree.

    Comment by wwinston — February 14, 2010 @ 8:53 am

RSS feed for comments on this post. TrackBack URL

Leave a comment

Powered by WordPress