A few weeks ago, I took a stab at the extremely difficult task of using statistics to measure coaching performance. One of the central challenges to this task is finding any numbers which can be attributed primarily to the influence of a coach. One area I did identify as a possibility was the distribution of minutes.
I’m talking about minutes played, and specifically, how minutes are distributed to the various players on a team. It’s one of a coach’s most basic duties – deciding who will play, when, for how long, and in what combination. Other than in cases of injury or suspension, this distribution is usually the sole discretion of the coach, and therefore provides a set of statistics which can be attributed directly to them.
In that first attempt I ran correlations between the effectiveness of each five-man unit and how many minutes that unit was played by their coach. The idea was that a good coach would recognize and understand the relative effectiveness of their different lineups, and with a few exceptions, give the most minutes to the most effective lineups.
Today I have a slightly different technique. I heard some criticism on my first piece, specifically about Doc Rivers and Phil Jackson ranking so high. The argument went that with three future Hall of Famers, lineup decisions should be easy to make. It’s no surprise then Rivers showed up as one of the most effective coaches, 2008 in particular. Good players make coaches look good, ergo the numbers I had put together weren’t really telling us much about coaching ability. Starting with that idea, I tried to create a method that compares the talent level of a roster to the quantity of time that a coach was able to get that talent onto the floor.
For each team, I looked at all the five-man units that played at least five minutes together and calculated what percentage of those units finished the season with a positive Net Rating. I then calculated what percentage of total minutes were played by those positive units. If a coach distributed minutes completely evenly, with no regard for a lineup’s effectiveness, those two percentages would be equal. Unsurprisingly, that was not the case in any of the seasons I looked at. So to restate, we’re comparing the percentage of lineups which outscored the opposition, with the percentage of time those lineups were on the floor. For our purposes here, we’ll consider the difference between those two percentages, positive or negative, as a representation of a coach’s ability to manipulate their roster.
Now this method is not perfect. I think it covers some of the holes in my first analysis, while creating plenty of its own. Obviously, a coach has more responsibility for a unit’s success or failure than just deciding what players make up that unit. Not all positive lineups are created equal, and neither are all negative lineups. Also a coach’s ability to create effective lineups can always be limited by injuries and roster changes. Despite the flaws, I think there is valuable information to be gleaned.
Since all the data again comes from Basketball Value, this analysis has some of the same limitations as my first piece. It only covers the last four seasons. Seasons where a coaching change was made during the year are not included. Aaron Barzilai who runs Basketball Value is a consultant for the Memphis Grizzlies which means their data isn’t listed on the site and therefore, is not included.
This first table shows the numbers for each individual season.
|2010||CHI||Vinny Del Negro||0.419||0.622||0.203|
|2009||CHI||Vinny Del Negro||0.531||0.434||-0.097|
|2011||LAC||Vinny Del Negro||0.427||0.483||0.056|
|2011||ORL||Stan Van Gundy||0.519||0.702||0.183|
|2010||ORL||Stan Van Gundy||0.558||0.749||0.191|
|2009||ORL||Stan Van Gundy||0.586||0.794||0.208|
|2008||ORL||Stan Van Gundy||0.67||0.870||0.200|
This second table shows the cumulative numbers for each coach over the past four seasons. For some coaches this is multiple seasons, for some it is just a single season.
|Vinny Del Negro||0.459||0.514||0.055|
|Stan Van Gundy||0.577||0.779||0.202|
With just a few exceptions, these numbers match up closely with the effectiveness correlations we looked at last time. Rivers and Jackson again find themselves at the top of the pack. With Jackson these numbers seem to support the argument of relying on terrific players. Over the last two season, he’s had less than 50% of his lineups outscore the opponents. However, he’s been able to give over 75% of the possible minutes to those positive lineups each season. But the bottom line is that basketball is not an individual sport. Jackson can’t just run Kobe Bryant and Pau Gasol onto the floor alone. He has to surround them with complimentary talent and find ways to keep the roster competitive when those two are resting.
I think what we’re seeing is more than just coaches of ambiguous ability who’ve been blessed with superior players. Looking at the 2008 Celtics in a little more detail might will help illustrate what I mean. That year 60.3% of their lineups, which played at least 5 minutes together, outscored their opponents. Those lineups were allocated 81.3% of possible minutes.
The Celtics were more than just the Big Three in 2008. They had a powerhouse starting lineup with Rajon Rondo and Kendrick Perkins playing alongside Paul Pierce, Kevin Garnett and Ray Allen. As was pointed out above, Rivers did ride that unit hard. Including only lineups that played at least 5 minutes together, the Celtics starting five played 30% of the possible total minutes as a group. That’s a significant chunk, but leaves 70% of the Celtics’ meaningful minutes for Rivers to formulate workable player combinations.
Obviously, with such talented players, Rivers was going to rely heavily on the best of what he had. 14 of their 20 most played lineups featured at least three starters. Altogether, 70% of the available minutes went to lineups that featured at least three starters. Those lineups were extremely successful. With three starters or more on the floor, the Celtics had a Net Rating of +15.2. With less than three starters the Celtics had a Net Rating of +10.8. The other important number is the Net Rating of +19.5 posted by the five starters together.
Broken down this way Rivers had three options. He was able to manage his rotations in such a way that 70% of the minutes went to the two better options. Just as important was the way he managed minutes, that pushing the team from great to elite, while still protecting the legs of his players. Other than his rookie season, Kevin Garnett had the lowest MPG of his first 13 seasons in 2008. For Paul Pierce it was the 3rd lowest MPG of his career up to that point. It was the same case for Ray Allen. Rivers cut minutes for his stars, but managed to keep what he needed on the floor for the greatest amount of time. I would argue that he did a masterful job managing his rotations that season. Perhaps the high level of talent on his roster made things easier. But he still needed to find a way to maintain effectiveness when any of those starters came off the floor. He did that in a big way.
Small changes in managing rotations make a huge difference. The Oklahoma City Thunder were a perfect example of this last season. 53.6% of their lineups, which played at least 5 minutes together, outscored their opponents. Those lineups were allocated 59.2% of possible minutes. The most commonly-used five-man unit for the Thunder last season was Russell Westbrook, Thabo Sefolosha, Kevin Durant, Jeff Green and Nenad Kristic. This was the Thunder’s starting lineup until Green was dealt to Boston for Kendrick Perkins at the trade deadline. This unit played 541.92 minutes, with a Net Rating of -0.94. When Serge Ibaka replaced Green, the Net Rating of that group jumped to +7.05. That unit, with Ibaka replacing Green played just 127.63 minutes together across the entire season.
Green was with the Thunder for 49 games before the trade. That means Scott Brooks had 49 opportunities to set his starting lineup, and each case he chose a less effective orientation with Green instead of Ibaka. All other things being equal, if the minutes allocated for each of those units had been flip-flopped it would have taken the Thunder’s season-long Net Rating from +4.18 to +5.03. That’s a significant difference, and all it would have taken to achieve is one change in the starting lineup.
A coach generally doesn’t have much control over their roster. They do have control over what players are on the floor at any given time. Looking at how many minutes are given to the best players and the best units is a rather simple and measurable way to analyze one area of coaching performance. The next step here at Hickory-High is to take these two techniques and apply them to some playoff numbers. Let’s see if we can find some clutch coaches.