“Analysis” is defined by Websters as “separation of a whole into its component parts”.
So when we listen to an analyst on television during a basketball game, we should hope to hear analysis that takes what we are seeing and breaks it down into the factors that are determining the observed outcomes.
On Saturday I was watching Connecticut play Pittsburgh. At one point in the contest, Connecticut was losing by more than ten points. At this juncture the TV analyst (I don’t know who this was) offered the following explanation of how Connecticut could get back into the game:
- Connecticut needs to “score the basketball” (i.e. as opposed to “scoring a football” or “scoring a tennis ball”?)
- Connecticut needs to “get stops” (i.e. stop Pittsburgh from scoring)
What exactly does this kind of analysis provide? We can clearly see that a basketball game can be broken down into two component parts. Teams win in basketball by scoring – or “scoring the basketball”, if you prefer – more than their opponents.
Looking back at the definition of “analysis”, this discussion of Connecticut’s path to victory qualifies. But one would think you could do better.
Let’s work a bit harder to break down winning in basketball into its “component parts”. We begin where the TV analyst seemed to stop:
- Wins are determined by points scored and points surrendered.
Now let’s take another step:
- Points scored are determined by shooting efficiency and shot attempts.
This is true by definition. If we know how many shots a team takes (field goal attempts and free throw attempts) and how efficiently these shots were turned into points, then by definition we know points scored.
So what determines shot attempts?
Shot attempts – as the following equation illustrates (this is from Stumbling on Wins) – are determined by how a team acquires the ball. So to increase shot attempts, you need to get rebounds, avoid turnovers, and encourage your opponent to commit turnovers.
FGA = c1 + c2* Opp.TO + c3* DRB + c4* Opp.FGM + c5*Opp.FTM + c6*TO + c7*ORB + c8*FTA + ei
These factors explain 98% of the variation in field goal attempts (the missing factor – as noted in Stumbling on Wins – is team rebounds that change possessions).
So what factors cause a team to lose like Connecticut did on Saturday? The above analysis indicates it could be inefficient shooting, a failure to get to the line (which would help a team score more efficiently), committing too many turnovers, not generating enough turnovers, and/or a failure to rebound (readers of Dean Oliver can recognize the classic “four factors”).
Using the box score, each of these factors can be linked back to individual players. And because these factors connect to outcomes, we can measure each player’s Wins Produced.
And one doesn’t have to stop with the components of Wins Produced. As was said in The Wages of Wins:
Knowing the value of each player is only the starting point of analysis. The next step is determining why the player is productive or unproductive. In our view, this is where coaching should begin. We think we can offer a reasonable measure of a player’s productivity. Although we have offered some insights into why players are productive, ultimately this question can only be answered by additional scrutiny into the age and injury status of the player, the construction of a team, and the roles the player plays on the floor.
In sum, there are many “components” that lead to outcomes in basketball. But most of these factors seem missing from the analysis that I typically see on TV. Here is what I tend to see (the Connecticut-Pittsburgh analyst is hardly an exception):
- First, wins are determined by points scored and points surrendered.
- Having observed that a team is losing (or winning), the analyst moves on to factors like “chemistry” or “energy”. Although these factors might be important, they can’t be measured very easily. So it is difficult to know if the analyst is focusing on the right components.
And by taking this approach, it appears the analyst is moving from the obvious to magic.
A better approach – at least “better” in the sense that the analyst is getting at more and more components that we know impact outcomes – is to focus on factors like shooting efficiency, rebounds, and turnovers. Then, having identified the factors that impact the outcomes we observe, turn to the factors that cause players to shoot efficiently, rebound, and commit turnovers. Is it just talent (often I think that is the most important story)? Is it injury and/or age (and I often think these factors are the second most important story)? Is it some coaching strategy, or perhaps the players’ attitudes?
In sum, let’s start with how productive the individual players have been (i.e. assign responsibility for outcomes to individuals). Then move on to why these players are productive.
Such an approach might actually tell the audience why we see the outcomes we observed.
Back to UConn and Pitt
When we apply this approach to Connecticut on Saturday, we see a bit more than a team losing because it was outscored. The box score tells us these teams were even on turnovers. With respect to shooting efficiency, there wasn’t much difference (whether we look at effective field goal percentage or true shooting percentage).
When we turn to rebounds, though, we see that Pittsburgh grabbed nine more boards. And when we turn to the players for Connecticut, we see a potential problem. The Huskies don’t have much size on their roster. Of the nine players who played on Saturday, only four – DeAndre Daniels, Tyler Olander, Enosch Wolf, and Phillip Nolan – are 6-8 or taller. And of these four, only Olander grabbed a rebound on Saturday.
So perhaps the problem was that Daniels, Wolf, and Nolan combined to play 37 minutes and failed to grab any rebounds. And that allowed Pittsburgh to take more shots, which led Pittsburgh to score more points.
Now why did this trio fail to rebound? Is it talent, injury, coaching strategy, attitude, etc…?
This is where we would hope to see some actual analysis. But you can’t get to these types of questions until you have done more than tell us that a team is losing because it is not “scoring the basketball” and not “getting stops.”