How Long Does It Take NBA Teams To Learn How To Play Together When They Lack Continuity?
Following an offseason as hectic as the one we just experienced, with tons of superstars packing their bags and heading off to new and exciting destinations, lots of people are wondering how long it will take these teams to jell. How long does it take a team made up of mostly new parts to become a well oiled machine?
Conventional logic tells us that it might take a while. There are new schemes to adapt to, new plays to learn. Players have to adapt to each other’s tendencies and playstyles, figure out how they play into the system.
Before I start, I just need to go over a bit of vocab.
Throughout the majority of the article, I’ll be referencing something called “Roster Continuity.” Basketball-Reference has a tool for measuring a team’s continuity, or how similar/dissimilar their team is to the year before.
Roster continuity is calculated as the % of a team’s regular season minutes that were filled by players from the previous season’s roster.
I’ll explain in the simplest way I can. Let’s say there’s only one, 48 minute game in a season. We have a fictional team. Last year, the team in question had five players, all of whom played the full 48 minutes. But two of them left, and were replaced by two new players. You take the total number of minutes between the three players who are left over from last season (48*3=144), and you divide it by the total number of minutes between the five players this season (48*5=240, 144/240=60%). And there you have it. The continuity rating of this team is 60%. Of course, it gets harder to calculate once you add in more players and more games, but in the end, it’s just the total number of minutes of returning players divided by total number of minutes played. Thankfully, Basketball-Reference does all of the calculations for this part, so we don’t have to (big s/o to them).
I chose 20 teams spanning the last 25 years of NBA basketball.
All these teams have continuity ratings of 50% or less. The average winning percentage of all teams evened out to a cool 48.6% (I chose both successful and unsuccessful teams to make sure the data was solid). There are a wide variety of teams here: old ones, young ones, good ones, bad ones, one fish, two fish, red f-
I took the average net rating through each fifth of the NBA season for each of the 20 teams to find how long it takes teams to jell. I figured I’d see an increase in net rating in whatever fifth of the season they “figured it out” in.
So it’s a bit unlike what someone would expect. It goes down, and then up. You’d think a team with low roster continuity would start the year with their worst rating, stagnate , then begin to go up. But hey, in the graphs, teams improve in the last 2/5 of the season. So that means something, right?
But look at the range. The difference between the lowest point and the highest point (net rating) is 2.14. And think about the story this graph tells. If it were true that it took time for teams to get better, it would make more sense for there to be an upward trend the entire time, right?
Let’s take a look at the Indiana Pacers, one of the most consistently average teams from the 2016–2017 nba season. They went 43–39.
Look at how close together the blue dots are compared to the orange Indiana Pacer dots. Despite being so consistent, they make the average of the 20 teams’ net ratings look downright monotonous in their continuity.
And finally, for reference, and because it’s fun (at least for me), let’s take a look at one of the more inconsistent teams from last season, the Miami Heat.
It’s a huge difference.
There were wild inconsistencies between each of the low-continuity teams, with some stretched being very good, some being very bad, and others being very normal. Therefore, they averaged out to be a fairly average team. You’d probably get a pretty similar graph if you took a random sample of any 20 teams in NBA history.
For any grouping of teams I could make —only teams with high winning percentages, only ones with low winning percentages, teams within certain ranges of continuity or certain types of players which were added, whether or not the team added a star player, etc. Nothing changed it.
Strength of schedule, injuries, and just sheer luck matter a whole lot more for a team’s improvement than their ability to jell over time.
This doesn’t mean teams don’t figure things out over time. It can happen more subtly. But usually, if it does happen, it gets worked out in preseason, or over the first few games of the season.
Why? Because other teams figure things out, and other teams grow too. And because they have all offseason to learn how to play together. Oh yeah, and they do this for a living.
Just because the net rating was fairly consistent for the average of the 20 teams doesn’t mean that because a team does poorly or well in the beginning of the season that they’ll stay that way. There’s way too much variation for that to happen. Teams get a lot better or a lot worse over the course of a season, and it has nothing to do with continuity.
Save for possibly the 2012–2014 Lakers, every single team on here had variation significantly greater than that of the average of all 20 teams.
Only 5 teams (20%) — The 2012–2013 Brooklyn Nets, The 2012–2013 Houston Rockets, The 2006–2007 Toronto Raptors, 2004–2005 Houston Rockets, and the 2003–2004 Miami Heat— improved throughout the season as one (one being myself) might hypothesize a team with low continuity would.
CP3 and Harden. Melo, PG-13 and Russ. Kyrie, Hayward, Horford. They might not have it figured out on day one of training camp. But you can bet they’ll be a team just like any other (except way better) by the end of the first week of NBA basketball. Take it with a grain of salt when people say they’re still trying to figure things out. They do this for a living, and have been working together throughout the offseason. If they improve, they improve. Lots of teams do.
