Warriors’ Dominance Leads to Little Wasted Effort
By Seth Partnow
One of the chief benefits conferred by the epicly great seasons currently unfolding in Golden State and oh-by-the-way San Antonio is the ability to set themselves up for a deep playoff run. Almost like a baseball manager juggling the pitching rotation to ensure maximum postseason impact for a team’s top arms, Steve Kerr and Gregg Popovich have the luxury of resting stars, experimenting with alternative lineup combinations and basically doing everything possible to ensure their respective teams are at their absolute peak come May and June.
Even the pressure imposed by the Warriors’ pursuit of the alltimes single season wins record[1. You might have heard tell of this prospective achievement.] is mitigated by their extreme degree of inseason dominance. As first discussed here, the Dubs spend more time leading by double digits (approximately 37% of game time as of this writing) than any team since at least 1996-1997. This naturally turns into a lot of blowouts, which in turn means a lot of 4th quarter rest for Curry. By comparison, a team fighting for home court advantage or even to secure a spot in the playoffs at all might have to run their starters a bit more ragged. If Jimmy Butler is the player who instantly springs to mind, there’s a reason for that – the chart below shows the time remaining, in minutes, for when Curry and Butler respectively have exited a game for the last time, as well as the final margin in each of those contests:
On the left side, we can see all the easy nights the Dubs have had, allowing Curry to kick up his feet and enjoy watching the final stanza. On the right side, all the times Butler has played to the bitter end. The gap from 12 to 6 minutes reflects standard NBA rotation practices to a degree. Like many stars, Butler and Curry are likely to grab a few minutes on the bench at the start of the fourth. If they are brought back into the fray at all, they will get several minutes of run, the stint lasting at least the period between under 9 and under 6 timeouts before departing for the evening.
Of course, the discrepancy in between the timing of the end of the night for the two players is only partially about coaching rotation patterns and decisions. The Bulls play more close games than the Warriors, so that’s part of what you get. But it isn’t just about the different circumstances. As the chart shows, there are plenty of instances of players remaining on the floor until deep in the 4th quarter despite a sizable lead. The basic conservatism at work – don’t blow a big lead and don’t give up on a game too early – is understandable, but it might also be costly.
According to data provided by Mike Beuoy of Inpredicatable.com, without knowing anything else[3. The relative strength of the teams, home or road, etc.], a 10 point lead with three minutes left gives a team a 99% chance of winning the game. A double digit lead at the 9 minute mark is around 89% chance of victory. Now those percentages possibly overstate the degree to which the remaining time in a game is superfluous in those situations, as the observed winning percentages reflect actual substitution patterns to a degree. If Curry were simply to not come back in the game if the Warriors led by 10 or more with 9 minutes left, their actual win percentage would likely drop, perhaps substantially. Still, we can see how relatively useless minutes might add up on star players.
To investigate how big a problem this actually is, I tried to estimate the number of “wasted” minutes played by each individual this season. Using Beuoy’s figures, I found the approximate win probability after every play-by-play event occurring in fourth quarters. I then created a “minutes leverage” index which assigned those minutes value from 100% in situations where win probabilities were 50/50 to 0% in situations where one team was certain of victory. Taking the average minutes leverage index for each player and multiplying it by fourth quarter minutes played, I derived the following “leaderboard” through Wednesday’s games:
In the chart, “Wasted 4thQ%” indicates what proportion of a player’s 4th quarter minutes have had relatively little chance to affect the outcome of the contest, while “% ‘Closing’ MP” reflects the percentage of the teams “Closing Time” minutes, where the clock was under 5 minutes remaining in regulation with the score margin under 10.[1. Note I’m only using fourth quarters, not overtime period, as those sorts of “accidental” minutes can easily defy the best laid minutes conservation plans.] To account for injuries and in-season player movement, this percentage is measured as playing time divided by the available time in games in which the player appeared.
A few notes. First, by limiting the study to 4th quarter minutes only it is only natural to see a few bench types atop the list. As evidenced by the previous graph, players such as Waiters, Turner and McCollum are often in the game at the start of the fourth regardless, and if the game gets out of hand, the starters might never come back in. Even if the game is over as a contest somebody is going to play out the string every night. Secondly, perhaps this suggests there might be some gains in terms of slightly altered sub patterns, going with the Spursian “three stints per half” method. Star players could perhaps be more efficiently used by resting near the end of the third, starting the fourth and then sitting from the nine minute mark or so if the team in the lead is able to maintain a double digit gap at the start of the fourth. Finally, it is worth noting how narrow the range seems to be for wasted minutes as a percentage of players’ 4th quarter court time. The following chart shows some of the most used players for the (very broadly defined) contending teams this season:
Griffin (and his truncated season) aside, there do not appear to be any players being massively exposed to extra mop up duty. Perhaps an Iguodala or Waiters ends the game with bench units a bit more often than some front line players. But the stronger lesson remains that the best way to avoid unnecessary play from stars in the fourth quarter is to have already put the game beyond doubt before the exact time at which to remove the main men becomes an issue.