Truth or Myth: Reviewing Roy Williams’ Key Lineup Decision Last Year

Posted by Brad Jenkins on October 17th, 2013

On February 13, 2013, North Carolina’s season was in limbo. That evening the Tar Heels would take a 6-4 ACC record into Durham against a highly favored Duke team, while coming off their worst performance of the conference season, a 26-point blowout loss at Miami. But for Roy Williams’ team, that UNC-Duke game was the debut of a new smaller lineup featuring P.J. Hairston starting in place of big man Desmond Hubert. Although UNC would lose that game by five points, 73-68, the new lineup had bigger Duke looking slow and confused for much of the night. From that point, the Tar Heels ran off  six straight conference wins, finished third in the league with a 12-6 record, and went on to make the ACC Tournament championship game before losing to top-seed Miami for the third time. National media types lauded the lineup change as a brilliant coaching move by Williams, while many local media and Tar Heel fans were left asking what took so long for the head coach to make the move in the first place. Before we turn our full attention to the upcoming season, let’s look back at some of the truths and myths concerning that North Carolina lineup decision as well as address some reasons as to why the move wasn’t made sooner.

PJ Hairston and Friends Survived the Villanova Comeback

Inserting PJ Hairston Was The Key Lineup Change Last Season

Truth or Myth #1 – The Lineup Change Had a Tremendous Positive Impact

This would be a big affirmative. UNC’s record improved from 6-4 in the ACC before the switch to 8-3 afterward, including the games in the ACC Tourney. But more than that, North Carolina just looked like a much better team and the stats back that up. Basically the net effect of the lineup change was to remove 20 minutes from post players Hubert, Joel James, and Brice Johnson and redistribute them among the perimeter players, with Hairston picking up more than half of what was left. As expected from a decision to take away significant minutes from big players, defensive efficiency was most negatively impacted.

In comparing the Tar Heels’ last 11 ACC games with the first 10, defensive two-point field goal percentage rose three percent and opponents’ offensive rebounding percentage rose seven percent. But the new and much quicker lineup forced more turnovers and defended the three much better. The net result was that the defense got slightly worse, allowing 1.02 points per possession (PPP) compared with 0.98 PPP in the first 10 ACC games, but the flip side is that the Carolina offense really took off. The smaller unit cut turnover percentage down by three percent, improved their effective field goal percentage by three percent, and raised the team’s overall PPP from 1.02 to 1.12.  They clearly were a much better overall team after the switch — even more than the improved record indicates — especially when you consider that the only three ACC losses after the switch were to either Duke or Miami, two of the top 10 teams in the country last season.

Another unexpected thing revealed itself with respect to the Tar Heels’ tempo. Always known for playing at a high pace, one might have thought going smaller and quicker would have resulted in a Tar Heel team playing at an even higher tempo. But the truth is that the opposite happened. In the 10 ACC games prior to the lineup change, North Carolina games were played at an average of 69.9 possessions per contest. That dropped to 67.1 possessions per game in the next 11 games. Part of the reason behind the change could be the drop in defensive rebounding chances that limited fast break opportunities, but it probably had more to do with the significant drop in turnovers experienced by the offense. To the naked eye, it seemed that early possession turnovers plagued the team often during the early conference season, frequently committed by the shaky hands in the frontcourt. With the change to a more perimeter-based offense featuring double screens to free jump shooters, those bad early trip turnovers seemed to disappear. TRUTH.

Truth or Myth #2  – North Carolina Went to a Four-Guard Lineup

In the college game,  guys like Reggie Bullock and P.J. Hairston, who are better rebounders than they are ballhandlers, are forwards. In the traditional UNC system they are “wing forwards” but they are forwards nevertheless. Think about this: Before the lineup change, did you ever hear UNC described as having a three-guard lineup? Maybe it’s just more fun to say four-guard lineup for dramatic affect. A more accurate description of the new lineup would have been a big forward (James Michael McAdoo), two wing forwards and two point guards. Yes, that’s right; Dexter Strickland functioned more as a second point guard than a traditional Roy Williams two-guard such as Wayne Ellington. Strickland led the team in assists, rarely took or made any outside shots, had a great assist/turnover ratio, and was the team’s best on-ball defender. What kind of guard does that sound like to you? A point guard. MYTH.

Truth or Myth #3  – North Carolina Bombed Away With Three-Point Shooting

While the Tar Heels shot threes considerably more than we are used to seeing a UNC team attempt, they were no Iowa State either. In the first 10 ACC games, 28.6 percent of the team’s field goal attempts were three-pointers. In the next 11 games, that figure jumped to 35.1 percent. While clearly a shift upward, that number would have ranked 115th in the country had the Heels shot treys at that rate for the whole season. More importantly, this increase in attempts coincided with an increase in makes, going from a solid 36 percent to a stellar 39 percent as a team. TRUTH and MYTH.

So What Took So Long?

So that leaves the big question, why did it take so long for Williams to pull the trigger? The common thinking seems to be that he was just too stubborn to change how his teams play. In retrospect, there could be a couple of other reasons that may have been part of the equation, and they both involve P.J. Hairston and Leslie McDonald. With each in the news this past summer for the wrong reasons, it’s fair to question whether character played a role. Perhaps Hairston was not invested in the team to the level he needed for Williams to feel he deserved to start. As for McDonald, another key perimeter cog, he had to serve a three-game suspension early in the conference season for not fulfilling his own responsibilities, so it’s probably fair to question his commitment as well. That suspension came on the heels of McDonald missing another three games with a knee injury, which leads to the other issue. When McDonald returned, Hairston was dealing with a concussion himself. He missed one game and only played 12 minutes in the next. So after the first ACC game at Virginia, UNC went seven straight games without one of McDonald or Hairston, followed by one game with a limited Hairston.

That brings us right back to where we started. The next game was the Miami debacle followed by the new lineup at Duke. The essential point is that Williams really couldn’t have made the switch any earlier in the conference season even if he had wanted to. It can certainly be argued that the switch should have taken place in early December when it became plainly obvious that Hubert and James were not going to be effective inside, but we will probably never know for sure if it was simply stubbornness or Williams’ standards for how his players act. Whatever the reason, Williams has stated many times that we are not likely to see it as an issue again this year. There is the as-yet-undefined suspension of Hairston to consider, the loss of Reggie Bullock, and only five real perimeter players anyway. Heading into this season, the days of Tar Heel “small ball” may be gone forever, but it is important to remember how and why UNC got itself into the situation in the first place.

Brad Jenkins (383 Posts)


Share this story

One response to “Truth or Myth: Reviewing Roy Williams’ Key Lineup Decision Last Year”

  1. ryan says:

    Great article, thanks for mentioning things others tend to forget, like controlling for opponents (8-3 did include losses to only duke and miami). that could also explain the pace decrease (3 poss/game) in such a small sample size. my only issue with the piece is this sentence:

    “But the new and much quicker lineup forced more turnovers and defended the three much better”

    Turnovers up, yes. Defending the 3? KenPom has consistently proven that this is a misnomer, and that a team’s defense has little effect on its opponent’s 3ptFG% (http://kenpom.com/blog/index.php/weblog/entry/the_3-point_line_is_a_lottery and http://kenpom.com/blog/index.php/weblog/entry/3_point_defense_should_not_be_defined_by_opponents_3p). How did the Heels’ opponents select their shots after the lineup change? There are lots of confounding variables to be waded through (was UNC winning by more forcing the opponent to play catch-up? did not having Hubert in as a shot blocker encourage the taking of 2’s?). These questions might lead to better answers of how the defense changed when the small lineup was implemented, as opposed to random noise in opponent 3pt percentage.

    Apologies if what you meant by “Defended the 3 much better” was that opponents took fewer 3’s because the shots were (in truth) defended better, i am assuming that you mean opponent’s 3ptFG% went down :)

Leave a Reply

Your email address will not be published. Required fields are marked *