This website is called RealGM, but the real general manager in college basketball is the head coach. And my content is going to include a lot of material discussing head coaches and which ones are performing at the highest level.

One thing you will not see me do very often is criticize a coach for lineup decisions. I believe that head coaches are smart, rational people, and they have a lot more information than we do. It is easy to point at a situation like Kansas where Josh Selby’s efficiency numbers are not great and say Kansas is better off without their star freshman. And I showed that as of the end of January the Jayhawks were performing at a slightly lower level with Selby in the lineup. (Since that time, Kansas crushed two teams with Selby in the lineup, crushed three teams with Selby injured, and lost a game where Selby returned to the rotation but did not appear to be 100%.) But I do not read Kansas’ success without Selby as an indictment that Kansas should not play their highly regarded freshman.

Bill Self is smart and recognizes two things. First, he realizes that for the team to perform at the highest level in the long-run, Josh Selby needs experience. And as Luke Winn showed last week, some elite freshman are dominant when they first step off the bus, but some elite freshman need a little time to sort things out.

Second, Bill Self realizes that part of the contract in recruiting elite freshman is guaranteeing them some playing time. Sometimes there may be a short-run hit to team performance, but by showing he will play a star freshman major minutes, Bill Self makes it more likely the next elite recruit will commit to Kansas. Were Kansas struggling to make the NCAA tournament with Selby, it might be more important to bench him and play the more experienced lineup right now. But that has not been a problem for Kansas. They have been winning and dominating games whether Selby is in the lineup or not. The team has only been minimally less efficient with Selby running the show, and that’s something Bill Self can live with to build Selby’s experience for the future.

• A New Framework

But even if I am reluctant to question lineup decisions, there are a lot of important questions to ask about college coaches. One of the key unresolved questions is how important recruiting is to success. How much of Duke’s success on the basketball court is because Mike Krzyzewski constantly has a line-up of McDonald’s All-Americans, and how much of the team’s success is because of his ability to develop players over time?

To think about this question from a new perspective, today I want to talk about how two coaches have maintained two of the most efficient programs in the nation over the last decade. Those coaches are Roy Williams of North Carolina and Bo Ryan of Wisconsin. I chose these two coaches because both have a reputation for winning in different ways. Roy Williams has a reputation for bringing in elite recruits, and giving them the freedom to showcase their talents. Bo Ryan has a reputation for bringing in players who fit his system, and grooming them into polished seniors who perform at the highest level. But do the numbers fit with their reputations?

Note: I apologize for the gory details about the numbers in the table. If you just want the conclusions of the analysis, scroll to the end.

In the table that follows, I include all the players to play under Roy Williams at North Carolina and Bo Ryan at Wisconsin except non-contributing Junior and Senior walk-ons, and four freshman that Bo Ryan inherited in 2001 that basically never played.

The first row shows the number of players by class, and what you see is substantially more freshman than seniors. The data includes some players who transferred into the programs after their freshman year, but it is much more common to see players transfer out because they are not contributing at a high level. Another reason there are fewer seniors is that North Carolina had nine early entrants selected in the NBA draft during this time period, while Wisconsin lost one early entrant (Devin Harris) to the draft.

The next row shows the number or RSCI Top 100 recruits in each class and not surprisingly North Carolina dominates this list. Wisconsin has usually had a few elite recruits on its roster, but not nearly as many as North Carolina. Recognize also that not every elite recruit leaves early for the pro leagues. Take Boo Wade for example. He left Wisconsin for “personal reasons” back in 2005.

The next row, Avg Pct Min, shows the expected playing time for players in each class. At North Carolina, the average freshman plays 27.5% of the team’s minutes. At Wisconsin, the average freshman only sees the floor only 18.2% of the time.

The next row, Avg Pct Possessions, lists the average percentage of possessions used by players in each class. North Carolina freshmen take a much higher percentage of shots when on the floor than Wisconsin freshmen.

Next, I present ORtg sorted multiple ways. (For uninitiated readers, let me indicate that ORtg is a Dean Oliver statistic which summarizes a player’s contribution to the offense. Ratings over 120 are elite, under 100 are often seen as poor, and below 80 are simply unacceptable at the D1 level.) The first row, Avg ORtg, shows the average contribution per player. This shows that freshmen that attend North Carolina and Wisconsin do not perform all that well. The average freshman efficiency rating is below 95. From a player perspective, that may be the variable that matters. No matter where you go, on average, you are not going to be “great” immediately.

But the more interesting question may be how freshman contribute from the team’s perspective. That is because all coaches will give more minutes and shots to freshman that play well. In the next row, Weighted ORtg weights by playing time and shots used. While freshmen are poor contributors on average, the ones who get to play major minutes and take a lot of shots, are not substantially worse than the other players in the rotation.

• Conclusions

1) Roy Williams gets all the credit in the world for being a top recruiter. But his average freshman recruit has not been substantially more efficient than Bo Ryan’s over this same time period. The average player’s offensive ratings, 93.3 and 92.9, are trivially different. 

If this seems odd to you, recall North Carolina players like Quentin Thomas and Leslie McDonald. They were elite recruits, but both struggled during their freshman seasons. North Carolina does look slightly better because of players like Brandan Wright (119 ORtg as freshman) and Marvin Williams (119 ORtg as freshman), but when Wisconsin can find someone like Alando Tucker (118 ORtg as a freshman) who was not highly rated, but who could flat out dominate at the D1 level, you are reminded that the recruit rankings are not everything. (Also, note that Wisconsin redshirts substantially more players which may give Wisconsin’s freshman more experience and a better chance to succeed.)

2) One of the key reasons North Carolina’s freshman occasionally struggle is that they get to play more minutes (27.5% to 18.2%) and take more shots (20.5% to 17.6%). North Carolina’s freshman are given more opportunities to take bad shots and commit silly turnovers, while Bo Ryan’s freshman are put in much more controlled situations.

But again, this is part of the implicit contract some coaches make with elite recruits. Roy Williams’ top recruits will get a chance to showcase their skills whether they are effective or not. But if you commit to Bo Ryan, you are not going to get a chance to play until you are ready. This implicit agreement allows Williams to pick up more commitments from elite players.

3) As we have seen with other player data, the biggest jump is always from freshman to sophomore year. I like to say that players learn to stop taking bad shots. As freshmen, players often try to make very difficult plays, and as they mature, they learn to only take shots they know they have a high probability of converting. But players also get better by practicing and honing their skills.

4) Both coaches do a tremendous job of making sure they have their best lineups on the floor. The weighted ORtg’s of 105.4 and 104.2 for freshmen are still very high. Freshmen who are allowed to play are not substantially worse than the upperclassman. For example, Devin Harris played record minutes as a freshman at Wisconsin because he was quite efficient in his time on the floor.

5) The North Carolina players in the lineup reach a higher peak. Bo Ryan receives all the credit in the world for player development, but the North Carolina rotation reaches a much higher peak in its junior season. North Carolina’s on-the-court players average 105.4 as freshman and then peak at 115.3 as juniors. Wisconsin’s on-the-court players start at 104.2 as freshman and then peak at 110.6 as seniors.

Now, before we jump off the deep end and start proclaiming Roy Williams the greatest skill developer in the country, keep in mind that there could still be a recruiting explanation here. North Carolina’s recruits may have always had more potential to improve over time. But certainly Roy Williams deserves some credit for making these players into stars.

6) Because of attrition to the NBA, the seniors at North Carolina and Wisconsin look very similar. While North Carolina’s rotation peaked in its junior year, many of those players leave early. The North Carolina players who stick around and see the court during their senior season have an ORtg of 110.9, very similar to the 110.6 for Wisconsin.

7) Finally, Bo Ryan makes up for his lack of elite NBA ready juniors by having more seniors on the roster, and by playing those seniors substantially more minutes, (53.0% of the minutes on average, to 43.9% for North Carolina).

• Punch-line

I have been struggling for some time with separating the impact of recruiting and player development, and I hope this column helps illuminate some of the dilemma. Roy Williams has had better recruits on paper, but because Williams gives his freshmen more chances to fail, Bo Ryan’s freshmen have not performed substantially worse.

And it also shows one of the enigmas about player development. Roy Williams is better at developing star rotation players, and particularly star juniors.  But because he loses so many of these players early, he tends to get less credit for how his players improve over time.

And if Roy Williams’ ability to develop players does not fit with your preconceived notions of him as a coach, think more carefully. Elite recruits can go almost anywhere. And while prestige and playing time are a large part of the equation, the bottom line is results. If Roy Williams could not turn a decent number of his elite recruits into stars, he would not have multiple national championship banners to his name.

Finally, even if Bo Ryan does not reach those same heights, by having more quality seniors stick around, and by giving those seniors more minutes, he maintains a consistently high level of play for his roster.

Both formulas are extremely successful, and it just goes to show that there is more than one way to build a consistent winner. And as much as this post was about Bo Ryan and Roy Williams, the main purpose was to build a framework for what successful coaches look like. In future columns, I hope to provide some nuggets about where other coaches match or fail to live up to these standards.