College basketball can be sloppy in November thanks to the large amount of roster turnover every year. Luckily, over the course of the season almost all teams learn to play better basketball and the level of play is elevated.

That has a pretty important implication at the team level.  If you want your team to have a great season, it isn’t enough to simply get better. You have to get better at a faster rate than your rivals. One thing I have said in the past is that teams that play a lot of freshmen have the potential to improve the most. This is based on the statistical sophomore leap. I.e. rising sophomores improve more than rising juniors or seniors. This doesn’t necessarily mean that freshmen are more likely to add a facet to their game. Freshmen are no more likely to become dominant shooters or drivers. But what freshmen do is make a lot of mistakes, and those can be corrected.

But should we expect teams that give a lot of minutes to freshmen to improve more within a season? In the following table, I took the 50 youngest teams from high major conferences over the last 10 years. I compared their offensive and defensive efficiency over the course of the season, adjusting for opponent and location as on kenpom.com:

Young offenses improve from 105.6 on average in the first eight games of the season to 108.9 at the end of the season. By correcting mistakes, making sure the right players take the right shot, and simply learning to play together, young teams improve substantially at putting the ball in the basket.

(Keep in mind that the magnitude of these effects may be slightly less meaningful than what you see here. I haven’t done a comprehensive study, but I believe the average D1 offense improves around one point over the course of the season as coaches begin to use tighter rotations. Regardless, the data suggest freshmen-led teams improve more.)

While improved offensive play seems obvious to anyone who watches the game, there was no guarantee we would see this in the data because there are mitigating factors. First, it often takes until the middle of the season for teams to develop scouting reports for freshmen. A freshman with a great first-step who was dominant at getting to the line in November and December, may suddenly get challenged to take three pointers in January and that may hurt his performance. The change in the quality of competition may also matter. The forward who looked great against low-major competition to open the season, now plays taller competition and gets the ball blocked back in his face. So certainly, the offensive improvements we see in this table were predictable, but perhaps not completely expected.

But as the Pythagorean standings show, late in the season young teams tend to play their worst basketball. Do the freshmen hit a wall? Is there a point in the season where they just run out of gas? Offensively things continue to get better. But there is a trend on the defensive end, and I think that trend deserves more explanation. Let’s look at a team like Indiana in 2009:

I think the thing we have to keep in mind with young teams is that losing takes a toll on the players. Notice how the Hoosiers defense falls off a cliff at the end of the season? The reality is that playing strong defense takes a lot of effort. And at a certain point, if you continue to put in a bunch of effort and not see a result from that effort, it is human nature to not try as hard.

It sort of becomes a cliché in February when a bad team hangs with a good team, that the coach deserves credit because his team doesn’t quit. But certainly for young teams, maintaining a level of defensive efficiency throughout a losing season can be very difficult.

But John Calipari’s young teams have not only been able to win, they’ve been able to improve their level of defensive play throughout the season. Look at the Kentucky splits from the 2010 season:

Despite one of the youngest teams in the nation, Calipari was actually able to improve his team’s defense late in the year. (His team’s offensive struggles early in the season were also corrected.) Talent had a lot to do with that season-long improvement. And there are plenty of folks who question whether John Calipari’s team has the talent to duplicate those splits this season.

But the data suggest we should not be too concerned if Kentucky struggles in a few early games. Almost all young teams improve offensively over time. And if a young team can maintain or improve its defense, it can get better at a faster rate than its rivals.

Opening Weekend Notes

Alabama’s Trevor Lacey opened the season with a bang, knocking down a game-winning three as time expired to beat South Dakota St. But I enjoyed the novelty of the Purdue-Bucknell ending even more. With Purdue trailing Bucknell by just 3 points with 10 seconds left, everyone in the stadium knew that DJ Byrd was going to get the ball. (Byrd was an outstanding three point shooter for the Boilermakers last season and is expected to step into a lead role for the team this season.) And the Bucknell players knew he was going to get the ball too. But in the final seconds two Bucknell players collided while trying to get in Byrd’s way, and collapsed to the floor. That left Byrd wide open from deep. Sadly for Purdue, his shot rimmed off. Even non-buzzer beaters can be filled with drama.

As much as I’d love to point out the things my model has gotten right and justify the things my model has wrong at this point in the season, we are definitely in the world of small sample sizes. Anything can happen in one game, and that is particularly true early in the season. I have major plans to evaluate my predictions, but not until at least January.  St. Peter’s upset Rutgers in the season opener, but we’ll have a better sense in January whether I was wrong about Rutgers, wrong about St. Peter’s, or whether this was a one-game fluke.

More than results, at this point in this season, I am most fascinated to see which players are breaking out and which players are treading water. Finding breakout players may be obvious. Everyone saw Alex Len breakout against Kentucky and Len’s dominant performance against an athletic front line may be a reason to substantially increase our expectations for the Terrapins this year.

Finding players who are treading water may not be as straightforward. But if you had to ask me what I’m looking for, I think UCLA’s Josh Smith personified it in a possession in the first half last Friday. Smith rebounded the ball, lackadaisically threw it to an Indiana St. player instead of his teammate, stepped into the lane to try to cut off the drive and was called for a foul. The look on Smith’s face said everything you need to know about why he still isn’t playing starters’ minutes despite being one of the most gifted players in the Pac-12.

Follow @DanHanner