This Friday Michigan State will begin yet another brutal nonconference slate by playing North Carolina on an aircraft carrier docked in San Diego harbor. Four days later, the team will face Duke on the opposite coast, at Madison Square Garden, before finally coming to home to play a more standard-issue nonconference opponent (Texas Southern). The remainder of MSU's nonconference slate includes two more top-level tests in Florida State (at home) and Gonzaga (on the road).
Spartan fans are right to be somewhat anxious about how MSU will fare against these nonconference opponents. The team returns only two players who played more than 30 percent of available minutes last season (Draymond Green and Keith Appling), and, of the ten players likely to form the team's core this season, five have never played a minute in Spartan green.
Further, Tom Izzo teams have something of a reputation for struggling early in the season, only to emerge in March as much-improved performers in position to contend for NCAA tournament glory. Over the last decade, Izzo's found regular-season dominance to be elusive. Michigan State's last No. 1 seed in an NCAA tournament was in 2001. Since then MSU's been seeded at the No. 5 line or below in nine of ten tournaments, the exception being a No. 2 seed in 2009. Yet the program's still managed an impressive 19-10 record in tournament play over that time, with three Final Four appearances to show for its efforts.
Izzo's employed a number of well-publicized strategies in order to propel his teams toward deep tournament runs: the masochistic non-conference scheduling he uses early in the season to test his players against the very best national competition; the "War Drill," which purportedly builds toughness as the season progresses; and the condensed-but-effective film study and game planning during the NCAA tournament. But if a brutal non-conference schedule, the War Drill, and slick film study were enough to get a team to the Final Four, it's a safe bet that more teams would do all of the above.
So the question is this: In what specific areas of play has Tom Izzo been able to generate improvement over the five months that make up a complete college basketball season? To answer that question, I've compiled data on average game-by-game statistical performance for MSU during the different stages of the season from 2002-03 through 2010-11, the period of time for which Ken Pomeroy has easily digestible "four factor" game data available.
The numbers, displayed below, are broken up as follows:
Games played against non-conference opponents ranked in the top 75 nationally by Pomeroy. The average ranking of these opponents is comparable to the average NCAA tournament opponent MSU's faced since 2003.
Games played against non-conference opponents ranked in the top 150 nationally by Pomeroy (including games against the top 75 opponents). This set of opponents comes closer to matching the typical range of opponent quality within the Big Ten.
Games played during the first half of regular-season Big Ten play (eight of 16 games early in the time period, nine of 18 later).
Games played during the second half of the Big Ten season.
NCAA tournament games.
(I've ignored Big Ten tournament games. Those numbers are U-G-L-Y. Also, the handful of games played against non-conference opponents in the midst of Big Ten play has been excluded.)
Slicing up the numbers this way allows for three sets of meaningful comparisons among the shifting statistical performances typical of the Spartan basketball team over the course of the season.
From non-conference to Big Ten: Righting the ship
Michigan State has, of course, beaten some pretty good teams in non-conference play. And, with a few embarrassing exceptions, the team's generally held its own against lower-tier non-conference opponents: 20-3 record vs. teams ranked between Nos. 76 and 150, plus a perfect 46-0 vs. teams ranked below 150. Still there are some serious weaknesses in MSU's non-conference statistical profile. The offense has turned the ball over on a fairly generous 22.5 percent of its possessions. Opponents have been allowed to post a healthy effective FG percentage of 49.5. And the defensive rebounding's been far short of dominant, as opponents have recovered 32 percent of their missed shots.
MSU's managed to make corrections in all of those areas during the first half of conference play, as shown in the first column of numbers in Table 2. Even as the pace of play slows in Big Ten play by about five possessions per game, the Spartans have improved their offensive efficiency by 1.5 points per 100 possessions. That improvement's been driven almost entirely by a 1.2 percentage-point reduction in the team's turnover rate.
Meanwhile, the team has tended to improve even more on defense when transitioning from non-conference to conference play, by 2.4 points per 100 possessions. That improvement's been driven by a slight improvement in FG defense, and by an even larger improvement of four full percentage points in defensive rebounding. Note, however, that part of the latter change results from the fact that so many Big Ten coaches eschew opportunities on the offensive glass, preferring instead to emphasize transition defense.
Those improvements have resulted in MSU winning an additional ten percent of its games in the first half of Big Ten play, compared to top-150 non-conference play. This is noteworthy given that, in the aggregate, the level of competition is similar across these two categories (although the non-conference sample is tilted somewhat toward games played at the Breslin Center, where MSU owns a distinct home-court advantage).
From first to second half of the Big Ten season: Stalling out
Hoops fans in the Midwest are well aware of what happens as the Big Ten season progresses. Final scores dip even lower. For MSU in particular, offensive efficiency has taken a hit in the second half of conference play, to the tune of about two points per 100 possessions (Table 2, second column of numbers). That decline's been driven by decreases in both offensive rebounds and free throw attempts. Those declines, combined with a slow pace, have resulted in MSU fans sitting through a number of aesthetically unpleasant contests late in conference play. Meanwhile the defense shows little or no change across the Big Ten season. Whatever powers Izzo has to coax better defensive execution from his players, they seem to lose their potency during the doldrums of conference play.
The change in performance within the conference season hasn't been enormous, but on average it's translated into half a win fewer for MSU in the second half of Big Ten play compared to the first half.
From non-conference to the NCAA tournament: Putting it all together
What is it MSU does so much better in March than in November and December? The final set of numbers in Table 2 gives us our answer. The average ranking of top-75 non-conference opponents is almost identical to that of MSU's NCAA tournament opponents. In both cases, we're dealing with a fairly random set of foes hailing from all areas of the country, playing in either neutral venues or on a balanced home-and-away basis.
MSU shows pretty dramatic improvement on both offense and defense: over 2.5 points per 100 possessions. On offense, it's all about turnovers. The team's TO rate declines from a dismal 22.9 percent in non-conference play to a much more respectable 19.4 percent in tournament play. On defense, it's mostly about FG defense. Opponents go from shooting over 50 percent from the field (an effective FG percentage of 50.8) to shooting well below it (47.4).
Put that all together and you get a dramatic increase in winning percentage, almost a full 25 percent, despite the average quality of NCAA foes being about the same as top 75 non-conference opponents. That increase is arguably skewed upward by the fact that naturally you can only lose one game per NCAA tournament, but it's striking nonetheless.
Summing up: Maybe we already knew this
One of the attractions of looking at actual performance is identifying counterintuitive trends. For example in a piece I wrote last year, I found that, despite Izzo's reputation as a defensive-minded coach, MSU teams have actually been more efficient on offense than on defense in the majority of seasons. In this case, though, statistical analysis confirms two conclusions many fans and analysts have reached regarding the keys to success for Tom Izzo's teams:
Limit turnovers. Controlling turnovers is often an issue for the Spartans. Izzo's tightly structured offense, with its emphasis on creating high-percentage shots through a series of complex set plays, takes time to come together. Early in the season, players can become flustered when an opponent does something on defense that takes the team out of those set plays, such as full-court pressing or aggressive man-to-man defense. By the end of the season, players have learned the rules of the offense well enough to adapt when the sets break down and avoid costly giveaways.
Let the defense evolve. Building an Izzo defense requires patience. At first glance, his teams play fairly orthodox man-to-man, but in fact it's a bit more complicated. MSU's defensive scheme relies on extensive "help and rotation," where defenders shift momentarily away from their primary defensive assignment in order to prevent opponents from getting good looks around the basket. Again, it takes time for MSU players to be able to react quickly and correctly in different situations. When it all comes together, the MSU defense is in position to contest shots more aggressively while limiting easy looks in the paint.
As much as rebounding's been Izzo's calling card, MSU's only out-rebounded its NCAA tournament opponents by about two percentage points since 2003. That's a solid showing, considering the size and athleticism of the typical tournament opponent, but -- "War Drill" stories to the contrary -- it has clearly been improvements in other areas that have driven the program's tournament success.
It's a cliched notion, but playing a diverse, high-quality group of non-conference opponents has paid off for MSU at the end of the season. The next trick for Izzo will be to get the train rolling more quickly, as he did during the 1999, 2000 and 2001 seasons. A Final Four run probably isn't in the cards for MSU this season, but history suggests Izzo's willingness to play anyone, anywhere (and the deck of a 100,000-ton naval vessel certainly qualifies as "anywhere") will pay long-term dividends for this new group of players.
Follow Kyle on Twitter: KJonthebanks. He extends his appreciation to Steve Hendershot for editing and suggesting improvements to this article.
This free article's an example of the content available to Basketball Prospectus Premium subscribers. See our Premium page for more details and to subscribe.