Designing the Perfect Schedule: More Art Than Science?

facebooktwitterreddit

How do you design a perfect non-conference schedule? It is a difficult question that I have grappled with here at Husky Haul over several posts previously. But, as the Huskies continue to struggle in the non-conference schedule over the last several years, one has to start asking; is there a better way to put it together? Sure, they can rack up wins against no-name opponents at home (save for the South Dakota State game last year), but they seem unable to win a big game on the road or at a neutral site. Look at the list of losses to major conference opponents and it is really starting to rack up in the Romar-era…

There is an old saying my mother used to say to me as a kid which is “You can’t please everyone”. So it seems to be the case with the Huskies’ non-conference schedules since Lorenzo Romar took over at the University of Washington back in 2002. Every summer when the Husky schedule is released there is a huge outpouring of disappointment and even frustration by many commentors on the various Husky blogs about the schedule. Depending on the year, it usually falls into one of three categories:

1) Not enough marquee names to get the national media (and elite recruits) excited.

2) Not enough true road games to test the players before the conference season starts.

3) Too many pansies and no-name teams coming to Hec Ed to justify the $$$ paid by season ticket holders.

There are  those who defend (or at least understand) the reason the schedules come out the way they do. And, anecdotally, it seems like Romar tends to be vindicated at the end of the season when UW ends up with a pretty solid Strength of Schedule (SOS) and several of the teams they played end up winning their conferences and joining the Huskies in the NCAA tournament. For instance, South Dakota State made the NCAA tournament, as did Saint Louis, Marquette, and Duke, while Nevada (15-1 in WAC) would have had they not been upset in their conference tournament.

I took a look at the non-conference schedules for UW over the past 6 years since the 2006-2007 season to see how strong or weak the schedule was and what impact it had on UW’s RPI. I only went back 6 years because the available data gets fragmented beyond that. Here are the statistics:

SeasonRecordRPINC SOSAve RPI of NC oppTotal SOS
2011/1224-11 (14-5)54136148.763
2010/1123-10 (14-7)3260122.543
2009/1024-10 (14-7)414012160
2008/0924-8 (15-5)147714519
2007/0816-17 (7-12)122224179.462
2006/0719-13 (9-11)84267177.365

Above you can see that there is a strong correlation between having a poor non-conference schedule and a poor overall performance of the teams. In fact, let’s look at that chart again, this time sorted by non-conference RPI.

SeasonRecordRPINC SOSAve RPI of NC oppTotal SOS
2006/0719-13 (9-11)84267177.365
2007/0816-17 (7-12)122224179.462
2011/1224-11 (14-5)54136148.763
2008/0924-8 (15-5)147714519
2010/1123-10 (14-7)3260122.543
2009/1024-10 (14-7)414012160

Notice that the Huskies three worst seasons in terms of overall winning percentage and overall RPI happen to also be the three seasons when they had the worst non-conference strength of schedule. So, this begs the question; Did Romar realize the weaknesses of his teams and schedule accordingly? Or, did the team perform badly precisely because it was not adequately challenged early in the season?

Granted, playing Duke and Marquette in 2011/12 were plenty enough of a challenge in November/December and it turns out Saint Louis was pretty good too. But, they also played Florida Atlantic, Houston Baptist, and Cal St.-Northridge. This team did go on to win the Pac-12 regular season title, but it certainly was not a convincing run with very close wins over lowly Utah and Arizona State and two come from behind wins over Washington State, an almost miraculous comeback against UCLA, and a couple of blowout losses (Colorado and Oregon comes to mind).

Of course, these numbers are only numbers unless you also compare them to other major conference college basketball teams. So, I took a look at who has been scheduling the most competitive and least competitive non-conference schedules the past four years. These numbers below are the averages of the non-conference SOS from 2008-2011 compiled by Joe Lunardi at ESPN. I also included the numbers for the 2011-2012 I gathered from Warren Nolen’s site to see if they were consistent this year with the previous pattern.

Worst NC SOS 08-112011-12Best NC SOS 08-112011-12
Oregon State318.8328Tennessee19.5173
Iowa State272.8184Georgetown33.587
Stanford265211Arizona36.8156
Nebraska256.8299Duke44.516
Colorado250.892Michigan State47.5125
N.C. State248.843Connecticut52144
Northwestern248.5185Syracuse5276
Rutgers240.8297Michigan55.8187
Auburn238.8322Wisconsin59160
LSU232.8236Pittsburgh61.5209
Penn State226238West Virginia61.5115

OK, let’s directly address the three objections raised at the beginning one at a time: To compare, UW’s average NC SOS over the same time period was 134.0, which while not anywhere near the toughest in the nation, is also nowhere near the worst of the major conference offenders out there. In fact, UW’s non-conference schedules are actually pretty high relatively speaking. But, if you want to look at one small factor hurting UW’s RPI each year, it is the horribly weak schedules Oregon State and Stanford are playing! What’s up with that?

1) Not enough marquee names to get the national press (and elite recruits) excited.

  • In 2011/12, the Huskies lost at Saint Louis, at Nevada, to both Marquette and Duke in New York, and then of course to Minnesota in the NIT semi-final in New York as well.
  • In 2010/11, UW played in the Maui Invitational against Virginia (W), Kentucky (L), and Michigan State (L). That was a clearly a marquee event, but UW did not take full advantage of it. However, outside of those games, UW’s only “marquee” opponents were a road game against Texas A&M (L) and home to Texas Tech (W).
  • In 2009/10, UW played Georgetown in Los Angeles and lost, lost at Texas Tech, and beat Texas A&M at home after Derrick Roland destroyed his leg.
  • In 2008/09, UW went to Kansas City to lose two games against Kansas and Florida, but beat Oklahoma State at home. None of the others were marquee.
  • In 2007/08, UW did play a marquee lineup, but it didn’t turn out that well. They lost to Texas A&M, Syracuse, Pittsburgh, Oklahoma State, but did manage to beat Utah and LSU.
  • In 2006/07, UW lost to Gonzaga, but beat LSU with no other games worth mentioning.

UW has established a pattern of playing just about 3-4 “marquee” games out of the 12 non-conference games. But, I suspect the bigger issue Husky fans have with these games is not the quantity of games, but rather that UW loses most of them. During this 5-year period UW is just 7-17. If the UW was 12-12 or better, something tells me people wouldn’t complain as much.

2) Not enough true road games to test the players before the conference season starts.

UW has generally only played one true road game in the non-conference schedule and Romar tends to prefer neutral court games. This past year Romar decided to roll the dice with two true road games and two neutral site games. Unfortunately, they lost all four. You might expect that to make Romar a little gun shy to try that again.

But, to his benefit, he will have a road game against UConn, in addition to the Hall of Fame Tip-off Classic at the Mohegan Sun in Connecticut against Seton Hall and then either Ohio State or Rhode Island the next day.

3) Too many pansies and no-name teams coming to Hec Ed for the season ticket holders.

People who pay good money want to see Duke come to Hec Ed, not Houston Baptist. I am not going to challenge this point. But, it should be realized that teams tend to schedule teams who they think they can beat and teams who want to come. Duke isn’t going to risk losing on the road any more than UW is. It’s hard to convince the major conference teams to come to UW for an almost certain loss. But, at least in 2012 Saint Louis and Nevada will be coming to town as return legs of last year’s road games and UConn will come to town in 2013. So, Romar is trying to beef up the home schedule.

One issue with scheduling is the difficulty of predicting which teams will be Top 100 RPI, Top 50 RPI, versus sub-200. Let’s take a look at a case study of the effects of scheduling on a team’s post-season fortunes. Last year Iona’s (25-8 overall, 15-3 in MAAC) overall SOS was 143rd, while Drexel’s (29-7 overall, 16-2 in Colonial) was 213th. However, this disparity doesn’t seem right when you compare the resumes side by side: both teams played two Top 50 RPI opponents and Drexel played nine Top 100 RPI opponents to Iona’s eight. Each team played 18 Top 200 RPI opponents.

The reason for the disparity can be seen when analyzing lowest RPI teams of each schedule. Drexel played 11 games against teams rated 250th or worse in the RPI, while Iona played just 6 games against such teams (and lost one of them to Hofstra).

What makes the SOS so misleading is that the RPI weights the difference between 200th team in the country and the 345th team (145 spots) the same as the difference between the 1st and 145th team in the country. But, we all know that team strength is not a linear thing. So, even though both schools faced an equal number of competent opponents, and Drexel faced more top 100 opponents, Iona has a large overall edge because the terrible teams Drexel played were slightly worse than the terrible Iona played.

In reality, any team hoping to make the NCAA tournament ought to be able to beat the 200th ranked team just as easily as the 300th ranked team. They should not receive just a huge RPI boost for it. But, given the reality of the situation, it makes no sense for a team like the Huskies to schedule teams who clearly will be 300+ RPI teams (such as Houston Baptist). It will only serve to cause the RPI and SOS to be lowered without any great increase in the likelihood of victory.