Coming into this season, Indiana fans had every right to worry about how this team's size disadvantage would translate into on-court play. And so far, the results on the court defending against top big men have not been great. Montrezl Harrell had a field day against us for Louisville, and when Josh Smith entered the game for the Hoyas during the second half, it completely changed the dynamics of last Saturday's contest. Venky Jois of Eastern Washington also had no trouble scoring 20 points in their upset victory over the Hoosiers last month.
But what I didn't foresee becoming such an issue was how opposing guards would light it up from mid- and long-range against us. In fact, it's the guard play that has been causing the Hoosiers trouble in our losses so far this season. And the numbers back it up: four guards from the three teams that beaten us have had offensive ratings that are statistically significantly higher than their average offensive ratings. In addition, these guards averaged an effective field goal percentage of over 10 percent higher than normal against IU than against their other opponents. These four guards who helped propel their teams to victories over IU with better-than-average games were Drew Brandon (EWU), Chris Jones and Terry Rozier (Louisville), and Indianapolis native D'Vauntes Smith-Rivera (Georgetown).
In our inaugural Indianalytics post, let's go deeper into the offensive numbers to see how guards on the opposing teams have made an impact in our three losses so far. For this analysis, we're going to look at two particular factors: offensive rating (ORtg) and effective field goal percentage (eFG%). To quickly recap these stats, offensive rating on a player-by-player basis is the number of points produced per 100 possessions. When discussed from a team perspective, it refers to the points scored per 100 possessions. Effective field goal percentage takes normal field goal percentage and brings three-point field goals into account, and effectively counting threes as being worth 1.5 field goals. In addition, many thanks to the good folks at KenPom and ShotAnalytics, for the wealth of data from their sites that I used to put this analysis together.
Want to see some raw data first? Well good, because here's a chart I compiled! As a side note, all of this data is through games on December 30.
Player |
Brandon |
Jones |
Rozier |
Smith-Rivera |
Team |
EWU |
Louisville |
Louisville |
Georgetown |
ORtg vs IU |
135 |
126 |
144 |
142 |
Avg ORtg |
111.9 |
94.6 |
111.9 |
122.7 |
Standard score |
0.57 |
1.77 |
1.48 |
0.65 |
ORtg percentile |
71.5 |
96.3 |
93.1 |
74.2 |
FG% season |
45% |
31% |
45% |
43% |
eFG% season |
46% |
37% |
50% |
51% |
FG% vs IU |
63% |
47% |
47% |
50% |
eFG% vs IU |
63% |
53% |
62% |
64% |
Now, let's take a closer look at the numbers from this chart and explain what everything means. Let's start with the first loss of the season, Eastern Washington. The Eagles' star guard, Tyler Harvey, continued his terrific season against us. He's averaging 22.8 points per game, which is good for 4th in the nation, and scored 25 against the Hoosiers. His offensive rating in the win was a 126, just above his season average of 125.6. EWU's other guard, however, was the one that hurt us. Drew Brandon, the point guard for the Eagles, has an O-rating of 111.9 on the year. Against the Hoosiers, his rating was 135. In addition, while his yearly eFG% has been at 46%, this jumped up to 63% at Assembly Hall.
Oh sure, you may say. One guy from a mid-major program caught fire against us. It wouldn't be the first time that happened and it won't be the last. But wait, let's see how the Louisville game shook out. Sure, Herrell was dominant from the post, but take a closer look at the guard play from Jones and Rozier. Chris Jones has been a subject of much conversation in Louisville as this season has progressed, and his poor shooting has resulted in an eFG% of 37% on the year. Against the Hoosiers, however, he caught fire, and this jumped to 53% on the game. The Cards' other starting guard, Terry Rozier, has shot better for the most part this season, averaging an eFG% of 50%. What did his eFG% climb up to in the IU game though? 62 percent.
Well sure, but that was Louisville. They're a top-10 opponent and we let them run up the score at the end. But wait - there's one more guard who had a stellar offensive day in a game that IU lost, and that's D'Vauntes Smith-Rivera of Georgetown. Last Saturday, DSR had a 64% eFG% rating against the Hoosiers, 13 percentage points higher than his average of 51%. As a result, there have been four guards in our three losses whose effective field goal percentages were over 10 percent higher than their averages for the season.
To get a better idea of how well these guards played, let's look at the standard scores (or Z-scores) of their offensive ratings. These scores, which measure how far from the mean an individual O-rating is based on standard deviations, can show us how much better each guard's offensive performance was against IU than on an average night. Jones's offensive rating of 126 against IU was 1.77 standard deviations above his mean of 94.6, meaning that according to a normal distribution, his performance against the Hoosiers would be predicted to outperform almost 96 percent of all of his games. At 1.48 standard deviations, Rozier's game against IU (144 ofensive rating) would project to being in the 93rd percentile of all his games, in which he's averaged a rating of 111.9. Drew Brandon and D'Vauntes Smith-Rivera's standard scores aren't quite as gaudy, but they performed well enough against IU to be in the 71.5 and 74.2 percentile of what we'd expect from their offensive ratings. So in short, it's firmly above their averages.
Finally, we can do a one-sample t-test to further evaluate Indiana's struggles against opposing backcourts. Comparing the overall mean offensive ratings of these four guards (105.98) to the mean of their games against IU (136.75) results in the offensive ratings of the IU games being very statistically significant, compared to the average game for the players. The results of the test below, obtained from graphpad.com, demonstrate this significance:
Basically, I just wanted to run some numbers to verify my theory that guards have shot significantly better against IU than other teams on their schedule. Thus, like opposing quarterbacks have done against the Hoosier football teams in the past few years, opposing guards have lit it up against the Hoosiers so far this season. Obviously, the biggest caveat of this takeaway is that this is a small sample size, but in a season of only 31 games that lasts for four months, there is little time to correct the mistakes. Because the success of opposing guards against IU has come in nonconference play, the opponents' offensives ratings come against other nonconference opponents, which mostly are not always the same caliber as teams in the B1G. In addition, this data only accounts for guards on teams who have lost to IU, but even in the victories, we've let up some big performances. Josh Newkirk of Pitt, for example, had a nice game against us in the Panthers' 81-69 loss at Assembly Hall earlier this season.
My theory for why this is happening is that the focus on the undersized Hoosiers defending the post has led to more opportunities on the perimeter for both three-pointers and long-range jumpers. However, I'd have to look at more game film to substantiate that. Regardless, the big games for opposing guards when they face the Hoosiers are a troubling trend. This team seems to have fixed one of its major woes from last season, in that they're able to put points on the board at a consistent pace. Now, let's see if they can handle the other side of the ball as conference play begins.