Basketball

By the Numbers Recap: How Accurate was the Data?

At the beginning of conference play, I set out to track ESPN Basketball Power Index’s projections about Texas Tech and the rest of the Big 12 throughout the season. I have to say, this recap will be a lot more fun now that the Red Raiders are conference champions. I enjoyed compiling these all season and hope you enjoyed following along. Now let’s see how accurate BPI was.

As of the end of the regular season for college basketball, Texas Tech is the No. 8 team in the country in BPI. Where did BPI have Tech pegged before the conference slate began? They were No. 8. That is borderline miraculous.

Below is a chart of how the Big 12 was ranked according to BPI at the start of conference play and at the conclusion. The top three and bottom two are the same, and the middle five are jumbled. That’s understandable since there seems to be quite a bit of parity in the middle of the conference. An obvious outlier is Kansas State, who split the conference title but is still pegged as a mid-tier Big 12 team.

Start of season
End of season
Texas Tech Texas Tech
Kansas Kansas
Iowa State Iowa State
Oklahoma
Kansas State
TCU Texas
Kansas State Oklahoma
Baylor Baylor
Texas TCU
West Virginia
West Virginia
Oklahoma State
Oklahoma State

 

Texas Tech’s final record of 14-4 had about a 15.5 percent chance of occurring according to BPI before conference play began. That might sound low but given all the possible outcomes prior to an 18-game conference season, it’s fairly high. It was the third highest probability behind 12-6 and 13-5 (between 19 and 20 percent each).

Another way to visualize this is that Texas Tech’s average win probability going into conference play was 69 percent. The Red Raiders wound up winning 78 percent of their games, so were within 9 percent of the original prediction.

At the beginning of conference play, Texas Tech and Kansas had nearly equal odds to win at least a share of the Big 12 title at 40-45 percent each. Iowa State was behind them in third. Texas Tech’s odds topped out at 74 percent when the Red Raiders were 4-0 out of the gates.

Of course, they lost their next three, which sent their title odds tumbling to 22 percent. After beating TCU and losing to Kansas to go to 5-4, their odds bottomed out at 16 percent. Interestingly, even at their low point, Texas Tech’s odds were higher than Kansas State’s at the time (13 percent).

So in terms of the title race, if you had taken a snapshot at random points during the season, BPI appears to have varying levels of accuracy. Starting out, BPI was higher on Tech than anyone else in the conference. And after four games, BPI was extremely high on Tech’s title odds. But if you had taken a snapshot two weeks later, you would have seen something that allegedly had just a 16 percent chance wound up coming true.

Overall, I’d say BPI mostly got it right. I’ll even defend their 16 percent title odds assessment once Tech reached 5-4. Winning nine straight games like Tech did is not a common occurrence.

BPI has Texas Tech No. 8 in the country. So does KenPom. The Red Raiders are No. 9 in the new NET rankings, and hovering around No. 8 in the AP poll as well. I’d say they’re a bona fide top 10 team in the country given all of this. And I think people could have seen it coming months ago if they had relied on the data instead of personal assessment.

Comments
To Top