/cdn.vox-cdn.com/uploads/chorus_image/image/65297809/usa_today_13396133.0.jpg)
A Brief Hypothetical...
I want to start this article with a question. If you had to choose between the following two options, what would you choose? A) Baylor plays in a visibly frustrating game in which they let a very bad Rice hang around but Baylor’s underlying performance is good, or B) Baylor comfortably wins against a bad Rice team but the score is misleading and the underlying performance is bad.
There is not necessarily a correct answer here; to each their own. However, if what you care about is Baylor being as good as possible in conference play, than you should choose option A.
Here’s the thing about statistics: they don’t necessarily win you the game. If you told me that Baylor made the conference championship game and gave me options between Baylor getting a very lucky win or Baylor statistically playing better but losing, give me the lucky win every time! It’s the conference championship game, who cares what the statistics might portend for the future.
But a non-conference game against Rice is different. It is not the end goal—all you need to do is win. What matters is how good Baylor actually is, because that will better predict how they play for the remainder of the season.
Baylor fans should be glad that this weird/unlucky/frustrating game came against one of the worst teams on it's schedule.
— Travis (@Travis_Roeder) September 22, 2019
The Game Was Not Fun to Watch. But Baylor Played Fine—Nowhere Near a Debacle
What Baylor Did Well
As I stated above, statistics don’t necessarily tell you who wins a game, but they are very good at predicting the outcome. You can start at the extremes: if I told you that this weekend Oklahoma will average 9 yards per play (YPP hereafter) against Texas Tech while Tech only averages 5 YPP itself, who would you wager won that game? Oklahoma, clearly. And here is what is crazy: only knowing that one stat, you would be correct 100% of the time. Here is a handy chart from Bill Connelly about the predictive power of YPP:
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/19222887/YPP_margin.jpg)
Against Rice, Baylor averaged 6.77 YPP while Rice only averaged a mere 3.77 (very bad!).
EDIT: @bawolfskill ran these numbers and found Baylor had a 6.92 YPP while Rice had a 3.67. I trust his numbers more because A) they’re better for Baylor and B) I calculated the YPPs on my phone calculator in the immediate aftermath of the game.
Some perspective: Baylor averaged 6.77 yards per play tonight, which would've been good enough for 8th in the country in the 2018 full season. Rice averaged 3.77 YPP, good enough in 2018 for... Dead last in the country!
— Travis (@Travis_Roeder) September 22, 2019
This, a margin of 3, means that from this one statistic, we know that Baylor wins somewhere between 95-99% of the time. Not only that, but most of the times that teams have a similar disparity in YPP, the scoring margin is closer to ~25 points. But wait! There’s more.
Another extremely predictive statistic is Success Rate (the percentage of plays that the offense gains 50% of needed yards on first down, 70% of needed yards on second down, or 100% of needed yards on third or fourth down). Success Rate is not yet available for the Baylor - Rice game, but from a close watch it was quite clear that Baylor had a decided advantage in efficiency. For example, in Baylor’s game against UTSA there was a 29% difference in success rate. I don’t think the disparity was quite as large in the Rice game, but I’d guess something around ~20%. And when that happens...
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/19222918/Baylor_rice_success_rate.jpg)
Again, with Baylor’s (estimated) performance against Rice, you expect a ~25 point scoring margin.
EDIT: @bawolfskill on twitter ran the numbers and found the success rate. My estimate was pretty close, as Baylor had an 18% advantage in success rate. Baylor had a 51.61% SR while Rice had a 33.33%. A 52% rate for Baylor would be amazing in conference but was probably a bit below what you want against Rice—For comparison, Baylor had a 58% rate against UTSA. But the 33% rate Baylor’s D held Rice to is awesome. The national average is 41%.
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/19223919/Baylor_rice_box_score.png)
What Baylor Did Poorly
There were a couple of related factors working against Baylor: field position and turnovers.
Baylor’s average starting field position was its own 24 yard line while Rice, on average, started at its own 35. Rice consistently kicked the ball short of Baylor’s end zone and yet Baylor struggled to get the ball out past their own 25. Besides the poor kick returns, the major reason for this discrepancy were Baylor’s two fumbles, which gave Rice the ball at its own 39 and at Baylor’s 25. This helped keep the game closer, as teams with such a large field position advantage usually win ~80% of the time. It makes sense when you think about it: if you need fewer yards to score, you can afford to be less explosive or efficient.
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/19222968/Baylor_rice_field_position.jpg)
And finally, turnovers, which are actually less predictive than you’d think. Baylor had 2 turnovers to Rice’s 1, giving Rice a +1 margin. Teams with a +1 margin win ~65% of games.
OK, So Baylor Did a Couple of Things Well and a Couple Things Poorly Against a Bad Rice—Isn’t That A Cause for Worry?
To put it simply, Baylor did the things that actually matter well, while the things it did poorly are much more correlated to luck and happenstance. Baylor dominated the game in both yards per play and efficiency. In generic football terms, Baylor moved the ball against Rice much, much better than Rice moved the ball against Baylor. When thinking about how good a team is, we generally think about a combination of how good they are on offense (how well they move the ball) and defense (how well they are stopping the opposing team from moving the ball). Baylor was very, very good at those two things against Rice.
The things that kept the score close against Rice—field position and turnovers—are much more susceptible to luck than how good a team actually is. Baylor’s -1 turnover margin made the game harder to win against Rice, but it doesn’t tell you much about what you expect the turnover margin to be against Iowa State. The negative turnover margin essentially happened because Charlie dropped the ball when he drew his arm back to throw—unless you think that is something that is liable to continue happening this season, there is zero reason to be concerned about it.
Succinctly, if Baylor had a smaller advantage in Yards Per Play or Success Rate I would be much more concerned than I am that Baylor had a bad result in field position and turnovers. What Baylor did well is much more predictive of how they’ll perform against Iowa State this coming Saturday than what they did poorly.
Baylor is Back in the SP+ Top 25
After a game that had many Baylor fans pulling their hair out, Baylor rose in SP+ back up to #25. This indicates that the underlying statistics of the game show that Baylor largely took care of business against Rice.
Baylor rises in both ranking and rating in SP+. This is mostly due to preseason projections being phased out more, but also because last night's performance against Rice wasn't nearly as bad as most Baylor fans thought it was.
— Travis (@Travis_Roeder) September 22, 2019
Also, the Big 12 is gonna be a mess. pic.twitter.com/7r8XBiGp5j