Tennessee (8-4) will play Nebraska (9-3) in the Music City Bowl. But, what can really be expected?
For those of you who have followed my evaluations since the Dooley era, you know the predictive power of recruiting evaluations. For those of you who haven't, a brief summary is that averaging four years of rivals recruiting produces an approximately 70% prediction rate between any two teams, almost 100% in national championship games (since 2005 only one less talented team has won), and the stability of these predictions is an even better indicator of seasonal predictions. For instance, something like 85% of the teams finish the season within a game of predictions (Ex: UT had a 9-3 seasonal prediction using this metric, and finished 8-4. This is well within normal variation). The real problem this season is that UT was wildly over-valued by many fans and pundits, but that didn't change the reality that talent for UT would orbit within a game of 9-3.
When I tell people that the B1G is the most over-valued conference, I get scoffed at. The reality is that the B1G is the least talented conference of the power5 having the lowest average team rating. The SEC has an average talent somewhere between Oklahoma and Michigan, the B1G has an average talent somewhere between Missouri and Vanderbilt.
Nebraska will enter the game with a talent average of 26.75 and UT with an 11.25. As a threshold matter this evaluation alone gives UT the 70/30 advantage. For comparison, Nebraska would be the 6th most talented team UT will face in 2016. Tennessee will be the 2nd most talented team that Nebraska will face.
But there is more to the evaluation than simply looking at talent, it is about looking at how a team performed in relation to that talent.
This chart shows the relative talent that both Tennessee and Nebraska faced in 2016. Both teams under-performed their predictions by 1 game. Tennessee's best win against talent was against an 8.25 UGA team, Nebraska's was a 23.26 Oregon team. UT's worst loss was against a Vanderbilt team at 43.75 whereas Nebraska's came against an Iowa team that is 53.5. In sum, it means that Nebraska has not illustrated an ability to perform well enough to overcome the talent advantage that UT has, but UT has lost to a lessor talented team than Nebraska (beating two more talented teams than Nebraska, and a third [Kentucky] that is virtually identical with Nebraska from a talent standpoint).
The conclusion is that talent favors Tennessee, and that both teams are performing the same way in relation to talent making it even less likely that Nebraska has the horses to overcome a far more talented UT. That said, as with any game the lessor talented team could win (30% isn't insignificant), but Nebraska isn't even as talented as SCAR, and has no indicia of seasonal over-performance in the way that Vanderbilt did. Note also how close Vanderbilt and Wisconsin are in talent.
Finally, for those who want to argue that the B1G is the better conference, I offer you two visualizations. The first is the combination of Nebraska and Tennessee's opponents ranked in relation to aggregate talent. The second is all of the P5 conferences teams ranked in relation to their conference talent and each other. Note that only three of Nebraska's opponents would fall between Alabama and Vanderbilt.
EDIT: I forgot a point of further comparison. If Nebraska was put into UT's schedule, their predicted outcome would be 7-5. Conversely, if UT was put into Nebraska's schedule, the predicted outcome would be 11-1.
For those of you who have followed my evaluations since the Dooley era, you know the predictive power of recruiting evaluations. For those of you who haven't, a brief summary is that averaging four years of rivals recruiting produces an approximately 70% prediction rate between any two teams, almost 100% in national championship games (since 2005 only one less talented team has won), and the stability of these predictions is an even better indicator of seasonal predictions. For instance, something like 85% of the teams finish the season within a game of predictions (Ex: UT had a 9-3 seasonal prediction using this metric, and finished 8-4. This is well within normal variation). The real problem this season is that UT was wildly over-valued by many fans and pundits, but that didn't change the reality that talent for UT would orbit within a game of 9-3.
When I tell people that the B1G is the most over-valued conference, I get scoffed at. The reality is that the B1G is the least talented conference of the power5 having the lowest average team rating. The SEC has an average talent somewhere between Oklahoma and Michigan, the B1G has an average talent somewhere between Missouri and Vanderbilt.
Nebraska will enter the game with a talent average of 26.75 and UT with an 11.25. As a threshold matter this evaluation alone gives UT the 70/30 advantage. For comparison, Nebraska would be the 6th most talented team UT will face in 2016. Tennessee will be the 2nd most talented team that Nebraska will face.
But there is more to the evaluation than simply looking at talent, it is about looking at how a team performed in relation to that talent.
This chart shows the relative talent that both Tennessee and Nebraska faced in 2016. Both teams under-performed their predictions by 1 game. Tennessee's best win against talent was against an 8.25 UGA team, Nebraska's was a 23.26 Oregon team. UT's worst loss was against a Vanderbilt team at 43.75 whereas Nebraska's came against an Iowa team that is 53.5. In sum, it means that Nebraska has not illustrated an ability to perform well enough to overcome the talent advantage that UT has, but UT has lost to a lessor talented team than Nebraska (beating two more talented teams than Nebraska, and a third [Kentucky] that is virtually identical with Nebraska from a talent standpoint).
The conclusion is that talent favors Tennessee, and that both teams are performing the same way in relation to talent making it even less likely that Nebraska has the horses to overcome a far more talented UT. That said, as with any game the lessor talented team could win (30% isn't insignificant), but Nebraska isn't even as talented as SCAR, and has no indicia of seasonal over-performance in the way that Vanderbilt did. Note also how close Vanderbilt and Wisconsin are in talent.
Finally, for those who want to argue that the B1G is the better conference, I offer you two visualizations. The first is the combination of Nebraska and Tennessee's opponents ranked in relation to aggregate talent. The second is all of the P5 conferences teams ranked in relation to their conference talent and each other. Note that only three of Nebraska's opponents would fall between Alabama and Vanderbilt.
EDIT: I forgot a point of further comparison. If Nebraska was put into UT's schedule, their predicted outcome would be 7-5. Conversely, if UT was put into Nebraska's schedule, the predicted outcome would be 11-1.
Last edited: