for those interested: https://www.footballoutsiders.com/stats/ncaa
utah state
overall: #21
offense: #20
defense: #34
special teams: #66
2nd order wins (differential): 9 (1.0 games overachieved)
for comparison:
texas tech
overall: #44
offense: #25
defense: #91
special teams: #6
2nd order wins (differential): 6.2 (1.2 games underachieved)
scheier really was a godsend this past season
note: 2nd order wins = number of games you should have won based on the stats, so a measure of under/overachieving. its a decent predictor of future success: if you are a significant underachiever, then the next season has decent probability of improvement, and likewise the other way around. obviously things like weather, injuries, etc have an effect on the following season's output, but it does have a pretty strong correlation when you get above 1.0 in either direction
here is a comparison of teams who had a similar level of "underachievement" and the W/L of the following season (2016 & 2017 seasons for brevity):
2016 season
team - underachieving year record - following year record
LSU - 8-4 (1.2 games underachieved) - 9-4 (1 win improvement)
ole miss - 5-7 (1.0 games underachieved) - 6-6 (1 win improvement)
TCU - 6-7 (1.4 games underachieved) - 11-3 (5 win improvement)
iowa state - 3-9 (1.2 games underachieved) - 8-5 (5 win improvement)
oregon state - 4-8 (1.2 games underachieved) - 1-11 (3 win regression)
missouri - 4-8 (1.2 games underachieved) - 7-6 (3 win improvement)
oregon - 4-8 (1.2 games underachieved) - 7-6 (3 win improvement)
arkansas state - 8-5 (1.1 games underachieved) - 7-5 (1 win regression)
tulane - 4-8 (1.0 games underachieved) - 5-7 (1 win improvement)
georgia state - 3-9 (1.4 games underachieved) - 7-5 (4 win improvement)
unlv - 4-8 (1.2 games underachieved) - 5-7 (1 win improvement)
kent state - 3-9 (1.1 games underachieved) - 2-10 (1 win regression)
florida atlantic - 3-9 (1.1 games underachieved) - 11-3 (8 win improvement)
fresno state - 1-11 (1.3 games underachieved) - 10-4 (9 win improvement)
umass - 2-10 (1.1 games underachieved) - 4-8 (2 win improvement)
so, of 15 teams that had a similar underachieving record to tech (+/- 0.2 games):
12 teams improved (80%), and improved on average 3.58 wins
3 teams regressed (20%), and regressed on average 1.67 wins
overall, on average, if you underachieved by 1-1.4 games in 2016, you could reasonably hope for an improvement of around 2.5 games in 2017
2017 season
team - underachieving year record - following year record
utah - 7-6 (1.0 games underachieved) - 9-4 (2 win improvement)
arizona - 7-6 (1.0 games underachieved) - 5-7 (2 win regression)
utah state - 6-7 (1.3 games underachieved) - 10-2 (4 win improvement)
eastern michigan - 5-7 (1.4 games underachieved) - 7-5 (2 win improvement)
umass - 4-8 (1.2 games underachieved) - 4-8 (no change)
BYU - 4-9 (1.3 games underachieved) - 6-6 (2 win improvement)
nevada - 3-9 (1.0 games underachieved) - 7-5 (4 win improvement)
coastal carolina - 3-9 (1.0 games underachieved) - 5-7 (2 win improvement)
rice - 1-11 (1.3 games underachieved) - 2-11 (1 win improvement)
note: i did not include idaho, who moved to the FCS between 2017 and 2018
so, of 9 teams that had a similar underachieving record to tech (+/- 0.2 games):
7 teams improved (77.7%), and improved on average 2.43 wins
1 teams regressed (11.1%), and regressed on average 2 wins
1 teams record remained unchanged (11.1%)
overall, on average, if you underachieved by 1-1.4 games in 2016, you could reasonably hope for an improvement of around 1.7 games in 2018
between the two years, teams with a similar 2nd order wins differential to tech won about 2 more games or so on average the following year
----------------
side note, for utah state we can see the metric at work during their 3-year losing streak:
2014 - 10-4 (1.0 games overachieved)
2015 - 6-7 (0.8 games underachieved)
2016 - 3-9 (1.8 games underachieved)
2017 - 6-7 (1.3 games underachieved)
2018 - 10-2 (1.0 games overachieved)
strong momentum between 2015-2017 that a breakthrough would happen, and it did in 2018
and texas tech over the same stretch
2014 - 4-8 (0.8 games underachieved)
2015 - 7-6 (0.0 games under/overachieved)
2016 - 5-7 (0.3 games underachieved)
2017 - 6-7 (0.9 games underachieved)
2018 - 5-7 (1.2 games underachieved)
looks like we have some momentum going in the same way now
TL;DR: texas tech will be going 7-5 next season
utah state
overall: #21
offense: #20
defense: #34
special teams: #66
2nd order wins (differential): 9 (1.0 games overachieved)
for comparison:
texas tech
overall: #44
offense: #25
defense: #91
special teams: #6
2nd order wins (differential): 6.2 (1.2 games underachieved)
scheier really was a godsend this past season
note: 2nd order wins = number of games you should have won based on the stats, so a measure of under/overachieving. its a decent predictor of future success: if you are a significant underachiever, then the next season has decent probability of improvement, and likewise the other way around. obviously things like weather, injuries, etc have an effect on the following season's output, but it does have a pretty strong correlation when you get above 1.0 in either direction
here is a comparison of teams who had a similar level of "underachievement" and the W/L of the following season (2016 & 2017 seasons for brevity):
2016 season
team - underachieving year record - following year record
LSU - 8-4 (1.2 games underachieved) - 9-4 (1 win improvement)
ole miss - 5-7 (1.0 games underachieved) - 6-6 (1 win improvement)
TCU - 6-7 (1.4 games underachieved) - 11-3 (5 win improvement)
iowa state - 3-9 (1.2 games underachieved) - 8-5 (5 win improvement)
oregon state - 4-8 (1.2 games underachieved) - 1-11 (3 win regression)
missouri - 4-8 (1.2 games underachieved) - 7-6 (3 win improvement)
oregon - 4-8 (1.2 games underachieved) - 7-6 (3 win improvement)
arkansas state - 8-5 (1.1 games underachieved) - 7-5 (1 win regression)
tulane - 4-8 (1.0 games underachieved) - 5-7 (1 win improvement)
georgia state - 3-9 (1.4 games underachieved) - 7-5 (4 win improvement)
unlv - 4-8 (1.2 games underachieved) - 5-7 (1 win improvement)
kent state - 3-9 (1.1 games underachieved) - 2-10 (1 win regression)
florida atlantic - 3-9 (1.1 games underachieved) - 11-3 (8 win improvement)
fresno state - 1-11 (1.3 games underachieved) - 10-4 (9 win improvement)
umass - 2-10 (1.1 games underachieved) - 4-8 (2 win improvement)
so, of 15 teams that had a similar underachieving record to tech (+/- 0.2 games):
12 teams improved (80%), and improved on average 3.58 wins
3 teams regressed (20%), and regressed on average 1.67 wins
overall, on average, if you underachieved by 1-1.4 games in 2016, you could reasonably hope for an improvement of around 2.5 games in 2017
2017 season
team - underachieving year record - following year record
utah - 7-6 (1.0 games underachieved) - 9-4 (2 win improvement)
arizona - 7-6 (1.0 games underachieved) - 5-7 (2 win regression)
utah state - 6-7 (1.3 games underachieved) - 10-2 (4 win improvement)
eastern michigan - 5-7 (1.4 games underachieved) - 7-5 (2 win improvement)
umass - 4-8 (1.2 games underachieved) - 4-8 (no change)
BYU - 4-9 (1.3 games underachieved) - 6-6 (2 win improvement)
nevada - 3-9 (1.0 games underachieved) - 7-5 (4 win improvement)
coastal carolina - 3-9 (1.0 games underachieved) - 5-7 (2 win improvement)
rice - 1-11 (1.3 games underachieved) - 2-11 (1 win improvement)
note: i did not include idaho, who moved to the FCS between 2017 and 2018
so, of 9 teams that had a similar underachieving record to tech (+/- 0.2 games):
7 teams improved (77.7%), and improved on average 2.43 wins
1 teams regressed (11.1%), and regressed on average 2 wins
1 teams record remained unchanged (11.1%)
overall, on average, if you underachieved by 1-1.4 games in 2016, you could reasonably hope for an improvement of around 1.7 games in 2018
between the two years, teams with a similar 2nd order wins differential to tech won about 2 more games or so on average the following year
----------------
side note, for utah state we can see the metric at work during their 3-year losing streak:
2014 - 10-4 (1.0 games overachieved)
2015 - 6-7 (0.8 games underachieved)
2016 - 3-9 (1.8 games underachieved)
2017 - 6-7 (1.3 games underachieved)
2018 - 10-2 (1.0 games overachieved)
strong momentum between 2015-2017 that a breakthrough would happen, and it did in 2018
and texas tech over the same stretch
2014 - 4-8 (0.8 games underachieved)
2015 - 7-6 (0.0 games under/overachieved)
2016 - 5-7 (0.3 games underachieved)
2017 - 6-7 (0.9 games underachieved)
2018 - 5-7 (1.2 games underachieved)
looks like we have some momentum going in the same way now
TL;DR: texas tech will be going 7-5 next season
