The Momentum of a Season Changes Quickly

How is Momentum Defined

One of the exciting aspects of the College Football Prediction Model I created is that not only does it show the odds of teams winning in a future game, say, next week, but it is also possible to see how momentum from a game can kill or make a season.

Throughout a season, teams can follow one of many trajectories:

• Consistent: Start strong and finish strong or start weak and end weak. Alabama and Clemson are great examples of teams that have stayed strong both this season and last season. Both teams collectively only lost three games altogether. On the other end, Kansas has just won four games so far this season and last year combined.
• Collapse: Start strong and end weak. After the 2006 season that saw contention with Oklahoma for the Big 12 Championship, Nebraska went into 2007 hoping to repeat its success. What followed was a 5–7 season, five straight losses, and the firing of its coaching staff.
• Rebound: Start weak and finish strong. Central Florida has been on a 20 consecutive winning streak that started last season under former HC Scott Frost. That marks a sharp contrast from its 0–12 record in 2015 in what became a remarkable turnaround.

Losing Momentum — Kansas State

Winning or losing a game or a series of crucial games has the capability of changing the trajectory of future games. The model visualizes this pretty well.

For instance, Kansas State (KSU) started this year as a favorite in most of its games. At the start of the season, the model gave KSU a 2 in 3 chance of beating Baylor and Texas, and a 7 in 10 chance of defeating Texas Tech.

The model also gave the Wildcats a slim advantage over Mississippi State but forecasted them as slim underdogs against Oklahoma State, Iowa State, and West Virginia. Overall, the forecast was a 7–5 season which would earn them another bowl game and a chance to repeat their 8–5 season in 2017.

That rosy picture did not last.

A variety of factors took place in the early weeks of this Season that changed KSU’s beginning season outcome — significant loses to Mississippi State at home and West Virginia in Morgantown. Meanwhile, the games against Baylor and Texas became more competitive than at the beginning of the season.

Both games against the Bull Dogs and the Mountaineers sent KSU’s ELO rating to 1,555 and 1538 by the Texas and Baylor games, respectively. Kansas States’ ELO at the beginning of the season was 1591.

Meanwhile, Texas and Baylor were able to lose and win enough games also to make their matches against KSU more competitive. Texas posted decisive wins against Tulsa, Southern California, and Texas Christian after losing to Maryland in the first week. Baylor, although falling to Duke and Oklahoma, won three other games against Abilene Christian, Texas-San Antonio, and Kansas. When KSU faced Texas and Baylor, the odds of Wildcat victories shrunk to 50% and 47%, respectively. KSU lost both games.

Kansas State did beat the forecast in one game — a 31–12 upset over the Oklahoma State Cowboys (KSU was a 42% favorite in that game) before falling two weeks later to the Oklahoma Sooners 51–14 (KSU was an 18% favorite in that game).

Now at 3–5 overall and 1–4 conference, Kansas State heads into November 3rd against Texas Christian (3–5, 1–4 conf), Kansas (3–5, 1–4 conf), Texas Tech (5–3, 3–2 conf), and lastly Iowa State (4–3, 3–2 conf). If they win at least three of those games, KSU still has a chance of making a bowl game.

The task will be tall, however, as the forecast predicts Kansas State only to win two of those games. As of this week, both forecasted wins are against Texas Tech and Kansas in Manhattan while Texas Christian and Iowa State are both on the road. If the forecast holds, that will leave the Wildcats with a 5–7 record just shy of a bowl game.

All eyes will be on Bill Snyder to see whether the Wildcats’ season will improve. Overall, the losses posted by Kansas State have slowed their momentum and decreased their winning odds against multiple teams. The remaining four games will be decisive in determining whether or not the dynamic changes for the better or worse.

You can interact with the dashboard above in this link here.

Correction: A special thanks to Sports-Reference for providing the data.