Lopsided games and playing the College Football Playoff semifinals on New Year's Eve for the first time led to a 13 percent drop in TV ratings for the New Year's Six Bowls from last year.

The Rose Bowl on Friday drew its lowest rating (7.9) since it became part of the BCS in 1999.

But don't expect any changes to future College Football Playoff schedules based on the one-year dip.

The overnight ratings for the big New Year's Day bowls were announced Saturday by ESPN, which broadcasts all the games.

Overall, the ratings for the six major games played Thursday and Friday, including the semifinals in the Orange and Cotton bowls Thursday, averaged a 7.1 overnight rating, down from 8.2 last season, when the first playoff games drew record-breaking audiences to ESPN.

"That decline, frankly, is not much of a surprise and it's modest," College Football Playoff executive director Bill Hancock said. "It's too soon to know how much was due to the lopsided games or how much what I think we all thought would be an inevitable decline from the excitement of the first year or the semifinals on New Year's Eve. I suspect it's a combination of those three, but I don't have any idea what the weighting is. ESPN is studying the numbers and we'll learn a lot more in the next few months."

The average margin of victory in the New Year's Six games was 24.2 points.

All three of Friday's major bowls were blowouts. The Fiesta between Ohio State and Notre Dame, earned a 6.2 rating. That was up 35 percent from last season's Arizona-Boise State game. The Buckeyes beat the Fighting Irish 44-28.

Stanford beat Iowa 45-16 in the Rose Bowl on Friday. The previously lowest-rated Rose Bowl since 1999 was the 2012 game between Stanford and Wisconsin, which drew a 9.4.

The Sugar Bowl, which Mississippi won 48-20 against Oklahoma State, drew a 5.3 rating. That was the lowest Sugar Bowl since the Bowl Championship Series was established in 1999.

ESPN did see big increases in digital viewers who streamed the game online through WatchESPN. ESPN said the New Year's Six averaged 776,000 unique viewers and 43,871,000 total minutes for the six games, up 54 percent and 67 percent, respectively, from last year.

"The College Football Playoff is a long-term, multiplatform play for us," Burke Magnus, ESPN executive vice president of programming and scheduling, said in a statement. "With that said, there are many variables that impact ratings results including what happens on the field and the numbers this year were obviously impacted by the unbalanced scores of these games."

Last season's College Football Playoff semifinals, played on New Year's Day, drew a record number cable viewers, and ratings that edged past 15 for both the Oregon-Florida State Rose Bowl and the Ohio State-Alabama Sugar Bowl.

Total viewership for the semifinals played Thursday between Michigan State and Alabama in the Cotton Bowl and Clemson and Oklahoma in Orange Bowl plunged 34.4 percent, from 28,271,000 in 2015 to 18,552,000.

"What that tells me is many, many fans found a new way to watch the games," Hancock said.

ESPN has the rights to all six games. This is the second season of 12-year contracts worth a total of $7.3 million to the major college football conferences. The Rose and Sugar bowls have separate television deals with ESPN that lock in their prime New Year's Day time slots. When the College Football Playoff semifinals are played in the other four games in the New Year's Six rotation, they will be played on New Year's Eve.

ESPN suggested moving this season's semifinals to Jan. 2, a Saturday with no NFL games to compete against, but playoff officials did not want to delay the start of what they hope can become a new tradition.

Next season, the Fiesta and Peach bowls will host the semifinals on New Year's Eve, which falls on a Saturday. The semifinals are back in the Rose and Sugar bowls, and on New Year's Day, after the 2017 season.

"There hasn't been discussion in our group at all about changing the dates," Hancock said.

___

AP College Football website: collegefootball.ap.org