This is why context matters in statistics. There's no real "trend" here when context is considered. This false "trend" was pointed out as a criticism when he was hired and for some reason it's getting repeated. He took over an extremely talented undefeated team his first season, went undefeated before losing in the Fiesta Bowl. That year the conference was fairly weak, with only UCF and Cincy ranked. They played to their expectation.
His second season they won 10 games and their conference had 4 ranked teams (Navy, Cincy, Memphis, UCF). They lost star QB Milton for the season and rolled with a true freshman. Easy to say they outperformed themselves with that situation.
Then they had a rebuilding year in 2020. This probably was a combination of recruiting class attrition from when Frost left and upperclassmen graduating. Covid also disrupted the season. They had to drop out of conference games, had multiple players opt-out, had a complete halt to fall practice due to UCF athletes expressing concerns with the AD. They went 6-4, beating Georgia Tech in Atlanta, losing to #7 Cincy (by 3), Tulsa, Memphis (by 1). Then lost to #16 BYU in a bowl. They were scheduled to play UNC and FIU and one other team. They could've beaten all three. Would have to call that a scratch.
The Heupel comes to Knoxville and everyone expected him to not win more than 4 games. We saw what he was able to put together. We went bowling. We outperformed expectation, and played above our means because of roster/depth issues and no cohesion prior to him coming. Went 7-6.
Based on the roster we have coming back, improved depth and leadership. Then looking at the schedule, the Vols should win between 7-10 games. 10 would be above expectation. 7 would be probably just below.
Then in 2023 I would expect a slight dip, then rise in 24, 25, and 26.
.
