There were eight returning starters and additional players with considerable playing experience. The unit had finished strong in the 2016 season, heavily impacting the season-ending four-game winning streak. There was plenty of reason to believe that Georgia Tech’s defense could make a breakthrough.

One could conclude that the Yellow Jackets did play better. In fact, defensive coordinator Ted Roof, whose job security is in question, will pick up two of the five performance bonuses in his contract tied to the play of his defense. Roof’s contract, which was extended after last season by two years through the 2019 season, calls for him to receive an extra $10,000 for finishing in the top five in the ACC (in league games only) in scoring defense, total defense, third-down defense, red-zone defense and takeaways.

Tech was second in third-down defense (26.6 percent) and third in total defense (358.5 yards per game), good for an extra $20,000 in Roof’s bank account on top of his $800,000 salary. (Tech didn’t do so well in the other categories -- ninth in scoring defense, tied for 12th in takeaways and last in red-zone conversion efficiency.)

However, it would be difficult to find many Tech supporters who saw reason to be encouraged about their defense this season. Not a small part of the disappointment lies in the way that those returning starters – a large part of the reason for the optimism at the season’s start – failed to progress and be impactful.

The secondary, which returned all five starters (cornerbacks Lance Austin and Step Durham, nickel back Lawrence Austin and safeties A.J. Gray and Corey Griffin), was not disruptive. Tackling was often an issue. Pass breakups were in short supply. Potential interceptions were dropped. The five players combined for eight interceptions last year but came up with five this season, four of them concentrated in two games.

A year ago, the Tech secondary was not an elite group, but had a knack for coming through with well-timed plays. This time, the defensive backs seemed cursed by unlucky bounces and repeatedly came up a hair short to get hands on balls or bring down tacklers.

In his second full season starting at linebacker, Brant Mitchell was fourth on the team in tackles with 51 despite missing two games to injury. However, for whatever reason, he couldn’t make the sort of impact plays that could have changed the fortune of a team that lost three games in the final minutes. He had three quarterback hurries, one more than last season. He had 1.5 tackles for loss, fewer than his three in 2016.

Defensive end KeShun Freeman, a four-year starter and captain, had 22 tackles, three tackles for loss, no sacks and one quarterback hurry. Statistically speaking, his freshman season was his best, when he had 54 tackles, 9.5 tackles for loss and 4.5 sacks. Over the final three seasons, he had 2.5 sacks and 11 tackles for loss. After starting seven games a year ago, defensive tackle Kyle Cerge-Henderson saw his role diminished in the rotation.

The numbers don’t say everything about a player. A scheme or play call may be designed for a player to play a role that frees others to make the play. Interceptions and pass breakups can be influenced by the pass rush and the coverage called. But it would be difficult to say any of the eight returning starters was a consistent impact player for the defense.

For the first time since 2010, there wasn’t one defensive player for Tech who didn’t earn at least honorable mention in All-ACC voting.

They have largely been outstanding representatives of the school and their team. Their effort could rarely be questioned. However, there are questions about much else – recruiting, development, scheme and execution. There is a reason that Roof’s future with the team is in question.

For the defense, the most lasting memories and feelings may be connected to the unit’s play in the final two games, against Duke and Georgia, when the Jackets gave up 81 points in 17 possessions, which excludes both teams’ final clock-killing drives. Fair or not, it obscures other moments, such as the game-changing interception by Lawrence Austin against Jacksonville State, a two-interception game by Gray against North Carolina and Mitchell’s nine-tackle game against Virginia Tech.

On the whole, the defense was either pretty good (a unit doesn’t rank in the top 15 nationally in third-down defense by accident) or ineffective, sometimes in the same game. But, it’s hard to shake the feeling that a considerable opportunity was lost, not least for the eight returning starters who could have helped shape a consistently effective defense but were unable.