If Gregory goes, where would Georgia Tech turn?

Three weeks ago, I believed Brian Gregory would be coaching Georgia Tech in 2016. I no longer believe as much. I’m not privy to athletic director Mike Bobinski’s thinking, but I can’t imagine that finishing 3-15 in ACC play in Year 4 under Gregory can be viewed in-house as anything but the final act of a failed stewardship.

Should Bobinski, who inherited this coach, choose to give Gregory one more chance, the only rationale would be that Tech has lost a bunch of close games and is overdue to win. My belief, stated here after the Jackets blew a 13-point lead against Louisville two weeks ago, was that some of those narrow losses can be traced to poor coaching, which isn’t a function of luck.

The only hedge: Jaylen Brown, the heralded senior from Class AAAAAA champ Wheeler, lists Tech among his possible destinations. He also lists Georgia, Kentucky, North Carolina, Kansas, Michigan and UCLA. Brown isn’t scheduled to announce his choice until April 1. If Bobinski waits to see what Brown will do, Tech will essentially have committed to Gregory — April is late to be firing a coach — for another year even if Brown winds up in Athens or Lexington or Westwood.

And let’s say Brown does choose Tech: He’d almost certainly be a one-and-done, and even if he powers the Jackets to the NCAA tournament next season, would that address the more systemic issues — a ponderous offense, substandard guard play and an over-reliance on transfers — that have beset Tech? Derrick Favors, another one-and-done, helped Tech to the Big Dance in March 2010; come March 2011, Paul Hewitt was gone.

To me, the choice seems obvious: A coach who’s 55-70 overall and 19-51 in league play has established a body of work, little of it good. Louisville, Syracuse, Notre Dame and Pitt have joined the ACC since Gregory arrived, rendering an already difficult league too difficult. I don’t know that one more year, even if the year is with Brown, or three more years will make much difference. I just don’t see Gregory winning big, or winning at all, in this conference.

Even as I say that, I go back to March 2011 and the weeks after Hewitt was dismissed. We all had our lists of candidates — Gregory was on mine, though not at the top — but it became clear that Tech wasn’t the destination of choice for many of those men. Gregg Marshall of Wichita State wasn’t interested. Chris Mooney of Richmond didn’t fly to Atlanta for his “interview,” doing it instead over the phone.

For a scary moment, it seemed the Jackets might wind up with Doug Wojcik, who’d never taken a team to the NCAA tournament and still hasn’t. He has since been fired by Tulsa for not winning enough and by the College of Charleston for verbal abuse of his players. If the alternative was Wojcik, then Gregory of Dayton was the better choice.

And that’s the point: The market for the best mid-major coaches keeps tightening. Four years after taking VCU to the Final Four, Shaka Smart remains in place and has spurned UCLA and North Carolina State. When finally Brad Stevens left Butler, it wasn’t for another college but for the Boston Celtics. The lure of a Power Five job, even one in the ACC, isn’t what it was. When Buzz Williams left Marquette for Virginia Tech, it was widely viewed as a downward move.

Tech needs to be realistic. While at Xavier, Bobinski hired Thad Matta and Chris Mack as head coaches — he was working in the development office when Sean Miller was promoted to succeed Matta — but it’s hard to envision Mack leaving the Musketeers for North Avenue. (To say nothing of Matta or Miller, ensconced at Ohio State and Arizona, respectively.) Xavier is a better basketball job than Tech, which has reached the NCAA once since 2007.

If Tech does go shopping for a new coach, it might be faced with a choice between a young assistant (Jason Williford of Virginia or Travis Steele of, er, Xavier) and a retread (such as Seth Greenberg, formerly of Virginia Tech). Gregory hasn’t been the man to lift the Jackets from oblivion; less clear is who might be able — and willing.