Georgia Tech cracked the top 10 in 2014. Don’t assume an exit in 2015.

Confused? Check out the advanced-stats glossary here.

As college football fans, we find glory in the smaller moments. This is the most beautifully strange American sport; of that, there is no question.

But for such a weird pastime, the end results are usually predictable.

Since 1992, only 13 schools have won a national title; only two -- Florida and Florida State -- won their first title in that span. Most of the list of champions emerged as national powers in the 1950s or earlier and have encountered only rare dry spells since then.

Alabama, Florida State, Oregon, and Ohio State were four of the five biggest favorites to reach the College Football Playoff when the year began, and all four made it. Sure, there were unique story lines along the way: Ole Miss and Mississippi State surging to the top of the polls, Duke remaining a Playoff contender into November, etc. But in the end, we got the result we expected.

Well, mostly. None of us saw crazy old Paul Johnson and his Georgia Tech squad running an SEC team off the field to win the Orange Bowl.

I’ve long been a defender of Johnson’s Yellow Jackets, but even I severely underestimated them. Since bottoming out in 2010, when they finished 6-7 and ranked 74th in F/+, they had clawed their way back to respectability. They ranked 37th in both 2012 and 2013, their 14-13 record held back by a 1-6 record in one-possession games. And in Justin Thomas, they appeared to find an exciting quarterback. Still, there were limitations, as I wrote in last year’s preview.

Johnson’s offense still works. I say it every year. The flexbone is not the best offense for drawing five-star talent, and it certainly isn’t great on third-and-8. At the same time, Johnson and Georgia Tech aren’t going to draw much five-star talent anyway, and almost no offense is particularly strong on third-and-8. When operated efficiently, the offense is the least of Johnson’s problems. Even with shaky operation in 2013, the Jackets still finished 31st in Off. F/+. You’re not going to win a national title with the flexbone option, but you can certainly win.

Perhaps because the end results are usually predictable, I delight more than I should in the surprises. Georgia Tech finishing 11-3 and No. 8 in F/+ tickled me beyond belief.

Eighth is high. Sometimes you can ride a stretch of close wins to a nice record, but the stats won’t be fooled. If you rank eighth in F/+, you’re really good, and you’re probably going to remain good. Of the nine teams that ranked eighth between 2005 and 2013, three ranked even higher the next year, and all but one remained at least in the top 20.

(The USF team that ranked eighth in 2007 turned out to be a flash in the pan, but if history has taught us anything, it’s that 2007 was the exception to every rule.)

In terms of final F/+ rankings, the Jackets beat No. 4 Georgia on the road, beat No. 13 Mississippi State and No. 14 Clemson by a combined 37 points, and nearly took down No. 15 Florida State for the ACC title.

Should we be taking them more seriously? When a team without a long history of success or without incredible recruiting rankings succeeds, we assume regression is coming. Should we? Or is the Ramblin’ Wreck, with an incredible option quarterback, a stellar line, and a more experienced defense, ready to be the best team in the ACC?

Even an 11-3 team is going to have glitches. For the first half of 2014, the Yellow Jackets were good, but they weren’t great.

They allowed bad Tulane and Virginia Tech offenses to average 5.5 and 5.9 yards per play, and they got torched by Georgia Southern, Miami, and North Carolina (combined: 34.3 points per game, 7.6 yards per play). And while the offense was fine, it let them down against Duke -- they gained 483 yards but had to settle for three field goals (missing one) and threw an interception from the Duke 10 in a 31-25 loss.

Following a frustrating shootout loss to UNC, Tech was 5-2 but a bystander in the ACC Coastal race. But then Duke began to stumble, and the Jackets caught fire.

Sometimes your stats improve by remaining about the same. In the first seven games, Tech averaged 6.8 yards per play and allowed 6.4. In the last seven, the margin was 6.6 to 6.3. But the last seven games included four top-15 opponents and no dead weight. The Jackets destroyed Clemson, holding the Tigers to 190 total yards. They outgained Georgia by 75 yards and survived pratfalls to win a thriller. They averaged 6.4 yards per play against Florida State but stalled out just enough to drop a two-point heartbreaker. And they averaged 7.9 per play against a Mississippi State unit that had allowed 5.5 per play for the season.

The Tech defense was an issue throughout the year, capable of slowing down mediocre units but lacking the firepower to do anything against good ones. But the Jackets only had to make a few stops a game, thanks to a terrifying option offense. And for the most part, it was good enough to do that.

Note: players in bold below are 2015 returnees. Players in italics are questionable with injury/suspension.

Heading into 2014, Thomas was tantalizing but unproven.

The former four-star athlete, a 100-meter track champion in Alabama, has the running down, but his early passing audition left something to be desired. He’s not a big specimen, but his potential as a playmaker is off the charts. If efficiency doesn’t suffer much, Tech’s explosiveness will probably improve. But that’s an important “if.”

As a freshman in 2013, Thomas averaged 8.9 yards per (non-sack) carry but completed just nine of 17 passes with two picks and two sacks. If you’re good enough at running the option, you’re going to find opportunities to hit receivers downfield, but Thomas wasn’t ready.

He was in 2014. His INT rate fell from 11.7 percent to 3.2, and his sack rate fell from 10.5 percent to 5.1. Part of that was a 2013 sample size issue -- your rates might be skewed when you only attempt 19 passes -- but part was obvious improvement.

The option is the most important facet of the Tech offense; Thomas was only once asked to throw more than 18 passes in a game. But aside from an implosion against Duke (6-for-15, 61 yards, two picks), he was somewhere between tolerable and excellent throughout. His passer rating was above 145 in nine games and above 200 in four. He threw for more than 1,700 yards, the first time a Tech QB had done that since the 2009 ACC title run. He was able to punish defenses for getting distracted by the run, and that’s all the Tech pass has to do.

And now we’ve got two more years to see what he can do. This is going to be fun.

I’ve long held a theory that the slotback in a flexbone-style offense has the simplest job in football. He must simply catch pitches and be fast, and he will succeed. Now, “simple” is not “easy,” mind you, but while Johnson doesn’t sign many blue-chippers, he can find fast guys to catch pitches.

This is the most plug-and-play offense in the ACC: the B-backs fall forward up the middle, the A-backs hit the corner, and the receivers block well and leap high for lobs. If the QB is up to snuff, the offense will succeed.

This theory will be tested. The return of Thomas and six linemen with starting experience (68 career starts) should assure the Tech offense functions. But there is drastic turnover at the other skill positions.

Tech must replace its top three B-backs (the fullback types) and four of the five A-backs who had at least 20 carries. Each of the A-backs averaged at least 6.1 yards per carry, and returning senior Broderick Snoddy averaged a hilarious 10.1, but the level of experience takes a major hit.

Johnson loaded up on running backs in February’s signing class, and it is realistic to assume some combination of Snoddy, redshirt freshmen, and freshmen will still thrive on the outside, and that Stanford transfer Patrick Skov, converted linebacker Marcus Allen, and youngsters will take advantage of space up the middle. But if anything, the lack of experience could cause more mistakes -- missed assignments, drops, etc. And while the Tech offense will be as explosive as ever, the efficiency could drop a few notches.

Everything I just said goes for the receiver position, too. A Tech receiver’s job is to block and occasionally spring downfield. Players like junior Micheal Summers, sophomores Antonio Messick and Ricky Jeune, and some interesting true freshmen could do that despite the loss of last year’s top two. But the mistakes could increase.

In 2013, Georgia Tech ranked a solid 12th in Adj. Line Yards, but heading into 2014, the Yellow Jackets had to replace three line starters who had combined for 117 career starts. Outside of all-ACC guard Shaquille Mason, the Tech line was awfully untested.

And the Jackets improved to third in Adj. Line Yards. Part of that had to do with Thomas and the experienced supporting cast, but the line did its job. No one allowed a lower percentage of run stops in the backfield than Tech, and for this type of offense, that might be the most important stat of all.

So now Tech returns four starters, six with starting experience (68 career starts), but Mason’s gone. Was he the lynchpin? Will seniors Trey Braun, Bryan Chamberlain, and Errin Joe assure more mistake-proof play? A good line can usually overcome one loss, but Mason was awfully good.

One of the flaws in my ratings is that I haven’t figured out an easy way to strip special teams from the offensive and defensive ratings. Since field position is part of the overall equation, that means the kicking (offense) and return games (defense) factor into the unit rankings in a way that isn’t directly applicable to offense and defense.

In that regard, Jamal Golden, with his efficient returns more than his decent safety play, may have had the most positive impact on Georgia Tech’s defensive ratings. Without solid field position and red zone stops, the Yellow Jackets’ had very little going for them when the defense was on the field. Bend-don’t-break can work, but Tech’s defense bent a lot.

While it’s been a while since Tech had an effective defense, youth dragged the Jackets down last year, as did the loss of lineman Jabari Hunt-Days to suspension. The ends were freshmen and sophomores, the leading tackler was a sophomore, and a few freshmen played roles in the secondary.

With only three starters gone and Hunt-Days back (and heavier), experience will be a strength, which means that one should expect some improvement. And if the defense improves enough to account for the offense being less efficient, then the Ramblin’ Wreck move on at the same pace. That sounds good in theory, at least.

Looking at the top returning members of the front seven, you notice something: they’re small. The top five returning ends average 6’3, 237. Including Hunt-Days, the top three returning tackles average 6’4, 286. Leading tackler P.J. Davis is 5’11, 218. This isn’t mid-major small, but it’s smaller than most power-conference fronts.

This can work if it results in extra speed and aggressiveness. You might get pushed around, but you can make up for it in your attacking. Havoc can become doubly important when you are a bit on the light side.

Georgia Tech didn’t create nearly enough havoc. The Jackets ranked 102nd in Adj. Sack Rate and a decent-not-great 62nd in stuff rate. There were some bright spots -- KeShun Freeman was Tech’s best pass rusher as a true freshman, tackle Adam Gotsis combined 6.5 TFLs with four passes defensed, and the top three linebackers combined for 20 tackles for loss and eight passes defensed. The secondary made plays, but early-down inefficiency was a big hindrance. Tech couldn’t stop the run and couldn’t rush the passer enough to make up the difference.

Hunt-Days will help. Plus, Freeman is no longer a freshman, and the experience is solid: only three primary members of last year’s rotation are gone. But experience alone doesn’t make you more disruptive. You need raw talent, too.

Opponents completed 63 percent of their passes, and in the Jackets’ three losses, the completion rate was 73 percent. Opposing quarterbacks had plenty of time to find open men and weren’t forced to throw too often on passing downs.

It’s miraculous that the Tech secondary made as many plays as it did.The top three safeties combined for nine TFLs, six picks, and 12 breakups, and the top two corners combined for 3.5, six, and 12, respectively. That’s not bad, and it hints at a potential strength.

Everybody but Isaiah Johnson is back, and the unit has quite a few seniors. If the front seven can make a few more run stops or get a few more hands on quarterbacks, Tech defenders should be capable of making more plays on the ball. But that’s a significant if.

So Harris Butker’s leg is a bit scattershot -- he made a higher percentage of his longer field goals than his shorter ones. But he wasn’t asked to attempt many FGs, and his kickoffs were strong. And while Golden wasn’t incredibly explosive in the return game, he consistently made the first guy miss and got upfield.

There were no extreme strengths or weaknesses, and with everybody back, that should be the case again.

Notre Dame and Georgia in non-conference play with trips to Clemson, Miami, Duke and Virginia? Yikes. If Georgia Tech is going to earn its fifth division title in a decade, the Jackets will have to earn it.

After a relatively soft start, with three opponents projected worse than 50th in the first four weeks, the final eight games of the schedule are rugged. If the offense is more mistake-prone with such young skill position guys, or if the defense doesn’t improve enough, Tech might not find a sure win after September 12.

But if the balance is still there, as I suspect it might be, then the Jackets have a chance at another really fun season. With Thomas and a strong line, the big-play spigot should remain on, and the defense probably won’t regress.

Even if Tech isn’t ready to be a top-10 team again, top-20 should be expected. That alone makes the Jackets the safest bet in the Coastal. Maybe Virginia Tech’s offense improves enough to get the Hokies back into the top 20, and maybe Miami’s defense and quarterback play do the same. But Georgia Tech already was a top-20 team and didn’t lose a ton of breakthrough talent.