clock menu more-arrow no yes

Filed under:

Despite What BillC Says, Virginia Tech Ain’t 10 Points Better Than WVU

New, 10 comments

I love advanced metrics but this time they’re wrong!

Via Virginia Tech Athletics

I love advanced metrics. I love reading about them, I love seeing how people develop them and I love seeing how they apply to the sports I love. Baseball is a sport that has been engulfed by advanced metrics. When I was a kid, you knew about a players batting average, on-base percentage and slugging percentage. You knew if the guy could hit 30 homeruns, score 100 runs, drive in 100 runs or steal 30 bases. You didn’t know about OBP [on-base percentage + slugging percentage], you didn’t know about ERA+, WHIP, OPS+, BABIP or any of the other countless metrics that have come out.

With baseball, advanced metrics are actually useful. They help explain a complex game of 9 players versus 1 player and boil it down to single instances. Football is different. Football is a game of 11 vs 11 doing 11 different things. There are certain things that hold true and certain things that just don’t really show up in numbers.

SBNation’s Bill Connelly, BillC for short, does the team previews using advanced metrics. He looks at a ton of different factors and applies them, giving him a look at the team as a whole. It’s a start. With 130 teams in the FBS, there are always going to be outliers. West Virginia has proven to be that outlier.

In BillC’s 2016 preview for West Virginia, his advanced stats indicated West Virginia was only predicted to win 6 games if you strictly looked at win probability. This is all based on what can be calculated in the spring, does not take into account injuries, coaching turmoil [*cough Baylor cough*] or progression/regression of players. The stats gave WVU a good chance to start 4-0, just as they did, but thought the Mountaineers would stumble down the stretch. Bill did hedge his bet by saying

I'm a little shaky about WVU this year. The offense returns a ton of last year's production but might not be any more consistent, and the defense might not be as reliable. That's a bad combination.

A few sustained drives or a few big defensive stops could go a long way. S&P+ gives WVU between a 42 and 57 percent chance -- a virtual tossup -- in five of 12 games and between 35 and 67 in eight. A little improvement could result in nine or 10 wins; a little regression brings 5-7 into view.

West Virginia saw more than a little improvement. Skyler Howard made the second year Air-Raid jump, Justin Crawford supplemented Wendell Smallwood as the leading rusher and Shelton Gibson kept making big plays downfield. On defense Rasul Douglas became an All-American cornerback while the defensive line used its experience to overwhelm offensive lines. The result was a 10-win season for the Mountaineers and a shiny new contract for head coach Dana Holgorsen.

2017 Season Previews

All of the 450 plus words I just wrote above bring me to this season. Bill has completed his West Virginia previews for 2017 and just finished his Virginia Tech preview for 2017. Bill uses a metric called S&P+, which uses three different metrics to produce a projection. S&P+ uses: recent history, returning production, and recruiting.

For 2017 West Virginia, returning production is a huge factor. West Virginia returns only 30% of its 2016 offensive production, the third lowest [and 7th lowest in the country] amount of any Power 5 team. On defense, WVU returns 25% of its production, the second fewest among Power 5 teams [and third fewest in the country]. The result of its returning production is that the offense is “expected” to regress by more than 5 points and the defense is expected to regress by 7 points.

These projections alone would be catastrophic for anyone. The Mountaineers scored 405 points last year, an average of 31.1 points. A reduction of 5.3 points per game would put the team at 25.8 per game. An offense that scores 336 points in 2016 doesn’t win 10 games. Just looking at the 2016 schedule, the Mountaineers would have lost to BYU.

The defensive reduction is even more drastic. Tony Gibson’s unit gave up 312 points last year, right at 24 points per game. Add 7.2 points per game to that number and the defense would have done last year’s offense no favors, allowing 31.2 points per game! Now you’re talking about losses to Kansas State, Texas and Baylor. Suddenly 10 wins has dropped to 6.

It gets worse, Bill’s metrics also use a five-year history. It makes sense to use trend data in metrics. One year can always be an outlier. Even two years or three years [2005-2007 is a great example of this] can be outliers. The Mountaineers five-year trend is using just their time in the Big 12, seasons from 2012 through 2016. That number unfortunately includes the dreadful 2013 season when the Mountaineers only won 4 games. The good news is that we’re trending up, but its slow and incremental. Until that 2013 season is off the books, it will drag down our rankings.

The last metric used is recruiting rankings. Last year’s contract debacle held down recruits for the Mountaineers. No one will make excuses but other coaches used the tenuous nature of Dana Holgorsen’s contract against him on the recruiting trail. So in Bill’s metrics WVU is penalized for a low recruiting average.

While Dana Holgorsen and the Mountaineers are contending with their low recruiting, massive turnover and history, Justin Fuentes is contending with Virginia Tech. Fuente took over a middling Virginia Tech team, once a thorn in everyone’s side, that had become a 7-or 8-win team the past four years. Good but not great. Fuente turned them around, winning 10 games his first year.

Fuente leaned hard on JUCO transfer Jerod Evans. Evans provided the spark to a zone-read offense, rushing for over 1,000 yards once sacks are removed. Fuente and the Tech offense possibly relied too heavily on Evans as the leading running back did not break 700 yards. Overall, the skill players for Tech helped push the team to double digit wins.

On the offensive side of the ball, Virginia Tech returns only 34% of its offensive production. According to Bill’s numbers, this should result in a 4.5 points per game reduction. Virginia Tech’s strength comes in that it returns nearly 80% of its defensive production, which should lead to a 2.3 point per game reduction over last year’s squad that allowed the same 312 points over 13 games as the Mountaineers.

Bill’s metric penalizes the Mountaineers for their dreadful 2013. Virginia Tech is able to escape that penalty as Frank Beamer settled into above average territory during his final years at Virginia Tech. The Hokies won 7 games in 2012, 2014 and 2015. They won 8 games in 2013. This means they have been a top-40 team over the past 5 years according to Bill’s numbers.

Thanks to a new head coach, Virginia Tech had an expected bump in its recruiting last year which also helps in the metrics. Virginia Tech’s two-year average has them right outside of the Top 25.

If you’ve made it this far, I applaud you because I’ve now spent over 1200 words just explaining all this stuff. Now let’s dig deep and tell Bill why all of his numbers are wrong.

Returning Production

I told you last year that quarterbacks in Dana Holgorsen’s system make a jump from year one to year two. I projected Howard to complete 62% of his passes for 3,800 yards, 31 touchdowns and 11 interceptions. Instead he did 61%, 3300 yards, 26 touchdowns and 10 interceptions. Those numbers are actually very good.

Will Grier is going to blow those numbers out of the water. Since becoming an OC and then a head coach, Dana Holgorsen-led quarterbacks have averaged 4,118 yards, 32 touchdowns and 65% completions. Even better, since coming to West Virginia, Dana has never had a quarterback who had been in the system a year before starting. You can sort of make the argument for Skyler Howard but even he got thrust into the starting role too soon. Of the quarterbacks that Dana has coached in his 12 seasons, only Graham Harrel at Texas Tech, way back in 2005-2007, got to spend a full year learning before taking over. In 2006, Harrel threw for 4,555 yards, 38 touchdowns while completing nearly 67% of his passes. I think Grier will be very close to that number.

Virginia Tech, on the other hand, has to replace Jerrod Evans, who decided to bolt to the NFL. Worse, his backup, Brendan Motley has also left the program after graduating. Virginia Tech’s options at this point are redshirt freshman Josh Jackson, a 3-star prospect, JUCO transfer AJ Bush, another 3-star prospect or 4-star true freshman Hendon Hooker. Bush is the only one with any collegiate starting experience and none have seen the 3-3-5, especially the havoc inducing version that Tony Gibson creates.

I have to believe a Parade All-American with six NCAA games under his belt are worth more than three quarterbacks who have never stepped foot on a collegiate field.

At receiver, West Virginia loses two players who led the team in receptions and yards. Still, it returns its next three receivers, Ka’Raun White, Jovon Durante and Gary Jennings. It is adding to that a 2015 receiver David Sills, who appears to have made an instant connection with Will Grier.

The four returning players for West Virginia combined for 1,210 receiving yards and 11 touchdowns. Virginia Tech also loses several receivers, including its leading receiver Isaiah Ford. The Hokies return Cam Phillips who caught 76 passes for 983 yards but their next leading returning receiver only caught 18 passes for 258 yards.

Where the Hokies have the perceived advantage is defense. The majority of the Hokies return on the top-20 defensive unit while the Mountaineers have to replace all three starting defensive lineman and 5 out of 6 defensive backs.

But is it as bad as it seems? The Mountaineers faced something very similar just last season where they had to replace all three senior linebackers and 5 out of 6 cornerbacks.

Many fans, rightfully so, are worried about a regression by the defense. A top Big 12 defense, led by seven senior and redshirt seniors, would appear to be due to falter. However, after reviewing the stats in Part I, I believe that the defense next year may be set up to be to match or exceed this years defense.

My belief was that the players who were not “returning starters” had benefited because the Big 12 requires a true two-deep to compete. Very few players play every snap. Players are rotated and subbed. Players who had not started still saw action in all 12 regular season games.

Defensive tackle Adam Shuler played in all 13 games registering 33 tackles. Reese Donahue played in 12. Toyous Avery played in 11. Elijah Battle in 11. On and on it goes with the Mountaineers. There are very few places where a player is slated to step in without having at least played the season before. I am a firm believer that this is important in the development of players.

Not talked about is the players that West Virginia returns, who would have been starters last year but missed significant time. Dravon Askew-Henry and Yodny Cajuste. Much has been made about how much of the offensive line the Mountaineers need to replace but Cajuste isn’t any replacement. He was the day one starter last year against Missouri before suffering a knee injury. Cajuste has NFL left tackle potential.

Dravon Askew-Henry has been talked about. He is on multiple preseason lists. Yet his return doesn’t appear to gain the Mountaineers respect. Askew-Henry has shown the ability to play deep safety and bring the wood in run support. Combined with Kyzir White, the Mountaineers could have the best safety tandem in the Big 12 and the country.

Five Year Trend

The big difference between the Mountaineers and the Hokies over the past five years is 2013. Where the Mountaineers suffered a slip, only winning 4 games, the Hokies won 8. This is where I disagree with Bill and this metric. The Hokies won 7, 8, 7, 7, then 10 games. The Mountaineers won 7, 4, 7, 8 and 10 games. The Mountaineers have the look of a team trending up. Three straight years they have won more games than the year prior. The Hokies showed the look of a team stagnating. I’ve always believed trends, especially trends in one direction or another should carry more weight.

Recruiting

Recruiting rankings are always a hard one to quantify. One service may have a guy as a 4-star where another has him as a 3-star. Based upon the number of camps he attends, how he performs his senior year and the competition he plays, that rating can increase or decrease over time. When recruiting sites rank teams signing classes, they rank the high school players and the JUCO players. Not included is the transfers.

West Virginia has taken it upon itself lately to keep spots open for transfers. This year alone, West Virginia has seen the inclusion of three different transfers, Jack Allison, Jovani Haskins and TJ Simmons. As one of our facebook commenters loves to tell people, West Virginia had the 56th ranked recruiting class according to 247Sports.com, but what if these transfer players were included in the last recruiting cycle?

Luckily for me, Chris Anderson over at EerSports.com did the legwork for me so I could save my brain power to write this way too wordy article.

So where does that put the Mountaineers? Well, West Virginia finished the 2017 cycle as the 56th-ranked class according to 247Sports with a 174.68 score. Adding these three players to that "class," gives West Virginia a score of 191.48, putting the Mountaineers 41st in the 247Sports Team Rankings. Those additions would push West Virginia past Texas Tech, Arizona, Missouri, and numerous others, as they settle in just behind Oklahoma State, Wisconsin, and Baylor. - Chris Anderson, EerSports.com

What if Bill C had West Virginia at 41st instead of 56th? Wouldn’t that have made a significant impact on his rankings? It certainly would have. That isn’t all though. What if Will Grier were included in this year’s class as he was finally declared eligible on West Virginia Day...I mean Griermas. According to Chris in the same article, including Grier’s 0.9758-rating would have seen the Mountaineers climb all the way to 32nd in the national rankings. 32 is much different in Bill’s rankings than 56.

I think Bill C has done a fantastic job with the team previews. He is a smarter man than I will ever be when it comes to these advanced metrics. His metrics are just wrong in this case because there is no way Virginia Tech is 10-points better than the Mountaineers. We will see come September 3 when the two meet at FedEx Field.

Poll

What will the margin of victory be for the Black Diamond Trophy?

This poll is closed

  • 40%
    WVU by 10 or more
    (325 votes)
  • 43%
    WVU by less than 10
    (348 votes)
  • 7%
    Virginia Tech by less than 10
    (63 votes)
  • 7%
    Virginia by 10 or more
    (60 votes)
796 votes total Vote Now