West Virginia’s 2017 football season is destined to be a bust if you believe experts like Phil Steele and Bill Connelly. Connelly has the Mountaineers finishing at 5-7. Steele is more generous with his prediction – he has WVU at 6-6.
Six and six seems to be the popular pick for the Mountaineers. USA today has them at 6-6, and the stalwarts at CBS Sports have them finishing no better than 4th with the consensus being a 6th place finish. Even some Big 12 beat writers have WVU 6th in the Big 12.
Why is a program coming off a ten win season getting so little respect? What do the experts know that we don’t? Because, the LGG has the Mountaineers as a dark horse to win the conference.
It’s not what they know, it’s what they think they know.
Connelly, Steele, Las Vegas bookies and an army of experts have turned to statistical analysis as the basis for their projections. Like oracles of old, these experts dig deep into the guts of the stat sheets, crunch the numbers, weight a few variables and divine the results.
The oracles of college football look at West Virginia and they see a program that lost 70 percent of its offensive production from a year ago. They see a team that must replace eight starters on defense. They see a program in flux about to experience a painful rebuilding year.
Predictions based on analytic analysis sound accurate. The more stats are cited, the more believable they become. What they don’t tell you is that an incomplete data produces invalid conclusions.
I’m the first to admit that predictive analytics is uncannily accurate when applied to baseball. In baseball, there’s a statistic for every facet of gameplay. Baseball tracks measurables as esoteric as curve ball spin rates and bat speed through the strike zone. The more data available, the better the accuracy.
What works for baseball doesn’t work so well for college football. Play away from the ball (non-linear play) is just as important in a pass or run but has almost no stats to quantify performance. Other factors such as the weekly deviation in the level of competition and roster turnover skew the data set making accurate predictions nearly impossible. There’s one other major flaw in the equation. Experience is valued more than talent, because “talent” is hard to quantify.
If the data set is superficial and the premise flawed the resulting predictions cannot be accurate.
WVU is the perfect example. The Mountaineer’s personnel losses would indicate a mediocre season and finish in the bottom half of the Big 12. However, that’s implying the Mountaineers have to rebuild, but what if they reload?
The buzz around the program is the 2017 Mountaineers are more talented, bigger, faster and stronger than last year’s 10-3 squad. The expectations inside the Puskar Center are high. How high? Nine wins would be disappointing.
WVU head coach Dana Holgorsen, defensive coordinator Tony Gibson and offensive coordinator Jake Spavital have been upfront with their praise of this year’s team. Holgorsen has openly mocked the Big 12 preseason media poll and said his Mountaineers are the most talented team he’s coached.
The fallacy of the premise that experience is more valuable than talent (first-year starters are rarely as productive as their predecessors) is best illustrated by the belief that WVU’s production at QB will decline because Will Grier is replacing Skyler Howard.
I’m not comfortable criticizing Skyler Howard, but to understand why the data model fails its necessary to look closely at Howard’s performance and his effect on the offense.
There’s no doubt that Howard was tough as nails and a great competitor, but his skill set and mental makeup did more to limit WVU’s offense the last two seasons than elevating it.
The stats suggest Howard was more than serviceable, but if you look at the four games Howard started against Oklahoma and Oklahoma State his stat line shows four fumbles and five interceptions with an average QBR of 27.9..
The statistics conceal the fact that Howard lacked the arm strength and accuracy necessary to run the traditional Air Raid offense. Holgorsen compensated for his QB’s deficiencies by abandoning his favored pass-first philosophy for a run-first scheme that utilized Howard’s prowess as a runner. Holgorsen’s pivot – more Don Nehlen and less Mike Leach – allowed Howard to be successful. There were indicators of inconsistent QB play harming offensive production. The most prominent being WVU’s problems with red zone scoring.
The Mountaineers were number 92 among 128 FBS programs in red zone scoring last season. The primary reason WVU struggled, according to Dana Holgorsen, was Skyler Howard.
“Our pass game down there has not been good,” Holgorsen said. “Skyler (Howard)’s strength was definitely not sitting in the pocket and throwing balls into very, very tight spaces. Which, in the red zone, the closer you get, the tighter the spaces and the tighter the coverage is.”
There isn’t any doubt in Morgantown that Will Grier has the tools to elevate the Mountaineers into an offensive juggernaut and banish WVU’s red zone woes.
“He (Grier) can make tight throws,” Holgorsen said. “In red-zone offense, you have to make tight throws. That field gets shrunk, so you have to make tight throws. Will can make those throws. Whether it’s him or whether it’s the receivers making those tight catches, I like what I’ve seen out of our red-zone offense so far.
The idea that WVU will have less production from the QB spot because a newcomer (Grier) replaces a veteran (Howard) is absurd and illustrates the problem is using data as a predictive tool for college football.
The Mountaineers are not just tasked with replacing their starting quarterback, though. WVU lost eight starters on defense including the entire defensive line and both starting cornerbacks. It would seem impossible for Tony Gibson to replace so many key personnel and come close to matching the production his unit had last year. Yet, Tony Gibson isn’t worried. He knows what the data doesn’t reflect; he has talent in the wings ready to step up.
“If you look at every position, I think we’re more athletic than we’ve ever been,” he said. “That’s counting up front, the linebackers and, obviously, the secondary. Gibson said.
Gibson is confident that Adam Shuler and Reese Donahue are more than ready to take over at DE for Noble Nwachukwu and Christian Brown. Maybe more than ready since both have a higher upside than their predecessors.
Shuler played in 13 games last season earning a spot on Athlon’s All Big 12 Freshman team (second team). Injuries forced Donahue into spot duty as a true freshman where he showed flashes of his talent with six unassisted tackles.
The Mountaineers have some concerns at nose tackle in replacing Darrien Howard. Sophomore Xavier Pugues won the spot in fall camp over five others, including prized recruit Lamonte McDougle. Pugues is a converted DE who Gibson says grew into a NT. While Pugues lacks experience, he has the mentality and physical tools to, at the least, match Howard’s performance.
At the linebacker position, WVU lost only one impact player in Justin Arndt. David Long (out with an injury until October) leads a talented and athletic group including the highly touted r-freshmen Dylan Tonkery and Brenden Ferns. Dependable veterans Al-Rasheed Benton and Xavier Preston round out a strong group.
Given the loses with the loss of Rasul Douglas, Nana Kyeremeh, Jeremy Tyler and Maurice Fleming, the experts consider WVU’s secondary a weakness going into 2017. That’s a lot of production to make up. However, there are reasons to believe the secondary may be a strength. Spur Kyzir White and free safety Dravon Askew-Henry look to be NFL talents and bandit safety Toyous Avery isn’t far behind.
While the talent of White, Askew-Henry, and Avery is well-known, both starting corners are a mystery to everyone outside of Morgantown. They won’t be a mystery for long. Hakeem Bailey won the left cornerback position over veteran Elijah Battle last spring by impressing cornerbacks coach Doug Belk with his cover skills and maturity.
“Hakeem’s got everything you’d want in a corner,” Belk said. “He’s got size, got athleticism, great ball skills and good physicality.
Joining Bailey is the hard-hitting senior Mike Daniels Jr. who won the right cornerback position over Syracuse grad transfer (and two-year starter) Corey Winfield.
Replacing eight starters on one side of the ball isn’t ideal, but West Virginia isn’t going to let that curb their expectations for 2017. The bottom line, while numbers never lie, stats don’t always paint the complete picture, and the lack of experience doesn’t mean the lack of talent.