Dissecting the Coverage: Spring Break!
Happy Friday, everybody. Unless you’re a cadet, in which case suck it down! You’ll be back at the Academy before the end of the weekend. Hahahaha!
I love it!
— Army Football (@ArmyWP_Football) March 8, 2019
Army Football finished Mat Drills last week and then held their first practice of Spring Camp right before the break. Beyond that, we’ve seen a fair amount of sports while the Corps has been away, and there’s been much, much more news than I’d expected. I hadn’t realized how many spring teams competed over the break, but they all do, apparently. Even the Swim Team has some folks at Nationals.
No Cancun for you!
Which college football coaches consistently produce overachievers or underachievers? (Football Study Hall)
Ten biggest 2018 overachievers, per postgame win expectancy:
– Army +2.9 (8.1 expected wins, 11 actual wins)
– Northwestern +2.8 (6.2, nine)
– Kentucky +2.2 (7.8, 10)
– Ohio State +2.1 (10.9, 13)
– Georgia Southern +2.1 (7.9, 10)
– Texas +1.7 (8.3, 10)
– Notre Dame +1.7 (10.3, 12)
– Liberty +1.6 (4.4, six)
– FIU +1.6 (7.4, nine)
Syracuse +1.5 (8.5, 10)
For a lot of teams, this was likely randomness. But it’s fair to note that Army also overachieved by 2.8 wins in 2017, and Northwestern overachieved by 2.1. For that matter, Kentucky (+1.9) and FIU (+1.6) also overachieved by healthy amounts.
[I]n the last two seasons, Army has produced two of the largest outliers in the data set. With a second-order win total of 15.1, they’ve gone out and won 21 games, overachievement of about three per year.
S&P+ ranked Army 91st in 2017 and 84th in 2018, and believe me, Army fans have noticed. There has quickly grown an “S&P+ just doesn’t know what to do with Army” consensus, and that consensus may be correct.
Rob sent me this last Saturday just to drive me crazy. And look, I like a lot of what Connelly does with S&P+. It’s an interesting model even when it’s not overly predictive. My only issue with it is mostly in the way they stick to the model’s prediction results even once actual game results become available after the fact.
Imagine this exchange in my office:
“The model said we’d get 50,000 amps of fault current at Bus X, which was well below the breaker rating of 63 kA. But in reality, the fault produced a 65 kA fault, which was well above the ratings. Unfortunately. The breakers couldn’t interrupt the fault as a result, and I’m sorry boss, but we’re looking at $300M worth of damage to the bus framework itself.
“I have no idea why real conditions didn’t match the model, but clearly, the real conditions were wrong.”
In a sense it’s correct, right? Real conditions didn’t match what was in the model, and that created a problem. What’s weird is the idea that it’s the real conditions that were incorrect. The model’s job, by definition, is to explain and predict real-world outcomes within an acceptable margin of error.
Yes, there will always be some error. However, what we’re seeing with S&P+ right now is systematic–repeated, predictable–error in reference to specific brands of football. That is the point of this article! And the obvious take-away that the model itself is simply not capturing some elements of what makes certain teams successful or unsuccessful. I said before that I think the issue pertains to complimentary football and offensive consistency (as opposed to explosiveness). I’m not sure how to model that, though, nor am I trying to sell you a competing model.
But we need to understand what we’re reading. Or, to put that another way, there is a reason why the previews we run on this site always focus on a few key statistics–yards/carry for and against, quarterback completion percentage, and 3rd down conversion rates for and against. Those things all speak to consistency more than explosiveness, and over the years, I’ve found them to be much more useful than simple yards/play or “success rate” calculations in evaluating how a game is going to unfold. Especially when watching Army Football.
You can check our work, but I think we’ve done okay.
Both lacrosse teams played twice since last week’s post, and both teams went 1-1. The Women started hot against Virginia Commonwealth but had to hang on as VCU made a blistering comeback attempt.
— Army W.Lacrosse (@ArmyWP_WLax) March 9, 2019
They ran into trouble against Villanova, though, falling 16-8.
Army led in total shots and in shots-on-goal through the first half, but they couldn’t quite find the net often enough to keep it even, heading into the locker room down by two. VU put the hammer down after the break, though, causing turnovers, taking more shots, and generally taking control of the game. VU was up 14-5 halfway through the second half, and though they fought gamely, the Black Knights could not pull off anything like another miracle rally.
Army 8 pic.twitter.com/8UmJhore6v
— Army M.Lacrosse (@ArmyWP_MLax) March 9, 2019
The Men got behind Lehigh early and were never able to mount a run to bring the score all the way back to even. I didn’t see a ton of this game, but the parts I saw showed Lehigh’s defense to be at least as good as advertised. The Black Knights didn’t play badly by any means, but Lehigh’s first quarter 4-1 run was enough to give them the victory even though it happened in the first fifteen minutes of the game.
That’s a tough way to lose, but what can you do?
Binghamton 6 pic.twitter.com/AKx3GidqTr
— Army M.Lacrosse (@ArmyWP_MLax) March 12, 2019
— Ohio Machine (@MachineMLL) March 10, 2019
Well, yeah, okay. You can smash Binghamton and then get your All-American defensemen selected in the third round of the Major League Lacrosse draft. That might help salve some wounds, sure.
Both teams are at home tomorrow for a Patriot League double-header. The Men take on Holy Cross at noon, and then the Women meet Loyola at 3 pm.
Both games are vitally important. The Men remain ranked at 18th nationally, but they definitely cannot afford another loss if they want a realistic shot at the NCAA tournament. The Women, meanwhile, have a legitimate chance to make their first Patriot League tournament, but only if they hold their early momentum going into conference play. Loyola in particular is not some pushover. The Women have looked good, but they still need to prove that they can hang with the best teams in D1 lacrosse.
Army completes the two-game sweep of Stephen F. Austin and heads into its weekend series at @Ohio_Baseball with a 7-7 record.
Giachin, Walden and Simoes with multiple hits, while Flannery is credited with the win after allowing just 4 hits, 2 runs (1 earned) in 4 innings of work pic.twitter.com/WtkDGBNMmB
— Army Baseball (@ArmyWP_Baseball) March 14, 2019
Army Baseball is 7-7 heading into a three-game set against Ohio before finally playing their home opener on Wednesday against Hofstra. A game against Bucknell gets next weekend going, and then the only Army-Navy home match-up of the year is next Sunday at noon.
My daughter really wants to go, so who knows? You might see us there.
No. 25 Army: QB Kelvin Hopkins Jr.
Hopkins displayed how just dynamic he can be in the Lockheed Martin Armed Forces Bowl, rushing for 170 yards and five touchdowns in a 70-14 demolition of Houston. The junior quarterback — who became the first in the academy’s illustrious history to throw and rush for 1,000 yards in the same season — is back to spearhead an Army option attack that returns mostly intact off a record-breaking 11-win season. — [Jake] Trotter
This is not a surprise. What’s a surprise–still–is that Army gets some love every time one of these crazy Top 25 filler articles comes out on the big Four-Letter Network.
H/T to Jordan Horvath for the link.
* * * Below the Fold * * *
This week’s post is brought to you by As For Football’s new Linktree mobile site aggregator.
What’s good about the Linktree thing is that it gives social media followers an easy way to see all of our content. Or, more accurately, whichever portion of our content our followers choose to consume. Do you like YouTube? Prefer the podcast? Are you here strictly for the written analysis provided by yours truly?
We don’t care! Take whatever you like.
It’s weird for me personally because I’m so used to driving content to a website for clicks, but that’s not the model at AFF. Which is good because that model was failing. So, however you choose to consume As For Football, enjoy!
But do us a favor and tell your friends.
Early predictions for 2019 college football season (Sporting News)
Oct. 12 will be awesome
That weekend features Alabama at Texas A&M, Oklahoma vs. Texas, Florida at LSU and USC at Notre Dame. The fallout from those games will be huge for different reasons. The Alabama-Texas A&M matchup should be a top-10 showdown regardless of what the Aggies do against Clemson on Sept. 7. Florida-LSU is always hot, and we outlined the stakes for Oklahoma and Texas. But Notre Dame-USC has a bigger step because we will learn what comes next for the Trojans.
Army is at WKU on the 12th. That also promises to be a good game, depending on how the Hilltoppers’ rebuilding project goes this season.
Michigan will be 10-1 (again) for The Game
We will have the same conversations about the Wolverines. Jim Harbaugh will negotiate another tough schedule to keep Michigan in contention, but the Wolverines must avoid early hurdles against Army and Wisconsin before a probable “White Out” against Penn State. The Wolverines get Notre Dame, Michigan State and Ohio State at home. It might take a sweep of those three to get Michigan in the Playoff, but the game against the Buckeyes will be the biggest referendum for the program once again. Can Harbaugh finally win The Game?
Folks are buying Michigan pretty hard this offseason. It’s amazing, really. Yeah, they return a lot of offensive players, but their offense wasn’t lights-out last season, and all those new guys on defense are kind of a warning sign, especially for early-season games. I get that folks like Harbaugh, but I would not bet on Big Blue being 10-1 for Ohio State, nor would I take them to win their rivalry game. Any number of these teams could knock the Wolverines off, and based on the way they finished last season, I think a few pull it off.
[Coach Ken] Niumatalolo wants [Navy OC Ivin] Jasper to design an offense around the talents of Perry…
“This scheme may have worked for so and so, but does it work for Malcolm? Keenan was good at this and Will was good at that, but those schemes might not fit Malcolm. So a big part of what we’ve done during the off-season is look at how we can help Malcolm be successful.”
Man, if I was a Navy fan, this line of thinking would scare the Hell out of me.
[New Offensive Assistant Billy Ray] Stutzmann Stutzmann was a four-year standout as a wide receiver at Hawaii then returned to his alma mater as an assistant overseeing the same position he played. Niumatalolo seems curious to find out if Navy can incorporate elements of Hawaii’s run-and-shoot attack.
“We’re excited about Billy Ray – a bright, young, energetic coach who has brought some really good ideas from Hawaii. I think Billy Ray is going to help us a lot with his knowledge of the passing game,” Niumatalolo said. “We’re trying to see how we can mirror those two worlds – a run-and-shoot passing game with what we do.”
So yeah. They’re putting Perry back there–again–and moving towards a run-and-shoot concept. Because of Perry’s noted accuracy throwing the football…
Georgia State basically did one thing well defensively last year: they hemmed in your run game most of the time. The Panthers ranked 30th in rushing marginal efficiency and 49th in opportunity rate (percentage of non-sack carries gaining at least four yards). That resulted in pretty good red zone defense, too — they were 36th in success rate between their 11 and 20, 21st inside the 10, and 38th on first-and-goal.
This is something to keep an eye on. I think most Army fans have the Georgia State game penciled in as a win, but the Black Knights have struggled at times against teams with good rushing defenses, especially when they’re backed by competent offensive play. The fact that GSU’s passing defense was actively terrible last year isn’t necessarily a salve for the problem, either.
The Alliance of American Football
— The Alliance (@TheAAF) March 13, 2019
— The Alliance (@TheAAF) March 13, 2019
Saturday’s games aren’t doing a lot for me. I mean, Memphis has looked a lot better of late, but they aren’t anybody’s idea of good, and Salt Lake is one of the least interesting teams in the league. The Stallions just don’t do anything well enough to be entertaining, and their stadium is always a ghost-town. Meanwhile, Arizona at Orlando might have had some sizzle early, but the Hotshots have regressed enough since QB John Wolford’s injury that I don’t give them much of a chance now.
Sunday’s games look better. The CBS Sports Sunday afternoon slot gets San Antonio versus Atlanta, and believe it or not, that one’s a must-watch. The Commanders have always looked like one of the most balanced teams in the Alliance, and I think they might also have the best team in the West division as well. Meanwhile, the Legends started as road-kill, but after–finally–settling their coaching staff and making a change at quarterback, they’ve won two straight. With the Birmingham Iron losing two straight, the Legends are now just a game back in the wild card race. Amazingly, a win this weekend would put them in legitimate contention. Moveover, their game last week was super-entertaining.
Alas, Birmingham at San Diego doesn’t do a lot for me. Teams have figured out how to score against the vaunted Iron Curtain, and the Iron offense just isn’t good. San Diego isn’t a bad team, but who wants to root for a squad with a bunch of ships on the sides of their helmets?
That’s all I’ve got this week. Enjoy the weekend!