Castricone: Which stats are most predictive of a Washington victory?
The voice of the Huskies presents his findings after digging deep on statistical trends.
When Washington play-by-play man Tony Castricone offers to crack open his spreadsheet vault and share some statistical analysis on UW football, you don’t say no. I hope you enjoy today’s guest column from Tony, who wondered which stats might be most predictive of a Huskies win or loss, and lays out his findings below.
By Tony Castricone
It was pretty early into my time on Montlake when then-coach Chris Petersen famously quipped: “Stats are for losers.”
While I laughed out loud, on the inside I was like, “aww, man!”
There’s no denying that on the jock-to-nerd spectrum, I’m much more in the neighborhood of Napoleon Dynamite than I am Napoleon Kaufman. And I get what Petersen was saying: it’s better to win games than it is to win statistical categories. We shouldn’t get so wrapped up in stats that you lose sight of the main goal, which is to simply score more points than the opponent over 60 minutes. And the game plan that gives a team its best shot of accomplishing that goal might require sacrificing certain stats, which can throw off a lot of rankings in a season with a sample size as small as 12 games.
Still, I’ve always liked digging for some numbers off the beaten trail that, within the proper context, might either a) give fans something interesting to keep an eye on as the game evolves, or b) might help explain why, exactly, a certain outcome manifests.
The problem with college football stats is, unlike the NFL, there are tons of blowouts, a wild variation in strength of schedules, and a plethora of styles that make controlling variables almost impossible. At the end of every season, you know without any doubt that service academies will be top-five in the nation in rushing, the teams with the highest completion percentages will all primarily run spread offenses, and the Oregon Ducks will still have zero national championships.
But I also wondered, if I could extract enough data from the box scores, whether there might be some clues as to when the Huskies are more likely to be successful, or when they might be in trouble.
After pulling all the data from every Husky box score from 2015-2023, here are the stats that are most predictive of a Husky victory.
Out-rushing an opponent by 100 yards (winning percentage: 100%)
With Jake Browning, Jacob Eason and Michael Penix Jr. slinging the ball around for the better part of a decade, you might think the Huskies are their best when the passing game is clicking. But if you’re looking for a sure thing, it’s actually when the running game outpaces the opponent’s.
Washington is 30-0 since 2015 when it outrushes the opponent by 100 yards or more. And to go even further, UW is 43-1 when outrushing the opponent by 50 yards or more, the lone loss being the 13-7 shocker suffered at Arizona State in 2017.
It’s interesting that being outrushed by 100 yards is not a death knell. It’s not great, but Penix in particular helped the Huskies make up for several occurrences in which they lost the rushing battle. UW is 5-9 in such games, with wins against top-10 Oregon teams in Eugene (2022) and Seattle (2023).
Having an average starting field position margin of plus-10 (winning percentage: 100%)
Washington is 23-0 when its average starting field position is 10-plus yards better than its opponent’s. The Huskies are just 3-7 when the opponent’s average field position is better than 10-plus, and 6-17 when the opponent starts five yards or more ahead of the Huskies on average. That’s how important the field position battle is. It might be the single most predictive stat that no one ever talks about.
Washington is 60-7 over the last nine seasons when winning the field position battle by any margin. Win it by a sizable margin? That almost guarantees a ‘W.’
Having a positive double-digit margin helps you either totally dominate opponents (2022 Michigan State, 2017 Apple Cup, 2019 Las Vegas Bowl, 2016 Pac-12 championship game), or helps you win a game you otherwise probably should lose (2023 at Stanford, 2023 at Oregon State, 2023 vs Utah).
A turnover margin of plus-2 or better (winning percentage: 97%)
Now this might be the most predictive stat that everyone talks about. The Huskies are 35-1 with a turnover margin of plus-2 or better. They are 3-12 with a turnover margin of minus-2 or worse.
To show you what a miracle the 2023 season was, two of those three wins occurred in back-to-back weeks last October, in close victories over ASU and Stanford.
Scoring two touchdowns in the first quarter (winning percentage: 96%)
UW is 27-1 when scoring at least 14 points before the end of the first quarter (with the lone loss being another ASU oddity in 2021, the game after Jimmy Lake’s suspension). The Huskies are just 1-3 when allowing an opponent to score 14 in the first quarter (with the 2023 shootout against an historically bad USC defense being the lone exception).
Averaging at least 5.0 yards per carry (winning percentage: 94%)
There’s something about the Huskies having success running the ball that makes them nearly unstoppable. They’re 32-2 since 2015 when averaging 5.0 yards per carry or better. The two exceptions were a close loss to Oregon in 2015 (5.53) and at Oregon State on a last-second field goal in 2021 (5.03).
Fortunately for the Huskies, giving up 5-plus yards per rush isn’t all that big of a problem. It’s only happened 21 times since 2015, and they went 13-8 in those games.
Having a two-possession lead at any point in the game (winning percentage: 93%)
Comebacks are always possible, but they are rare once you’re down two scores. Washington is 69-5 (.932) since 2015 in games where it’s taken a two score lead at any point. The five losses all included unique circumstances: the bizarre lightning game against California in 2019 (blew a 10-0 lead); the game with Lake suspended against ASU (14-0); another road loss at ASU in a 2015 turnover fest (17-0 lead); and back-to-back games at home against ranked teams in 2019, when they relinquished double-digit leads against Oregon (14) and Utah (11).
Conversely, the Huskies are just 2-26 (.071) since 2015 once they’ve fallen behind by two scores, with the lone comebacks being at home against Utah in 2020 (21) and at a 1-11 Arizona team in 2021 (13).
Outscoring opponents by double digits in the ‘Middle 8’ (winning percentage: 100%)
One of the keys for former coach Kalen DeBoer was to win the “Middle 8” — the last four minutes of the first half through the first four minutes of the second half. The idea was that if you won the coin toss, you’d defer to the second half and try to time out your final possession before halftime to score with as little time remaining as possible, so you could essentially get back-to-back possessions once you received the kick to open the third quarter.
If you are able to string two scores together that way, it’s massive.
Washington is 13-0 when outscoring teams by 10-plus in the Middle 8, and 0-3 when having it go against them. The most notable examples of victories are the 2016 rout over Stanford and the 2016 Pac-12 title game, along with the 2022 Apple Cup (75-yarder to McMillan on the first play after halftime) and the 2023 win at USC (when the crucial ZTF strip sack turned the tide).
The three losses were all within UW’s grasp, but the Middle 8 swung the momentum in home defeats to Cal and Utah in 2019, and at UCLA in 2022.
Now, here are some things that I think we probably have overvalued in terms of their importance.
Scoring first
Obviously, it’s better to score first than not. But if it doesn’t happen, there’s no need for panic. Since 2015, Washington’s opponent has scored first in 38 games, and the Huskies have a 21-17 record in those contests. Notable games in which it’s quite easy to forget the Huskies started with an early deficit include:
2015 vs. Washington State (W, 45-10)
2016 at Cal (W, 66-27)
2017 at Colorado (W, 37-10)
2017 vs. Oregon (W, 38-3)
2022 at Washington State (W, 51-33)
2023 vs. Boise State (W, 56-19)
Obviously, UW’s record is better when it scores first (58-15). But, here are some games in which it’s easy to forget that UW took the early lead:
2016 vs. USC (L, 26-13)
2016 vs. Alabama (L, 24-7)
2017 at Stanford (L, 30-22, but in fairness, we probably forget because we were watching truck races)
2021 vs. Montana (L, 13-7)
2022 at Arizona State (L, 45-38)
As we’ve demonstrated, it’s much more important to not let any deficit turn into a two-possession game than it is to score first.
Limiting penalties
Did you know Washington has committed double-digit penalties in a game nine times since the start of the 2015 season? Care to guess what its record is in those games? Try 8-1, with the lone loss coming in a virtual road game against No. 9 Auburn, 21-16, in the 2018 opener.
In fact, the Huskies are 18-4 when committing eight penalties or more.
It’s one of the interesting things I’ve noticed over the years: sometimes, really good teams have really high penalty numbers because of how aggressive they are.
For example: the Huskies ranked 125th nationally in penalties per game last year.
In 2022, Alabama ranked 126th and Tennessee ranked 129th.
And I remember from prepping for the Rose Bowl a few years back that Ohio State came into New Year’s Day ranked second to last in the country in penalties per game.
Clearly, no one wants to see those drive-extending third-down penalties on defense, or the careless false start early in an offensive drive that puts you behind the sticks. But DeBoer flat-out said last year that he was OK taking some penalties of aggression. That worked out just fine through New Year’s Day.
Meanwhile, limiting penalties was by no means a guarantee of success. The Huskies had just one infraction in their aforementioned 2021 home loss against ASU, and drew just two flags in their 2021 loss at Oregon State.
What do you think? And what are some other metrics you’re interested in? Drop a comment and I’ll see if there’s a way to look them up.
Husky Legend Tony Castricone
I love the final sentence of paragraph five. These are the stats coaches look at, but these are not really stats that you can coach to. I wonder why explosives are not on that list.