FanPost

SPAM - 2021 Edition - Pre-Season Preview


I am once again pleased to present the SUPER PREDICTION ACTION MACHINE for the upcoming 2021 NCAA Football season. Today's post will provide the SPAM season preview for the 2021 season.

Background (skip if you're familiar with SPAM)

The Super Prediction Action Machine ("SPAM"), it is a rudimentary statistical algorithm I developed many years ago to predict the winner of Iowa football games. The idea was to remove the subjectivity and bias imparted by polls, brand, and even win/loss records. The algorithm ignores all of those things, disregarding polls, rankings, recruiting, roster talent, and record. Football is a team sport, and Iowa as a whole is often greater than the sum of its parts. Thus, SPAM looks only at offensive and defense unit statistics..

The concept is simple: what is the point-value of a yard earned by the offense, what is the point-value of a yard surrendered by the defense? These values are weighted by opponent quality and SPAM projects how many yards Iowa is expected to gain (and thus, how many points Iowa will score) against the opposing defense (again, adjusted for opponent quality) and how many yards Iowa's defense is expected to surrender (and how many points that will turn into), also adjusted by opponent offensive unit strength. I then apply this across the entire schedule and use a basic regression analysis to determine the win percentage of each game based on the expected point differential (which I haven't updated in a few seasons now and probably should). Finally, I simulate the season 100,000 times to get an idea of the chances of Iowa going anything from 0-12 to 12-0.

This approach is not without its shortcomings. Because roster composition is ignored, issues like roster change year over year are not factored, and it ignores matchup problems and play style. This means that SPAM is generally the least accurate at the beginning of the season, when we have little or no data for the current team. To smooth this out, I blend in old data from the prior season, giving more and more weight to the current season as Iowa plays more games. This helps reduce the impact of Iowa's early season performance against non-conference foes, which doesn't usually tell us much.

2021 SEASON PREVIEW

Below is SPAM's projection of Iowa's estimated margin of victory (or loss, for negative numbers) in each game on the schedule for the 2021 season, and the corresponding estimated win percentage in each game based on that margin.

Game by Game


Opponent Margin Win%
Indiana 6.1 66.5%
@Iowa State -2.1 42.2%
Kent State 7.1 68.6%
Colorado State 17.6 86%
@Maryland 11.0 75.7%
Penn State 9.4 72.9%
Purdue 12.0 77.4%
@Wisconsin 0.5 54.1%
@Northwestern 1.3 56.0%
Minnesota 12.1 77.6%
Illinois 17.7 86.1%
@Nebraska 11.0 73.9%

Overall, the Hawkeyes have a pretty favorable schedule, but I should note that some major caveats are in order. First, these numbers are based on how these terms were performing at the end of last season, which was, to put it mildly, an unconventional season heavily influenced by the COVID-19 pandemic. So, while the numbers are overall not outrageously unrealistic for even a typical Iowa season, they are derived from a data set of questionable applicability.

But, taking this data at face value, a few things jump out at me.

Iowa State. This game is going to be weird and awful. This game is weird and awful in the best of circumstances, and is nationally famous for how weird and awful it is, having garnered the well-earned nickame,¡El Assico!. The 2021 edition of ¡El Assico! promises to deliver strongly on the assiness factor, but in a way that we Big 10 fans may actually enjoy. Let's catalog the weirdness factors.

The first weirdness factor is the gap. The last notable event in the Iowa-Iowa State rivalry was MarchingBandGate in 2019, an absurd, ridiculous, and avoidable event that resulted in Iowa and Iowa State each predictably blaming the other school for it and, in some cases, saying very stupid things. Tensions were poised to be high for 2020 at Kinnick, but, of course, that game didn't happen, marking (I think) the first break in the series in my living memory, and I was born during the Ford administration. I have no idea whether that gap will have cooled things down or heated things up. I certainly don't think it will matter to the players on the field, they're all going to show up and play to win. But I do wonder whether the pandemic-imposed cooling off period will have any impact on the home environment for the Cyclones, or if the gap in time might actually heat things up between the fan bases. Beating Iowa is the only thing Iowa State fans want out of Matt Campbell that he hasn't delivered yet.

The second weirdness factor is that Iowa State is good. Real good. We all know that the Cyclones had one of the best seasons in the history of the program last year (and they were close to a Big 12 championship, which, full disclosure, I went to my Iowa State friend's house, wore red and gold, and was openly rooting for). While some of us in the Iowa fan community have sneered at their record, the bottom line is that not many teams have beat Texas and Oklahoma in the same season, they had an excellent season by any reasonable standard, and the 2021 incarnation hasn't lost much. It has been rare in this series that both Iowa and Iowa State are good at the same time, and while Iowa is certainly no juggernaut in 2021, the Hawkeyes will still be competitive and could easily rattle off a 10 win season. To the extent these teams have been evenly matched in the past, it's been mutually assured assiness. Now, we have two rankable programs facing off. This is unprecedented, and I have no idea how that will play out.

The third weirdness factor, which relates to the second, is that Matt Campbell has been building something in Ames, but it's not the usual Big 12 Frankenteam. No, Coach Campbell has been recreating Iowa. The gimmick is no gimmick. The scheme is to be bigger, stronger, tougher, and kick your goddam ass up and down the field. The strategy is to win at the line of scrimmage. Field position. Ball control. Game management. Defense. Unlike prior installments, Iowa is not going into Ames to play Big 12 football. Iowa is going into Ames to play a mirror image of itself. And not a pale imitation, but an honest-to-God championship caliber team. This is new, and it's going to be weird.

The fourth weirdness factor is related, and it could either fuel or undermine the Cyclones. Iowa State hasn't been here before. When was the last time the Cyclones were pre-season top 10 and expected to compete for a Big 12 championship? They could even be in the running for a playoff berth, the committee rankings were extremely charitable to them last year despite the ugly early season loss. Are the players ready to stand on that platform? Will Campbell have them ready to be the hunted? Will they rise to the moment or will it overwhelm them? This is new, and it's going to be weird.

The fifth weirdness factor is that Iowa will have already played a Big 10 opponent, and one that presents nightmarish matchup problems. Iowa's last game against Indiana was also weird, with Brian Ferentz running up the score for some reason and as I recall it got a bit frosty on the field, too. I get the impression the Ferentzes don't like Tom Allen, I don't really know why, but Iowa could be 0-1 coming into Ames having already had to deal with a P5 conference opponent. This is new, and it's going to be weird.

Throwing all of that into the mixing bowl, SPAM has this game a virtual dead heat, with Iowa State getting the nod due to their considerable home field advantage. This is an ¡El Assico! for the ages, and I beseech you all to acquit yourselves respectably with the regimented decorum for which this rivalry is nationally famous.

I think, realistically, Iowa goes 1-1 against the combination of Indiana and Iowa State, and if I had to pick which game I'd rather win, it's Indiana. That game impacts Iowa's road to the conference championship. The Iowa State game does not. And, barring an inexplicable collapse by the Cyclones, a loss to Iowa State this year is unlikely to be a major stain on Iowa's record.

Now watch Iowa State win this game 38-20 and go 2-10.

Other Hashtag Noteworthy Thoughts.

Maryland #NOTEWORTHY for only one reason. There are only two Big 10 venues where KF has never won a game: Ohio State, and Maryland. Iowa has only played at Maryland once (2014) and lost that game. The Hawkeyes are at Ohio State in 2022.

The margins against Wisconsin and Northwestern clearly establish the Triangle of Hate in the West. As badly as Minnesota wants to make it a quadrilateral, they remain on the outside looking in.

Chinks in Wisconsin's armor have been forming over the last few years. Is their run of dominance in the West coming to an end? If so, who will fill those shoes? Because right now, it's Northwestern's division, they've got Iowa and Wisconsin's number.

Don't sleep on Nebraska. Never sleep on Nebraska. They hate Iowa with the fiery passion of 1,000 suns and despite the significant gap between the teams on paper, Scott Frost has kept these games close. In 2018, Iowa got a big lead but then went to sleep and Frost's Cornhuskers came back and made a game out of it. Iowa had to squeeze out a last minute field goal to avoid overtime. In 2019, Iowa again got a lead but it evaporated quickly and Iowa again had to pull a rabbit out of its hat with a walk-off field goal. Last year, Nebraska took the first lead it's had in this series in regulation in almost a decade, Iowa came from behind, but the Cornhuskers had the ball and a chance to win in the closing minutes of the game. Nebraska has been in these games, and although they probably won't admit it (I wouldn't, either, no judgment,) this game is their Super Bowl for now. Beating Iowa in what is probably their last game of the year sure would feel great for their fans going into the offseason. TLDR: Nebraska is the new Iowa State.

Record Projection

SPAM has done the math, and SPAM has predicted how likely every possible outcome is in this season based on the above winning percentages. Here's the odds.

Record
Odds
4-8
1.0%
5-7
2.0%
6-6
5.0%
7-5
9.0%
8-4
24.0%
9-3
36.0%
10-2
20.0%
11.-1 2.0%
12-0 2.0%

Probably no major surprises here. Iowa is solidly in the 8-4 to 10-2 range, and you can see how. The Hawkeyes are probably looking at 2-2 against the likes of Indiana, Iowa State, Wisconsin, and Northwestern, and it's not find to hard another loss or two lurking in the data. Of note, Iowa's pre-season prediction here is usually more heavily weighted towards 8-4, so on paper at least, the program is improving.

DIVISION CHAMPIONSHIP ODDS

SPAM also projects the seasons out for the rest of the Big 10, and, based on that, can project who is most likely to be the division champion for the East and West divisions.

Big 10 West

Team
Champ. Odds
Iowa
29.73%
Wisconsin
29.42%
Northwestern
23.93%
Purdue
4.00%
Minnesota
3.11%
Nebraska
2.19%
Ilinois
0.20%

For I believe the first time since I've been doing these, Iowa is favored to win the West, though by the slimmest of margins. Typically Wisconsin is about a 30-40% favorite. Iowa's 29% is relatively low, which I think actually reflects well on the competitiveness of this division.

Big 10 East

Team
Champ. Odds
Ohio State
47.57%
Indiana
27.89%
Penn State
10.37%
Rutgers
3.04%
Maryland
2.35%
Michigan
2.29%
Michigan State
0.50%




Well that got weird fast. If I told you a few years ago that Michigan and Michigan State would be rounding out the bottom of the Big 10 East, and that Ohio State's usual 70-80% chance to win the East had been eroded away by Indiana you might have had legal grounds to force me to submit to a psychiatric evaluation.

WHAT'S NEXT FOR SPAM?

Let's see how this season goes!

Unless otherwise expressly indicated by BHGP editors, this FanPost is strictly the viewpoint of the author and is not endorsed by BHGP in any way.