Metuchen takes a team photo with its trophy after beating arch-rival Highland Park 29-0 on October 5, 2024. The Bulldogs met our projected minimum wins to make the playoffs, but missed this year. (Source: @metuchenhsfoot1 on Twitter)
Every year before the start of the season, we may our way-to-early playoff projections for the high school football season.
Sure, we look at who might be considered a contender – or even a favorite – to win a championship, but we also go by the numbers, looking at Strength Index and potential OSI numbers in the preseason, to see how many games each team may need to make the playoffs.
After so many years, we’ll assume most people have a fair knowledge of Strength Index, OSI, and the difference between the two. But, essentially, every team has a Strength Index value that changes throughout the year as games are played. Teams are then given an OSI value – Opponent Strength Index – for each game played, getting the full SI value of an opponent they beat, and half for an opponent they lost to.
Using this, we determine a team’s maximum and minimum OSI for the year, based on the average of all their opponents. The maximum would be for a team that beat all its opponents, the minimum would be if the team went winless. We then average out the OSI value per game to figure out how much each win is worth.
Then, we look at a continuous average of the minimum OSI needed to get in the playoffs. There’s no set number, and it varies from section to section, group to group, and year by year. Here are the averages per supersection since 2021, the first full year of OSI (since there were no playoffs in 2020 due to COVID, and the inaugural SI year in 2019 had numbers based off the one-year experiment with the Born Power Index, a slightly different formula).
We call this the “Target OSI”:
- North 5: 40.18
- South 5: 42.44
- North 4: 43.04
- South 4: 42.91
- North 3: 43.46
- South 3: 42.32
- North 2: 40.63
- South 2: 39.42
- North 1: 36.96
- South 1: 36.20
While larger schools tend to have slightly higher OSI targets, the teams tend to have larger SI numbers, so we have found in the last several years that Group 4 and 5 schools tend to have more wiggle room in making the playoffs. This calls for careful scheduling by leagues, who could hand a team a schedule that might require them to win more games to make the postseason.
In our projections, we give the number of wins a team would have to win to get closest to and over the minimum OSI, and then one more win for a full “step” over.
For example, Piscataway in North 5 had to reach 40.18. We projected they needed just three wins – going 3-6 – to reach just over that number, and four to be a playoff “lock.” They won six games, and not only made the playoffs, but got a first round home game. We projected Bridgewater-Raritan would need to win two, three for a lock. They won three, and also got a home game, due to their strong schedule. (Every team on the Panthers’ schedule made the playoffs this year.)
You can find our original projections in our story from August 29th by clicking here.
So, in our book, we got those “right.” How many did we get right this year?
All but one: Metuchen.
We got 56 of the 57 public school teams right, leaving out the two non-publics (St. Joseph and St. Thomas Aquinas) who are seeded by committee.
That’s a 98.2 percent rate, and on track with prior years. We got 54 of 57 last season, 55 of 57 in 2022, and 55 of 58 in 2021.
Overall, that leaves us at 220 out of 229 in four seasons, a 96.1 percent success rate.
Who did we miss? There were six teams that met the “closest” criteria, but didn’t meet our “lock” number. We consider those ones we got right, since they didn’t meet the number we projected would guarantee them a berth; that first win number could be a toss-up. Those schools included Linden, Franklin, Highland Park (which was ineligible, regardless), North Hunterdon, Sayreville and Voorhees.
So what happened with Metuchen?
The Bulldogs’ average OSI before the season was 49.01, meaning if they had won all their games – and SI values didn’t change – they would have a 49.01 OSI. A winless season would have gotten them a 24.51 OSI. Needing a 39.42 OSI to get in the playoffs in South Group 2, as we projected, we expected they would have to win five games to top that, and six to “lock” in a playoff berth.
But SI values do change, and five of the nine teams Metuchen played had SI values drop this year, the most notable of which was Brearley, falling from a 53.06 off an 3-5 season in 2023 to a 33.00 with a 0-10 campaign this year.
Metuchen met our six-win “lock” number, but missed the playoffs, finishing 17th. The top 16 teams make it in each supersection.
Ultimately, the best Metuchen could do – with an unbeaten season – would have been 46.19, down from 49.01 at the start of the year. That was enough to leave them out, as they finished with a 37.1 OSI. That was lower than Spotswood, which got in at 37.82, and was the lowest OSI team qualify. This year’s preseason target OSI to get in was 39.42, so it ended up being lower, but Metuchen just missed.
Here’s how far their opponents had fallen. In the preseason, their opponents were an average 3.05 SI points lower. By the end of the year, they were 20.31 points lower.
Spotswood was another issue. We figured they needed five wins to get in, six to make it a lock. As it turned out, they went 8-1 and still had to sweat it out. Their maximum OSI dropped from 49.31 to 41.02m while the teams they played were an average 1.48 points lower than them before the season, but 21.1 lower by Cutoff Weekend.
One of the issues we have found to be responsible for the wild variations is the lack of common opponents on a schedule.
When one team beats another during the year, SI values change. If Team A blows out Team B, Team A’s value generally shoots up, while Team B’s value plummets. A team that wins four or five games one year and none the next will drop down a lot.
The problem is, the fewer common opponents there are, the more variation you can have.
Consider if Team A was in a three-team division. Teams B and C have identical 80 SI values. Wins over both of them would be worth 80 points, total 160, and the average – or the OSI – would be 80.
Now let’s say Teams B and C played, Team B won and rose to a 100, while Team C fell to a 60. Now you might think, “Oh, no, Team B plummeted.” But if Team A beat both, they would get 160 total points, averaged out to the same number as before, an 80.
But now let’s say two other 80 teams, Teams D and E, played each other, with the same results: one rose to 100, the other fell to 60.
If Team A only plays Team E, they expected an 80, but only got 60. And they have no recourse to get those points back, because they don’t play Team D. If they played Team D, the loss would be balanced out by the gain.
In real life, Spotswood played South Hunterdon in a crossover this year. The Eagles fell from 63.80 in the preseason to 38.6 by Cutoff Weekend. Belvidere blew them out 36-0 on October 10th. But Spotswood doesn’t play Belvidere. So when South Hunterdon lost 6.08 SI points in that game, so did Spotswood. Had they played – and beat – Belvidere, that would have balanced out, but they don’t.
So, one solution to this issue would be to potentially have larger divisions and fewer crossover matchups, at least for the smaller schools, who are much more susceptible to swings in SI values than larger schools tend to be.
We’ve heard some say one-year scheduling cycles could prevent that, too, but even from one year to the next – look at Montgomery, which went 8-1-1 last season, but 1-9 this season – there can be many variables.
Larger divisions – built in some respect on the SI values of the teams in it – would be a big help, and would save the conference a lot of time and energy scheduling crossovers as well.
Discover more from Central Jersey Sports Radio
Subscribe to get the latest posts sent to your email.