The NJSIAA's playoff formula for 2019 is still being finalized, but a minor tweak will make it very different from what it was in a very controversial 2018.
The basics of the formula were presented Monday night to the New Jersey Football Coaches Association at its monthly meeting at the Hale Center at Rutgers University. While the NJSIAA is not releasing anything official, WCTC Sports was able to get some details by talking with some of the coaches, whose names aren't being used because they don't speak for the NJSIAA.
The main two points in a presentation by Westwood Athletic Director Dan Vavino are that spreads are out of the formula, and that it will be shared with coaches and the public. Vivino is a member of the New Jersey United Committee, the group that was tasked in 2018 with fixing New Jersey's playoff system, and includes coaches and ADs from throughout the state.
During a 2018 season where the playoff formula included the Born Power Index - a formula that takes into account margin of victory in rating teams, and rewarded teams with higher playoff seeds for winning by larger margins - Piscataway volunteer assistant coach John Thompson suggested using the formula a different way: as a strength index.
A new formula that's based on the Born Power Index is being proposed to replace it: the OSI, or Opponent Strength Index.
In 2018, 60 percent of a team's Born Power Index ranking position - based on its rating - determined its playoff seeding. But in 2019, the team's own rating - affected by margin of victory - would not determine it's seeding. Rather, the total points the team accumulated based on the strength of its opponents - and whether it won or lost - would determine the team's seed. That removes the incentive to "run up the score" or win by larger margins.
The plan is still being tweaked, and may change in small ways between now and April, when the NJSIAA Executive Committee is expected to have a final say on its implementation. But here's the proposal has been explained to WCTC Sports:
- The UPR remains. In 2018, the UPR was 40% power point ranking and 60% Born Power Index rating. The Born Power Index will be replaced by the OSI (Opponent Strength Index). The OSI would be maintained by the state, likely through the website Gridiron New Jersey, as opposed to by Bill Born.
- Power points remain unchanged. The formula for power points would remain the same. It would still be calculated as an average so that all games count. Multipliers would remain; the values - so far - would remain unchanged.
- The Born Power Index is gone. The BPI is replaced by the OSI, which is similar, but used in a different way, to truly discourage wider margins of victory.
The NJSIAA has admitted that no formula will be perfect and please everybody, but Vivino said in an email Monday night he believes the new formula satisfies the major themes of transparency and promoting sportsmanship.
How is the OSI work?
The Born Power Index was never publicly released, but analysis by WCTC Sports found that margins of victory in games mattered. The difference between two teams' ratings was called a "spread" or "projection." In our observations, how a team did against that projection determined whether its rating rose or fell, and by how much.
That's not exactly how Born's formula works, on paper. But the relationship between the numbers is the same.
According to coaches who attended Monday night's meeting, the NJSIAA now proposes using a similar method to what WCTC discovered in its OSI formula.
In 2018, WCTC Sports found that for most of the season, a team's BPI rating increased or decreased by 1/6 the margin of victory against the projection. If Piscataway's rating was a 90, and South Brunswick was an 80, the projection was 10. If Piscataway won by 16, they beat the projection by 6. One-sixth of 6 is 1. Piscataway would go up 1 point to 91, while South Brunswick would drop 1 to 79.
The NJSIAA's working proposal would use something similar: 1/5 of the margin vs the projection, all season long. But the main difference is Thompson's initial proposal.
In 2018, Piscataway's 91 rating, influenced by their margin of victory, would determine their seeding. If they'd won by more, their rating would go up and their seeding would go up.
In 2019, Piscataway would get 79 points for beating South Brunswick. Had they lost, they would get half of whatever South Brunswick's new rating was. Their winning margin has no bearing on their ranking. South Brunswick's number is the one that matters to Piscataway.
The total points accumulated, averaged out over the number of games played, would determine a team's ranking, and account for 60% of their UPR.
Vivino says the hope is this formula encourages schools to schedule teams of a similar strength. Scheduling lowly-ranked opponents for easy wins would hurt their playoff chances.
The OSI Formula
Here's the premise of the OSI formula, as explained by the coaches we spoke to (keep in mind, the formula has not been finalized, and anything could be tweaked or changed between now and a formal presentation to the NJSIAA Executive Committee, which is expected to have a final say in April):
- Teams are centered at the start of each year. The BPI used historical data each year. That is, a team began 2018 with the same number they ended 2017. The new proposal would center teams around a midpoint of 60, and reduce them by a fraction of 1/3 of the difference to that midpoint. Simply: Piscataway ends 2019 as a 90. The difference to the midpoint is 30. One-third of 30 is 10. Piscataway becomes an 80. This replaces the "in-season" adjustment period of the Born Power Index, which lasted 3 or 4 games, but was murky, at best. Now, the adjustment is uniform across the board.
- The Margin vs. the Projection is still used. Simply, a team's OSI is calculated by 1/5 the margin against the projection. If Bridgewater is an 80 and Westfield is a 78, the projection is 2. If Bridgewater wins by 7, they beat the projection by 5. One-fifth of 5 is 1. Bridgewater rises 1 to 81. Westfield falls 1 to 77. The same calculation is used all season long, unlike the BPI, whose formula changed at some unknown point around weeks three or four.
- The OSI totals up throughout the year, and is averaged. A team's OSI rank is determined by it's total OSI points accumulated, averaged out. Say Bridgewater had a 2-game season. Bridgewater beats Westfield with an OSI of 81 and loses to Piscataway with an OSI of 90. Bridgewater gets 81 for beating Westfield, but 45 for beating Piscataway. The total is 126. The average (of two games) is 63. Where the 63 places Bridgewater in the OSI standings would determine 60% of Bridgewater's UPR.
- The OSI numbers change throughout the year. A team's OSI average will be based on the opponents' final OSI on cutoff weekend. So, if Bridgewater gets 81 for beating Westfield in Week 1, but Westfield drops to a 70 by cutoff weekend, Bridgewater gets a 70. This way, every team gets the same amount for winning or losing to Westfield, regardless of what point in the season they play Westfield and what their rating was at the time.
- No Blowout Cap. This is what started the controversy last year: The Born Power Index had some kind of cap, but not the 35 points that the NJSIAA has where a mandatory running clock goes into effect in the second half of games. Due to this reveal by WCTC Sports, Born began using a cap throughout the season. But with OSI, it doesn't matter. If Team A beats Team B by 60 points, Team A's playoff seeding doesn't depend on its margin of victory. Rather, it depends of the point value of Team B. In fact, if it beats Team B by only 20, Team B's value is higher to Team A. In this way, it's better for Team A to win by fewer points.
Other Proposed Tweaks
There are a couple of other minor tweaks that have been proposed.
While it's highly unlikely this will ever occur, a safeguard has been suggested that all 8-0 teams should automatically qualify for the postseason. Likewise, no winless teams would be allowed to qualify for the playoffs. That nearly happened last year with Cherokee - with a record of 0-7 at the cutoff in South Group 5 - narrowly missed making the playoffs. In the end, they were edged out by Hillsborough.
The NJSIAA also plans to "snake" brackets in an S-format, just like the NCAA does with it's March Madness basketball tournament. In that tournament, the top four teams get the four seeds in each region, rather than being seeded on pure geography. It's expected the state will still take the best 16 North teams and the best 16 South teams, then snake them into two divisions each. Some have proposed taking the best 32 teams statewide, but that could potentially lead to long travel times for some schools.