I know this is starting to look like The Providence Blog but they're the most interesting story of the 2021-22 basketball season and no one else seems to realize it. Anyway, my point is that if this were the 80's or 90s even, Providence at 25-3 would be #1 or maybe just #2 in the AP poll. Instead, even after 7 top 10 teams lost over the weekend they moved up to #9 from #11.
The difference is the influence of power ratings over the last few decades. Until the mid-80s debut of Sagarin's ratings they had no influence among the AP sportswriters obviously, and even through the 90s that influence was limited. When Pomeroy debuted in the early 2000s it took a while to catch on but especially in the last decade it's clear that the AP voters look at power ratings to guide their judgement of teams.
With Providence at #39 in Pomeroy no one's putting them close to #1. And that's not a bad thing, but it's pretty clear that in the pre-power rating era no one would have hesitated to give the Friars a 1st place vote.
Back then the RPI might have been somewhat influential. Not for the rankings themselves—no one ever thought they were good—but maybe for the SOS rating. That's how some people judged a team's schedule. More often, they just went by the team's conference. And being the regular-season Big East champion would be a strong argument for being ranked #1.
There have been teams in the past that ranked high in the AP poll while their power ratings were very low. 1982 West Virginia was 21-1 in late February and ranked #6 in the AP. They were only #76 in Strength, so that's roughly where Sagarin and Pomeroy would rank them, too. Their SOS in the RPI was a dismal #139, so it's easy to see why they didn't top the charts.
In 1983 Villanova was 19-4 and ranked #7. They're a lot more similar to Providence: same conference (in name at least, it carried more weight back then), and roughly the same SOS (52). This would seem like evidence that Providence would be similarly ranked, but in that poll all six teams ahead of Villanova had 3 or fewer losses; in this week's AP, six of the 8 teams ahead of Providence have more losses, and the #1 and #2 teams are the ones that have the same 3 losses.
So it seems obvious that 24-3 Providence would be up closer to where #1 Gonzaga (24-3) and #2 Arizona (25-3) are, if this were 1983 (or 1988, or 1993). And since their SOS (#47 via RPI, #59 Pomeroy) is better than either Gonzaga (#62, #112) or Arizona (#80, #71), it's a decent bet they would be #1.
Furthermore, Gonzaga would probably be the team lolling around at #9 instead. The 80s-90s precedes Gonzaga's flight to the top of the college basketball world, so they'd by the team out of place. Their tough early schedule would probably be held against them—look, they lost to the only really good team they played, Duke, and also to 10-loss Alabama! Ok, they did beat a good UCLA team, and Texas Tech isn't bad (though Texas is unimpressive at #38 RPI). So the Zags would be the team considered suspect for a top ten ranking, not Providence. Their weak conference would only have multiplied the effect. Are they even the best team in the West Coast Conference? People might have wondered in the 80s—after all, St. Mary's has a far better RPI (#5 to #13).
Arizona would have the backing of the Pac-12 to "legitimize" them of course. But they lack big wins, too. Illinois is their best out of conference win. Meanwhile Providence beat Wisconsin (RPI #6) and Texas Tech (#17), and the Big East outranked the Pac-12 in the 80s (and even this season!)
One could say, no it's not the influence of power ratings, the Friars just don't pass the "eye test," you can see Providence isn't as good as Gonzaga and Arizona. But can you? Without this year's expectations in play, I really doubt anyone could tell these three 3-loss teams apart visually. They all just win and keep winning. If you know Gonzaga was unstoppable last year and was expected to be unstoppable this season, that will affect how you "see" them. Arizona wasn't supposed to be great either, but 25-3 tells a different story. And Providence was supposed to be just plain mediocre. But here we are. And does anyone watch enough basketball to really know that much about a large number of team? Certainly that would be tough to do in the 1980s. And all you would see is three teams winning consistently.
So why do the power ratings make such a distinction to teams that have had similar results vs. (roughly) similar aggregate schedules? It comes down to winning margin basically. Providence doesn't blow teams away; they win just as consistently as Gonzaga and Arizona but not by as much. The computer can keep track of this much more easily than a human which is why, before power ratings, the win-loss record and the team's SOS was pretty much all that mattered to people. It's easy to point to certain games and discount them because of the opponent, but it's hard to keep a regressed tally of point spreads. On paper—wins and losses and opponents only—Providence rates just as well as Gonzaga and Arizona, maybe better: they're #1 in my Success ratings this week following last weekend's carnage.
It's true the Friars have had some close calls: they won three overtime games within a 2 week span! But in the past that was just taken as a team's "grit" and called "they know how to win." No one would have counted that against them in the olden days. It would have been a plus!
NCAA Tournament Seed: then vs. now
1982 West Virginia got 5-seed at 26-3, while 1983 Villanova got a 3-seed at 22-7. It looks like that's Providence's range for this year, too, but for different reasons. The Mountaineers only played one Top 50 team so that sank them with the Selection Committee, and Villanova ended up with more losses than any of the 1-seeds and two of the 2-seeds (the other two also had 7 losses). The top seeds went to the top four teams in the final AP poll.
Which means that this year, Providence would be in the running for a #1 seed if 80s criteria were used that would put them at #1 or #2 in the AP.
Is it unfair that they probably won't be considered? It's probably better that power ratings have some influence; after all, you want the teams that are really the best to be the top seeds. But you have to reward winning, too. What's the balance? That's something the Committee has wrestled with for decades. But in the past they didn't have Pomeroy numbers staring at them—they could make arguments based on schedule and results, and that's really it. Now that won't happen, especially since the NET employs some power rating features of its own (which is how Gonzaga is #1). Providence is #25 there, not a good omen for getting a high seed.
But if the Friars beat Villanova tomorrow and then win the Big East tournament, it would be hard to say they don't deserve at least a 2-seed, power ratings be damned.