Pitchers and catchers report to Spring Training today, and while the focus around Major League Baseball is on the larger bases and the pitch clock, Joe West allegedly fighting editors of his own Wikipedia page is the most compelling story of the week so far.
There will be plenty of time to say plenty of things as the MLB season approaches, and we don’t have anything interesting to say right now, so we’ll let the game practice on. The first robin has sung, though. Pitchers and catchers have reported. Summer will come again.
Kansas Is Winning the Big 12
It’s just about official. After sweeping the Oklahomas by an average of 17 points apiece and watching Texas fumble the metaphorical ball in Lubbock, KU is back tied for first in the Big 12 with only five games to play. The five aren’t easy—they host Baylor, visit TCU, host West Virginia, host Texas Tech, and visit Texas—but the Jayhawks are set to be favorites in the first four, and Texas still has to go to Waco while Baylor still has to visit all three of Lawrence, Stillwater, and Manhattan. TCU’s fallen off due to injury. Iowa State and Kansas State have shown who they are. KenPom projects a three-way tie, still, but it’s hard to bet against Bill Self in the Big 12, especially with Texas no longer holding serve.
On the topic of TCU, Iowa State’s hosting the Frogs tonight, trying to bounce back having lost four of their last five. It’s unclear still if Mike Miles, Damion Baugh, and Eddie Lampkin are playing, but even if all three are on the court, they 1) shouldn’t be at full strength and 2) shouldn’t make TCU better enough than Iowa State to make the Cyclones an underdog. This isn’t a must-win—we’re past that point for a while, unfortunately—but it’s an attainable good win ahead of a tough road trip, and we’re at a point in the season where holding a solid spot in the AP Poll could make a difference in NCAA Tournament seeding. ISU’s probably two wins away from locking in a bid. Go make that magic number one, fellas.
Can the Eagles Regroup?
Shane Steichen’s in Indianapolis. Jonathan Gannon’s in Arizona. Jason Kelce’s rumored to be retiring, something like eight other Super Bowl starters are free agents, Jalen Hurts could be due for a big payday. Eagles head coach Nick Sirianni and general manager Howie Roseman are tasked with rebuilding a team that not only has to deal with the Super Bowl hangover, but has to deal with massive turnover and a tough division and a tough interdivision schedule.
Still…
Look at the medium term.
The Eagles might not be back to the Super Bowl next year. The Eagles might not win their division next year. But compared to the rest of the NFC East, who are you taking over Sirianni & Hurts? At some point, you’d assume they’ll figure it out. It might require a step back in 2023, and things can always fall apart, but right now it’s hard to find many franchises in better shape than the one in Philadelphia.
Also, while we’re here: I’m not sure the Colts hiring Steichen based on his plan to develop Bryce Young necessarily means they’re trading to draft Bryce Young. For two things, there’s so much time and evaluation ahead of everyone, and that headline about Steichen’s plan for Young wasn’t based on an explicit plan for Young (it was about a generic young quarterback, as I understand the reporting). For another, Jim Irsay is not unbending. If he was, Jeff Saturday would still be the head coach in Indianapolis.
Fresh SP+
Bill Connelly released his preseason SP+ rankings yesterday, and this is what we’re working with nationally:
- The top seven are no surprise at all: Georgia, Ohio State, Michigan, Alabama, Penn State, Tennessee, and LSU, in order. The top six aligns, almost in perfect order, with where Movelor ended 2022 (LSU finished 9th for us, and we had Alabama/Michigan and Tennessee/Penn State flipped by fractions of points). It’s the SEC and the Big Ten’s world, and the gap between those is lesser than it appears, largely because the narrative has yet to grasp how darn good Penn State is (Penn State whose road games this fall are at Illinois, Northwestern, Ohio State, Maryland, and Michigan State, at least giving them a path to 11–1 even if that finish isn’t their most likely outcome).
- Oregon, USC, and Utah are the highest-ranked teams in the Pac-12, coming in at 8th, 10th, and 13th.
- Texas and Oklahoma are the highest-ranked teams in the Big 12, sitting at 9th and 14th.
- Florida State and Clemson are the highest-ranked teams in the ACC, at 11th and 12th.
- Notre Dame is 15th. TCU is down at 19th. Washington, Texas A&M, and Mississippi are between them.
Projecting this onto the playoff race, with schedules also set, it seems Georgia is as close to a lock as a team could be in February. Ohio State, Michigan, and Penn State should contribute at least one playoff team between the two of them, and the SEC West stands a good chance of getting a team in as well. Call it three from those top seven, with Tennessee facing the tough situation of having to host the Dawgs and play the Tide on the road. On the ACC front, Florida State’s game with LSU is huge, and it’s tough that the Seminoles play Clemson away from Tallahassee. Clemson also hosts Notre Dame this year, rather than going to South Bend or avoiding the Irish altogether, making them a strong candidate. USC has to go to Eugene and South Bend, and Oregon has to go to Salt Lake City, so the Pac-12’s cannibalization of itself is likely. As for the Big 12? Texas should be favored in every game except for possibly the one in Fort Worth. This doesn’t mean Texas will roll through the field (and Oklahoma should be favored in every game except for the ones in Arlington, putting them in the picture as well), but it sets the stage.
Overall, the baseline expectation is that three of UGA/OSU/Michigan/Bama/PSU/Tennessee/LSU should make the field, and that the last one should come from Texas, Oklahoma, and Clemson. That’s just a starting point, though. Obviously. It is merely February. (For Connelly’s take on this: He has Georgia, Ohio State, and Michigan as the only schools with at least a ten percent chance of starting 12–0, with I believe only eleven teams holding at least a ten percent chance of starting 11–1.)
In the Big 12, the fourteen teams stack up in Connelly’s computer as follows:
- Texas: +21.4
- Oklahoma: +18.9
- TCU: +15.4
- Kansas State: +14.8
- Oklahoma State: +10.8
- Texas Tech: +10.6
- Baylor: +9.8
- UCF: +9.6
- Cincinnati: +7.6
- Iowa State: +6.3
- West Virginia: +5.3
- Houston: +4.3
- Kansas: +3.1
- BYU: +1.6
If BYU’s your worst team, your league’s in great shape. Depth in more sports than just basketball these days.
Translating this onto Iowa State’s schedule, the Cyclones are looking at two games as a heavy favorite (three or more scores), three games as a medium underdog (two possessions), two games as a small favorite (one possession), and five games as a small underdog. That’s a 4–8 record if the Cyclones take care of business, with a need to win more of those toss-up games than they lose if they’re going to make a bowl. I would guess, given that they aren’t a heavy underdog at all, that their projection is closer to 5–7 than 4–8, but that’s the territory we’re in. The goal is making a bowl. Start 3–0 and they’ll be in position to go 4–2 at home, needing only one road win beyond the one they’d have already gotten at Ohio. Meaning? As is often the case, the objective is to beat Iowa.
Model Talk
We’ve gotten some good traffic off our bracketology update from earlier this week, so a few thoughts on it:
I’m optimistic about this year’s Paymon Score (the score Bracket Matrix uses to grade final bracketologies). I love that the formulas have three straight years up around the market average rather than only two. At the same time, though, our final performance has exceeded my lowest expectations each of the last three Selection Sundays, so I should probably chill out.
It’s hard to bracket good mid-majors. This is probably because they’re rare enough that each is unique, making it difficult to draw comparisons year over year. Saint Mary’s jumped four seed lines between Friday and Monday, almost all from us changing the relevant formula. Florida Atlantic dropped substantially. There was little movement from high-majors, low-majors shuffled a little but they’re so stratified it didn’t matter much, mid-majors are wildcards.
The next three steps are building out our simulators, which involves a straightforward step, a mixed step, and a convoluted step.
The straightforward one is figuring out how to run our model “lukewarm.” We ran it “hot” during the tournament last year, meaning we updated KenPom ratings after every game proportionally to how they normally adjust after every game, reflecting how teams get better and worse in correlation with their strong and weak performances. This ended up overdoing the randomness. Why? The best way I can explain it is that KenPom ratings two or three or four games ago still retain some predictivity. This isn’t like Movelor, where ratings could go in any direction. KenPom ratings are usually pretty correct. They move, but they often move around a long-term mean. So, we’ll need to tone down just how hot our model runs, which will involve some trial and error but shouldn’t be too difficult.
The mixed step is handling conference tournament seedings, given all the tiebreakers. The ones that really matter for this are at the top of leagues, especially mid and low-major leagues where NIT automatic bids are at stake. We’ll probably end up hard-coding individual tiebreakers rather than building the system to handle them automatically. This is a pain in the ass, but figuring out how to code in every league’s tiebreakers has thus far proved more painful in the gluteal region.
The convoluted step is forecasting changes to NET, KPI, and SOR. We have some good data on these from a few years ago, but it’s bulky to work with. Making matters worse, to estimate SOR we probably need to estimate BPI as well. Our thought right now is to try the following:
For NET, we’re going to take the individual KenPom changes and add and subtract them from the NET score. Our understanding of NET is that it’s kind of like a recency-less KenPom, with a few added tweaks. We’ll ignore the tweaks and treat it like a sum of every KenPom over/underperformance, then check how that turns out in backtesting.
For KPI, we’ll do a straight elo adjustment with no margin included. We aren’t sure what else to do with this one. We have the poorest understanding of how it works out of any of the relevant systems.
For SOR, we’ll do a margin-based elo adjustment for BPI and then try to build an SOR system based off the new BPI ratings.
It’s possible all of this will prove too bulky to run as many simulations a day as we feel we need. If it does, we’ll likely revert to some simpler proxy. In the end, that might be just as accurate as our attempts. We need a week on this stuff in the summer. At least a week. Golly.
**
What’s happening tonight:
College Basketball (the big ones)
- 7:00 PM EST: Alabama @ Tennessee (ESPN2)
- 7:00 PM EST: Xavier @ Marquette (CBSSN)
College Basketball (the good ones)
- 8:30 PM EST: Kentucky @ Mississippi State (SECN)
- 9:00 PM EST: Indiana @ Northwestern (BTN)
- 9:00 PM EST: TCU @ Iowa State (ESPNU)
- 9:00 PM EST: Arkansas @ Texas A&M (ESPN2)
College Basketball (national team of interest)
- 7:00 PM EST: Virginia @ Louisville (ESPNU)
NBA (the best two games, plus the Bulls)
- 7:00 PM EST: Bulls @ Indianapolis (League Pass)
- 7:30 PM EST: Cleveland @ Philadelphia (ESPN)
- 9:00 PM EST: Dallas @ Denver (League Pass)
NHL (the best two games)
- 9:30 PM EST: Detroit @ Edmonton (ESPN+)
- 9:30 PM EST: Colorado @ Minnesota (TNT)
NASCAR
- 8:15 PM EST: Daytona 500 Qualifying (FS1)
Premier League
- 2:30 PM EST: Manchester City @ Arsenal (Peacock)
EFL Championship (games of promotion/relegation consequence)
- 2:45 PM EST: Wigan Athletic @ Bristol City
- 2:45 PM EST: Huddersfield Town @ Stoke City
- 2:45 PM EST: Luton Town @ Preston North End
- 2:45 PM EST: Blackpool @ Swansea City
- 3:00 PM EST: Blackburn Rovers @ West Bromwich Albion
- 3:00 PM EST: Middlesbrough @ Sheffield United (ESPN+)
