2019 Week 4 Recap: Sunshine Day!

In retrospect, we all should have seen this coming. The signs were all there.  We know that MSU has a dominant defense.  On offense, the Tulsa game was not good. I was harsher than most following that game. But, the offense was much, much better against Western, and at least in the box score, it was better against Arizona State. Once the initial sting of last week's lose wore off, I was still pretty guardedly optimistic about the direction of this team.  Some might even refer to this belief as "sunshine blowing."

Many people, myself included, pointed out that the ASU game looked like an anomaly.  Sure, MSU put up only 7 points, but you change the outcome slightly of one of out of 10-12 different plays, MSU wins that game.  If you chance 3-4 of those plays, MSU wins going away.  Anomalies are frustrating because they can feel like reality. If you spend too much time staring at the fun-house mirror, you start to squint and think, "no, maybe that is actually what I look like..."  If you watch too many TV sitcoms, you start to think that is what life is really like. But, it's really just fiction.

Anomalies are simply low probability events, and if you play enough games, eventually every team is going to have a bad day.  Those bad days may or may not also correspond to days when you have Pac 12 officials... just ask Ole Miss.

But that was last week and this week was quite the opposite of a bad day.  Today was a very, very good day. Despite the blustery weather in Evanston, the sun seemed to shine on MSU today.

Now, don't get me wrong, I was about as stressed out before this game as any MSU game in recent memory. It felt like the whole season, and perhaps even the entire Dantonio legacy was on the line.  When White fumbled the punt early in the game, my heart sank.  As the 1st half drew to a close, it seemed like Northwestern was gaining momentum. It felt like just the kind of game where MSU finds a way to lose.  But then, Josiah Scott picked off a pass, Lewerke drove the offense down the field for a score just before the half.  After that, MSU never looked back.

By the end of the game, everybody was smiling. Coach Dantonio went for a walk to midfield to receive congratulations from Coach Fitz, my personal 2nd favorite coach in the Big Ten.  Then coach went for another walk back to the locker room to celebrate with his team, the Spartan Bunch, as the all-time leader in wins in MSU history.  In might have been overcast in Chicago, but it was certainly a Sunshine Day.

By the end of the afternoon, when all of the scores from around the Big Ten rolled in, everybody was laughing. It seems that Older Brother (more Eddie Haskell than Greg Brady, in my experience) had a rough day in Madison. So sad for them. But in East Lansing, everybody seems so happy today.

The question is now is how long with this happiness last? MSU's performance, on balance, over the first four games has given us the road map. MSU's defense is as good as advertised, and now ranks 3rd overall in yard/play (behind... Wisconsin and Ohio State... gulp).  MSU can win A LOT of games with that defense.

As for offense, they have shown that when they don't shoot themselves in the foot, they can be a potentially potent unit as well.  We all said that they just needed to be average for MSU to have not just a good, but a great season. Well, MSU currently ranks 75th in total offense (yd/game) and 84th in yards/play.  That's not great, but it is progress, and I think there is the room and potential to get a lot better.  This game and this weekend, much like the summer sun, has shown the way to be happy now. Here's to hoping that we continue on that path.

Big Ten Metrics Update

Considering the results from Week 4, here are my updated odds to win the Big Ten Divisions and overall Conference:

As for the expected win totals and games where a team is favored:

A more visual representation of the data, along with the trends for the years are shown here:

Honestly, the results from this week do not actually change much of the big picture. Ohio State is still the huge favorite in the East at 92%, while Wisconsin is a huge favorite in the West (80%). As for the overall race, the Buckeyes hold a fair lead over the Badgers to win in Indy, 56% to 38%.  Part of the reason for this is that the 2nd place team in the West, Iowa, still has reasonable odds to make it to Indy at 19%.  In contrast, my algorithm still likes MSU as the 2nd best team in the East, but the Spartans have only a 6% chance right now of unseating the Buckeyes.  

Ironically, MSU's numbers have been quite stable this year. overall.  MSU actually went up this week in my power rankings, from 17 to 13.  Penn State also rose from 30 to 20, despite the fact that they took the week off.  Pitt's upset of UCF probably helped that, while MSU's opponents, did not fare as well this weekend.  

MSU's expected win total is now at 8.3 plus or minus a game.  I give MSU a 9% chance to win at least 10 games, a 44% chance to win 9 or more, and an 81% chance to win at least 8.  Another way to look at the data is that I project that there is a 72% chance that MSU wins 8 or 9 games total.  That seems pretty reasonable and is in-line with most of the preseason predictions.  As for my projected lines for the rest of the season, I have:

• MSU (-16.7) vs. Indiana
• MSU (+17.1) at Ohio State   
• MSU (+14.1) at Wisconsin   
• MSU (-6.8) vs. Penn State   
• MSU (-26.3) vs. Illinois   
• MSU (-5.2) at Michigan   
• MSU (-30.3) at Rutgers
• MSU (-15.8) at Maryland

Some of those lines may look a bit odd, but keep in mind that I said all week that my algorithm had Wisconsin as a 25-point favorite over Michigan.  How did that one turn out? Speaking of Michigan, my spreadsheet predicts more bad news for the Wolverines.  I only give them a 37% chance to finish with 7 wins or more, and I only have them favored in 3 more games this year, Rutgers, at Illinois, and at Indiana, while the road game at Maryland looks like a toss-up. 

Other metrics (like the FPI) are clinging to the idea that Michigan might still somehow be a Top 20 team, but this seems like it is based more on recruiting rankings and nostalgia with the winged helmets than on any relevant data. My power rankings have them at 35 overall, but that number is still being influenced by their Top 5 preseason ranking. Based on what we have actually seen on the field this year, Michigan is basically a 0.500 team. Until they prove otherwise, that is what they are.

National Overview

As usual, here is my weekly comparison of the week's results to the opening Vegas spread.

I should note that I did not include Ohio State's blow-out of Miami (OH) because it was so far off-scale. Other than the Buckeyes, the teams that over-achieved this week include Iowa State, Kent State, UAB, Washington, Wisconsin, MSU, Florida, and Syracuse.  On the flip side, Baylor, Virginia, and Miami (FL) were well below the line.  

As for my conference-by-conference hot takes:

• Honestly, nothing really surprising happened in the Big Ten this week for those that are actually paying attention. Wisconsin and Ohio State still look really good. Michigan is over-rated, and MSU has potential.
• In the SEC, we mostly learned about which team is the 3rd best in each division: Auburn and Missouri. Other than that, it looks like an Alabama-LSU show down in the West and a Georgia-Florida show-down in the East. Also, Arkansas and Pac-12 officials still suck.
• The gaps between Clemson and everyone else just keeps growing in the ACC. Virginia still looks like the second best team, and they trailed Old Dominion by 17 in the first half. On the bright side, Boston College avoided embarrassment in Piscataway, and Pitt had a nice win over UCF.
• In the Big 12, I guess it is notable that Texas managed to beat Oklahoma State, who has given them fits for years. I can't say the same thing about TCU, who went down against SMU.  More than anything, that just begs the question of how bad is Purdue?
• It seems like every Saturday night, I go to bed and I tell any remaining Pac 12 team that is still awake to behave themselves. Serious guys, I can't take you anywhere... This week's offenders were Washington State and Arizona State, who both suffered upset losses well into Sunday morning on the East Coast. That, coupled with USC's upset of Utah on Friday and, as usual, the conference is a mess.
• In the Group of Five, UCF looked to by on the verge of lapping the field. But, they stumbled in Pittsburgh this weekend, and now things are a bit murky.  While it still seems likely that the AAC Champ will play in the Cotton Bowl, I must point out that Boise State is undefeated, and Utah State had a nice win over San Diego State. Furthermore, Appalachian State is also 3-0 after their upset win over UNC this weekend.

Upset and Performance Against the Spread (ATS)

Overall, my algorithm has a respectable weekend. It went 24-25 ATS (49%) bringing the year-to-date tally to 95-93 (50.5%).  I can live with that.  ESPN's FPI did very slightly better at 25-24 (51%), but its total is only 80-108 (43%) for the year.

As for my recommended bets, things look much better, for my algorithm specifically.  Here is the summary table:

It was a good week to place some bets.  My spreadsheet was an impressive 5-1, while the FPI struggled this week to a 1-3 record.  Overall, the combined performance was 6-4 (60%) which I will take any week of the year.  In total, my data is 14-11 for the year (56%), the FPI is 11-10 for the year (52%), while the combined strategy is 24-20 (54.5%).  That is pretty much right where I thought it would be (53%) based on my historical data.

As for straight-up upsets, my math did not have a great week.  The summary table is shown below.  

I count a total of 10 upsets relative to the opening spread, including two pretty big ones: San Jose State over Arkansas and UCLA's Pac-12 After Dark upset of Wazzu.  Despite the fact that, once again, my simulation correctly predicted the number of upsets within 2, my spreadsheet only went 1-5 with actual upset picks this week.  That brings the year-to-date total to 6-9 (40%) which is right at my historical average. The FPI did slightly better at 2-2 (50%), bringing its total to 4-8 (33%) YTD.

Finally, here is my updated Power Rankings Top 25:

1. Georgia
2. Ohio State
3. Wisconsin
4. Clemson
5. Alabama
6. Oregon
7. LSU
8. Auburn
9. Oklahoma
10. Notre Dame
11. Florida
12. Iowa
13. Michigan State
14. Missouri
15. Texas
16. Washington
17. Iowa State
18. UCF
19. Texas A&M
20. Penn State
21. Utah
22. TCU
23. Arizona State
24. Kansas State
25. Mississippi State

Again, some of these look a bit odd, but it is really any less valid looking than the AP poll?  In any event, the influence of the preseason rankings is starting to fade as more data accumulates. Right now, I am still using that data.  I think that I will still use it for next's week's projections, but after that, only data from this year's actual on-field performance will be considered. 

That is all for now. Keep a look out for my Week 5 preview a bit later in the week.  Until then, enjoy, and Go Green.