I'm going to restrict myself to one hour here. This is a topic where I could wander off on tangent after tangent so I want to draw some boundary lines. 60 minutes on the clock, have to publish before I hit 60:01. Starting... NOW. (You know I'll never pull it off.)

Several times over the last month, Bret Bielema has stated that he thinks this team is ahead of last year's team. And he consistently follows it with "I don't know if that will mean more wins than last year, but..." (followed by whatever statement he's making). I want to focus on the "I don't know if that will mean more wins than last year" part. To some, that's just a coach not wanting to get caught in a trap. If he comes out and says "I think this is a better team than last year", and then we go 3-9, it wasn't a better team than last year.

Or was it?

This is a topic that always bothers people. At the first sign that an article is going in this direction, many will skip to the end: "if you're going to tell me that a 3-9 team is 'better' than a 5-7 team then we have nothing to talk about; wins and losses, end of story." And that's true. Some baseball team is going to finish with 105 wins, and then some 92-win team will win the World Series, and someone will claim that the 105-win team was actually better, but it won't matter. One team has a trophy and the other does not.

But when it comes to a rebuilding football program, how much do wins matter? I mean, obviously, you can sell your rebuild to recruits when you go 6-6 and you'll struggle to convince any recruit that you're destined for greatness if you go 0-12. I'm more talking about 5 wins in a lucky season and 3 wins in an unlucky season. I should just get to talking about it.

As you probably know, my favorite NERDstat is Postgame Win Expectancy (PGWE). Football games have weird things happen. You probably know the example that comes next. A game that is going to end up a 28-27 win can be a 34-21 loss with one play: a 99-yard fumble return for a touchdown. You're about to score 7, and not only do you not score 7, they score 7. 14 point swing on one single play.

Because football is so volatile, it's easy to get lost in wins and losses. And you probably know the example that's coming NOW as well. Statistically, the 1994 Illini team was better than the 2001 Illini team. Replay those seasons and 1994 probably goes 9-2 and 2001 probably goes 8-4. The sun shined on 2001 and we won very close games against Penn State, Wisconsin, Northwestern, and Northern Illinois. The sun didn't shine on 1994 and we lost very close games to Penn State, Wisconsin, Michigan, Purdue, and Washington State.

Again, wins and losses tell you everything. Kurt Kittner, Brandon Lloyd, and Rocky Harvey made it happen in 2001 while the 1994 offense just couldn't do enough to help that amazing defense. They played the games, 1994 went 6-5, and 2001 went 10-1. They don't pick the national champion based on Postgame Win Expectancy. The team that wins the playoff wins the title. Even if it's the fourth-best team statistically.

But for a rebuilding team like Illinois (and let's face it - we've been rebuilding since 1994), PGWE can tell us if there's progress even if the wins aren't there. It's a stat that might back up Bielema's "we're better, but I don't know if the wins and losses will show it". Let me give you an example.

Once October gets here, we face Wisconsin, then Iowa, then Minnesota. Those games will tell us a lot about this team. Go 0-3 and we'll scream "Same Old Sorry Illinois". Go 2-1 and we'll start making bowl plans.

Fans don't just rest on wins and losses, though. Scoring margin matters. Losing all three games 55-3 is obviously different than losing all three games 28-27. Keep the games close and the debate will center around the inability to close games out, not the disaster of three straight 55-3 losses.

PGWE simply takes that a step further. In some cases, you can actually feel "better" about a 34-30 loss than a 35-27 win. I know some of you choked on that, so I'll give you a moment to compose yourselves and then I'll continue.

Example: Indiana beat Penn State in 2020 in one of the flukiest games of all time. Total yards were 488-211 Penn State and yet Indiana won. PGWE for Indiana was 5%. Play that game with those exact stats 100 times and history says that Penn State wins 95 out of 100 times. This was one of the five times Indiana would win. They needed, like, nine different things to happen and all nine happened.

Which is why I say that PGWE is a good indicator of where your rebuild is going. Are you winning a bunch of 5% games? Or are you winning both on the scoreboard and on the stat sheet? Bill Connelly adds up his postgame win expectancies to see if your record matched your stats (calling it "second order wins"). And sometimes it tells you there's good on the horizon. Best example: 2nd Order Wins for the 2006 season: 5.5. Actual wins: 2. That team was accelerating (as we saw the next season).

I probably should address the turnover thing before we move on because so many people think that this stat is just "got unlucky with turnovers". PGWE is NOT "Illinois was going to win but the four super unlucky turnovers meant that we lost." It looks at expected turnovers (you're expected to recover a certain percentage of fumbles; you're expected to intercept a certain percentage of passes defended), and if you turn the ball over way more than expected (meaning, say, you fumbled six times and lost five of them), it means you were a little unlucky and it probably won't happen again. The best example of that: the Oklahoma team 15 years ago that fumbled 18 times on the season and lost two. I don't care how good you are at diving on the ball once it's on the turf - a team is getting 1-in-500 lucky if they recover 16 of 18 fumbles in a season.

I bring all of this up because I tweeted this the other day:

BillC has the advanced box score up. Postgame win expectancy on Friday: 64.8%. Even with the mistakes, that’s a game Illini win 65% of the time.

(Also, please accept no Postgame Win Expectancy numbers bedsides Bill. I studied them all last year. This is the only one that holds.) https://t.co/wpEvZPjdTn pic.twitter.com/hNc03CxcGp

— Robert Rosenthal (@ALionEye) September 6, 2022

(I'm over an hour now so I need to wrap this up. But I've only now gotten to my point.)

Even with all the bad stuff that Illinois did on Friday night, it was still a game that Illinois wins 65 out of 100 times. It's why Illinois jumped from 79 to 70 in the SP+ this week after a loss. The numbers. They look pretty good.

I've seen this game compared to Maryland and Purdue last year. Chance to win, lose it right at the end. But this game wasn't anywhere close to Maryland and Purdue last year. Here's the PGWE for all three games:

• Maryland 20, Illinois 17. Expected scoring margin given the stats produced in that game: Maryland by 16. Actual scoring margin: Maryland by 3. PGWE for Illinois: 3%. In a game with those stats, historically, the team on Maryland's side of the stat sheet wins 97 out of 100 times.
• Purdue 13, Illinois 9. Very similar game to Friday: Illinois holds off Purdue until the very final drive... and then Purdue drives the length of the field to score the winning TD. Expected scoring margin in that one: Purdue by 8. PGWE for Illinois: 17%.
• Indiana 23, Illinois 20. Expected scoring margin given the stats produced in that game: Illinois by 3. PGWE for Illinois: 65%.

That's quite encouraging to me. Maryland and Purdue dominated us in the very-simple-yet-very-important Yards Per Play statistic. And we dominated Indiana. All three were losses, but this was the most encouraging loss. ("Encouraging" in the long term, or course, not "I slept great that night" encouraging.)

What's even more encouraging: after those 3% and 17% PGWE performances last September, the Illini finished the season above 50% in four of the last five games:

Penn State: 87%
Rutgers: 54%
Minnesota: 75%
Iowa: 29%
Northwestern: 100%

And then this season...

Wyoming 100%
Indiana 65%

That's where I'm hanging my hope right now. September last year: UTSA 13%, Virginia 0%, Maryland 3%, Purdue 17%. Now we've started 100% against Wyoming and then 65% against Indiana with Virginia coming to town.

Let's maybe flip last year's 0% over to 100%.