San Francisco beat the best team in football by twenty in Melbourne, and our model moved two full wins
Updated
Twenty-seven to seven. On the other side of the world, at the Melbourne Cricket Ground, against the team our model had at fifteen wins in July. That is how San Francisco opened the season, and it is the kind of result that forces a projection to sit up and pay attention. In the summer, when we ran the simulations, the 49ers came out at six and a half wins, more than three below the market, because a defense our grades rated 30th in the league kept giving back everything the offense built. One week of football does not erase a grade. But this was not an ordinary week, and when we re-ran the season with the Week 1 numbers seeded in, San Francisco climbed to nine wins, a 9-8 finish with a 42 percent shot at the playoffs.
Start with what actually happened. The 49ers held an offense we still project to lead the league in scoring to seven points and held Matthew Stafford to four completions in the first half. Renardo Green picked off a deep ball meant for Puka Nacua at the 14. Keion White recovered a Tyler Higbee fumble. Osa Odighizuwa stopped Stafford on fourth and goal. Fred Warner piled up eleven tackles and passed Patrick Willis as the franchise's all-time leader. The secondary that looked so thin on paper in July did not get carved up, and that is the single most surprising sentence anyone could have written about this team a month ago. Our model does not treat one game as a season, but it does treat the stats from that game as real, and those stats describe a defense that is closer to average than to awful. It is worth noting the Rams played that game without Aaron Donald, who was held out of the trip, and with Myles Garrett on a knee that has since needed surgery.
The numbers moved accordingly. In July our projection had San Francisco allowing 424 points, one of the worst marks in the league. After Week 1 that figure drops to 363, which sits 17th, right in the middle of the pack. That is the whole story of the two-win jump. The offense was never the problem, and it still is not. Brock Purdy went 25 of 34 for 205 yards and three touchdowns in Melbourne without being sacked once, spreading the scores to Demarcus Robinson, Mike Evans and Deebo Samuel. Our model projects 375 points scored, 13th in the league, and the pieces are the same pieces. Purdy sits at 4,074 yards and 27 touchdowns with a passer rating near 98, a touch fewer yards than the summer projection but a cleaner, more efficient line. Christian McCaffrey remains a two-way weapon, our model handing him 1,023 rushing yards, eight scores and 74 catches for another 626. Mike Evans is the number one on the outside at 81 catches and 1,037 yards. George Kittle adds 56 grabs and six touchdowns, and Deebo Samuel, back in the fold, chips in 59 catches. That is a top-half offense with a quarterback who does not turn it over.
The defense is where the re-run tells a different story than the preview did. Nick Bosa did not register a sack in Melbourne, but the pressure he generated showed up in Stafford's 61 passer rating, and our model now projects him for nearly sixteen sacks, up from just under eleven in July. Keion White, an afterthought in the summer, now projects for ten of his own. Two edges getting home changes everything behind them. A secondary that has to hold up for four seconds looks thin. A secondary that has to hold up for two and a half looks fine. Warner still projects for 133 tackles and three interceptions, and the unit as a whole comes in at nearly 13 picks, eighth in the league. That is a takeaway defense, not a sieve.
Honesty requires the other half of the sentence. One game against the Rams is a small sample, and a defense that grades near the bottom of the league in July does not become a top-ten unit because of one Sunday. Our model still has San Francisco allowing more points than it scores in a meaningful share of simulated seasons, and the depth behind Bosa, White and Warner is the same depth it was. The climb to nine wins is the model saying the defense is competent, not that it is good. Competent is enough to matter with this offense.
The division is the reason nine wins is a ceiling as much as a floor. The Rams are still the best team in football by our numbers even after losing in Week 1, and Seattle still grades as a contender on both sides of the ball. The 49ers have already banked one of the two Los Angeles games, which is worth more than a normal win, but the return meeting and two Seattle games are still coming, and the Rams will have Donald and, from Week 6, Garrett for those. When we ran the rest of the season, San Francisco split those matchups more often than it swept them, and that is the difference between a wild card and a division title.
The market opened at ten wins, and in July we were three and a half under it. Now we are a little more than one under, and the direction of the disagreement has not changed even if the size has. The market bought the names. Our model bought the whole roster, and the whole roster still has a defense we would call adequate rather than a strength. The gap is narrower because the 49ers went out and proved that adequate is at least on the table. It is not closed because one game does not make a defense.
The path to blowing past nine runs through the pass rush staying healthy. If Bosa and White combine for 25 sacks the way our re-run suggests, the secondary is protected for a full season and the 49ers are a ten or eleven-win team with a real shot at the division. The path to falling short is the July version of the defense reappearing once the schedule brings a run of good quarterbacks. Both are live, and our model splits them almost down the middle, which is what 9-8 and a 42 percent playoff chance actually mean.
So the read on San Francisco has changed, and we will say so plainly. In July this was a team we called overrated on reputation. After one week it is a team our model calls a coin flip for the playoffs, carried by an offense that was always good and a defense that just showed it can be enough. We take the number, respect what happened in Melbourne, and wait for the second sample before we believe the defense is anything more than average. Nine wins is the honest projection, and it is a very different sentence from the one we wrote in the summer.
