Predicting the 2025 Cal Football Season: Results!
Another 7-5 prediction that can easily veer into disaster or triumph.
How would you feel about going to Vegas and wagering a large sum of money on the outcome of five coin flips? Getting 1 correct should be trivial (unless you’re severely unlucky and manage that 1 in 32 chance of getting all five wrong), and getting 2 right is highly likely. But would you push your luck and bet on getting 3 correct, maybe even 4? Or are you an absolute madlad who thinks you’ll get that 3% chance of getting them all correct? Does this exercise sound stressful? Intruiging? Perhaps rather exciting? However it sounds is how you should feel about the 2025 Cal football season, because this team is five coin flips away from outcomes ranging from utter disaster to runaway success.
A couple weeks ago I asked you all to predict Cal’s chances of winning each game on the 2025 schedule. Over 100 of you participated, and today we tally the results.
2025 Season Predictions
The following table shows our average prediction of a win for each game and a standard deviation showing how much variability is in our predictions (larger SDs indicate more uncertainty in our predictions while smaller SDs indicate greater consensus).
Interestingly, we favor the Bears in 9 of the 12 games. Of course, being slightly favored is no guarantee of a win, so those 9 favorable predictions will not necessarily translate to wins (likewise, those 3 unfavorable games are not guaranteed losses).
I tend to think of these games as five classes of outcomes: chances under 15% are certain losses, 15%-39% are probable losses, 40-60% are toss-ups, 61-85% are likely wins, and 85+% are certain wins. According to that, we have 0 certain losses (yay for a weak ACC!), 2 probable losses (Louisville, SMU), 4 probable wins (OSU, SDSU, Virginia, LSJU), and 1 certain win (Texas Southern). That gives us 4-5 wins. And then we have our five, season-defining toss-ups: Minnesota, Boston College, Duke, North Carolina, and Virginia Tech. Better yet, these five all occur within a six-game stretch. After the Virginia Tech the next four games are in the “probable” range (Cal likely splits those and goes 2-2). So after that stretch of toss-ups we should know the fate of the season (and perhaps the fate of the Wilcox Era) following Cal’s Friday night game in Blacksburg, VA. But before we venture into where this season takes us, let’s have a look at each individual prediction.
Below I show the distribution of our predictions as boxplots. These reflect how much variation we have in our predictions. The box captures data between the 25th and 75th percentiles, and the horizontal line represents the median (or the midpoint of the data).
From here we can visualize our different classes of games ranging from the trivial win (Texas Southern) to the two probable losses (Louisville and SMU). Interestingly, several of those toss-up games have smaller interquartile ranges (the length of the box); that is, we tend to agree that we don’t know what to expect in those games.
Simulating the Season
To figure out how likely each outcome is for Cal, I took our predictions and simulated the season 1 million times. The simulation process was straightforward: I drew one Oregon State prediction at random and used that to generate an outcome (for example, if I drew a 70% then Cal would have a 70% chance of drawing a win and a 30% chance of drawing a loss) and then moved to Texas Southern, where another outcome would be drawn at random. And so on for all 12 games. Adding up all those wins and losses gets us our win-loss record, and then we repeat the process 999,999 more times. More interesting than the final win-loss outcome is how likely Cal is to be at each possible outcome after each game (1-0 vs 0-1 after OSU, 2-0 vs. 1-1 vs. 0-2 after Texas Southern, and so on).
The plot below indicates how likely each outcome is after each game. After Texas Southern there’s a 3% chance of being 0-2, a 36% chance of being 1-1, and a 62% chance of being 2-0. I’ve highlighted the most likely outcome after each game. Cal looks poised to start 2-0, lose a game, and then head into conference play with a 3-1 record. And then the fates of the coin flips decide the season. Over the next 8 games Cal is predicted to alternate wins and losses, culminating in a 7-5 record. But what if an extra one of those coins flips in our favor? Well, Cal has a 22% chance of finishing 8-4 and a 13% chance of going 9-3 (which would likely require winning all five coin-flips). If those coins start flipping unfavorably, Cal has a 19% chance of finishing 6-6, a 10% chance of going 5-7, and a 4% chance of ending up with a 4-8 record. The coin flips giveth, the coin flips taketh away.
Do you want to add an extra layer of uncertainty into these predictions? Consider the concept of correlated errors in our predictions. That is, we may be systematically underestimating or overestimating this Cal team. So once those coins start flipping in one direction, they may be more likely to continue flipping in that same direction over the remainder of the coin flips (that is, a loss portends future losses and a win portends future wins). The data and predictions do not account for correlated prediction errors, but it’s worth considering if you want an extra bit of uncertainty in your expectations for this season.
My main takeaway from all this? This season is probably going to spiral in one direction or another. And it’s worth examining whether you trust Wilcox to lead a downward or an upward spiral.
Awards
Before we wrap things up, it is our tradition to hand out some awards to our most optimistic, most pessimistic, and most reasonable predictions.
Sunshine Pumpers
First, we have the most lofty predictions of all.
Like last year, Calamo and LABear predict perfect seasons, followed by a bit of a gap to those who think all these coinflips will go in our favor (and then some).
Old Blues
At the other end of the spectrum are our most pessimistic predictions: those who think all those coin flips will go the wrong direction.
Amusingly, the expected number of wins if we lose all the coinflips is 4.41, so several of these are spot on for that worst-case scenario.
The Voice of Reason
Finally, we have those whose predictions for each game were closest to the community average.
Thanks to all who participated! In a few days we can start turning these predictions into real, on-the-field outcomes. Go Bears! And may our coins flip favorably this season…
Interesting that Old Blues are not far off from Voice of Reason, while Sunshine Pumpers are in their own alternate universe
I am a sunshine pumper before every first game of the season. This is my short term prediction. I am going to drive to Gold Beach, fish for salmon, and drive to Corvallis and watch Cal play football. I am going to catch three big fish and Cal is going to beat the Beavers. My buddy is going to pay me money for his team's loss. Go Bears!