With Justin Wilcox and friends at the helm, many of us are optimistic about the future of Cal football. While the medium-to-long-term future may hold a return to consistent respectability, the Cal fanbase is less certain about what to expect in the immediate future. Will we sustain high-quality QB play? Will the defense turn things around quickly? Will our incredible streak of injury woes finally come to an end?
To get a better sense of how we’re feeling about the upcoming season, I asked you all to give us the Bears’ chances of winning each game of the 2017 season. Today we pore over the results and try to extract some wisdom. Let’s start with the table below, where I have listed the average win chance for each game and the standard deviation (or uncertainty) for each game.
Opponent | Avg. Win Chance | Standard Deviation |
---|---|---|
at North Carolina | 44.3% | 19.6 |
Weber State | 93.0% | 10.4 |
Ole Miss | 46.8% | 18.4 |
USC | 26.0% | 20.0 |
at Oregon | 41.2% | 19.2 |
at UW | 24.3% | 18.5 |
Wazzu | 51.2% | 17.2 |
Arizona | 65.3% | 15.2 |
at Colorado | 43.2% | 16.1 |
Oregon State | 66.8% | 16.2 |
at LSJU | 41.6% | 27.5 |
at UCLA | 44.5% | 19.7 |
Total | 5.88 wins | 1.68 |
Our projected win total is a reasonable 5.88 wins, but the standard deviation for our total wins is 1.68. We have not seen a standard deviation that high since the 2014 season, when we were coming off an absolutely dreadful 1-11 campaign. SDs are usually around 1.25 for these season predictions. That elevated standard deviation suggests that we’re fairly uncertain about what to expect from the team this year. Win totals were all over the place: we saw plenty of win totals of 4, 5, 6, 7, and even several 8s. Some are optimistic, some are pessimistic. Uncertainty abounds.
One feature of the results that immediately stands out is that we favor the Bears in only four games: Weber State, WSU, Arizona, and OSU. Achieving 6 wins will require some upsets. However, every game except USC or UW is a toss-up or better. With 5 games where the Bears’ win chances are in the 40s, an upset or three is not unlikely. On the other hand, those wins against WSU (51%), Arizona (65%), and Oregon State (67%) are hardly guaranteed. Achieving bowl eligibility will require a little luck, but a streak of bad luck could keep us to only three or four wins.
To help better visualize the results, I have created plots of our win chances below. Let’s start with the home games. The x-axis lists our win probability and the y-axis provides an estimate of the number of people who made that prediction. The plots are a bit lumpy because your predictions are a bit lumpy—we tend to post predictions that are divisible by 5 (50%, 65%, etc.).
The Weber State game is obviously the most likely win. The Oregon State and Arizona games are pretty close, although the Oregon State game is looking slightly easier. Avenging last year’s utter disaster in Corvallis would be quite pleasant. Ole Miss and Washington State are pretty close to toss-ups. I’m surprised we weren’t more optimistic about the Ole Miss game. They lost most of their starters from a team that stumbled to a 5-7 record. A loss to them would substantially decrease our chances of getting to a bowl game (and perhaps temper our expectations for the rest of the year). At the toughest end of the spectrum is the USC game. I pray to Oski we do not suffer a 14th consecutive loss to them. 14. Ugh.
Next up are the surprisingly uniform-looking road games.
Colorado, Oregon, UCLA, and North Carolina all look like toss-ups. Splitting those four games would put us in a great position to earn a bowl berth. The Big Game and UW game look like more daunting matchups. I am thankful that year after year, there is always a segment of the participants who cannot abide by anything but a 100% chance of beating the Lobsterbacks.
Simulating the Season
In addition to tallying up the results from your predictions, I decided to simulate the season to see how likely we are to achieve each win total. The simulation process is pretty straightforward: I start by selecting one prediction at random for North Carolina and use that to predict the winner. If I draw an .45, then the Bears will have an 45% chance of drawing a win. Based on that percentage, I then draw either a win or a loss for the Bears, move on to the Weber State game, and repeat the process for each of the 12 games. At the end, I tally up the number of wins and start the process over again. I ran this process 1,000,000 times. My computer is still mad at me about that.
I plotted the results below, with win totals on the x-axis and the chance of attaining that number of wins on the y-axis. I even colored the bars according to the official Cal Athletics color guidelines (R4 G30 B66). That’s dedication.
6 wins is the most likely outcome, followed by 5, then 7. The following table illustrates the win probabilities exactly.
Wins | Probability |
---|---|
0 | 0.00% |
1 | 0.16% |
2 | 1.30% |
3 | 5.23% |
4 | 12.91% |
5 | 21.30% |
6 | 24.10% |
7 | 19.19% |
8 | 10.61% |
9 | 4.03% |
10 | 1.00% |
11 | 0.14% |
12 | 0.01% |
5% chance of winning 9 or more games? Yes please! If you’re still not optimistic about the upcoming season, perhaps this will lift your spirits: according to our predictions, we’re more likely to beat USC than we are to finish with a worse record than we had last season. The future may be uncertain, but it’s unlikely we’ll take a step back under the first year of the Wilcox era.
Finally, we have our usual array of awards to hand out. We’re in a new coaching regime, so we’re retiring the Sonny Delight and Sonny Yikes awards. I’ve decided that instead of sticking with fixed names for the awards, we’ll rotate through award names this time, although they’ll all have a common (and fairly obvious) theme. Our first set of awards recognizes those with the most optimistic predictions for the season.
Justincredible!
Name | Wins |
---|---|
1. Goffense | 12.00 |
1. Panoramic Hill Sues For The Win! | 12.00 |
1. Vol Crush | 12.00 |
4. alpha1906 | 11.88 |
5. DavidShawIsMyDogsCloggedAnalGland | 11.07 |
6. OCBear1983 | 10.21 |
7. Old Bear 718 | 10.20 |
8. Oski Disciple | 10.00 |
9. 1031michael | 9.42 |
10. shelby03 | 8.58 |
Three-way tie for first! Goffense, Panoramic Hill, and Vol Crush all predict nothing but roses this season. alpha1906 predicted a much more modest 99% chance of Cal victory in each game. In fifth place is my favorite regular contributor to these predictions and someone whose creativity is outmatched only by his/her dislike of David Shaw.
Next up we have those who are most pessimistic about the upcoming season.
Justinfuriating
Name | Wins |
---|---|
1. Bear | 0.00 |
2. DRIVETIME | 1.00 |
3. POOHBEARS | 2.70 |
4. calsuckscock69 | 2.80 |
5. texashaterforlife | 3.00 |
5. kchin | 3.00 |
7. Uthaithani | 3.17 |
8. rollonyoubears111 | 3.30 |
9. puresilence | 3.44 |
10. Calbears03 | 3.47 |
1 win? Well, that would be on par with our last coach’s inaugural season. Fortunately Wilcox inherited a program in better shape than the one Dykes inherited (although hiring Buh did him no favors whatsoever). Even though these are among the most pessimistic predictions, I can envision how the Bears could win only three games this season: QB struggles, no depth on defense, a tough schedule, a longer-than-expected transition to the new coaching system. I wouldn’t be shocked if Cal finishes with only 3 wins, and I picked the team to win slightly more than 6 games! Again, uncertainty abounds.
Finally, we have old reliable, both in nomenclature and in category. This final set highlights those whose predictions were closest to the community average.
The Voice of Reason
Name | Standard Deviation |
---|---|
1. joltimpact | .038 |
2. 0.4 | .046 |
3. 1988goldenbear | .055 |
4. Young_bear | .056 |
5. pierrezod | .057 |
6. JDub12 | .061 |
7. Oski Manifesto | .064 |
8. Thepro | .068 |
9. Umbertok | .069 |
10. GoBears!!! | .071 |
joltimpact leads the way followed by 0.4 and 1988goldenbear, who regularly finishes among the most reasonable during our weekly report card series. Sadly, we still have three months to go until that series returns and, more importantly, until we can watch college football again. Before that we’ll have one more round of these season predictions, probably about halfway through fall camp. Hopefully we have a QB by then! If not, we’ll likely continue to see plenty of uncertainty about the upcoming season.