Dr. Bob’s Week 1 NFL Survivor Strategy
Dr Bob Sports is excited to announce that we will be starting a weekly NFL Survivor Pool blog series. Each week, we will break down a couple of key matchups with a tilt towards game theory and the meta of survivor, the game behind the game. We will show you how to optimize for the “expected equity” of your entry (in both single- and multi-entry pools), but at the end of the day it only matters if the team you pick wins. You’re playing the long game, until you aren’t.
Methodology: The Three Keys to Survivor
Every survivor pick must take (at least) three variables into account.
- Win Probability: Our proprietary model, which has been trained to predict game outcomes, has consistently outperformed the betting market since we began incorporating play-by-play data and advanced metrics in 2016. Our Best Bets have achieved a 57.5% win rate over that period.
- Future Value: We calculate each team’s future value based on its predicted win probability for all remaining weeks. This calculation is weighted more heavily in weeks where high-probability games are scarce. Considering Thanksgiving and Christmas weeks are split in some contests (and each only have 3-4 games to choose from), banking reliable teams for those condensed weeks is something to heavily consider.
- Pick Probability (Ownership Percentage): New to this season, we have trained a model on the past three years of pick data from the Circa Survivor contest to project how the field will pick. By feeding our model each team’s win probability, future value, and what percent of entries have that team available, we can accurately forecast how the field will pick.
Combining these three variables, we run a Monte Carlo simulation[1] to calculate the EE (Expected Equity) of every potential pick. This allows us to weigh the probability of victory against the value of picking contrarian. Expected Equity is calculated as “your share of the prize pool:” (your remaining entries) / (total remaining entries) * (total prize pool).
Scenario 1: Single-Entry League Strategy
In a single-entry pool, your goal is to maximize your chances of advancing while taking leverage –the expected selections of the other people in your pool –into account. Our analysis identifies two viable approaches for week 1:
The EE-optimal Selection: Denver Broncos
(Win Prob: 82%, Pick Prob: 34%)
In a 20,000 entry pool (Circa Survivor), a $1,000 entry on Denver carries an expected equity of $1,073. Even though the Broncos are projected to be far and away the most popular team selected, a bet on them still carries value. As noted on the Dr. Bob Sports NFL Free Analysis page, Denver possesses a documented early-season advantage at altitude.
Listeners of the Dr Bob Sports Podcast could point out that we discussed betting the Titans Season Total over 5.5 wins, and we do believe that Cam Ward and the Titans will exceed expectations this season. That being said, we don’t expect Tennessee to pull the upset in Week 1.
The Contrarian Play: Jacksonville Jaguars
(Win Prob: 66%, Pick Prob: 2%)
According to our model, only three teams show positive expected equity as single selections this week: the Denver Broncos, the Washington Commanders, and the Jacksonville Jaguars. Our numbers give the Jaguars just 66% win probability, but projected ownership sits below 2.5%, and that’s enough value to make Jacksonville a compelling contrarian play. Carolina’s offense remains limited and will struggle to score enough points to beat Jacksonville on the road. For larger or squarer pools, consider taking the Jags.
Scenario 2: Multiple-Entry League Strategy
Being able to place multiple entries can change the game theory around survivor picks. You can choose to maximize EE, or you can choose to take a more balanced approach that limits downside in the least likely scenarios. Picking Chalk can and will backfire (given a large enough sample), so balancing exposure with contrarian picks is something to consider.
For this section, we will take you through a strategy built for the Circa Survivor contest – 10 entries, $1,000 each and a pool size of ~20,000 entries.
The EE-optimal Selection: All-in on Denver
Pure expected equity optimization would allocate all 10 Circa Survivor entries to Denver Broncos. Even with expected 33% ownership, their 82% win probability means this approach generates the highest expected combined equity at approximately $10,700. However, this strategy carries risk of complete and total loss if Denver is upset.
Last season’s week 1 Bengals’ pick provides a cautionary example. If we had written this article last year, the EE-optimal play would have been on Cincinnati, who was upset by New England 16-10.
Portfolio Optimization Monte Carlo Analysis
Our Monte Carlo simulation models various game results and calculates equity distribution in all scenarios, including low-probability ones. If we optimize for equity in the bottom ~10% of outcomes (the scenarios where Denver loses), we can see an all-Denver selection is no longer optimal.
Our simulation identifies an optimal portfolio that protects against worst-case scenarios while maintaining strong expected returns:
- Denver Broncos: 4 entries
- Washington Commanders: 2 entries
- Arizona Cardinals: 2 entries
- Cincinnati Bengals: 1 entry
- Jacksonville Jaguars: 1 entry
This approach generates expected equity of $10,200 compared to $10,700 for the all-Denver portfolio. In the bottom 10% of outcomes, this portfolio still maintains average equity above $9,000, as opposed to $0.
You could say that any non EE-maximizing choices are sub-optimal, and you would have a point. But to that, I say that Dr. Bob Sports leads the industry by being right 57% of the time – which means that we are wrong 43% of the time. Long-time subscribers know as well as anyone else that there is no such thing as a “sure thing” in NFL Football.
See you next week.
Appendix:
Expected Team Pick Probability for Week 1 in Circa Survivor:
| Team | Predicted Pick % | Team | Predicted Pick % |
| DEN | 33.89% | MIA | 0.38% |
| WAS | 14.76% | ATL | 0.29% |
| PHI | 14.50% | BAL | 0.07% |
| ARI | 13.91% | DET | 0.07% |
| CIN | 6.13% | CHI | 0.06% |
| NE | 3.07% | SEA | 0.06% |
| JAX | 2.47% | LAC | 0.06% |
| PIT | 2.46% | TEN | 0.06% |
| LA | 1.56% | LV | 0.06% |
| SF | 1.06% | CAR | 0.06% |
| GB | 0.99% | NYJ | 0.04% |
| TB | 0.99% | HOU | 0.02% |
| KC | 0.80% | DAL | 0.02% |
| MIN | 0.79% | NYG | 0.02% |
| IND | 0.75% | CLE | 0.01% |
| BUF | 0.50% | NO | 0.01% |
[1] A Monte Carlo Simulation involves running a large number (typically 10-100k) of hypothetical season scenarios to calculate which team selection strategies perform the best across the full spectrum of potential futures.