Dr. Bob’s Week 6 NFL Survivor Strategy
Welcome back to Dr. Bob’s Weekly Survivor Strategy blog. Last week was a bloodbath, with the top two most selected teams (LA Rams and Arizona Cardinals) both losing, and with that more than 60% of the Circa Survivor field was eliminated. We got caught in the carnage, with our first EE-favorite of the year loss on the LA Rams and our first Contrarian Pick loss of the year on the Seahawks – both of which lost on last-minute game-winning drives. This brings our overall record on suggested picks to 8-2 (4-1 on EE-Favorite, 4-1 on Contrarian). Those two picks accounted for 3 of our remaining 7 entries (2 on the Rams, and 1 on the Seahawks), and we lost two additional entries (one on Arizona and one on Kansas City), bringing us down to just 2 entries out of our starting 10. Since a large portion of the field was eliminated as well, those two entries still account for over 80% of our original $10,000 equity.
After the Los Angeles Rams lost on Thursday, there was significant discourse on Twitter and other platforms about sources of variance (ie. chance of an upset) in the NFL, and the benefits of staying away from variance in survivor contests. There are three major topics cites as leading to a higher chance of an upset: 1) road favorites (betting against home underdogs), 2) early week games (due to lower prep time), and 3) divisional games (due to team familiarity). The Rams fell into two of these categories (early week game and divisional matchup). Let’s check if the public discourse is based on fact or bias.
For this sample, we will only consider favorites of 6-10 points, since most survivor picks fall into that spread range.
Let’s start with road favorites. Since 2010, favorites of 6-10 playing at home (avg spread: -7.6) win 75.8% of the time, and favorites of 6-10 playing on the road (avg spread: -7.3) win 74.5% of the time. It may seem that there is a ~1% higher chance of an upset against road favorites until we consider the difference in average spread for those scenarios (1 point of spread is worth ~3.5% win probability on average; the difference of 0.3 points of spread (-7.6 for home versus -7.3 for away) is worth the ~1% win probability difference that we observe). We can conclude there is likely no difference between the win probability of a 6–10-point home favorite and a 6–10-point road favorite.
Next, let’s look at Thursday night games. Since 2010, favorites of 6-10 points playing Thursday night games (avg spread: -7.5) win 74.1% of the time, and favorites of 6-10 playing non-Thursday games (avg spread: -7.5) win 75.5% of the time. This suggests there may be a slightly higher chance of upset on Thursdays – but the effect is not statistically significant due to low sample size (only 81 games for the criteria for Thursday night).
Finally, for divisional games: since 2010, a 6–10-point favorite in non-divisional games (avg spread: -7.5) wins 73.7% of the time, while in divisional games the 6–10-point favorite (avg spread: -7.6) wins 78.0% of the time. The 3-4% increase in win probability for divisional games conflicts directly with the narrative that the underdogs are more likely to upset their divisional rivals.
In conclusion, when considering survivor matchups, given a similar spread, there likely is no difference between being a home or road favorite, there may be a small increase in variance when playing early in the week (although the sample size is very small), and there may also be a small decrease in variance (against popular belief) when playing a divisional game. All three effects are small and likely insignificant, and I think the narratives surrounding them are a product of confirmation bias: people tend to notice upsets that happen on Thursdays, or on the road, or against divisional rivals – and they tend to ignore the games that get “as planned”.
These findings make sense intuitively: if there was a statistically significant difference between expectation and the true outcome, the market would identify and account for it. For the betting public, the best prediction for who will win a game (and with what probability) has been, and will continue to be, the closing moneyline. If you want an edge up, you can subscribe to our picks here; we have a history of consistently beating the market for over 35 years.
On to our picks for the week:
Our EE-Optimal Selection: The Philadelphia Eagles (@NYG)
(Win Prob: 82%, Pick Prob: 9%)
We expect a large portion of the field to be riding with Green Bay this week (~40%), and that, combined with our expectation of people staying away from the Thursday night game after last week’s LA Rams loss, means that there is good value this week on the Eagles. Last week, we covered the uncertainty surrounding Jaxson Dart and the New York Giants, and so far, we have been unimpressed by his production, especially from the pocket. In this issue, we covered different sources of variance because the Eagles fit all 3 categories (Thursday night, divisional game, on the road) – which we don’t believe have a significant effect on the outcome – and we think much of the public will avoid this game. The Eagles have a good statistical matchup across multiple categories versus the Giants, and we think Philly will win convincingly on Thursday.
The Contrarian Play: The Los Angeles Chargers (@MIA)
(Win Prob: 70%, Pick Prob: 3%)
The Chargers got rolled last week versus the Commanders, but we expect them to bounce back this week on the road against the Dolphins. We don’t think that the market has a bad read on either of these teams, but the Chargers do have a few matchup advantages that make them a compelling choice. You can read more about these matchup advantages on our NFL Free Analysis page. The public focus this week will likely be on other games, and that leaves value of the table for those willing to forego the “safer” options.
The 90% Optimal Portfolio
For details on our selection strategy for our 90% Optimal Portfolio, see our Week 1 Blog Post and our Week 4 Blog Post.
| Selection | Team |
| 1 | PHI |
| 2 | GB |
| 3 | LAC |
| 4 | DEN |
| 5 | LAR |
| 6 | GB |
| 7 | PHI |
| 8 | PIT |
| 9 | LV |
| 10 | WAS |
Since we have 2 entries remaining, this will be our portfolio:
| Team | Num |
| PHI | 1 |
| LAC | 1 |
Appendix:
Expected Team Pick Probability for Week 5 in Circa Survivor:
| Team | Predicted Pick % | Team | Predicted Pick % |
| GB | 37.6% | BAL | 0.2% |
| LA | 15.1% | JAX | 0.2% |
| PHI | 8.6% | CAR | 0.2% |
| PIT | 7.0% | NO | 0.2% |
| LV | 7.3% | TEN | 0.2% |
| DEN | 6.3% | ATL | 0.2% |
| IND | 5.8% | CLE | 0.2% |
| DAL | 2.6% | NYJ | 0.2% |
| LAC | 2.6% | NYG | 0.2% |
| WAS | 2.4% | KC | 0.2% |
| NE | 1.6% | CHI | 0.1% |
| BUF | 0.2% | MIA | 0.1% |
| DET | 0.3% | SF | 0.1% |
| TB | 0.2% | CIN | 0.0% |
| ARI | 0.2% | HOU (BYE) | 0.0% |
| SEA | 0.2% | MIN (BYE) | 0.0% |
| NO PICK | 0.001% |