Two NFL upsets eliminated 1,604 entries from a $25 million contest in one day

On October 18, two NFL results transformed the position of entrants in the largest Survivor contest within a few hours.

In Circa Survivor, a player selects one NFL team each week. That team must win: a loss or tie immediately eliminates the entry. A team cannot be selected twice, and each entry costs $1,000.

The 2026 contest attracted 25,017 entries and a $25,017,000 prize pool. There were 2,908 entries still alive before Week 6. More than half selected just two teams.

Both lost.

How 1,604 entries disappeared in one day

The Rams were the week’s biggest favorites at about −12.5. Arizona nevertheless led for much of the second half and won 18–16 on Chad Ryland’s 43-yard field goal in the final seconds.

The Colts closed as approximately −6.5 favorites against a Tennessee Titans team that had started the season with five losses. Indianapolis led 13–0 at halftime, but Tennessee scored the next 20 points, including a touchdown after a blocked punt, and earned its first win of the season, 26–20.

There is a discrepancy in the published data worth noting: Covers listed the Colts at −11.5, while other closing-line sources show approximately −6.5. It does not affect the contest mechanics: the Colts only needed to win, not cover the spread.

Why the favorite is not always the best Survivor pick

In a conventional single bet, the task is straightforward: find a price that is better than your own estimate of the probability. A correct Survivor selection does not pay fixed odds. Its value also depends on every other entrant’s decisions.

The primary objective does not change: an entry must survive the current week and ultimately the entire contest. The selected team’s probability of winning therefore remains the most important factor. Passing on a clearly safer favorite solely to make an unpopular pick usually adds an unjustified risk of immediate elimination.

Popularity matters when several options offer comparable reliability. If two teams have similar chances of winning, but half the entries select one and only a small share select the other, the less popular team may offer a better combination of risk and potential position after the week.

When a popular favorite wins, a large group of competitors advances with you. When it loses, that group is eliminated and entries on other winning teams improve sharply. That is a possible reward for an alternative selection, not an independent reason to choose a weak team.

Survivor strategy therefore requires at least two probabilities:

1. the probability that your selected team wins;
2. the probability that competing entries select other teams and survive the week.

The most likely winner and the most valuable selection are not always the same, but moving away from the safest option should be justified by analysis, not merely by a desire to oppose the majority.

What pick popularity means

The share of entries selecting a particular team is commonly called ownership.

The Rams had 31% ownership and the Colts 24.2%. Together they accounted for 55.2% of all active entries. Anyone who selected another winning team did more than survive Week 6: they outlasted more than half of their competitors in a single day.

This effect is known as leverage, a strategic advantage relative to popular selections. It does not mean always choosing a rare underdog or deliberately sacrificing a strong chance to advance.

An unpopular team with a low probability of winning remains a poor selection. First, you must estimate the chance of surviving the week and the rest of the schedule. Popularity can then help choose between close alternatives or determine whether a small reduction in reliability is justified by the potential reduction in competitors.

Future upsets will gradually eliminate opponents before the late stages anyway. An entrant does not need to force that outcome immediately. Survival probability is the foundation of the strategy; pick popularity is an adjustment.

How the entry’s notional value changed

Before the week, simply dividing the prize pool by the 2,908 active entries gave:

$25,017,000 / 2,908 = approximately $8,603

After 1,604 entries were eliminated by the Rams and Colts losses alone, no more than 1,304 remained:

$25,017,000 / 1,304 = approximately $19,185

By this rough calculation, each surviving entry’s share more than doubled in a matter of hours.

However, $19,185 is neither a market price nor the entry’s exact expected value. The simple division assumes that every remaining entry has the same chance, which is not true.

The actual value depends on:

• teams already used;
• strong favorites available in future weeks;
• competitors’ selections;
• the number of entries controlled by one owner;
• the mandatory separate Thanksgiving and Christmas legs explained below;
• the possibility that several finalists split the pool.

Other entries may also have been eliminated during the week. The calculation shows the scale of the change but does not replace a complete valuation of the position.

Why strong teams need to be saved for later

Each team can be selected only once for each Circa Survivor entry. Using a strong favorite therefore has an opportunity cost.

The Rams may have appeared to be the best way through Week 6, but selecting them meant Los Angeles could not be used in a more difficult future week. A slightly less reliable team can sometimes be more valuable today if it preserves a powerful favorite for a week with few good options.

Circa has up to 20 separate selections: 18 regular weeks, a special Thanksgiving Eve/Day/Black Friday leg, and a separate Christmas leg. Each holiday leg requires an additional selection only from teams playing on those days: Thanksgiving offers 10 teams from five games, while Christmas offers eight teams from four games.

The one-use rule still applies. If an entrant has already used every team available in a holiday leg, they cannot make a valid selection and the entry is eliminated. A strong team may therefore be saved not for an abstract future benefit, but for a specific leg with a limited schedule.

Reaching the end is not simply a matter of picking 20 winners. Entrants must allocate a limited inventory of teams across the schedule. During the holiday periods they must also survive two selections in quick succession: the separate holiday leg and the next regular NFL week.

The optimal selection therefore depends on the entire future schedule, not only the next set of games.

Could the Rams and Colts have been avoided in advance?

After an upset, it is easy to call a popular favorite a trap. Before kickoff, the decision was much less obvious.

The Rams were favored by about 12.5 points. The market price implied a very high probability of winning. Rejecting them solely because they were popular meant knowingly accepting a greater elimination risk with another selection.

The Colts had a smaller advantage in the market but were at home against an 0–5 team and led 13–0. Even a sound pregame decision can lose: Survivor does not turn a probabilistic judgment into a guarantee.

The right question is not: “Could these losses have been predicted?”

It is: was the Rams’ or Colts’ probability of winning high enough to justify 31% or 24.2% ownership, given the future schedule?

The answer requires a model, not knowledge of the final score.

Why multiple entries change the strategy

Circa allows one participant to hold multiple entries. There is no need to repeat the same selection across all of them.

A player can:

• advance some entries with the most reliable favorite;
• use a different team on other entries to gain leverage;
• preserve different sets of teams for future weeks;
• reduce the risk of the entire portfolio being eliminated at once.

This begins to resemble portfolio management, although the outcomes still depend heavily on a small number of NFL games. Repeating the popular selection on every entry increases the potential share of the prize but does little to reduce the risk of losing them together.

The final prize, not merely one survived week

A central Survivor strategy error is optimizing every selection in isolation. The safest team maximizes the chance of reaching the next week, and without that there can be no prize. But the contest does not pay for one survived week, so the entire sequence of future decisions must be evaluated.

The objective is to maximize the probability of receiving a share of the final pool. That requires considering:

• current win probability;
• team popularity;
• the future value of unused teams;
• the possible number of remaining entries;
• the probability of sharing the prize with other winners.

These factors will often still lead to the obvious favorite. When the options are close, they may point to a less popular team with a slightly lower win probability. The process should become neither an automatic search for the largest spread nor an attempt to differ from the majority at any cost.

Conclusion

The Rams and Colts losses do not prove that selecting popular favorites is wrong. They demonstrate the cost of correlated decisions: 1,604 entries independently converged on two choices and were eliminated on the same day.

Entrants who selected other winners received more than another week of life. Their notional share of the $25 million contest more than doubled while the number of competitors fell sharply.

That is why Survivor is not an ordinary sequence of winner predictions. The first priority is maintaining a strong probability of completing the whole contest, followed by planning around the future schedule and the behavior of other entrants. Team popularity may adjust a decision between close alternatives, but it cannot replace the central requirement: the selected team must win.

Sources: Circa Survivor’s official page and rules, Circa contest and prize-pool information, Covers’ report on Week 6, and VegasInsider’s weekly results and closing lines.