Weekly vs. Season-Long Pools: How The Length Of Your Football Pick’em Impacts Strategy
Your strategy should be different depending on if you are in a weekly contest or one that goes all season.
Tre'Davious White demonstrates a different kind of pick popularity (Photo by Mark Alberti/Icon Sportswire)
We love value picks, where you get more “bang for your buck” with a pick due to the public undervaluing a team’s real chances of winning. The best way to determine value is to compare objective win odds, like those we discussed earlier, with public pick popularity.
At TeamRankings, we gather public pick data from a variety of sources to get a good sense of how people are likely to be picking games, allowing us to identify those values.
But how much weight or importance pick popularity should have compared to win odds depends on a lot of factors. Primary among them is the length of a contest, or to be more precise the number of games used over the course of that contest.
NFL Weekly and Season-Long Contests: Opposite Ends of the Spectrum
At TeamRankings, we offer pool advice for a variety of types of contests, ranging from NFL survivor pools to bowl pick’em pools to NCAA Tournament bracket pools, in addition to regular season football pick’em pools.
Each of these types of contests has different considerations, often driven by the nature of the pools. Even within a given type of contest there can be major rules variations that affect strategy. One big one in NFL pick’em contests is whether you are competing for a weekly prize, or are in a contest that goes all season with the prize determined at the end. These two varieties are at opposite ends of the spectrum when it comes to how pool length impacts strategy.
Here are some examples of the size/length of pool contests:
- 272 games = full NFL regular season contest that goes 18 weeks
- 90 games = “Pick 5” Pool for season-long NFL (like the Westgate SuperContest)
- 63 games = total number of picks required in a NCAA Tournament Pool
- 40-45 games = college football bowl pool
- 32 games = first round of NCAA Tournament pool
- 13-16 games = weekly NFL prize contest
What stands out?
A weekly NFL contest where you can earn a prize based on just that single week’s results has the fewest total games impacting the outcome. A season-long NFL contest where you pick every single game, on the other hand, has by far the most, compared to these other types of pools.
Let’s talk about why this matters.
Optimal Pool Strategy Requires Considering Both Risk and Reward
Optimal strategy in pool contests requires balancing both risk and value. To win, you must score some points that your opponents do not get. That’s the value part, where picking unpopular teams can give you an edge in your pool in situations where your opponents may be under-picking a particular team relative to their chances.
On the other hand, missing out on too many points that your opponents do get can be deadly to your chances. And therein comes the risk part of the evaluation. In order to find value, you have to make some unpopular picks. But if you get too risky in the process of doing that, you could be giving away points to your competition.
If you do more of the former (gain points on the field) than the latter (take zero points when others are winning), then you have a pretty good chance to cash out and finish near the top of the pool. But how do you decide the optimal mix of risk vs. reward? That’s where it gets tricky. How those scales tip depends on the length of the pool.
We are going to walk through some examples to illustrate how the number of opportunities impacts your value versus risk outlook.
Let’s say you have a pool where a single game will determine the outcome. Everyone who picks the winner correctly goes to a blind drawing to determine first place. In our game between the Chiefs and the 49ers, the Chiefs have 70% win odds, while they are being picked 80% of the time.
Looking at that, then, you can see that there is some “value” in picking the 49ers in this single-game hypothetical. With 100 entries in this pool, your chances of winning are 1.5% if you take the 49ers (30% chance of winning the game, divided by 20 entries from which the winner will be drawn). At the outset of the contest, you would have had a 1% chance (1 out of 100), so picking the underdog in that scenario would have improved your chances.
Thus, in this example where the result of one game decides the outcome, paying attention to pick popularity matters a lot. In fact, knowing the value proposition is the driving force behind most strategic decisions.
Now, let’s expand it to a hypothetical where there are 10 games, and every game features a team with 70% win odds and 80% pick popularity. Each individual underdog pick, then, provides value just like above, with the upset being picked less frequently than the odds it occurs.
Let’s say you are in a pool with 10 other entries, and each of them are picking exactly two upsets each, but distributing the upset picks equally across the options.
Here is an example of what the results for those ten entries could look like in an “average” week where the win odds match the results, so that three different upsets happened, each picked by two entries:
|1 entry got both upset picks right||9-1 record|
|4 entries got one upset pick right||7-3 record|
|5 entries got no upset picks right||5-5 record|
If you add those all up, you would get an average number of correct picks at 6.2 out of 10 for this group.
Here, picking every upset because it has some value would be a poor choice. Your average expectation is 3.0 wins. You could still win a week where things went truly crazy, but the odds of seven or more teams (out of 10) with 30% win odds winning in this week is just over 1%, for example, so it’s not the best strategy here.
On the other hand, picking every favorite gives you a better average win expectation (7.0 wins). But in a contest where these 10 games alone decide the winner, someone in the pool is likely to catch a good combination of upsets. Finishing better than half of the entries in the pool isn’t the goal when trying to win an individual week.
The best strategy lies somewhere between those two extremes.
Weekly Contests and Finding the Value Picks
In the previous hypothetical, every game had the same win odds and pick popularity. In real life, though, you will get plenty of variation, and opportunities will arise to make some value plays without going crazy and picking all upsets.
Here are 10 of the games from Week 2 of the 2019 NFL season. They average out to be very similar to our hypothetical 10-game exercise: 69% win odds and 81% pick popularity for the group. But as you can see, there was plenty of variation.
|Team||Opponent||Win Odds||Pick Popularity|
Identifying Value Gambles for Weekly Contests
When we look at the potential upset picks we might want to take, where we consciously pick teams with under 50% win odds, because they present a value opportunity, a few stand out from this group. We refer to these types of teams as “Value Gambles.” You are taking a risk, but it is a calculated one that presents value in a weekly contest, because of the pick rates.
The Detroit Lions (picked 12% of the time) and Indianapolis Colts (picked 13%) both had win odds over 40%, but were being picked infrequently. Simply hitting those two upsets would give you two wins when the average (if your pool resembled the public) would be 0.25 wins. Tampa Bay over Carolina was another one to consider, depending on whether it made sense to take on additional risk. Their win odds were lower (just under 30%) but Carolina was as popular as three other favorites on this list with far better win odds.
Identifying Value Favorites for Weekly Contests
Another group of teams we like to identify are “Value Favorites”. These are teams that are the favorite to win the game, but the public is picking them at a rate lower than their actual win odds. These are important picks in both weekly and season-long pick’em contests. That’s because you get to pick the favorite to win, while a large chunk of the public is going the other direction, providing the best of both the risk and value worlds.
Buffalo, favored over the Giants, but with more than half the public going the other way, was one example. The Steelers versus the Seahawks was another, and the Packers versus the Vikings nearly qualified as well.
Four favorites from this week had win odds between 57% and 60%. If you took the value choices of DET, IND, GB, and PIT, you would have expected only about 0.5% of all entries to match your picks for those four games. Take the opposite in all those games, and over 11% of the public was expected to pick that same combination.
Yet the combined win odds of those two groups was nearly identical (50.5% average for the undervalued teams, versus 49.5% for the overvalued teams). Picking either group was essentially flipping four coins. The difference was that with the first group, if the flips all went your way, you’d have gained a lead over 99%+ of your opponents.
That’s the key in a weekly contest — to find the best values, and balance your risk and value profile so that you can be in great shape to win a prize if your specific handful of key results hits.
Season-Long Contests and Minimizing Risk
Now think back to that 10-game hypothetical, where all other entries in your pool picked two upsets. There is a chance each week that someone will pick the right combination of upsets, and have a good week.
If your goal is to beat them just for that particular week, you need to adjust your risk versus value profile to take on some more educated risks, as we showed with the exercise about value favorites and value gambles.
If your goal, though, is to outlast them for the full season, then the better strategy is a conservative one that limits risk, and reduces the number of bad weeks compared to the rest of your pool.
Overall, the group averaged 6.2 wins, while simply sticking with the favorites would have given you 7.0 wins. While any one (or handful of entries) might have a great stretch of games, they have to repeat it again and again to prevail over the course of the full season using that same strategy. Your expected wins over time by playing a conservative strategy that relies on proper win odds evaluation will gain you the advantage.
More Games Means Results Tend Toward the Mean
As the number of games increases, the results tend to move toward the expected averages. You might be able to outperform the market in a small number of games, but it’s more likely to catch up to you over the course of a 272-game schedule.
This means that if you want to have the best chance to take down a season-long pool, you need to think more like the tortoise than the wild hare. Slow and steady wins the season, while volatile (often) wins the week.
To see how this concept has played out in real life, check out our article discussing what winning entries look like, which shows where entries that eventually won their pools were in the standings at various points in a season.