August 30, 2018 - by Tom Federico
In this second post on football pool strategy, we’re going to dive into how the smartest football pool players use game theory to get a big edge in football pick’em contests.
In its most basic form, a football pick’em contest involves predicting the winners of group of football games each week. You get a fixed number of points (e.g. 1 point) for every correct pick you make. The most popular pick’em pools involve picking the winners of every NFL game, every week; some pick’ems incorporate college football games, either exclusively or with NFL.
In a confidence pool, in addition to making a pick for each game, you assign a number of “confidence points” to each pick. If there are 16 NFL games to pick in a given week, for instance, you would assign each of your picks a confidence value of 1 (lowest confidence) to 16 (highest confidence). If you get a pick right, you earn the number of confidence points you assigned to it.
Prizes in pick’em pools are typically awarded to the player(s) with the highest score(s) at the end of the season. Additionally, some pools offer a weekly prize to the highest scoring player of each week, which opens up a completely new dimension of strategy.
Other popular rule variants of pick’em and confidence pools include:
Last season, 81% of our subscriber survey respondents reported winning a weekly or season prize in a football pick’em contest, thanks to our Football Pick’em Pool Picks.
Casual NFL pools such as pick’em contests can offer fantastic profit opportunities to smart players. Here are three reasons why:
As a result, football pools can offer much higher expected returns than traditional sports betting. If you have what it takes to be a sharp pool player (spoiler alert: we can help), you should seek out and enter as many football pick’em pools as you reasonably can.
Whether you are a loyal participant in the Westgate SuperContest or just play in a small NFL pick’em on a site like RunYourPool every season, applying the concept of game theory to your picks is probably the #1 thing you can do to win a prize more often.
In short, sharp NFL pick’em players don’t just try to figure out the most likely team to win every game. In addition to seeking out accurate game predictions for every team, they put a lot of effort into gathering data that helps them project how their pool opponents are going to make their picks.
They look at pick popularity data published by big national pick’em contests, and they think about the potential biases of players. If a pool is sponsored by a bar in Kansas City, for example, you can bet that most players in it are probably going to overestimate the Chiefs.
Smart players then use this competitive intelligence to adjust their picks in a way that increases their expected profits from the pool. In some cases, they may even determine that making highly risky picks — picks that look crazy to most people — is well worth the expected reward.
Let’s drill down into a simplified example to show how important it is to anticipate how your opponents will make picks.
With Week 1 of the 2018 NFL season coming up, four buddies — Larry, Moe, Curly, and Shemp — have decided to compete in a pick’em pool. Here’s how it works:
The rules are simply to pick NFL game winners, with one point awarded per correct pick. The pool lasts just one week, though, and it only includes the following three games:
Based on Vegas lines, the win probabilities for those three games looked roughly like this:
|Game||Favorite||Win Odds||Underdog||Win Odds|
|Game||Larry's Picks||Moe's Picks||Curly's Picks||Shemp's Picks|
|1||Saints (80%)||Saints (80%)||Saints (80%)||Buccaneers (20%)|
|2||Patriots (70%)||Patriots (70%)||Texans (30%)||Texans (30%)|
|3||Rams (60%)||Raiders (40%)||Rams (60%)||Raiders (40%)|
Looking over these picks, notice that:
So who made the best picks?
Do the math, and in retrospect, Larry made the best picks.
By playing it safe in such a tiny pool, he let his opponents shoot themselves in the foot by picking too aggressively. Consequently, he’s given himself a 42% chance to win the pool, compared to baseline win odds of 25% (1-in-4).
Larry’s expected payout from the pool is therefore $168 — the $400 pot times a 42% chance to win — for an expected return of 68% on his $100 buy-in fee. Warren Buffett would be jealous.
Moe has a 28% chance to win, Curly has a 19% chance, and poor Shemp only has a 10% chance. At those odds, Moe still has a positive expected return of 13%, but Curly and Shemp are expected to lose money.
Let’s trace this oversimplified example back to game theory. In a small pick’em pool, where you only need to beat a few opponents, taking even moderate risks with upset picks can be too aggressive of a strategy, and decrease your odds to win. Picking all or almost all favorites may feel wimpy, but there are times when it completely makes sense.
There is one important contingency in the example above, however. Larry is clearly in the driver’s seat, but a big reason why is because he was the only player in the pool to pick all the favorites. What if another player like Moe had exactly duplicated Larry’s conservative picks?
As it turns out, Larry’s pick strategy is still the best call. Although his profit expectation would be diminished, Larry (and now Moe, with the same picks) would still have the highest expected profits from the pool, while Curly and Shemp would still be expected losers.
Translation: Picking a bunch of upsets in a small pool is usually the kiss of death.
Something fascinating and fun happens when you expand pool size, though.
Let’s imagine a pool with not four entries, but 400 entries — the Four Stooges and 396 of their closest idiot friends. Winner take all, same games to pick.
In pools that big, an expected pick distribution might look something like the following:
So who made the best picks for this 400-entry pool?
The Shemps, by a significant margin. In bigger pools, public bias toward favorites often presents great opportunities to increase your expected profits by aggressively picking unpopular, and typically quite risky, teams.
(This is especially true in pick’em contests with weekly prizes only. In those pools, who cares if your picks totally bomb in a given week? You either win the weekly prize or you don’t, and the slate is wiped clean the next week. You just need to have the guts to trust the numbers, and ignore the inevitable ridicule of your opponents when they mock your insane looking picks every week.)
By picking all underdogs, the Shemps still only have a 10% chance to end up with the highest score in the 400-person pool. But when that scenario happens, since there are only 10 Shemps in total, each of them has a 1-in-10 chance of winning the straw draw and taking home the pot. That means each Shemp has a 1% chance to win the pool — a 10% chance to get the highest score, times a 10% chance to then win the straw draw — and take home the $40,000 pot.
Correspondingly, there’s a 99% chance that each Shemp doesn’t win the pool, and thus loses his $100 entry fee. Combine those probabilities and outcomes (a 1% chance to win $40K, a 99% chance to lose $100), and in the long run, each Shemp is expected to get $301 back from the pool — an expected profit of more than 200% on his $100 entry fee.
By picking all favorites in a bigger pool, the Larrys will still get the highest score way more often than the Shemps do. But when that happens, each Larry still has to win a straw draw against 199 other Larrys in order to take home the pot. That’s a crushing blow to expected profits.
In the end, in a 400-person, 1-week pick’em pool where player picks are distributed as explained above, the Larrys and the Curlys are now both expected to lose money.
The Four Stooges case study illustrates the general strategy of applying game theory principles to NFL pick’em pools. Sharp pick’em players not only understand each team’s odds to win, but they also project how their opponents will pick, and balance the risk/reward profile of their own picks accordingly.
When it comes to maximizing their odds to win a football pool, the pros know when picking like a Larry is a better strategy than picking like a Shemp, and vice versa.
In fact, a pick’em pro might submit two very different looking pick sheets for two different pools, knowing that there is no such thing as a universal set of “best picks” that applies across all different types pick’em contests.
As it happens, after researching football pool strategy for over a decade, we engineered the only product that algorithmically optimizes your weekly NFL pick’em pool picks. Among other things, it factors in your pool size, projects your opponents’ picks, and adjusts for your pool’s scoring system.
In fact, it is so advanced yet simple to use, that it took us years to come up with an appropriately awesome and wildly creative name: Football Pick’em Picks.
2018 Football Pool Strategy Series
Printed from TeamRankings.com - © 2005-2019 Team Rankings, LLC. All Rights Reserved.