TeamRankings.com
Get Smart About Sports
September 4, 2014 - by David Hess
Note: This post is taken from our NFL Survivor Pools product, which includes our picks, analysis, and tools for survivor/eliminator style contests. We’re posting it on the blog to provide an example of the depth of Survivor pool analysis we do. If you like what you see, we invite you to check out the full product.
A quick review of our picking methodology is in order since it’s Week 1. Our goal is to maximize your expected winnings in your survivor pool. We assume your #1 goal is winning your pool — not just surviving for a certain number of weeks so you can tell your friends you made it that long.
Maximizing your odds to win your pool most often requires a slightly riskier picking strategy than most of your opponents will use. In most weeks, we’re looking to avoid the most popular public pick in favor of a less picked team, and rooting for an upset to take out a large percentage of your pool competitors.
Put another way, even though we may not recommend the safest pick, we still only advocate taking calculated risks that increase your expected value. To make sure of that, our analysis incorporates three key factors:
As for the underlying data we use:
The Report Card page is the heart and soul of our weekly Survivor analysis. You should check that page first every time you use this product. That one table contains all of the data that informs all of our weekly advice, and has for years, including during our 17-0 undefeated season in 2011.
These weekly write ups are essentially interpreting the Report Card for you. We understand a lot of people like to read words more than interpret data tables, so we’re happy to do them. However, there’s a good chance that things change throughout the course of a week, as win odds and public picking trends shift. This writeup is based on the information available to us right now (Wednesday afternoon), but the Report Card updates multiple times per day with the latest info.
Consequently, we’ll try to note any major post-publication changes with updates at the top of these writeups, but we can’t stress it enough: If you want to base your pick(s) on the most up to date information, come back and check the Report Card on the day of your pool deadline to get the latest info.
With that said, here is how things are looking as of Wednesday afternoon, along with a glossary of terms for the stats we quote on each team.
Week 1 of 2014 is going to test our mettle right off the bat. We’ve got an overwhelmingly popular pick, Philadelphia, that is worth avoiding unless you’re playing multiple entries, and a couple top options that will probably make a few of you gulp.
Detroit Lions (vs. New York Giants)
TR Elimination Risk: 25% (#3) — Money Line: -242 (#7)
Popularity: 3% — Future Value: 0.6 to 1.1
The Lions are currently projected with the third highest win odds of the week according to our models (75%) but are being picked by only 3% of the public, and they don’t have a ton of future value (best future options: Week 5 vs BUF, Weeks 13/14/15 vs CHI/TB/MIN). By our numbers, they’re the best pick this week for all pool sizes. However, it’s worth noting that their money line is only the 7th highest of the week, so this is a pick that relies somewhat on trusting our models, which are high on Detroit. If you go strictly by the betting markets, the Lions are still a solid pick, just not the very best.
New York Jets (vs. Oakland Raiders)
TR Elimination Risk: 31% (#6) — Money Line: -252 (#6)
Popularity: 5% — Future Value: 0
The Jets are a bit riskier than the Lions according to our models, though the betting markets put them as basically even with Detroit. New York is also a touch more popular. However, this appears to be the Jets’ easiest game of the season, unless the Raiders end up much better than everyone expects, and they have essentially no future value after Week 1. According to our models, the Jets are the second best pick. Ignoring our models and looking only at the Vegas money line, the Jets jump into the top spot.
Chicago Bears (vs. Buffalo Bills)
TR Elimination Risk: 27% (#4) — Money Line: -300 (#3)
Popularity: 18% — Future Value: 0
The Bears have little to no future value (Week 11 vs MIN and Week 12 vs TB seem the only games that might end up as easy as this one), and are a relatively low risk pick, especially according to Vegas. However, their popularity hurts them compared to the Lions and Jets. In short, the Bears are modestly less risky, but 3-6 times as popular. That hurts the most in large pools, where avoiding popular teams and taking smart risks is most important.
Philadelphia Eagles (vs. Jacksonville Jaguars)
TR Elimination Risk: 19% (#1) — Money Line: -510 (#1)
Popularity: 43% — Future Value: 0
The Eagles are clearly the lowest risk pick of the week. Based on what we know now, this is also their easiest game of the season, and our models don’t give them much future value either. However, being picked by over 40% of the public is a big, big negative. As a result of their popularity, they have only the 9th highest Immediate EV (a measure of their value as a Survivor pick just for Week 1) in large pools. However, in very small pools, where your pick(s) can actually account for a decent chunk of the overall pool picking percentages, the Eagles look like a reasonable choice.
Denver Broncos (vs. Indianapolis Colts)
TR Elimination Risk: 22% (#2) — Money Line: -330 (#2)
Popularity: 3% — Future Value: 5.9 to 7.5
The Broncos are a low risk pick, and they’re unpopular, but they have tons of future value. Right now we project five future games where they’ll have higher win odds than this one, and several other games where they might still be a legitimate pick option. So it’s generally going to be better to save them than to use them this week.
Arizona Cardinals (vs. San Diego Chargers)
TR Elimination Risk: 34% (#9) — Money Line: -160 (#12)
Popularity: 1% — Future Value: 1.1 to 1.3
The Cardinals are highly unpopular, and don’t have a ton of future value, which means they are worth a look. But they are also a significantly higher risk pick than any of the teams above.
St. Louis Rams (vs. Minnesota Vikings)
TR Elimination Risk: 38% (#12) — Money Line: -179 (#11)
Popularity: 1% — Future Value: 0.3 to 0.6
The Rams are even riskier than the Cardinals according to our models (though not Vegas), but that is balanced out somewhat by their lower future value.
New England Patriots (at Miami Dolphins)
TR Elimination Risk: 30% (#5) — Money Line: -218 (#8)
Popularity: 3% — Future Value: 3.6 to 5.1
Although unpopular, the Patriots aren’t any safer or less popular than our top options, the Lions and Jets. However, the Patriots have much more future value, which means you’ll almost certainly want to save them.
Pittsburgh Steelers (vs. Cleveland Browns)
TR Elimination Risk: 35% (#10) — Money Line: -285 (#4)
Popularity: 9% — Future Value: 0.2 to 0.5
The Steelers are relatively low risk according to the betting markets, but our models are more pessimistic about their chances against the Browns. But even ignoring our models and going by the betting markets, the Steelers are a worse option than the Jets this week due to their higher popularity and higher future value.
New Orleans Saints (at Atlanta Falcons)
TR Elimination Risk: 36% (#11) — Money Line: -149 (#13)
Popularity: 1% — Future Value: 1.9 to 2.5
The Saints are unpopular, but they are a high risk pick, and have significant future value. There’s almost certainly no reason to take on that level of risk this week.
Seattle Seahawks (vs. Green Bay Packers)
TR Elimination Risk: 33% (#7) — Money Line: -260 (#5)
Popularity: 2% — Future Value: 4.0 to 5.1
The Seahawks are one of the most valuable teams to save, so they would have to be much less risky (and yet still somehow unpopular) in order to be a good option this week. They are not.
This week, going by our projection models, the Detroit Lions are the best pick if you want to maximize your odds to win your 2014 survivor pool. According to our models they have the third lowest elimination risk behind Philadelphia (too popular) and Denver (too much future value), but are not popular (3%) and have a little future value, but not a ton.
Ignoring our models and looking solely at the betting markets to determine elimination risk, the New York Jets are the best pick. They’re only the 6th-safest team of the week according to their latest money line, but safer teams are more popular (Philadelphia, Chicago, Pittsburgh) or have too much future value (Denver, Seattle). Pittsburgh would be a close second.
In small pools, playing it “safe” (FYI, nothing is ever safe in the NFL) and picking Philadelphia isn’t a terrible option, but it’s not the optimal pick either. Just think this scenario through in your mind: If you pick the Lions or Jets this week based on our advice, and they lose, and the Eagles win, will you still be able to hold your head high knowing that it was a good risk to take? There is a decent chance (about 15%, actually) that the Lions or Jets win this week, and the Eagles lose. And if that happens, you’ll be at a huge advantage in your pool. It’s far from a guarantee, but the risk/reward tradeoff is worth it.
Your maximum expected value decision in Survivor is rarely going to be the safest pick that everyone else is picking too. That’s also a great segue, because if you’re really serious about winning something in a survivor pool and reducing your risk of early elimination, then you should be…
When submitting multiple entries, one point to keep in mind is that very popular teams that you’d want to avoid in single-entry strategy can be become acceptable picks, if your pick “portfolio” still features a team at a lower rate than the overall pool.
For example, as of Wednesday afternoon Philadelphia is being picked by 43% of the public. So if you have five entries in a pool, picking Philadelphia with one of them isn’t a terrible option, as they’ll represent 20% of your picks, but 43% of your opponents’ picks. A Philly upset would still hurt your opponents more than you.
The reason you might want to do this is that spreading your portfolio out too thin, and taking on too many risky picks, is a quick way to end up no longer having a portfolio. The goal of multiple entries is to insulate you from losing all your entries in one week, via only one or two upsets. However, if you spread your picks too widely, you will almost certainly lose one or a couple entries each week, and gradually bleed your way out.
So from a risk/reward perspective, you need to strike a balance between those two extremes (picking only one team with all five of your entries, and picking five different teams).
Our advice this week for multiple entry strategies is:
For each of those, weight the earlier teams on the list a bit more than the later teams. For example, a 7 entry strategy might be 3 DET, 2 NYJ, 2 PHI.
If your pool has a tiebreaker based on the combined number wins of all the teams you pick (i.e., the lowest total wins the pool), then all else being equal you’ll want to pick teams with poor overall season outlooks. We’ll use win totals from our season projections to help you do that.
Here are our projected wins for the teams mentioned in the main write up this week:
Given that the Jets were nearly as good of a pick as the Lions in standard pools (and possibly the best pick if you ignore our models and just look at betting market lines), their lower win projection is enough to bump them ahead in pools will total season wins tiebreakers.
Also, if you were for some reason considering taking Denver, New England, or Seattle still … those high win projections should change your mind.
If your pool extends into the playoffs, that can affect optimal pick strategy. All else being equal, you’ll want to avoid teams that stand a good chance of being favored in a playoff game. A good way to get an idea of whether that applies to a given team is to use the odds from our season projections.
If a team has a high chance of winning their division, then they are likely to host a playoff game. And because of home field advantage, it also stands to reason they might be the favorite in that playoff game. As we get later in the season and begin to figure out exactly playoff matchups that rule of thumb may not work so well, but for now it’s good enough.
You’ll also want to avoid teams that have a high chance to make the playoffs (having to pick an underdog in the playoffs is better than having no team left to pick), and especially avoid teams with the highest chances to win the Super Bowl.
Here are the approximate division win odds (Div), Wild Card odds (WC), and Super Bowl win odds (SB) for the main options this week:
As with the season wins tiebreaker, dropping from the Lions to the Jets is probably worth it, in case Detroit does come through on that roughly 1 in 4 chance to win their division.
Also, Philly’s reasonable chance at winning their division is another reason to avoid them, beyond their high popularity. And of course stay away from Denver, New England, New Orleans, and Seattle, who are are likely to make the playoffs.
Strategy can get quite fun with multiple-strike pools, where you’re not eliminated with your first loss. In Week 1, though, the best strategy for strike pools is not too much different than for traditional pools. The main difference is that your opponents will likely make riskier picks than in a typical pool, since they know they won’t be eliminated with a loss.
However, many strike pools use the number of strikes as a tiebreaker, so in large pools (hundreds of people) where somebody is likely to survive the year without a loss, a single strike can essentially mean you lose the pool. So we do not recommend getting all crazy just because you have that strike in your back pocket.
Later in the season, when you or your opponents have some strikes, then strategy can change more drastically. But in Week 1, we would still suggest picking the Lions or Jets. Those two teams already have low popularity and low future value, so there’s no real benefit to dropping to a riskier pick.
If you can pick a team as many times as you want (meaning you could pick the Seahawks every week if you so desired), then you don’t need to consider future value at all. In order to maximize your long term expected winnings, you should simply pick the team with the best balance of win odds and popularity in your pool. In TR terminology, that is the team with the highest “Immediate EV” (expected value), which balances win odds against popularity, but ignores future value.
Because of the different rules in your pool, our public pick data probably won’t be representative of what people are actually picking. So the best course of action is to make some guesses as to which teams will be popular, put those into the EV Calculator, and see which teams have the highest Immediate EV.
Philadelphia is by far the biggest favorite of the week, so they ought to be the most popular. After the Eagles the next highest money lines belong to Denver, Chicago, and Pittsburgh, then a roughly equal tier of Seattle, NT Jets, and Detroit.
So, using the EV Calculator, we plugged in 45% on PHI, 20% on DEN, 15% on CHI, 10% on PIT, 5% on SEA, and 2% on DET/NYJ. Here’s the resulting ranking for a 200 person pool, starting with the best pick (highest Immediate EV):
This depends pretty heavily on our public pick estimates being correct, and on the size of your pool. So if you can make better estimates for your specific pool, we’d recommend using the EV Calculator yourself to come up with a pick ranking.
If you have to make multiple picks per week at some point in the season, then future value is more important to you. That means sliding teams with low future value up the rankings, and teams with high future value down. The rearranged order this week would look something like:
If you’re in a Loser Pool, where your rules are just like a standard Survivor pool, except you have to pick a loser instead of a winner, the basic analysis works the same. You’ll want to strike a good balance between low current week win odds, low popularity, and low future value (though here future value means having difficult games in the future, not easy ones).
Each week we’ll do a very quick run down of the teams with the lowest win odds, and try to find one that we think will be unpopular, and doesn’t have much future value. We don’t have good public sources for pick% in loser pools, so we have to go with our gut here.
Moving from lowest win odds to highest:
Mirroring our regular Survivor advice, picking the Giants to lose at the Lions looks like the best Loser pick this week. The Colts have a higher chance of losing, but will probably be super popular, and the Jaguars have many tougher games left, so we’d suggest saving them.
As mentioned earlier, the weekly Report Card, which is the basis all this analysis, will continue to update throughout the week and into the weekend. As betting lines move, our predictions for the games may move as well, and we’ll continue to update the public pick info to reflect the latest numbers. Those changes can have an effect on the ordering of the picks, and on which teams look like the best options this week.
We’ll post occasional updates at the top of this post if something major happens (e.g. if Peyton Manning gets injured, for example), but more for the most part you should use the Report Card as your source for our latest and best advice. If you’re playing a single entry, we suggest you check the Report Card right before you submit your picks, and take the top ranked team available to you at that time.
Some users find it frustrating that the Top Options on the Report Card may be different than the Top Options earlier in the week. But if you think about it, it’s the only way to go. Why would you want to make a key decision based on stale information? Things do change during a week’s time; news breaks, reliable predictors like betting odds shift. So you want to take advantage of the latest information before your weekly pick deadline.
Finally, our Q&A Forum is available each week to paid subscribers for any questions you may have after reading this write up. (It’s free to everyone in Week 1, so you can get a taste of the interaction that goes on.)
Just one courtesy note here: If you ask questions that are already answered here in the writeup, we will need to refer you back to the writeup. We get a lot of questions each week, so we have to be efficient in answering them.
Our Q&A Forum response policy is as follows (it’s also noted at the top of each forum page):
If you’ve gotten this far, your next step is to check our our full NFL Survivor Pools product.
Printed from TeamRankings.com - © 2005-2024 Team Rankings, LLC. All Rights Reserved.