Week 4 Survivor Final Update: It’s Down To The Wire Again

posted in NFL, NFL Survivor Pools

It’s late on Friday and I know people want to know our final verdict for the week, so let’s get right to it. Guess what? We’ve got a photo finish for the third week in a row.

First, let’s take a look at the updated data table. As a reminder, this helps us answer the three main questions we ask ourselves about each team when making our Survivor pick.

1. How likely are they to win? (Vegas Line at -110 payout odds & TR Odds from our NFL win picks page)

2. How popular is this team? (Average public Pick % from sites like Yahoo! and OfficeFootballPool)

3. Should I save this team for later? (Future Value: the number of future games where the team is expected to have win odds of 75%+ from our NFL Survivor Tool, modified by some manual fudges for Week 17 and special-case weeks. Near Value: a modified version of Future Value that assesses near term value and is more relevant to players in very small pools)

Week 4 NFL Survivor Decision Factors (Updated)

Teams are listed in order of how attractive we think they are as a choice this week. They’re also separated into rough tiers. If two teams are in the same tier, you may want to choose among them based on which pros and cons are more important to your particular situation.

TeamOpponentLineTR OddsPick %Future ValNear Val
Tier 1: Top Options
Houstonvs Tennessee-12.081%13.0%5.51.2
Green Bayvs New Orleans-7.577%3.0%4.01.3
Denvervs Oakland-7.069%3.7%2.80.7
Tier 2: Might Be Worth A Look
Baltimorevs Cleveland-12.081%54.8%42.1
San Franciscoat NY Jets-3.561%1.2%1.50.8
Atlantavs Carolina-7.068%8.6%73.8
Arizonavs Miami-5.566%9.5%3.52.2

Teams We Already Picked: Houston Texans (WIN), Cincinnati Bengals (WIN), Dallas Cowboys (WIN)

What Has Changed Since Wednesday?

Since we posted our preliminary Survivor article a couple days ago, like last week, there have been changes in information that impact our analysis.

First off, we have continued to make minor refinements to our “Near Term Value” calculation, adjusting the weights to put even more importance on the next six weeks. It’s doubtful we’re going to see many pools that are already cut down to a small size carrying on past Week 10 or so.

Secondly, as we usually do, we spent more time analyzing and adjusting Future Value for teams, and charting likely paths through the next few weeks. As it turns out, those exercises revealed a key insight that could impact your best pick for this week, in most situations.

Before we get to that, though, let’s see how our top considerations and other popular teams have fared over since Wednesday:

1. Houston has become a slightly more popular pick, but still not even close to as popular as Baltimore. With no major changes, the Texans still rate as the best pick of the week.

2. Green Bay has become slightly more popular and slightly less safe. The Packers’ money line is down to -320 on Pinnacle and upon further study, our models have tended to overestimate win odds for teams like Green Bay in similar situations. Still, the Packers are still the safest pick this week after Baltimore (duh) and Houston.

Perhaps more interestingly, prompted by an insightful discussion with fellow Survivor analysis ninja Vegas Watch, we took a much deeper look at the Packers’ future value. We’ll get into the details below, but our initial assessment on Wednesday likely underestimated Green Bay’s future value by at least a bit. But the Packers remain a top consideration.

3. Denver has experienced a favorable line move of half a point, and is now favored by a touchdown over Oakland. The Broncos are also -305 on the money line, which is close to as safe as the Packers. In addition, another revelation from the Green Bay investigation was that (again, details below) we may have overestimated Denver’s future value by at least a bit. All told, Denver took a solid step up in the pick ratings.

4. Baltimore beat Cleveland on Thursday night, in a game that was closer than the Vegas spread implied, with the Browns having a couple shots either to force OT or to win the game in the closing seconds of the fourth quarter. Man, it feels weird not to have the huge favorite that everybody was picking get eliminated!

5. San Francisco has experienced an unfavorable line move of half a point and are only -197 on the money line. Their TR model odds have also dropped to the point where we would no longer consider them at all for Week 4, even though their future value looks fairly limited. The risk is just too great now, so the Niners are officially dropped to AVOID status.

6. Atlanta is actually a decent pick this week in isolation, but given the alternatives, at -7 and -310 on the money line (less safe than the Packers), and with tons of future value, the Falcons are an AVOID except for some potential special cases. If you have several entries in your pool(s) and want to diversify your risk, Atlanta might be worth considering. If you have only one entry, it’s a pass.

7. Arizona has experienced an unfavorable line move of half a point (spread down to -5.5) and are -241 on the money line, which is risky. We also realized our initial future value estimate for the Cardinals was too optimistic. The Cardinals have a couple of favorable matchups in the next few weeks, and they are relatively popular this week, so picking Arizona is a no-no.

What It All Comes Down To: Who Dat?

We’ve used the Texans, and except for recommending the Ravens in pools where only you and one other person remained, we passed on Baltimore due to it’s overwhelming popularity this week. And now that we know the Ravens won, it’s definitely not time to take any crazy risks.

So it’s down to Green Bay and Denver, and man, it’s close. For both smaller and larger pools, our numbers still give the Packers the ever-so-slight edge — but there’s a catch, and it’s a pretty major one. We’re not sure how much we should trust our ratings of New Orleans at this point.

To make a long story short, the ratings currently used in our Survivor Predictor tool to project future win probabilities are largely based on current season performance. In short, here’s what they know about the Saints:

  • They were a good team at the end of last year
  • They have been a horrible team so far this year

Despite giving the Saints some hang-over credit for last year, these ratings still see New Orleans as a very bad team right now. Not slightly bad. Like, bottom of the barrel bad. And the Saints have been really bad…for three games.

However, if you look at Vegas futures odds and other types of ratings that factor in preseason expectations, the prognosis for New Orleans right now isn’t nearly as bleak. In that world, the Saints still project as a decent (but not elite) team that has just gotten off to a horrible, fluky start. And with a healthy Drew Brees at QB, it is indeed a bit difficult to believe the Saints will end up being terrible this year.

The outcome of this thought process could have a major impact on your pick this week. In Week 8, for example, the Packers (at home) play the Jaguars while the Broncos (also at home) play the Saints. According to our predictive ratings that focus on current season results, those two games will be relatively similar in terms of win odds. That projection implies that the Saints will be about as good a team as the Jaguars when Week 8 rolls around.

If you believe this will be the case — either because you do believe the Saints will be awful this year, and/or you think the Jags are indeed better and will continue to improve — then Denver should be a great pick in Week 8. And by that logic, you would probably double down on that opinion by picking the Packers against New Orleans this week.

If you think that the Packers over the Jaguars sounds a lot better than the Broncos over the Saints, then Denver looks like a better pick this week, especially if you think Week 8 could be critical in your pool.

There are other factors at play here, but the more we looked into the numbers and did sensitivity analysis, the more it became clear that the prognosis for the Saints was by far the most important decision factor this week.

Official Week 4 NFL Survivor Picks: Green Bay Packers over New Orleans Saints

We’re going to stick with the Packers as our official pick. In midsize to larger pools, for which our official pick is geared, our weekly picks tend to be driven by value and relative safety, vs. picking the team most likely to survive. We do think our ratings are underestimating the Saints, but even if the Packers end up being the most likely winner of Week 8 by a decent amount, there appear to be a few promising potential alternatives there from a value standpoint.

In addition, the Broncos have a favorable mid-range matchup, home vs. Tampa Bay in Week 13. The Packers, on the other hand, after Week 8, may not play another home game against a cupcake until Week 16, and a bunch of pools may not get that far. Based on what we know now, the Packers may also not turn out to be the dominant team that most people expected this year.

This pick is a risk, though, in that the Packers are likely at least a slightly better team than the Broncos, and we’re burning them this week for not much of an increase in safety. The biggest concern, especially for smaller pools, is that we pick Green Bay now, and by the time Week 8 rolls around, there are only one or a few other people left in your pool, and Green Bay is the safest pick that week by a significant margin. And you don’t have them, but your opponents do. You’d be at a disadvantage.

If you think overall safety over the next 3-4 weeks could be key to winning your pool, it’s worth going Broncos this week and sacrificing a wee bit of safety to preserve the Pack.

As always, feel free to ask questions about your specific situation in the comments section.

  • ZenBuddhi

    I have Houston Denver and Green Bay available in my pool. There are 30 people left who is my best option for this week?

  • http://www.teamrankings.com TeamRankings.com

    Assuming typical picking percentages, Houston.

  • Mason

    I posted in the earlier Week 4 article (I’m playing in a pool with 5 people left, including myself. Of the 5 people left, 3 people have used Houston, including myself. Knowing this info, who would you guys pick, Green Bay or Denver?

  • http://www.teamrankings.com TeamRankings.com

    My guess is the other 4 all go BAL or HOU, so it’s sort of a wash, especially if you’ve got a feeling one way or the other. Week 8 may be important for you, though, so I’d probably lean Denver, unless you think the Saints will really suck this year.

  • Mason

    Thanks Tom! Yeah, that’s pretty much what I was thinking, and I went with Denver. I have no idea about New Orleans, but you’d figure they’d show at least some signs of life this year . . .

  • Surviving

    I’m in a very large pool where 18.5% of the participants are left. I have 2 entries. In one I have used Hou, NYG and Chi, and the other I’ve used Det, Hou, Dal. My question involves the fact that this pool goes into the playoffs if a winner is not decided in the regular season. If a player in the pool does not have anybody left to pick, he is eliminated that week. Does a rule like that make it more favorable for me to pick Denver this week?

  • http://www.teamrankings.com TeamRankings.com

    Our playoff odds for GB and DEN aren’t too far off one another right now, and neither of them look like a lock. I don’t know how big your pool is — if it’s gigantic like thousands and thousands, you need to take some big risks to win, so I’d double up on which ever team you like better. Otherwise, if you split DEN/GB you’ll cut your chances of getting eliminated totally from ~25% to ~6%, which is nice insurance.

  • Tommy (NY)

    Amazing analysis as usual Tom!!! I am in a large pool that started with 5,000 and is down to 768 people. Amazingly I still have my original 13 picks thanks a large part to what you guys have been doing on this site. My question is how do I play this week with all my picks. I can still take Houston, Packers and Denver just not sure how I should divide. Then a small part of me is saying to put all my picks on Houston which seems foolish. Please help if you can. 367/768 already took the Ravens on Thursday night.

  • http://www.teamrankings.com TeamRankings.com

    13 entries averages out to about 1 per 60, so I don’t think you should be playing super aggressively anymore. Putting all your picks on Houston is the riskiest thing you can do from a downside perspective. You maximize the number of entries that are expected to survive, but you also give yourself a 20% chance of losing all your entries this week.

    If you divide two ways between HOU and GB/DEN, you’ve got about a 5% chance of losing all your entries — way less.

    If you divide between HOU, GB, and DEN, you’ve only got a little more than a 1% chance to lose all your entries, but you are also the least likely (about a 45% chance) to have all your entries survive. So if you do this, you would expect to lose some entries this week.

    One factor here is that we know BAL won, so the ideal outcome of the week is now gone. So I would play it more conservative, I think, and split entries between HOU and GB/DEN (personal preference on the second part). Might be tempted to throw one entry on the third team as extreme worst case insurance just for kicks.

  • Matt

    Tom…this is truly great. Absolutely worth waiting Friday for!

    My question is I am in one of those double-elimination pools. Only 24 out of 55 entrants are left and of those 24 only 8 have no strikes. I am the only person to have 2 entries in those 8 with no strikes and I also have 1 entry with 1 strike.

    So far I’ve picked HOU/NYG/DAL/BAL, CHI/CIN/NO/BAL (1 strike with NO), and the pick I need to make this week is with a HOU/CIN/DAL/? entry.

    In this league, with 8 players still with no strikes, it has always come down to at least past week 10 depending upon how much people mess up the next few weeks.

    Keeping that in mind, does Denver become a better pick because I think I will be picking into the final weeks? Or should I just stick with Green Bay?

    Thanks in advance!

  • http://www.teamrankings.com TeamRankings.com

    Not sure it really matters. With two entries already through and the strikes thing, my guess is by Week 8, there will still be around 15 people left in the pool, and we may find that there is more value on avoiding GB then if they are really popular. Still, with two entries already through and the potential for the pool to go long, I would probably lean toward burning the team with less future value (DEN) this week.

  • MonkeyBoy

    My pool of 44 had a buy-back clause for the first 3 weeks, and even though I was one of only 4 remaining last week, there are now 28 people still in the pool. Of those, 11 chose BAL this week. I have GB and HOU available. Of the remaining 17, 9 others also have HOU remaining, and all have GB. Seems like I should be taking GB this week instead of HOU, since I would think most people who have HOU will use them. What do you think?

  • http://www.teamrankings.com TeamRankings.com

    That sucks, they should have let you buy a second entry too!

    If 7 take Houston, that’s 25%, and that’s still not enough to NOT make HOU the best pick according to our numbers, especially given your pool is pretty small and you only have one entry. With BAL already through as well, I’d play it safer and go HOU.

  • MonkeyBoy

    OK, sticking with HOU. Thanks Tom!

  • http://twitter.com/swamiluv Russ Aragon

    thanks guys for the great work, 4200 start in my pool 635 now, I am torn between Pack & broncos, you are saying Pack for the bigger pools correct?

  • http://www.teamrankings.com TeamRankings.com

    Our numbers say yes for such a big pool, but it’s darn close. If you had a strong feeling either way, I’d say just follow it.

  • LPG

    I am in an (Illinois) pool that started with about 65 and is down to 11. No one picked Baltimore. I have gone with your official pick each week (Thank you). It is a traditional pool – one loss and you are out. Apart from knowing that no one picked Baltimore, I have no idea who the other survivors have chosen in past weeks or who they are picking this week. Do you think GB is still the best choice? I am not sure if others will pick GB given our proximity to Wisconsin. Thank you.

  • http://www.teamrankings.com TeamRankings.com

    Tough call here. If even one other person picks Green Bay in a pool as small as yours, Baltimore becomes the best pick, followed by Denver. So if you think there is a good chance of some regional love for GB, Denver would probably be a better bet.

  • Lolo

    Have 40 entries out of 110 left. I have 2 entries have used Houston in one will use them in my other entry. 20 used Baltimore. Half have used Atlanta. Next most popular is cards then packers. Would u go packers or broncos? Only 1 pick for them so far.

  • http://www.teamrankings.com TeamRankings.com

    ATL looks great in Week 6, so that bodes well for you, assuming the Falcons pick up even more of your opponent picks this week. If Packers have been relatively popular already, with that Week 8 matchup on the horizon in your fairly small pool, I’d go Broncos.

  • NYJoe

    I have 2 picks in a pool that has about 1700 left (680 or so took the Ravens). Pick 1 I have used: Texans, Giants and Bears – Pick 2 I have used: Bears, Bengals, Cowboys. I was thinking of using Packers for Pick 1 and Texans for Pick 2. Or do you think I should save the Texans and use Denver for Pick 2. Curious of your thoughts. Thanks, you guys are great!

  • http://www.teamrankings.com TeamRankings.com

    Wow — that’s only 40% picking the Ravens in a really big pool, which is unexpected. I know you don’t want to hear this now, but in that case, your best play was Texans/Ravens.

    With Baltimore already through, I think you still want to use the Texans for safety this week, and choose between GB/DEN for the second pick.

  • NYJoe

    I’m in the same boat, must be the same pool

  • http://www.teamrankings.com TeamRankings.com

    Ha, funny. Honestly, any quick analysis I do isn’t going to be all that great for a situation like yours with the double and triple picks. But at a high level, I think it means that the probability of having to use a team in a not-optimal situation after Week 10 (e.g., playing on the road) because of the multiple-picks thing goes up significantly. And in that case, having better teams in your pocket means a lot.

    GB is probably a better team than Denver, so if I was choosing between the two, I think I’d go Denver.

    Next question is do you play Houston or not. That’s a huge pool and holding on to HOU would have tremendous value. If you are splitting your picks across two teams, you won’t gain that much additional insurance against losing both your entries. So I would save them.

    In a pool this big, I’d be considering GB/DEN or DEN/DEN. Second option is risky, and without more time to analyze, I’d probably go with the first.

  • Ian

    Oakland always plays divisional games tough, and usually win, no matter how bad they are. I would avoid this game at all costs.

  • http://www.teamrankings.com TeamRankings.com

    The problem with that logic is, there’s a certain probability Oakland has to beat Denver. Everybody in the world has access to the information about how OAK has done in divisional games recently, and the betting markets have settled at certain numbers, which imply certain win probabilities.

    So should we trust the global betting markets, which represent millions of real dollars being bet on this game, or should we trust that you are better than the markets at processing public information relevant to this matchup?

    Can OAK beat Denver? Sure. And if you’ve got a bad feeling, by all means avoid it. That’s what makes sports sports.

    Not that one game means anything, btw, but OAK has played one divisional game so far this year, against SD, at home, and they lost by 8 points as 1-point favorites.

  • rcc22

    ok…here is the situation…i have 2 picks…

    just over 100 left in the pool…about half have picked the Ravens.
    far another 20 have houston…funny..nobody i believe except one other
    has Green bay…but still 40 or so havent put there picks in…
    Im thinking Green Bay/Houston? Denver?
    any thoughts?

  • http://www.teamrankings.com TeamRankings.com

    In a “typical” pool, only 2-3 would pick GB, so that’s about what you’d expect. I’d split HOU and GB/Den, take your pick on the latter.

  • rcc22

    Thank u for the insight…looks like Houston is the 2nd highest chosen in the pool…followed by texans, cards,falcons, and than packers…

    note…Cards ? would they be a better option than the Pack?

  • rcc22

    thanks…looks like u got me to through week 4…

  • rcc22

    any last minute udates for today?
    Is Green Bay still ur pick of choice?

    Much appreciated..

  • JO’C

    80 out of 514 left. Only 30 went with Bal Thursday night. Around 90% of the pool had their picks in before the game so I assumed (incorrectly :)) that Bal would be more popular then they turned out being. Having missed out on that game I’m going to hold my breath and go with the Pack. I also feel NO is a better team than they have showed and this game scares me but Oakland @ Denver scares me even more. McFadden has had some big games vs the Broncos the last 2 years and this is a ‘rivalry’ game. Good luck today. Something tells me we’ll need it.

  • JimmyV

    You survived! Congrats, but you could have saved some deep heart pounding had you taken Denver :))

  • Sunnie

    Nine left. 4 took Baltimore, 3 are taking Houston, one Atlanta, and me. I have GB penciled in, but I can change up to kick off (of whomever I take.) So, do I stay GB or go Denver? No diff to me. thx! (I’ve already taken Houston.)

  • http://twitter.com/swamiluv Russ Aragon

    Pack by the hair of my chinny chin chin…ty guys!!!

  • http://twitter.com/swamiluv Russ Aragon

    We are starting a new pool for week 5, so you still do the #1 pick even if they have been used. That will help with the new pool.

  • JimmyV

    You guys provide some great analysis, Thank you…I do think you are way off on future value though…How does SF have a FV of 1.5 and Arizona more than double that. SF has 3 home games ( Buff, Sea, STL) where they will be favored between 7-12 points. I cannot say the same for Arizona (even though they are not bad). Id much ratger have SF in my pocket but acording to FV its better to have Arizone. Can you explain? Thanks

  • http://twitter.com/swamiluv Russ Aragon

    I can’t see the comments.

  • http://www.teamrankings.com/ David Hess

    Yeah, Disqus was having technical difficulties earlier. They say everything should be fixed going forward.

  • http://www.facebook.com/daniel.silberfarb Daniel Silberfarb

    I sense a 49ers pick coming on…

  • http://www.teamrankings.com/ David Hess

    Yeah, a quick 60-second analysis leads me to believe you are correct. Highest win odds, and we’ve already used 2nd & 4th highest (GB/HOU). 3rd highest is the most popular team this week (NYG). Looks like this may be a no-brainer. Of course, I’ll dig deeper tomorrow.

  • Chris

    Still 20% of people taking them though, and near future value seems high. I’m liking the looks of a NWE pick. 5th best win odds, and less than 1% of public on them. Only Hou (taken), GB (taken), and NYG/SF (top 2 public picks – account for over 55% combined) have better win odds.

  • http://www.facebook.com/daniel.silberfarb Daniel Silberfarb

    The dropoff on win odds for NE is steep though. 11% on this site, and advancednflstats.com actually has Denver higher on their efficiency rankings than NE. I’ll pass.

    I don’t like wasting future value with SF, but I think there are other options in weeks 7 & 10. Hopefully the pool will be over by 14.

  • http://www.teamrankings.com/ David Hess

    Yes, Daniel is right — there is a pretty large penalty in current week win odds for picking New England.

  • Chris

    hmmmm…I guess I’ll wait for the official post, but at this point I’d probably rather run with the sheep and take NYG rather than SF.

    My only hesitations are:
    1) I lost with NYG on week 5 last year, and
    2) NYG always seems to play to the level of their opponents and tend to be worse when favored by a TD+. That being I don’t know much about football at all, so should just rely on the hopefully efficient odds market rather than my own presuppositions.

  • rcc22

    Mr. Hess is BACK?
    ooohhh….I actually like the replacement picker…lol


  • http://www.teamrankings.com/ David Hess

    Yeah, Tom definitely has more flair than I do.