NFL Survivor Picks

Circa Survivor goes to week 16 with all 10 surviving week 15.

GaryA took a huge hit when his Chiefs pick for week 17a went from 60% to 40% against Broncos as Mahomes is done for the season. Chiefs may still be his best play as he does not have Commanders available and Vikings project to be 37%.

ACCORSI, KICK YOUR KNEES UP and REAL BRO all have Texans available this week. They also have Cowboys available for 17a, projecting at 58% against Commanders.

of the remaining 7, all have Eagles (72%) available this week except DYLAN W, who can go with either 49ers (70%) or Saints (66%).

BRAD LEE and JUICY KEWCHI may steer away from the likely popular Eagles pick with either the 49er or Saints as well.

C3 PICKZ has the Bills available (81%), a solid contrarian play.

Projected selections:

Eagles: 4
Texans: 3
49ers: 2
Bills: 1

I got REAL BRO the current favorite with Texans this week. He saved Jags for week 18, and he has a reasonable counter to Patriots in week 17b with Bengals (69%).

JUICY KEWCHI looks a solid 2nd choice. He has Cowboys in 17a and Patriots in 17b, with the option of saving Eagles in 18. If he can get by this week he should become the favorite.

Circa Survivor Week 16 selections are in:

Saints moved up in value during the week to about 71% win equity against the rock-bottom Jets, making them a contrarian counter to Eagles for Caveman6470. (DYLAN W already played Eagles last week) But Eagles winning earlier today took that away.

Odds of all contestants surviving this week is 35%, with 49ers the weak link.

REAL BRO should have little problem getting thru with Texans @ 87%, but JUICY KEWCHI is already looking forward to 17a and is right on his tail. Both will be using Cowboys 17a, but JUICY KEWCHI has Patriots at 80% vs Jets in 17b.

Tua's benching has made Buccaneers a 75% play in 17b, and they will likely be the most popular pick as every contestant except REAL BRO has Buccaneers available. REAL BRO will likely go with Bengals at 70% as a contrarian play in 17b to counter Buccaneers and Patriots.

One way or another week 17a will see multiple contestants go down. BRAD LEE, C3 PICKZ, DYLAN W, GaryA, and PARTZ1 will all need to pick a winning dog in 17a if they get thru this week. While ACCORSI, Caveman 6470, JUICY KEWCHI, KICK YOUR KNEES UP, and REAL BRO all have Cowboys at 63% in 17a.

The final week promises to be very interesting with teams sitting players, other teams tanking for draft slots, etc. Eagles may be close to worthless, while Jags might be over 80%. Other possible picks include Bengals, Falcons, and Buccaneers, but a lot depends on who sits and playoff implications.

I got JUICY KEWCHI the current favorite. If Texans win REAL BRO is close, and both will be about equal after this week. KICK YOUR KNEES UP is right behind the top 2 and also sets up well if he gets thru with Texans this week.

All contestants survived week 16 and have moved on to the wall that is week 17b.

Some big line moves early in the week now have Cowboys 74%, Commanders 26%, and Vikings 25%.

ACCORSI, Caveman 6470, JUICY KEWCHI, KICK YOUR KNEES UP, and REAL BRO should all pick Cowboys, with a huge EV advantage over the other five contestants.

BRAD LEE, C3 PICKZ, DYLAN W, and PARTZ1 should all pick Commanders, with GaryA stuck with Vikings.

Below are the current expected values, ignoring future weeks' team availability:

Cowboys $2.65M
Commanders: $1.15M
Vikings: $819K

If one of the Commanders picks switched to Vikings, it would cost that contestant about $400k in EV, as noted below:

Cowboys: $2.57M
Commanders: $1.46M
Vikings: $740K

GaryA being stuck with Vikings, the best he can hope for is Vikings win, with either 5 or 6 surviving to week 17b. The 4 Commanders contestants would move on if Commanders win, with either 4 or 5 surviving to 17b.

For Commanders/Vikings to be GTO neutral Vikings would need to be around 38% win equity. If Vikings win equity were > 38%, then one (and only one) contestant should switch to Vikings. If < 38%, then all 4 should stick with Commanders.

Looking at it using Commanders win equity, GTO neutral is Commanders 16% win equity. That is, one of the 4 contestants should switch to Vikings if Commanders win equity were < 16%.

Thanks to Housenuts who pointed out the best EV picks in SE NFL thread.

Vikings need to be about 10-12% better win equity (non-linear) than Commanders for one of the 4 Commander contestants to switch to Vikings.

I have no doubt you are correct, but Washington is really imo undersold on just how bad they actually are.

You may be correct, and if that's your view you should be hammering the Cowboys ml/spread in size.

He could scoop with a MIN win and DAL/WSH tie. That would be the nuts.

Yeah, I have been a bit of a slow pony on this. I thought about the tie after I posted the calcs last night and I was going to note it this morning.

I use binary results in my calcs which models a lot easier than tertiary results which is obv the more precise way of doing it.

I got it at 0.13%. Certainly possible, and it adds about $23k to GaryA's EV.

Hedging Cowboys makes sense for Commanders contestants. If they could have gotten -240 earlier in the week even better. The problem for them is hedging with the Cowboys is very expensive (-350). Cowboys contestants are in much better situation to hedge, as their EV with a win is roughly the same, but the Commanders hedge price is about 90% cheaper (+300).

Ties can also be hedged, but in general that prop is terribly priced (50:1 in most shops).

Hedging 5% of unrealized EV gain, Commanders contestants should hedge about $320k laying Cowboys -350. That leaves about $3.4M on the table as potential realized gain, with $100k paid out if Commanders lose.

Cowboys contestants hedging 50% on unrealized gain requires a relatively small bet of about $200k. That leaves about $2.9M on the table as potential realized gain, with a nice payout of $600k if Commanders win.

GaryA's situation is similar to Commanders contestants, i.e. Lions hedge is very expensive, and EV with a win is roughly the same.

Hedging on unrealized gain is relatively straightforward and leaves room for future hedging, but is somewhat subjective and only an approximation. Optimal hedge size is determined by each contestant's financial situation (i.e. bankroll), tax implications, hedge price, future hedging costs, and their chances of winning $ prize going forward with their available plays.

There's 3 weeks of picks left.

Hedging gets pricey.

Picks go 5-4-1 so if you hedge cowboys or commanders, and your team advances, yes that's nice and your equity has doubled, but you lost your hedge bet and there's still 4-6 ppl left in the contest. Then you hedge again 17b and 18. Dollars add up.

But yes, to each their own and can certainly realize some amount of $ at this point if so inclined.

That's why I showed hedging against potential unrealized gain. It is simple and avoids most of what your example illustrates, but it certainly is not the optimal hedge.

There is far too much $ at stake to ignore hedging at this point. Notably at week 17a where everybody's unrealized gain is ~ $1.5M if they win. Or, their EV goes to zero if they lose.

I agree that the dog players Thursday are up against it as their hedging costs are 10x the Cowboys players.

If any of the 10 think to just let it ride and wait for the big payoff that's fine, but from an EG perspective that would be a huge mistake.

It is similar to a 15 game parlay ticket holder where the first 13 hit, and the potential payoff is far above that person's bankroll. I am aware conventional wisdom advises to not hedge, instead scolding the bettor that he/she should have simply played a 13 game parlay if they intended to hedge. But that is a false dichotomy.

Of course hedging with +EV bets is a lot better than -EV bets. But unless the hedge bet is awful (like 50:1 for tie) a good hedging strategy to increase EG is the way to go.

And in this case EG is appropriate for bankrolls in the 6 figure range. One with a 7+ figure bankroll who does not want to waste money on -EV bets makes sense because that person's EG won't go up much, if at all.

Insane picks

FWIW I have not yet seen the official Circa X selections post.

Current Pinnacle no vig win chances (money lines):

Cowboys: 77.2%
Commanders: 22.8%
Lions: 75.4%
Vikings: 24.6%

Revised EV calcs using current win chances (if picks went as assumed above):
Cowboys: $2.77M
Commanders: $1.01M
Vikings: $801k

Now, using the selections Housenuts posted:
Commanders: $3.43M
Cowboys: $2.57M
Vikings: $604k

C3 PICKZ has .228*.754 = 17.2% chance of scooping $18,718,000 TODAY. What a gift, now he can hedge about 3x what he could assuming the picks went as expected.

Fair to say these guys don't read 2p2.

I posted it above from their Director of Operations. And CircaSports posted shortly after

Thinking seems to be a deal was made by all Wash/Min people.

This distribution minimizes DAL ev.

Yeah, I was thinking something similar.

If all 5 went Vikings then Cowboys EV goes down a bit to $2.53M, so maybe C3 PICKZ did not sign on?

btw those EV calcs above are way off for Commanders, s/b 1.83x. Using scoop chance alone they are 1.72x.

This distribution is best combined ev for wsh+min and minimizes DAL to lowest it can be.

vs

I see what they did, they got Cowboys 79.6% win chance. Using that figure reduces Commanders EV to $3.06M, or 1.63x. Still not quite right, but they are closer than I thought.

Using their win probs:

All 5 go Vikings:

Cowboys: $2.89M
Vikings: $847k

4 go Vikings:

Commanders: $3.06M
Cowboys: $2.65M
Vikings: $601k

3 go Vikings:

Cowboys: $2.70M
Commanders: $1.62M
Vikings: $652k

And they might be factoring in ties (which I don't), which could explain the difference. That would push up dog pick EV, and lower fave pick EV.

Thanks for that Bully.

If there were an agreed split of the five dog contestants, Commanders winning likely yields a 5 way split of $18.7M regardless of Lions/Vikings. But ties are possible, about 0.5% chance each game.

So if they agreed to a split, it only makes sense for them to agree to split a hedge with Cowboys, and a share of ~$300k. I'm guessing they are kicking themselves for not hedging more, but wtv.

Hindsight is 20-20, but they should have taken care of this on Monday when Cowboys were about 70% win chance and -260 ML instead of -400 (or higher) today.

The proper amount to hedge depends on their bankroll, and their initial investment (I'm assuming $50k total for the five). House credit must be paid back, so that does not add to their bankroll.

With $16M bankroll, ties 0.5%, Cowboys -260/70%, and Lions 65%, max EG hedge on Cowboys -260 is $11.7M. Any bankroll over that does not yield a larger hedge bet. with $5M bankroll the Cowboys hedge is very close to all their funds, $4.94M. With $2M bankroll the hedge is $1.98M.

Of course I'm guessing the contestants did not want to risk $2M or $5M with a chance of TWO ties, 0.0025% probability. Their other bad outcome is a Cowboys tie and Lions win, which amounts to them losing their initial investment as the Cowboys hedge pushes.

I did not calculate a Lions hedge, which makes sense knowing the Cowboys tie. I'm guessing the Lions hedge would be a similar amount, perhaps a bit less because it would only take one tie to crush their bankroll.

Week 17b selections are in: