The difficulty people have with expected value - in particular for NFL Survivor pools

[StackinGreen]

This could be a thread on EV in its own right, but obviously would be too large. I wanted to ask you guys what you think about the idea of expected value for making selections in NFL survivor pools. I've encountered people over the years that can't seem to grasp that proper selections have everything to do with the guess you make about public selections in your pool, or the proxy that you can get from survivor pools that publish who is on which team (yahoo, runyourpool, office, etc). Obviously, surviving to the next week is the goal, but the nuance here that is commonly misunderstood is that surviving, per se, is not the goal. More distinctly, winning is the goal. And there is only 1 winner. Being that surviving week to week is a phenomenon for that week, and winning the pool is a combination of surviving and taking advantage of others' picks, mishaps or bad luck or frailties, we have our topic.

I don't think this concept is super hard to understand to most involved handicappers, but I've noticed a few smart friends of mine who really don't get it. The obsession is with surviving as opposed to taking calculated risk and surviving, with more knockouts increasing your chance of winning at the end. A simple way of saying this is via recognizing susceptible favorites. In this sense, it is very much like horizontal wagering in horse racing where the payout at the end is nothing to get excited about if chalk hits over and over; too many people kept "surviving" through the final leg and split the pot.

All in all, there are 3 factors for making a selection:
1. Win odds, which you can utilize a standard book line for approximation, or handicap yourself, or both.
2. Public play % on each team. Again, your best guess or sources as above
3. Future value.

Why do you think a certain number of otherwise intelligent thinkers can't grasp #2? Afraid to stay away from the "sure favorite" or the not go against the "bad team" they are playing against? But they still can't understand that by not picking a big public selection you gain the absolute most if your mitigate risk otherwise?

Also, why do people spread so much? I think the plays on the favorites are correlated with "spreading"... you want to survive to the next week but don't grasp that getting picked off week by week with multiple entries doesn't get you value. Only having multiple entries LATE in the contest does.

Thanks for any input, it's more of a psychological question I imagine, and is probably emotional since I don't believe it's rational at all, but you'll get arguments back at you when you bring it up, which is amusing.

[semibluff]

I don't think there's much, if any, value going against the norm in the 1st 13 weeks. I'm not necessarily going to take the biggest favourite on any given week because i'm looking at the most likely combination of successful picks over a 3-4 week stretch. History has taught me not to look further ahead than that. A team that's good now might not be in 4 weeks time. Hoarding good teams for the final stretch hasn't proven to be +ev for me. By all means make strategic picks in the final 4 weeks. I could see the value in taking the 2nd best pick in the final week if everyone else was on the best pick. The immediate money-line loss value is compensated by the possibility of winning the pot outright.

Off-topic but this is the 1st year in memory that I haven't committed to Pick'em and Survivor. My motivation levels are low and i'm busy on game days. Last year I ran a reverse-survivor competition, (IE pick a team to lose), with only 20 of the 32 NFL teams eligible for picks. That was the most fun NFL game i've been involved in and it's the only 1 I miss this year.

[gojetsgomoxies]

commenting on the very general idea,

i think it's a combo of 1) lack of understanding; 2) behavioural factors (people aren't trying to maximize $$EV even though they may not know it)

i think people get utility from being in the pool itself. sort of like a poker tournament. you want to hang around awhile. it's fun just being the tourney vs. not

i think for +EV you do want to think about grabbing non-dominant teams early on as i can foresee tons of people making it to week 5 or 6 with similar teams and then getting badly stuck as to a team to choose.

a good rule for survivor pool would be "cant repeat losing team" for the first 4-5 weeks.. probably too much admin though

[Waterstpub87]

I tend to see this as a similar situation to a manager picking a pitcher to go out to pitch an extra inning. I don't care about who has to go out in the 11th, so I put my best guy out there in 10th, to maximize survival.

However, you could instead think of this like a march madness bracket. The correct strategy being predicted on picking the best upsets.

But with an nfl survivor pool, if it is small, less than 200, it makes more sense to think of it as situation A. If it is super large, situation B might make more sense, as you need to survive and have better teams left than others who survive.

Mostly luck either way.

[vivaporto]

good read on expected value

as for survior, after not winning for 3 years (i have been playing only 5 years, this is my 5th season), i decided to do 3 entries in my 4th year last season and i ended up winning out among 88 players to take home 2000 pounds. in one of the entries i did exactly what semi bluff said. for the 2nd entry and 3rd entry i did a lottery system. for this lottery, i would write down every team that i predicted to win that week on a sticky note. then i separated all teams that were -7.5 or better favorites from the rest of the pile which also included some underdogs. then fold all the sticky notes and mix it all up (without mixing the two groups), then pick one out from each group.

using this method, i encountered my 1st loss in the entry with the random picking from the -7.5 or better lines during the 4th week. after this week, i combined the -7.5 or better lines with the the other one. during the 8th week i lost my regular picking of 1 game entry. the lottery entry kept going till i won in the 14th week and i ended up winning that week.

I really think that doing the lottery way after predicting all the winning teams for the week helps because at some moments we make crazy decisions or try to play it too safe. doing the lottery way takes out all emotion and makes it random. if you are decent at picking outright winners throughout the board, i recommend giving this method a try.

obviously lot of luck involved too but you need a lot of luck in survivor anyways.

[StackinGreen]

Good posts, thanks guys.

[gojetsgomoxies]

as per the one comment, a survivor pool is quite different from a basic ncaa march madness pool.

to win a survivor pool, you will have to pick non-top teams to win many weeks. so why not try to find the best path and if it means picking an average team day 1 then do it.

one remedy to my "people are maximizing personal enjoyment, not $EV" is to have multiple teams and to perhaps have one entry which is consensus and built to last a long time with everyone else)

[Waterstpub87]

as per the one comment, a survivor pool is quite different from a basic ncaa march madness pool.

to win a survivor pool, you will have to pick non-top teams to win many weeks. so why not try to find the best path and if it means picking an average team day 1 then do it.

one remedy to my "people are maximizing personal enjoyment, not $EV" is to have multiple teams and to perhaps have one entry which is consensus and built to last a long time with everyone else)

You didn't read my post. I compare this picking of non top teams to picking upsets, which is a more optimal bracket strategy when there is a large pool.

[StackinGreen]

This past week was a classic example: While you may have been on one of the bigger favorites and got blown up, you would have been lucky to just pick the one surviving double digit favorite (Baltimore) when considering the long term. It was also a divisional game and road, beyond that (to a team that was a mess but played them to a 6 point game at Baltimore earlier this year). It turns out that spreading didn't work, either for the "multiple picks" players.

I do agree, the likely pitfall is that people want entertainment more than to actually win. I find that stupid, since I can bet ATS each week, anyway. I have tried to convince the others in question that if you aren't handicapping to some degree in the first 10 weeks (and I never say don't take a large favorite, just look for susceptible ones), then you will be relying on total luck to win later, and will still have to take -3, +1, or -2.5 games, for example --- but now you are forced to do so, and you have less teams and bullets available. Why not take good plays early that aren't the ones everyone else is taking, but are still good winners (-5 or -6 at home)? Again, if the public is overdoing it.

[semibluff]

A lot of smaller Survivor pools don't go the distance so it's all about getting through the first 13 weeks. If you're in a huge pool, for example the ESPN Pigskin pool, then you have to pick tactically...and people do. People tend to look for weaker or mediocre teams who play the really poor teams. Picking against the Bengals, Dolphins, and Jets this season has generally been a great strategy...and such a strategy normally is. No-one in their right mind is picking the best teams against good, solid teams because it would be a waste. The other sound strategy is to pick winning but possibly questionable teams whilst they're hot and against weaker opponents. If you didn't pick the 49ers or Seahawks by now then you probably aren't going to pick them due to their remaining schedule.

[StackinGreen]

Everyone picks on the "bad" teams. This is usually what I see with the huge spread guys I've sorta complained about in this thread. "Atlanta" sucks, "Miami sucks" ... until they hand you your ass. People act like the 3-4 worst teams in the league are all going to go 0-16 or 1-15, lol

As an aside, I've noticed a lot of teams can get quite rusty coming off a bye. The Bills were with the Dolphins and squeaked it out, but NO didn't at all. Notice these are divisional games too. Usually they score very little in the 1H in those games off the shelf.