SoSH Survivor Pool - Week 1 Discussion

[tonyandpals]

I hear what you are saying, but couldn't you say it's going to be hard to pick those games right on a tough week?
 
Would anyone entertain the picks split on two teams we would likely never take, that are playing in the same game? (I'm not saying I'd do this, just want to hear thoughts)

 

[FL4WL3SS]

No, that's just burning money. I would rather take my chances and try to get lucky, by doing that you're just giving away picks.
 
Maybe if we're trying to survive later in the season with only a few teams still in the hunt. 

 

[Cumberland Blues]

I'm not going to chime on which games we pick as I don't really pay close enough attention to the NFL to add anything of value there.  But I will express an occasional opinion on the game theory aspects of this.  Regarding the current debate over how much to spread around the first week picks - I think there's a ballance to be struck.  You don't want to spread it out over too many picks as that gives you a small chance of keeping all the tickets alive, it's very hard to pick 6-7 winners in a given week.  But too few choices and you run the risk of wiping out a bunch of tickets in week 1.  I'm guessing 3-4 teams is the sweet spot early on - find the 3-4 "best" games and allocate between them, we are going to have 17+ tickets, so we could do something like 5-6 tickets on each of the two "best" games as determined by the community, and then 2-3 each on the next teir.  We should still have a fair shot at keeping all the tickets alive, and a fairly small chance of facing week 1 ruin.
 
And as a horseplayer (who hasn't played the horses in far too long) I can whole-heartedly endorse the idea some folks are espousing of looking for the losers rather than the winners.  Knowing who to bet against is just as important as knowing who to bet on (perhaps moreso as it's a bit contrarian and might offer some advantage).
 
 

I hear what you are saying, but couldn't you say it's going to be hard to pick those games right on a tough week?
 
Would anyone entertain the picks split on two teams we would likely never take, that are playing in the same game? (I'm not saying I'd do this, just want to hear thoughts)

I always figured the dream scenario is to enter week 17 with two tickets left and teams who are playing eachother still available on both.  But before week 17, I agree with the folks who see this is a waste of resources - the group should arrive at a community opinion and be willing to back that opinion 100%.  Maybe if we still have a bunch of tickets going in week 15 or so, but not before then.
 

 

[glennhoffmania]

No, that's just burning money. I would rather take my chances and try to get lucky, by doing that you're just giving away picks.
 
Maybe if we're trying to survive later in the season with only a few teams still in the hunt. 

 
 
I think I agree.  That's basically a surrender in blackjack and I'd rather try to win more than 50% of the entries.

 

[Stitch01]

There's no reason to bet both sides in week 1 (that's probably a little strong, but for this picking format its pretty much true)

Going to the extreme of splitting 18 picks like 3/3/3/3/3/3 or something like that doesn't seem right either, this isn't a great slate but its not that terrible and there are a few teams in better spots than anyone else. Like I don't think putting 1/6 of our shares on some -2 favorite where there isn't a real strong consensus among the community that the line is wrong helps us a lot at reducing risk. If I was doing this myself on this slate, Id probably have 75% of my shares split between 2 or 3 teams (depending if Brady plays) and then 20-25% where I diversify more depending on my specific view on a team where I look to start some potentially interesting/contrarian trees where future value is still available to me and might be closed off to other teams. Need to do more research on the games though.

 

[FL4WL3SS]

I love it.
 
The idea of having a few risky picks so that later on they're in a much better spot is great. Go with 75% "safe" picks and 25% diversity with risky picks.

 

[wade boggs chicken dinner]

I love it.
 
The idea of having a few risky picks so that later on they're in a much better spot is great. Go with 75% "safe" picks and 25% diversity with risky picks.

 
If only there were a couple of "safe" picks.
 
If Randall Cobb doesn't play for the Packers (sprained AC joint that may keep him out of week 1) and Brady doesn't play for the Pats, are there really any safe picks?

 

[Stitch01]

I think at first glance Dallas and Green Bay are still safer than anything else on the board. They arent locks or anything, but its impossible to avoid risk. Its obviously a matter of degree, but I think deciding to split games in a way where we pick Dallas equally as much as a couple of other teams to diversify is a strategic error.

 

[glennhoffmania]

Even if Brady plays is NE really a safe pick? I guess Bell being out helps but it's not like Pit is a shitty team.

 

[FL4WL3SS]

What if every week we have a poll of all the games and we divide the picks up based on the highest percentage of picks to win?

So we could divide the picks up to the highest 3 games not already picked.

 

[FL4WL3SS]

I used the word pick a lot in that post. I like to be confusing.

Picks.

 

[Stitch01]

Even if Brady plays is NE really a safe pick? I guess Bell being out helps but it's not like Pit is a shitty team.

Safer bet this week than anything on the docket if we don't account for future value. We are dealing with unknowns here, but Pats track record at home is about as safe as we can hope to find on this slate although they'll almost certainly have a spot where they're a bigger favorite.

 

[dynomite]

Safer bet this week than anything on the docket if we don't account for future value. We are dealing with unknowns here, but Pats track record at home is about as safe as we can hope to find on this slate although they'll almost certainly have a spot where they're a bigger favorite.

As I said above, the Steelers offense is down not just one but 2 All-Pros (Bell and Pouncey) in addition to Bryant.

If Brady plays, that's the safest play on the board in my opinion.

 

[Kenny F'ing Powers]

That team has offense to throw away and we have no idea what type of D the Pats have. The Pats could hang 28 on them and lose.

Also, I know Gronk/Edelman haven't played much/at all, but the O has been poor in preseason.

The Dolphins are for real. Top 10 D (or better), a solid QB, and a solid running game. How people don't see that game as a layup is beyond me.

 

[Ed Hillel]

Honestly, I kind of like Minnesota this week. With Peterson back, Bridgewater showing promise last season, and most of San Fran's talent retired or facing legal issues, I think it's a solid option. It's a road game, unfortunately.

 

[loshjott]

That team has offense to throw away and we have no idea what type of D the Pats have. The Pats could hang 28 on them and lose.

Also, I know Gronk/Edelman haven't played much/at all, but the O has been poor in preseason.

The Dolphins are for real. Top 10 D (or better), a solid QB, and a solid running game. How people don't see that game as a layup is beyond me.

 
I agree on the Dolphins. DC is in free fall, and Kirk Cousins isn't the answer, unless the question is "Who's not the answer?"

 

[glennhoffmania]

That team has offense to throw away and we have no idea what type of D the Pats have. The Pats could hang 28 on them and lose.

Also, I know Gronk/Edelman haven't played much/at all, but the O has been poor in preseason.

The Dolphins are for real. Top 10 D (or better), a solid QB, and a solid running game. How people don't see that game as a layup is beyond me.

 
Yeah I agree.  The Pats aren't a bad pick especially compared to the rest of the games.  But do we know who will be healthy for them on offense in week 1?
 
As for the Dolphins, I'm a Miami fan so I'm not the most objective person when it comes to that game.  I don't love road teams but Washington looks terrible.  I think I'd be more surprised if Washington beat Miami than if Pit beat NE.

 

[Stitch01]

I like the Miami pick, but less likely for Miami to lose on the road than for the Pats to lose a home game if Brady plays is a bold call and certainly not one I could make personally (ignoring future value for the moment)

 

[FL4WL3SS]

So no love for creating a poll for all the games and using the top vote getters as a basis for our picks?

 

[tims4wins]

So no love for creating a poll for all the games and using the top vote getters as a basis for our picks?

 
I think this is ultimately the way to go - we list all the games, and make it a multiple vote type of poll. Since we are going to have 15+ entries I would just let everyone select as many games that they would like to pick as they want. So if I want to cast votes for 8 teams, I can. If you want to cast votes for 4 teams, you can. Then toward the end of the week, we will see how the %s fall, and make a group decision on how many different teams to use based on what appears to be a reasonable cutoff point.

 

[FL4WL3SS]

I think it would work better if everyone voted on every game, that way we know which way folks are leaning and can make picks instead of just getting a lot of votes for the 'trendy' picks.
 
You can't enforce it, but it would be ideal if we could.

 

[tims4wins]

If we voted on every game, how would you go about selecting picks? Based on the highest winning %s?
 
Edit: I'm not convinced something like that would work, just because of situations like Carolina, or Seattle - the vast majority of us will pick those teams to win in week 1, but I'm not sure anyone would want to use a pick on them.

 

[FL4WL3SS]

Well if 75% of folks pick Seattle to win, then that's a pretty good confidence indicator, no?

 

[tims4wins]

Well if 75% of folks pick Seattle to win, then that's a pretty good confidence indicator, no?

 
True, but it doesn't mean that we want to use them. There is strategy to think about. I guess a poll could work for week 1, if that is what you are suggesting. It likely won't work going forward, however, as teams are used up.
 
Edit: I am having difficulty articulating it, but I see a lot of games where we will all likely agree on who we would pick to win the game, but where we still wouldn't have a ton of confidence.

 

[FL4WL3SS]

I don't understand that logic. If, say, 90% of the folks pick Seattle to win and that ends up being the highest percentage, how is that not indicative of the confidence we have in the team? Why would we then say 'yeah, but we don't want to pick them'.
 
Your strategy point is valid, however. There could be situations where a team is at 90%, but we want to save them for later in the season. I just figured we could use it as a starting point for discussions. Maybe it won't work, however.

 

[tims4wins]

I don't understand that logic. If, say, 90% of the folks pick Seattle to win and that ends up being the highest percentage, how is that not indicative of the confidence we have in the team? Why would we then say 'yeah, but we don't want to pick them'.
 
Your strategy point is valid, however. There could be situations where a team is at 90%, but we want to save them for later in the season. I just figured we could use it as a starting point for discussions. Maybe it won't work, however.

 
It's because these things aren't binary. If I think Seattle has a 51% chance of winning, I will select Seattle. Everyone else could have the exact same thought. And then 100% of us will pick Seattle even though none of us are extremely confident they will win. So it will look like we all have a ton of confidence in them when we may not. In short, your idea means we have to pick a team in every game even if we don't like the matchup.
 
Whereas if we go with selecting however many teams we each want in a given week, we can basically build in a confidence interval by only picking the games where we feel confident. If there are 50 of us voting, and the Pats get votes from 38 of us, and Miami from 32 of us, and Philly from 28 of us, and then no one else gets over 10 votes, then we know where to cut it off.
 
YMMV I guess.

 

[ipol]

Also, how would picks be weighted? For instance, I suck at prognosticating winners. No way my vote should count as much as someone who is more in tune with the league.

 

[FL4WL3SS]

Makes sense.

 

[j44thor]

Also, how would picks be weighted? For instance, I suck at prognosticating winners. No way my vote should count as much as someone who is more in tune with the league.

 
Don't vote if you don't think you can contribute to overall success, nothing wrong with going along for the ride.

 

[Kenny F'ing Powers]

I think it would work better if everyone voted on every game, that way we know which way folks are leaning and can make picks instead of just getting a lot of votes for the 'trendy' picks.
 
You can't enforce it, but it would be ideal if we could.

 
I think it makes sense to NOT have a poll for the first few days every week. Give people a chance to parse through the games and make their case. Otherwise you'll have a bunch of people voting on day 1 or 2 that may have their opinion swayed by facts in day 3-4.

 

[j44thor]

I agree with KFP, probably best to have people nominate teams MON-WEDS and have people vote on the nominated teams Thurs-Friday.  

 

[Stitch01]

That's how we did it last year IIRC.

 

[tonyandpals]

That's how we did it last year IIRC.

 
Yes, and we avoided the "poll" feature and tallied them in a spreadsheet to keep people from voting that didn't have shares.

 

[glennhoffmania]

I agree with KFP, probably best to have people nominate teams MON-WEDS and have people vote on the nominated teams Thurs-Friday.  

 
 

 
Yes, and we avoided the "poll" feature and tallied them in a spreadsheet to keep people from voting that didn't have shares.

 
This sounds perfect to me.

 

[wade boggs chicken dinner]

One thing about the way we did it last year.  It seems like we as a group are pretty similar to the population of people playing survivor pools.  If I recall correctly, our votes ended up basically mirroring the everyone else's picks - in other words, our internal voting percentages were pretty close to the voting percentages on survivorgrid and some other websites.
 
Again, I don't have an answer for this but I'm just wondering out loud if that's the best way to allocate our picks. 

 

[wade boggs chicken dinner]

If anyone is interested, while the 2015 version of survivorgrid isn't open, they do have information on 2014 knockouts here:  http://www.survivorgrid.com/knockouts/2014
 
Turns out that the CHI-BUF game where we lost 3/8ths of our picks was the 9th biggest upset in terms of pick percentage.
 
In terms of teams that were picked that ended up losing, the top ten by percentage were: 


CHI (19.9%)
SEA (13.0%)
NO (11.0%)
PIT (9.1%)
DAL (6.1%)
SF (6.0%)
TB (5.9%)
DET (4.3%)
CIN (4.3%)
KC (4.0%)
 
In terms of opponents that caused the above-referenced teams to lose, the top ten by percentage of times they defeated a picked team were:


BUF  (24.1%)
STL  (11.4%)
CLE  (9.9%)
TB  (9.7%)
DAL  (7.3%)
WAS  (6.2%)
CAR  (5.2%)
CHI  (5.0%)
TEN  (3.7%)
MIA  (3.3%)
 

 

[j44thor]

I think what we can glean from the above is A. Don't pick against strong D and B. Never trust Jay Cutler

 

[ipol]

This just might be the KoD for EdH's pick but I'm down with The Vikings. Have the 49er's looked at all competent? Who needs my money?

 

[tonyandpals]

Alright, we gotta start talking shop here.  Should we get the most confident pick from each player, bolded?

 

[weeba]

I think so. Last year's format seemed to work well, sorry if it's more work on your end Tony.
 
But, we're 1 week out. Does it make sense to start the confidence picks on Sunday/Monday?

 

[tims4wins]

Let's wait to get confirmation Brady is playing week 1...

 

[FL4WL3SS]

Even if Brady is playing, I don't like that game. The Patriots started off slow last year and I can definitely see them starting off slow again this year given the shitty offseason they just had and the superbowl hangover.

 

[tonyandpals]

Brady, in.

 

[j44thor]

I'm sticking with GB over CHI. Even with no Jordy CHI has no answer for Lacy and Cobb/Adams et all.  Not to mention Cutler no longer has Trestman at OC and a less than 100% Jeffrey.  They are going to struggle on both sides of the ball in this one.
 
The other road favorite I would take is MIA over WAS.  

 

[Stitch01]

Even if Brady is playing, I don't like that game. The Patriots started off slow last year and I can definitely see them starting off slow again this year given the shitty offseason they just had and the superbowl hangover.

They have the best chance to win the game of any NFL team this week based on the data we have available. Not saying we should use them, but there aren't any locks this week.

 

[FL4WL3SS]

What do folks think about Cleveland @ NYJ? I think the Jets will have a very strong defense this year and the Browns offense will not be in top form going into this game.

 

[fiskful of dollars]

I like Miami, GB, NE and  (God help me) NYJ in that order. 
If we're voting I say Miami.

 

[tonyandpals]

In reviewing how we voted last year, each share voted as if all the picks were there's. Example, if we had 20, mine could (not saying it is) be
 
GB(12)
NE(7)
MIA (1)
 
Then, I tallied them all to come up w/ the %s...
 
Should we stick to that?

 

[Stitch01]

What do folks think about Cleveland @ NYJ? I think the Jets will have a very strong defense this year and the Browns offense will not be in top form going into this game.

For 1 or 2 picks out of 20 or whatever, sure.

 

[tims4wins]

In order of confidence I would go with
GB
MIA
MIN