SoSH Survivor Pool - Week 1 Discussion

j44thor

Member
SoSH Member
Aug 1, 2006
11,137
tonyandpals said:
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?
 
I wasn't involved in this last year so excuse my ignorance.  Are we shortlisting the teams down to say 3-5 and then picking the shares or is this a free-for all and then you pick based on %?
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
I personally find it hard to understand how people are voting a week out from the first game and more than a week out from the bulk of the games.
 
j44thor said:
 
I wasn't involved in this last year so excuse my ignorance.  Are we shortlisting the teams down to say 3-5 and then picking the shares or is this a free-for all and then you pick based on %?
 
Everyone picks how they see fit; Tony adds up the percentages; and then allocates votes on these percentages.  For example, in Week One last year (we only had eight entries), the voting went like this:
 
PHI had 52 votes or 46.43% so they received 4 of 8 entries.
CHI had 39 votes or 34.82% so they received 3 of 8 entries.
PIT had 10 votes or 8.93% so they received 1 of 8 entries.
NYJ had 9 votes or 8.04% but there were no more entries left so they didn’t receive any entries.
DET had 2 votes or 1.79% which wasn’t enough to register so they didn't receive any entries.
 
And to emphasize my point about matching the general population, I will note that according to Survivorgrid, the general public picked teams at the following percentages:
 
PHI = 42.8%
CHI = 19.3%
PIT = 10.6%
NYJ = 6.5%
KC = 3.5%
DET = 3.2%
 
Here's the thread for anyone who wants to peruse it.  http://sonsofsamhorn.net/topic/85448-sosh-survivor-525k-week-1/
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
Hey Tony - one strategy question.  I remember last year the rule was that if you didn't pick the Thursday game, we didn't have to get our entries in until prior to the Sunday games?
 
Assuming that is still correct - because we have 18 entries, could we put a certain number on the Patriots before the Thursday game but not put in the rest on until the following Sunday?  If that's true, then it would be useful to figure out how many of our entries (if any) are going to go on the Pats and then wait to allocate the rest after that game.
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
wade boggs chicken dinner said:
Hey Tony - one strategy question.  I remember last year the rule was that if you didn't pick the Thursday game, we didn't have to get our entries in until prior to the Sunday games?
 
Assuming that is still correct - because we have 18 entries, could we put a certain number on the Patriots before the Thursday game but not put in the rest on until the following Sunday?  If that's true, then it would be useful to figure out how many of our entries (if any) are going to go on the Pats and then wait to allocate the rest after that game.
 
That is correct. Each pick is independent of the others, so we could go X on Thursday and not lock in the rest until Sunday.
 

glennhoffmania

meat puppet
Lifetime Member
SoSH Member
Jul 25, 2005
8,411,726
NY
FL4WL3SS said:
Why would we waste the Patriots this early in the season? I would think we'd want to hold onto them as long as possible.
With so many picks I don't think it's a bad idea to use NE for a couple of the entries. Besides them as of now I also like GB and Miami. The Jets don't give me any comfort. I could see Fitzy having a rough time.
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
So survivorgrid is up: http://www.survivorgrid.com/. Looks like top five picks from the general population are:

GB (16.2%); CAR (12%); MIA (10.7%); SEA (7.8%); and IND (6.5%).

It seems to me that there are three questions about which we need to come to a consensus.

(1) How many teams are we going to pick?
(2) Are we going to put any entries on NE for the TRS game?
(3) What are the other teams that we want to pick?

What's the best way to figure this out?
 

DanoooME

above replacement level
SoSH Member
Mar 16, 2008
19,939
Henderson, NV
I wouldn't touch Seattle for multiple reasons:
 
1. On the road, against a decent team
2. The O-line, despite Pete and Tom Cable's praise of them recently, is still a work in progress.  The Rams' D-line could dominate this game
3. The offense still seems out of sync because of #2.
4. Divisional game
5. Special teams coverage has been meh
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
wade boggs chicken dinner said:
So survivorgrid is up: http://www.survivorgrid.com/. Looks like top five picks from the general population are:

GB (16.2%); CAR (12%); MIA (10.7%); SEA (7.8%); and IND (6.5%).

It seems to me that there are three questions about which we need to come to a consensus.

(1) How many teams are we going to pick?
(2) Are we going to put any entries on NE for the TRS game?
(3) What are the other teams that we want to pick?

What's the best way to figure this out?
 
I think we need to start with everyone giving there picks if all 20 brackets were there own.  I'll break them down as usual and we can hash out any of the ones on the cusp. I don't see another way that works across such a large group.  If anyone objects, lets here the reason and anther suggestion. Bold your picks so they jump out at me.
 
I'll kick it off, if these were my 20 picks, I'd be
 
GB(12)
Pats(6)
Dallas (2)
 
edit: And they will be factored by the # of shares that person has.
 

tims4wins

PN23's replacement
SoSH Member
Jul 15, 2005
37,764
Hingham, MA
Here is my vote for the 20:
GB: 8
Miami: 6
Philly: 2
Minnesota: 2
NYJ: 2

Edit: I realize this is 4 road teams which is pretty scary, but I really like GB, Miami, and Philly, and then at least one of the Jets and Vikings should win, if not both.
 

FL4WL3SS

Member
SoSH Member
Jul 31, 2006
14,930
Andy Brickley's potty mouth
tims4wins said:
Here is my vote for the 20:
GB: 8
Miami: 6
Philly: 2
Minnesota: 2
NYJ: 2

Edit: I realize this is 4 road teams which is pretty scary, but I really like GB, Miami, and Philly, and then at least one of the Jets and Vikings should win, if not both.
I can get on board with these votes.

GB: 8
Miami: 6
Philly: 2
Minnesota: 2
NYJ: 2

Can someone bold, on mobile.
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
Here's some more information from USA Today if anyone is interested: http://www.usafootballpools.com/pool_format/survivor/picks/website_weekly_top_survivor_picks.php?top_pick_week=1.

As of right now, the survivor picks for their website are:

1 Green Bay Packers 12%
2 Miami Dolphins 12%
3 Carolina Panthers 9%
4 Tampa Bay Buccaneers 7%
5 Seattle Seahawks 7%
6 Dallas Cowboys 7%
7 New York Jets 7%
8 New England Patriots 6%
9 Cincinnati Bengals 5%
10 Indianapolis Colts 5%

And for those interested, the picks in their loser pools go like this:

1 Chicago Bears 26%
2 Pittsburgh Steelers 11%
3 Washington Redskins 11%
4 Baltimore Ravens 10%
5 Buffalo Bills 7%
6 New York Giants 6%
7 St. Louis Rams 5%
8 Oakland Raiders 5%
9 Jacksonville Jaguars 4%
10 New England Patriots 2%
 

tims4wins

PN23's replacement
SoSH Member
Jul 15, 2005
37,764
Hingham, MA
What is most interesting to me is the relatively small %s for the top picks. Very spread out

Also, lots of love for the Panthers in week 1. Very good D, won't be phased by hot weather, so from that perspective I can see taking them. But I do think Jacksonville is going to be feisty this year and Carolina started quite slowly last year
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
What is most interesting to me is the relatively small %s for the top picks. Very spread out

Also, lots of love for the Panthers in week 1. Very good D, won't be phased by hot weather, so from that perspective I can see taking them. But I do think Jacksonville is going to be feisty this year and Carolina started quite slowly last year
Yeah, I can't see picking them with Benjamin out and at least JAC's offensive line should be pretty good.

I also can't see picking TB, Seattle, or Denver.

Pats are currently biggest favourites of the week. I wonder if their pick percentage is so low because people put their picks in before the ruling came down. But no matter what the reason, it's unusual that the biggest favourite would get such a low percentage of picks.
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
QUESTION:
In looking at last year, we (I) did not factor the # of shares into the pick counting. IE I have 3 shares and go 12/6/2, that turns into 36/18/6.
 
My question is, should it? Should the # of shares you have weigh your vote? Or is the share really only come into play when paying out?
 
My thought is, the more you bought in, the more say you have and I should apply the factor. I want to get the feeling of the group.
 
I'm putting together the spreadsheet, so I need to know if I should build in this logic or not.
 
Looking back @ last year, in week 1 we had 112 picks made out of a possible 192, if that matters to anyone.
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
tonyandpals said:
QUESTION:
In looking at last year, we (I) did not factor the # of shares into the pick counting. IE I have 3 shares and go 12/6/2, that turns into 36/18/6.
 
My question is, should it? Should the # of shares you have weigh your vote? Or is the share really only come into play when paying out?
 
My thought is, the more you bought in, the more say you have and I should apply the factor. I want to get the feeling of the group.
 
I'm putting together the spreadsheet, so I need to know if I should build in this logic or not.
 
Looking back @ last year, in week 1 we had 112 picks made out of a possible 192, if that matters to anyone.
I only have one share but I'd defer to what you feel is easiest for you to tabulate. Last year we had 8 entries but still I imagine the tabulation took some real time and I can see this year being exponentially more difficult to administer. Whether or not you factor in shares.

I also have a question. Right now, people have voted for seven teams. How are we going to decide how many teams we are voting for? Or are we going to put at least one share on any team that has enough votes for one share?

(This is like trying to deciding on a constitution for a country).
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
We set a threshold last year (I forget what it was). Lets look at what the data shows, but I think it's not going to be much of an issue.
 
As far as tabulating, # shares or not, same work for me. I would just need it noted in each persons selection, so I don't have to keep looking it up.
 

tims4wins

PN23's replacement
SoSH Member
Jul 15, 2005
37,764
Hingham, MA
I think the more shares you have the more your vote should count. For the record I only have one share. I am fine either way.

Also worth remembering that some folks bought in but will not be voting, so that could skew the vote counts of the multi share people. Just something worth thinking about.
 

DanoooME

above replacement level
SoSH Member
Mar 16, 2008
19,939
Henderson, NV
I can see both arguments about weighted voting.  People that pay more should get their fair share of representation.  On the other hand, that may negatively influence the opinion because one person's opinion counts two or three times.
 
If I had to split my 20 votes:
 
NE - 10
Mia - 4
GB - 2
SD - 2
Phi - 2
 

glennhoffmania

meat puppet
Lifetime Member
SoSH Member
Jul 25, 2005
8,411,726
NY
tims4wins said:
I think the more shares you have the more your vote should count. For the record I only have one share. I am fine either way.

Also worth remembering that some folks bought in but will not be voting, so that could skew the vote counts of the multi share people. Just something worth thinking about.
 
I agree with this and I only have one share as well.  I could see the argument both ways.  I think those with more shares should say whether they want more votes.  One vote per guy is easier and arguably the correct way but they may disagree and that's fine.
 
As for picks I think mine are:
 
Miami-5
NE-5
Dallas-5
GB-5
 

bostonbeerbelly

Member
SoSH Member
Apr 26, 2008
2,240
San Fran
NE - 4
GB - 4
MIA - 4
DAL - 4
NJY - 2
SEA - 2 
 
x2 shares 
 
I am fine with my buy-in only counting one time, I do not do as much research as many on this board so I don't need it to count twice. My thought for the first week is to spread out the picks and hopefully at least 4 of the 6 games above we win. 
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
Stitch01 said:
Do we have to have all picks in by Thursday or just if we are picking that game, I cant remember
 
No, only if you are picking it.  We could pick the Thursday game on 5 sheets and wait until Sunday to pick the other 15.
 

Hambone

will post for drinks
SoSH Member
Jul 15, 2005
2,822
NE - 4
GB - 4 
AZ - 4 
MIA - 4
DAL - 4
 
I had 2 shares, but I think it would be a pain to have to figure out how to calculate points so don't need them counted that way.
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
FL4WL3SS said:
So we're looking at ~$10k/share?
 
We're up to 63(21) now (havent updated that page).  So around that if history hold true w/ the last minute registrations.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
43,134
AZ
tonyandpals said:
QUESTION:
In looking at last year, we (I) did not factor the # of shares into the pick counting. IE I have 3 shares and go 12/6/2, that turns into 36/18/6.
 
My question is, should it? Should the # of shares you have weigh your vote? Or is the share really only come into play when paying out?
 
My thought is, the more you bought in, the more say you have and I should apply the factor. I want to get the feeling of the group.
 
I'm putting together the spreadsheet, so I need to know if I should build in this logic or not.
 
Looking back @ last year, in week 1 we had 112 picks made out of a possible 192, if that matters to anyone.
Depends on whether we think crowd sourcing this leads to better decisionmaking. Giving more control to those with more at stake is certainly equitable. But the number of shares does not make a person x times more insightful. If we think there is value in diversity of opinions -- maybe there isn't but if we think there is -- you may lose something. Seems easier to see the point if we were down to one share.

Ultimately, it doesn't matter too much to me. There are enough shares that I think we will get a diversity of views. I only have one share and probably won't weigh in most weeks because I'm not as well informed as others in the group. But it seemed worth trying to get behind the question.
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
Bill Barnwell has some candidates for teams that could do substantially worse than they did last year: http://grantland.com/the-triangle/nfl-2015-season-predictions-part-2-the-falling-stars/. Perhaps this might give us some insight so we can avoid doing what we did last year and staking 1/2 of our picks on teams that eventually won five or fewer games.
 
Kenny F'ing Powers said:
Miami - 20

2 shares.

Yes, I'm that comfortable with the pick.
I understand your pick and it seems to me that from a strategy POV, putting all (or almost all) of the entries on one's best pick has the highest potential reward, particularly if the goal is to get 20 entries to week 2. As Tony and Pals said upthread, it's far easier to go 1-0 or 2-0 versus 6-0 or even (I would surmise) 5-1.

OTOH, you're basically putting five grand on the Dolphins to win.

I'm going to vote putting 1/2 our picks on NE on the theory that it's not often that the game with the highest spread gets only 10% (as of right now) of the picks. So

NE: 10.

I'll vote on the other shares later.
 
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
Depends on whether we think crowd sourcing this leads to better decisionmaking. Giving more control to those with more at stake is certainly equitable. But the number of shares does not make a person x times more insightful. If we think there is value in diversity of opinions -- maybe there isn't but if we think there is -- you may lose something. Seems easier to see the point if we were down to one share.
I have an idea.

I don't think crowdsourcing really helps - because if we really wanted to crowd source from a larger sample, we could use the %s on the Survivorgrid or USA today websites. And last year our picks effectively mirrored those %s and I don't think it worked out too well.

In every pick 'em pool I've been in, it seems that the people who win (or come close to winning) are those who have an intuitive sense of what teams are going to be bad early on in the season - they can wrack up 10 and 11 win weeks early in the season because they have a better feeling for teams that are going to suck than the general public does. Bill Simmons (I think it was him) has said something similar - that people can make a lot of money betting the NFL in the early weeks if they can identify the really bad teams before Vegas can do it (the lines will be out of whack).

So my idea is this: each week, we keep track of our own results (TonyandPals has enough administrative burden already), and when we post our vote from week to week, we include those past results. Not trying to embarrass anyone but my theory is that people whose votes would have done better have a better feel for the relative strength of the teams.

We don't have to weigh votes according to past results, but hopefully all of us will take past results into account when making our future picks.

Seems like the best way to make use of our collective "expertise."

Hope this makes sense. Thoughts?
 

Stitch01

Member
SoSH Member
Jul 15, 2005
18,155
Boston
NE: 4 (or 1/5 of the shares). Pats should have good future value where they are double digit favorites against teams we can be fairly confident aren't contenders (Jags, Skins, Titans) so even as the biggest favorite this week I don't want to use that much of them.
 

Stitch01

Member
SoSH Member
Jul 15, 2005
18,155
Boston
wade boggs chicken dinner said:
I have an idea.

I don't think crowdsourcing really helps - because if we really wanted to crowd source from a larger sample, we could use the %s on the Survivorgrid or USA today websites. And we did that last year and I don't think it's very helpful.

But in every pick 'em pool I've been in, it seems that the people who win (or come close to winning) are those who have an intuitive sense of what teams are going to be bad - particularly very early on in the season. Bill Simmons (I think it was him) has said something similar - that people can make a lot of money betting the NFL in the early weeks if they can identify the really bad teams before Vegas can do it (the lines will be out of whack).

So my idea is this: each week, we keep track of our own results (TonyandPals has enough administrative burden already), and when we post our vote from week to week, we include those past results. Not trying to embarrass anyone but my theory is that people whose votes would have done better have a better feel for the relative strength of the teams.

We don't have to weigh votes according to past results, but hopefully all of us will take past results into account when making our future picks.

Seems like the best way to make use of our collective "expertise."

Thoughts?
I see what you are getting at, but Id probably lean against it. There's small sample size issues involved, people look like their dividing shares four or five ways so its going to be sort of cumbersome to go through everyone's record, there's game theory type reasons to make picks beyond "I think this team is most likely to advance", and I personally wouldn't want people to get a week or two wrong and then feel marginalized for picking in future weeks.

I actually think the best way to do this, jumping off your survivor grid point, probably isn't for each person to vote on specific games with their shares but rather to come to a consensus for overall strategy for the week that gets as much support from the group as possible but I don't know if its feasible to pull off. Voting shares is always going to trend towards picking primarily the favorites for the week and underweight the game theory/contrarian aspect needed to win giant pools like this one (similar, but not identical, to the idea of finding low ownership plays in GPP's for those of you that play daily fantasy)
 

wade boggs chicken dinner

Member
SoSH Member
Mar 26, 2005
30,880
Courtesy of Ed Hillel in the Pats pre-game thread, apparently there's an inch of rain in the Foxborough forecast for tomorrow in case that affects anyone's votes.
 

CantKeepmedown

Member
SoSH Member
Jul 15, 2005
2,597
Portland, ME
I only have 1 share
 
But I'd go something like:
 
Green Bay - 8
Miami - 6
New England - 4
Dallas - 2
 
I am also more than willing to ride along with the others.  I'm not going to throw a fit if certain picks aren't made the way I want.  
 

glennhoffmania

meat puppet
Lifetime Member
SoSH Member
Jul 25, 2005
8,411,726
NY
wade boggs chicken dinner said:
Courtesy of Ed Hillel in the Pats pre-game thread, apparently there's an inch of rain in the Foxborough forecast for tomorrow in case that affects anyone's votes.
 
That would affect my thinking, but I still think at least some entries should go to NE. For the moment should we only focus on how many picks we use on NE and then resume the rest of the debate Fri?
 

tims4wins

PN23's replacement
SoSH Member
Jul 15, 2005
37,764
Hingham, MA
What % of picks are the Pats currently in our votes? Agreed that we need to decide how many picks they are getting before we allocate the rest.

Also this is my thinking on a general schedule each week:
Mon-Tues: nominate teams, discuss each game
Weds-Thurs: bold nominations, determine how many picks (if any) to use on Thursday game
Friday: allocate remaining picks

It is tough allocating all of these entries in a week, but hope we have this problem for a while!
 

tonyandpals

Well-Known Member
Lifetime Member
SoSH Member
Mar 18, 2004
7,867
Burlington
Picks so far:
 
5.6 MIA
5.1 GB
4.2 NE
1.1 DAL
.65 MIN
.59 SEA
.35 NYJ
.35 PIH
.24 SD
.24 AZ
 
edit: Yes I did miss AZ for a silly reason...These should be correctish.
 

JoePoulson

Well-Known Member
Gold Supporter
SoSH Member
Feb 28, 2006
2,755
Orlando, FL
There an AZ vote in there from Hambone that's missing from those numbers, I think.
 
I have two shares, but am not sure of the final breakdowns.  I'll just do 50% GB, 35% NE, 15% MIA.  If that's OK, of course.