Read this before starting a "Who Do I pick in Survivor" Thread

[kyro]

[ QUOTE ]


Two outcomes:

1) You go into week two 70% of time with 100 people (We'll assume everyone else wins which is worst case for calcs). Your EV is ~.70
2) You go into week two 65% of the time with 70 people. Your EV is ~.98

Variance isn't really much higher at all. You're out of the pool 7% more often, but have increased your EV 40%.

[/ QUOTE ]

I've been playing around with this a lot. I don't understand a few things. I ran the equity numbers myself and I got:

Scenario 1: .705
Scenario 2: .734

That's assuming Team A wins 70% of the time, Team B wins 65% of the time, 30% of the league takes Team A, and 2% of the league takes Team B. It is also assuming that all the other teams win 100% of the time. If we adjust the numbers to something like the other teams win 80% of the time, then it's something like .732 vs .800.

It seems you left out the chance that the team you don't pick ends up winning, which is a pretty crucial part.

[g-bebe]

I guess I agree with all of your points, and I'm by no means a football expert, but I usually just find the weakest teams, and see who they play. If I can find a great matchup, then I'd probably revert to your scheme.

Then again, I wasn't aware how close (supposedly) SF was to ARI when I went with the 49ers in week 1 of my survivor league. Thank god they pulled that out...

[Thremp]

Kyro,

Alright. There are 4 outcomes.

45.5% of the time you both win.
24.5% they win and you lose.
19.5% they lose and you win
10.5% you both go down.

This can be compared to.
70% winnars
30% losarz.

If we assume everyone else wins... By tagging along we only have .700 equity.

Now for example 2.

45.5% of the time we all goto round 2 EV = 1%
24.5% of the time we don't move on and they do = 0%
19.5% you move on without them = 1.42%
10.5% you are all busto

EV = .7319

Now... Correct me if I'm wrong, but lets get a bit more complex...

We'll go 30% win .7 and then 2% win at .65 and then the rest are morans and we'll assume 35 of them carryover due to random spread out picks (most of which are likely correct).

45.5% 75 people left EV = 1.3%
24.5% 0EV
19.5% 40 people left EV = 2.5%
10.5% 0EV

EV = 1.079

As opposed to (I'm gonna assume you buddies go 2-2 in this on random teams):

70% 77 people EV = 1.29
30% = 0EV

So you can see here giving up the equity in a group of 5 isn't worth it.


Though it does appear my initial numbers were incorrect. No clue what happened.

[Tyler Durden]

7) DO pick a home team coming off a bye

[Tyler Durden]

Interesting weekend for Survivor leagues w/ the Bengals and Saints being upset. I suspect a bunch of people took the Panthers over Houston as well.

Glad I chose the Ravens today.

Next week it's looking like the Patriots at home versus Buffalo.

[Russ M.]

2/3rd of my league took the Bengals. I took Chicago.

I'm also almost surely taking NE at home over Buffalo.

[jedi]

Denver. At home. Vs. Oakland.

Sure thing, right??? Almost oops. I guess I forgot rule 5

[dkgojackets]

I had Pittsburgh. Seemed to fit all the rules.

Everyone with Carolina and Cincy deserved to lose.

[DonkeyKongSr]

Well, Cleveland + people forgetting to pick has dropped my league from 20 to 10.

For next week (PIT & SEA used), I see these as my two possibles:
- Philly over Detroit
- New England over Buffalo

New England should be an easy pick, but should I go for Philly (or someone else) for the "seperation concept"? FWIW, only 2 of the other 9 teams left have used their NE pick.

[rafiki]

I would take NE and feel pretty good about it. It's a rather tricky week, there's no doubt. Right now I don't feel Phili has proved that it'll stop that detroit offense If Kitna plays 4 quarters. Megatron is just too awesome.