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

[kyro]

week 9 is actually worse. Here's to everyone being knocked out by then.

[Assani Fisher]

[ QUOTE ]
7. Look at Pinnacle.

[/ QUOTE ]

Agreed. I'd use the "biggest favorite that I havn't yet used" method against OP's method any day.

[splashpot]

[ QUOTE ]
[ QUOTE ]
Kyro,

You're leaving out the separation concept which is really the only concept that matters in survivor pools.

[/ QUOTE ]

I have never heard of the separation concept. Any chance you could explain it either here or via PM?

I mean, if I'm going to suck at FF, I might as well try to pwn at Survivor.

[/ QUOTE ]
The separation concept is that you should make picks different from everyone else in your pool whenever possible.

I'd also like to add that looking ahead to week 7 is retarded.

[Thremp]

[ QUOTE ]
[ QUOTE ]
Kyro,

You're leaving out the separation concept which is really the only concept that matters in survivor pools.

[/ QUOTE ]

I have never heard of the separation concept. Any chance you could explain it either here or via PM?

I mean, if I'm going to suck at FF, I might as well try to pwn at Survivor.

[/ QUOTE ]

Basically you don't want to pick the same teams as the majority of your your league.

If you are picking outright winners, and 30% of the league is on a 70% favorite. You increase you expectation by picking a 65% fave if there are only a few people on it. You can carry this out to the extremes. Its very similar to winning NCAA pools (since it applies to all pools), you want to choose upsets that not everyone chooses.

The advice you gave is fine, but winning pools is kinda like BJ tournies etc its not about making the play the wins the most often, but that has the highest expectation.

[tdarko]

lastchance, nez,

Well first thing I already took Pitt this weekend.

I also think that the 6-3 comment about the bears/raiders game is a little bit of an exaggeration, what you saw last week was the Bears against the Chargers--not the Raiders.

So on to the future. Would you go with lets say Atl @ Car instead on Week 10 and go with Chi this week? B/c the Atl/Car game looks to be next best at week 10.

[kyro]

[ QUOTE ]
[ QUOTE ]
7. Look at Pinnacle.

[/ QUOTE ]

Agreed. I'd use the "biggest favorite that I havn't yet used" method against OP's method any day.

[/ QUOTE ]

while I'm sure the books are smarter than me in regards to this stuff, I'd rather do my own work on it. Most of the time we're going to pretty much agree, anyways.

[kyro]

thremp,

Interesting. I'll take that into account from now on. Thanks.

Your concept is pretty high-variance for an event that occurs once a year, though.

[kyro]

td,

In your shoes, I would plan on taking Car over Atl Week 10, barring Atlanta suddenly becoming good.

Unfortunately, it pretty much handcuffs you into taking SD Week 4, I think.

[Nez477]

[ QUOTE ]

The separation concept is that you should make picks different from everyone else in your pool whenever possible.

I'd also like to add that looking ahead to week 7 is fun .

[/ QUOTE ]

[Thremp]

[ QUOTE ]
thremp,

Interesting. I'll take that into account from now on. Thanks.

Your concept is pretty high-variance for an event that occurs once a year, though.

[/ QUOTE ]

Umm... Your variance comment means exactly nothing. If you take picks that someone else takes you will never win a league. At best you can chop it up. Here I'll do some basic math from my example.

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%.