Suicide Pool

Tough week for me in my local suicide pool. For those who don't know, in a suicide pool, one needs to pick one game right each week, straight up. The one caveat is that you can only take each team once.


Possible choices for this week are Rams/Giants, Vikes/Lions and Bear/Cardinals.


Any thoughts?


Danny

I had the same thoughts in my pool. I took MIN as there should be ample opportunities to take STL later in the year. Of course this may be the week DET wakes up, but still think it's a decent chance.


Oh, and I'm _still_ not convinced CHI is for real. They should beat ARI but I am not ready to risk losing yet.

What percentage of your pool is left? We are down to 40%


Danny

In the pool I was in we were down to 50% after the first week... yup, half the participants picked Minny in the first week. The pool is now down to 19%. Unfortunately, my buddy and I got booted this past week after picking Philly.

My pool has a rebuy feature - you're allowed one loss in the first 3 weeks, but have to pay in another entry fee (and you still can't reuse any teams). Now that the rebuys are over...


Entries: 161

Rebuys: 82

Left alive: 97, or about 60%


Of course some contestants are much bolder during the first 3 weeks when they can rebuy once. I'm not convinced that is a good strategy given the double entry fee. My picks were TAM, GBP, PHI, OAK... other than week 1, the biggest strength disparity I could find, with the home team. I _hate_ taking road teams in these things, they seem to lose more often than not.


Actually I think the best way to do these pools is to sit down before the season with an NFL schedule and list the possible picks each week, then try to optimize your selections for maximum chance of survival, weighting it more toward the early weeks (figuring that surviving through week 13 may be enough to win e.g.). Of course your set of 'good teams' will change as the season goes on, that is another reason I like to grab the known strong teams early and watch people fall by the wayside.

I'm in (was in - Min week 1) one of these and I had the idea of writing a program to simulate the season using dynamic power ratings and then seeing if I could come up with a technique for optimizing selections beyond random iterations. The problem was I didn't have any way of knowing what optimized success because I don't know how long I'm aiming to stay alive. This changes the picking considerably as one would imagine. My league has less than 20 people (now 3 left), so maybe making it to week 7 would be good. My brother in law is in one in Chicago with thousands of people. One thing I could do would be to sim everyone else picking and see what the distributions are for various sized fields for contest duration, but that brings up the problem of using my own techniques to generate the conditions upon which I will generate techniques thereby creating an error feedback loop I'm interested in hearing how long everyone's leagues have gone in the past for various numbers of entrants.


JG

There is so little skill involved when dealing with 1000 entries that I would just take some chances early with teams you don't see using for the 1st half of the year because in this you are bound to get unlucky with a big favorite at some time. If your early season gambles pay off then you are in good shape facing a lot fewer competitors with limited options compared to you.


I had never heard of this pool until this year, but it does sound like fun. However as I say even 200 entries and it all becomes luck. Sure there is some skill, but considering one pool a year you are likely to die before your expectation comes in.