Suicide pools.

For those who don't know, a suicide pool is a last-man-standing contest, generally based on the NFL, although any sports league will do. Here are the fundamental rules.

1. Every player pays the same amount to play.
2. Every week every player must predict the winner of one game that week. All games are scored straight up, not against the spread.
3. If a player is ever wrong he is immediately eliminated from the contest.
4. No player may choose the same team more than once.
5. The last man standing wins the pot.

There may be additional rules, like the person running the pool getting a cut, and splitting the pot once it gets down to a few people, if and only if by mutual agreement.

Before going further, here are the results from last season from officefootballpool.com.

Week	Team	Number	Spread
1	Phil	54	-10.5
1	Chic	29	-7
1	Pit	14	-6.5
1	Det	6	-6
1	KC	5	-3
1	NYJ	5	-5.5
1	NE	1	-4
1	SF	1	-4.5
1	StL	1	-3.5
2	Den	38	-12.5
2	GB	23	-8
2	NO	7	-6.5
2	SF	5	-7
2	TB	3	-5.5
2	Wash	2	-6
2	Cin	1	-5.5
2	NE	1	-5.5
3	NE	44	-14
3	NO	16	-10
3	Atl	3	-6.5
3	Cin	1	-6.5
3	Ind	1	-6
4	SD	34	-13
4	Ind	20	-7.5
4	Pit	10	-7.5
4	Bal	1	-3
5	NO	21	-10
5	GB	13	-8
5	Den	6	-7.5
5	Det	5	-6.5
5	Phil	3	-6.5
5	Dal	2	-6.5
5	Pit	2	-6
5	SD	1	-6.5
5	Sea	1	-7
5	SF	1	-5
6	Sea	21	-8.5
6	Den	13	-9.5
6	SD	7	-7
6	Cin	5	-6.5
6	Ten	2	-4
6	Atl	1	-3
6	SF	1	-3
7	NE	9	-9.5
7	Sea	9	-6.5
7	GB	3	-6.5
7	Bal	2	-6.5
7	Buf	2	-5.5
7	Wash	2	-5
7	Dal	1	-6.5
8	Dal	12	-9.5
8	KC	5	-7
8	Clev	1	-7
8	Mia	1	-6
9	SF	3	-10
9	Cin	2	-10.5
9	Sea	2	-14.5
10	Bal	2	-9.5
10	Cin	1	-6
10	Sea	1	-9
11	GB	1	-5.5
11	Mia	1	-6
11	NO	1	-7
12	Ind	2	-14
13	Bal	1	-6.5
13	Det	1	-7
14	NO	1	-10

The "number" column refers to the number of players who picked that team.

Note that not once did anybody ever pick an underdog. If the 497 picks made, the average point spread was 9.3.

I do not know why they recorded a pick in week 14, because there was only one player left at that point.

The preamble out of the way, I open the thread up to discussion of suicide pool strategy. I'll start with my own basic strategy. These are all just general guidelines and not hard and fast rules you must follow every week.

1. Remember that your enemy is the other players. Don't focus too much on handicapping the NFL but on beating your competition.
2. Don't pick the biggest favorite. That is what most other people will do. Hopefully that big favorite will lose and the field will be significantly thinned down. You don't want to be a survivor rather than a victim when a huge favorite loses.
3. Don't forget the rule that you can't pick the same team twice. That said, you can't always pick great teams. At times you should pick average teams playing awful teams, to preserve the easy wins on the best teams.
4. Don't bother handicapping. Just use the point spread as an indicator of your changes to win any given pick.
5. If a week has exactly two huge favorites, I would avoid both of them. Remember that you're trying to balance picking games likely to win while not following the masses. As a very general rule, I think picking the third highest spread is a good idea, especially if you're not wasting one of the best teams.

What do you think?

The question for the poll is, had this been week 1 this season (2014), which team would you have picked? The spread on the game is indicated in parenthesis.

I've participated in these types of contests.
Actually won it one year, 10 buck entry, about 45 players, winner take all, won about 450 bucks.
I pick the 2nd favorite, PIT, to get a little separation from other players picking overall favorite.
Do that 1st few weeks picking 2nd or 3rd favorite, then pick overall favorite if I haven't picked that team before for the rest of the season.

It depends on how many people are in the pool. I won mine this year (I picked the Steelers week 1) that had 17 people in it.

My typical strategy is to pick the 2nd highest home favorite. I know you're playing against everyone, but some weeks if I'm nervous about the 2nd or 3rd biggest favorite, I'll just pick the biggest favorite so I can survive.

If a team is a 10 point favorite, they are historically a lock. Early in the season, I play very conservatively, choosing big favorites, or whoever is playing the weakest teams (which is often the same pool). Barring an early upset, the field thins until you run out of eligible opponents for the Jaguars or Raiders.

I am not that big an advocate for "saving" a good team. It won't do you any good if you are eliminated before you can use it due to a losing pick on a closer matchup.

One thing to consider is how the winner is determined when multiple contestants are all eliminated in the same week. An even division of the pot could make a more chancy pick worthwhile if you are still in contention when there are only 10 or so teams left.

Thanks for the comments. Any others?

Quote: Wizard

Thanks for the comments. Any others?

Need a lot of luck.
The year I won, won 1 game due to a miracle comeback.
Won another game due to a blown refs call at the end.
My last pick that won it, last game of season I think, got incredibly lucky. Had to go with an underdog, Vikings with a rookie quarterback against Vick and Philly the year Vick did very well. Vikings won.
2010 season, came in 1st out of 64. Just looked it up :-)

The ones I play are not blind. Everybody can see everybody elses picks. Clearly in this format it is bet to pick last.

Quote: DRich

The ones I play are not blind. Everybody can see everybody elses picks. Clearly in this format it is bet to pick last.

The ones I played were blind.
We used Yahoo fantasy survivor.
Don't know everybody's pick till kickoff Sunday.
Yahoo does give stat info on overall percentages of teams picked, but those numbers are compiled from all leagues.
I recommend using Yahoo for this, very easy convenient system.

1) I agree most players will indeed pick the top top favored team each week. That's usually how it goes.

However, depending upon how many opponents you are up against in your pool and THEIR overall picking strategy, you certainly might wish to at least consider picking the top favored team, and not avoiding it completely, at least on some weeks.

The idea is, if many of your opponents are also utilizing the same strategy of NOT picking the top team, it would then certainly be okay for you to consider doing so.

Consider Week 2 for example. Denver was the top favored team (-12.5) and 38 players picked this team. But notice MORE than 38 players (42 in this case) picked a team other than Denver. (Of course, some of these players may have been forced to pick another team, if they picked Denver the week before.)

Week's 5, 6, and 7 also all saw more players combined NOT picking the top team, than the total number of people who did pick that top team. (And Week 4 was close... 34 vs 31)

(So my point is, the strategy of "not picking the top team" shouldn't be a hard and fast rule.)


2) "If a week has exactly two huge favorites, I would avoid both of them."

I would be very careful with this rule too, and also not have it set in stone. If that week all of the other games are much harder to handicap (for example if all of them were essentially pick 'em games), I think it would be okay to avoid these other games and just go ahead and pick one of those two top favorites, just be assured you're surviving another week. (Let someone ELSE use that strategy of not picking either top team, and let them gamble with another game.)


3) I also agree with Ayecarumba of not saving a good team for the reason he mentioned. But in mid season, when you're down to just a few other players, it can't hurt to look ahead at the next week or two's schedule, and attempt to determine ahead of time who you might wish to pick for those upcoming weeks.

For example, assume you have used the Ravens nor the Packers yet, and you're considering picking either of them this week. If the Ravens have two tough games coming up in the next two weeks, and the Packers do not, it should be obvious you almost certainly want to pick the Ravens this week and the Packers next week.



Has anyone ever done a computer simulation to see what type of strategy is best? I don't think it would be that hard to program at all. As far as determining a random winner for each game, we would just use your wonderful "Probability of Winning in the NFL by Point Spread" chart. If the spread is 7 points for example, that team has a 74% chance to win and the simulation would give a win to the underdog 26% of the time.

Just pick the team that you feel has the greatest chance to win each week, without worrying about "using up" that team, or what anyone else is doing.

If you pick the greatest favorite in any given week, you gain by other people picking lesser favorites. Obviously if everyone were playing optimally, everyone would pick the same team each week.

In real life, people don't do that. So just go with the biggest chance to win every week.

About "using up" teams -- the NFL is a fast-changing league. Injuries, unexpected contenders, unexpected bad teams, etc. Your goal is to maximize the information you have when it's at the highest value. That's why you simply need to pick your best option each week.

Quote: sodawater

Just pick the team that you feel has the greatest chance to win each week, without worrying about "using up" that team, or what anyone else is doing.

If you pick the greatest favorite in any given week, you gain by other people picking lesser favorites. Obviously if everyone were playing optimally, everyone would pick the same team each week.

In real life, people don't do that. So just go with the biggest chance to win every week.

About "using up" teams -- the NFL is a fast-changing league. Injuries, unexpected contenders, unexpected bad teams, etc. Your goal is to maximize the information you have when it's at the highest value. That's why you simply need to pick your best option each week.

The classic problem with following the crowd, is that you will all drown if you stay on the Titanic with them when an upset occurs. I have seen more than half the players eliminated when the biggest favorite was upset. Part of the strategy is knowing when to get in a lifeboat. Icebergs are everywhere...

Quote: Ayecarumba

The classic problem with following the crowd, is that you will all drown if you stay on the Titanic with them when an upset occurs. I have seen more than half the players eliminated when the biggest favorite was upset. Part of the strategy is knowing when to get in a lifeboat. Icebergs are everywhere...

That's all well and good with regard to your narrative. But the biggest favorite is still the best play. I could just as easily frame the biggest favorite as the "lifeboat," and everyone else who picks lesser favorites in hopes of an upset as going down with the ship.

There's more value in picking the biggest favorite and surviving while smaller groups lose than being on the other side and hoping for a big upset.

My position remains the same is that you should save the best teams for when they play awful teams plus wait until at least week 6. To win a big suicide pool you'll have to take some risks at some point and hope for some big upsets to take down large percentages of your competition. The reason for delaying is that later in the season your opponents will have already wasted the top teams so you can coast through some weeks while they have to take chances on weaker teams.

Wizard, your approach assumes that the consensus top teams (and bad teams) will remain the same up to two months later. Information in the NFL expires very quickly.

I think from a game theory point of view, the correct move would be to maximize the current week -- assuming like I do that predicting the future in the NFL is tough to do.

Coming somebody that won one once, 1 out of 64.
Have to admit was incredibly lucky.
Also have to admit that all the subsequent ones I played in, I went out early lol
But anyway.
One of my strategy is separation from the pack.
I'm in it to win it.
I'm either going to go out early or the pack goes out early due to an upset.

Quote: Wizard

My position remains the same is that you should save the best teams for when they play awful teams plus wait until at least week 6. To win a big suicide pool you'll have to take some risks at some point and hope for some big upsets to take down large percentages of your competition. The reason for delaying is that later in the season your opponents will have already wasted the top teams so you can coast through some weeks while they have to take chances on weaker teams.

In week one of this past season, the top five most popular picks on Yahoo were:

PHI (vs. JAX) 48.25%
CHI (vs. BUF) 20.89
PIT (vs. CLE) 8.70
NYJ (vs. OAK) 4.84
KC (vs. TEN) 3.97

Which would you choose? Going with the crowd, the Eagles were the obvious choice. Fortunately, they won, but the Bears and Chiefs lost, taking out almost a quarter of the field in the first round.

Granted, early picks are dicey, as the pre-season doesn't really reveal much nowadays.

How about week 2?

There were two popular picks, and a third within striking distance:

GB (vs. NYJ) 28.92%
DEN (vs. KC) 23.83
NO (vs. CLE) 19.11
TB (vs. STL) 6.78
SF (vs. CHI) 4.92

Since there are two leaders that are pretty close, and a third that was pretty close to them, avoiding them would be recommended. Unfortunately, the next two in popularity (TB and SF) both lost taking out 11.7% contestants who survived week 1. It should be noted that the third most popular pick, the Saints, also lost.

If you were using a "Biggest Bandwagon", or "Whoever is playing Jacksonville" strategy, you would at least still be in the fight... and it is only the second week.

Week 7 is where the iceberg wins. SEA was a huge favorite over STL, 29.07 of the remaining player picked them to win. The next most popular pick was CLE over JAX (the, "Whoever is Playing Jacksonville" party) at 16.66%. Surprisingly, both of these teams lost, so a blended strategy would have been best, but honestly, I don't see how anyone could have seen the Seahawks upset coming.