Continuing my series of trying to figure out which team is best to pick for survival football and then ignoring it, I present my week 3 analysis. I used the same method as the past 2 weeks, and didn't make any updates to it since last week.
Here we go:
NE (53%), BAL (20%), WAS (8%), and MIN (7%) are the most common teams picked in the Yahoo! leagues. They are also the top four teams according to my metric in a different order.
This week has NE favored by the most that they will be favored by for the rest of the season and by quite a bit so it makes sense that they are ranked at the top. Maybe the biggest difference between this analysis and the Yahoo! distribution is that Baltimore has 20% of Yahoo! but is just fourth best here. I would have to say that my numbers make more sense because Baltimore has three more games with a spread of 10 or higher and couple more with a spread higher than 6. I am surprised that Washington is up on both lists because they are only favored by 4 and playing an away game. Although they only have one more playable game and that is a 6 point favorite.
I would definitely pick New England this week if perfection is your goal.

So this is late, but I already did the analysis and I wanted to share my results for posterity. I used the same method as last time to try to evaluate who should be picked in a survival football pick 'em. This method basically only tries to figure out which teams, or combinations of teams, give you the best chance of getting all of your picks correct for the entire year. Obviously you want to pick a team this week that has the best chance of winning. But you do not want to pick a team that also has a lot of future value, where they will be favored by quite a bit in their remaining games. To reiterate, the way I did this is that I randomly picked a teams for the remainder of the season. I then used the Las Vegas point spreads of all the games to give determine the probability of winning each game and the rest of the games. Finally, I compared the average probability of winning every game when picking team X next week to the average probability of winning every game when picking a random team next week. I express this as a ratio - the higher, the better.

There are a couple of differences to how I did this last week.

• I modified the sampling of teams so that it does not choose teams on a bye week. This saved a lot of wasted simulations, and helps to make the results more stable. Therefore, I only simulated 100,000 seasons worth of picks compared to 1,000,000 last week.
• I omitted the team I picked last week from consideration of being picked for the rest of the season. Last week I picked Chicago, so they are not included. 
• The bar charts on the graphic start at zero. I committed a cardinal sin of graphing last week because of Excel's defaults.

Here is what I came up with for Week 2:
What this is saying is that you are about 48% more likely to go 16 - 0 for the rest of the year if you choose Green Bay than if you pick a random team.
One aspect this method does not address is that your goal is not always to have a perfect season. If the league is small enough, you just want to outlast your opponents. My league only had about 10 people to start. If you looked at Yahoo!'s dashboard, you'd see that 56% of people chose GB this week. If they go down and you didn't pick them, your chances of winning increase dramatically. That is why I didn't choose GB and why I lost.
I chose DAL, because the spread was high, the percent taking them was low, and they had less future value. It was almost entirely on the recommendation of Vegas Watch. Maybe I should have looked at my ranking and picked Oakland, Atlanta or Cleveland since there was even a lower percentage picking those teams and a higher ranking than Dallas. With hindsight, we now know that Green Bay, Oakland, and Atlanta won and Cleveland and Dallas lost. But there was a strategy to it because 4 of the 9 teams that picked took Green Bay. If they lost and my team won, I would have had a much better chance of winning.
I will brainstorm about a way to factor the pick distribution into the rankings, but I do not think it is possible the way it is currently set up.

The NFL season is starting tomorrow night and I am in a survival league this year. If you are not familiar, in a survival league, each week you pick one team to win their game. Once you pick a team, you can no longer pick them for the rest of the season. So you have to balance pick who you think has the best odds of winning this week and what the rest of the team's schedule looks like.

There are some sites online that are very helpful for deciding which team to pick this week. One of them is http://www.survivorgrid.com/. I was wondering if there was a way to use this information to automatically decide which team to pick. This is an optimization problem with a lot of moving pieces, so simulation seemed like a natural choice.

I downloaded the spreads for each game from that website (using some expert Excel skills). From the lines, I was able to discern the probability of winning the game from this paper. (I found out about this method from the book Mathletics.) Basically, I just assume that margin of victory for the game is normally distributed with a mean equal to the spread (meaning the bookies are on average right) and a standard deviation equal to 13.86 points (as found in that paper). I finally calculate the probability that the margin of victory (>0) is in the team's favor. There are obviously a lot of assumptions here (I doubt the spread is normal plus we are dealing with discrete amounts), but at least this gives us a ballpark.

So now I have the probability of each team winning each of their remaining games. Next, I simulated. If I wanted to spend more time, I would have used a genetic algorithm or some other AI to try to find the best combination of picks. But I decided to go brute force because I didn't want to spend a lot of time. Here was my process:

• I told R to randomly sample which teams to pick each week (SD for week 1, IND for week 2, etc.)
• Using the probabilities I created from the spreads, I calculated the probability that this combination of picks goes undefeated the whole season. The probability of going undefeated is the product of winning all of the individual games.
• I wish there was an easy way, but I had to waste a lot of trials here. Whenever a team had bye and the random sample picked that team on that week, they obviously have a 0% chance of advancing. I wanted to prevent the random sample from selecting teams on a bye but it became a little hairy so I gave up.
• Do this 1,000,000 times
• Analyze the picks with the highest probability of going undefeated.

 Here are the 20 best simulations I came up with. You can see that the even the best have pretty low odds of going the distance.

Finally, I took the average probability of winning it all when each team was selected for the first game. I then compared that to the average overall probability of winning it all. This should give you a good idea of what team to select this week.


This is probably the best way to use this data rather than looking at the best combination of picks since the lines for the rest of the year will see a significant change as they get closer. This also agrees with Vegas Watch's analysis, which I like to see.