NFL Survivor Week 1

Survival pools are one of my favorite things to do during the NFL season. I compete in the simplest variety- pick one team to win straight up each week (of course the team can't be used more than once). More competitive pools force contestants to pick two teams and in some cases the team(s) must win against the spread.

For my purposes, I have a clear strategy that I've used the past two years. I compile an Excel spreadsheet of all the point spreads for each game of the season, then I find the combination of picks that fit the criteria (one team per week, only use each team once) that maximizes the number of points I have in my favor over the course of the season. This is the perfect thing to use the Excel Solver add on, but the free version won't accept this many inputs and I don't want to pay $200 for the paid version; I think I do a good enough job eyeballing it.

Here's my strategy going into week 1:


A couple of notes:

• Just pick whoever is playing the Cardinals or Panthers. 
• Every division is picked except the NFC West which is picked against. 
• It's looking more and more like Peyton Manning won't play in Week 1 vs Houston. The game has been taken off the board. Once Manning's status is clear, the Texans could be in the mix for the week 1 selection. 
• Looking at every single point spread is a good exercise. For example, the Bengals, my week 16 selection, are favored by seven points in that game against the Cardinals. I don't think the Bengals should be favored by seven points against a high school team. They'll probably win because the Cardinals won't be very good, but I except that line to be much tighter come kickoff. 

StarStreet

I first heard about StarStreet at the MIT Sports Analytics Conference in March. Some of you may remember Protrade; StarStreet is trying to do the same thing except it uses real money. When StarStreet first started trading, there was no backing to the players. It made no sense. The only way to make money was to buy a player and hope the site grew and people would want to buy popular players. It seemed like a short lived idea and within a month or two, the market had stalled because the site had stopped growing. I know of at least one person who made money though.

On August 1st, StarStreet announced a new valuation system based on player performance (here). There is still one aspect I'm a little unclear about and that's the idea of proportional value. In the actual stock market everything can go down or up independently of each other. If Apple blows out its earnings report, if will go up even if Cisco does too. For StarStreet, if Jose Bautista hits three home runs, he will only go up if the other offered players do not increase their value to the same extent.

Since the new valuation system was started with around 100 games of data, the final payouts based on relative fantasy points were already priced into the StarStreet determined IPO. The only way a player could see a substantial change in value would be an injury and since StarStreet doesn't allow short sales (yet?), there's no way to make money off that.

Nevertheless, here is a chart of the StarStreet offered players with their final projected price vs. last traded price. I used the ZiPS RoS projections available at Fangraphs.



It's not surprising to see the players (Ubaldo and Liriano) with good track records but underperforming this season be projected to return the most value. As I write this, Ubaldo is getting shelled in Detroit so I wouldn't go buying his shares just yet. As I wrote above, I wouldn't expected any drastic returns from the MLB market.

NFL Season

With the lockout over, I guess I have to get back into football. I'll have a large post about survivor pool strategy coming up later this week, but I thought I'd throw up some basic stuff. I used Sbet's lines for each game this season.

First, here is a ranking of the teams based on average point spread. This DOES NOT represent who the best team is; rather, it is a combination of team skill and strength of schedule.


Next, projected wins for each team based on the individual game spreads vs. win totals offered at Sbet .


I used a difference of two wins to determine which were favorable bets.