Monday, July 6, 2015

Home Runs or Small Ball? Part 2 – Moneyball and the Small Ball Approach


People who have seen the Brad Pitt movie Moneyball might remember Billy Beane discussing on-base-percentage (OBP) as a key indicator of a team’s ability to score runs.  This is part of the small ball strategy - a lower variance, more reliable style of play.   
 
In Head-to-Head play, a small ball strategy is the correct approach.

Head-to-Head – A contest where two contestants pick lineups and the highest fantasy score wins the prize.  

·        You Need to Cash Most of the Time - Unlike tournament play where maximizing your score and your variance is the best strategy, (see our post on tournament play here) having a consistently good score and winning more often will lead to long-term gains in head-to-head contests.  Let’s look more in depth with this example:


  • $5 MLB Head-to-Head contest (https://www.draftkings.com/contest-lobby)
    • The winner gets $9
    • DraftKings takes $1 in rake (their fee for running the contest)
    • The loser gets $0
    • If you want to be a winner in the long run, you need to win more than 55.5% of the time. By setting your expected value (EV) equal to zero, you can solve for the breakeven win percentage in these contests.
      • EV = $0 = 55.5%*($9 - $5) + 45.5%*($0 - $5)
·        Pick Players with a High Floor – In tournaments, where we’re screening for players with high ceilings who will differentiate our lineups from the field, we’re not so concerned with the floors of the player. By floor, we are referring to a player’s worst performances on a given night. 
  • In head-to-head games, we’re only up against one player.  We need to beat him to succeed.  In order to accomplish this, taking a lower variance route is a superior approach.  Draft players with low downside risk and who have low deviations in their performance.  You need consistently good fantasy scores every night.
    • For example, let’s create two hypothetical fantasy NBA players who have the same average fantasy points (let’s say 25) but player A has a standard deviation of 15 points while player B has a standard deviation of 10 points. I.e. Player B is much more consistent than player A. Let’s also assume they have the same salary for a given night.
    • Intuitively, you might think it’s a no brainer to pick player B for all types of contests. You might be thinking “consistency = good.” If you were running a pro team in real life, this in fact would be true. The ability to bank on results from a player is a huge benefit to a real life team. But a daily fantasy tournament is completely different type of game. You are looking for players who are more prone to feast or famine nights, and you hope for a feast on the nights when you draft them. If enough of your high variance players have good nights, you will have a confluence of outsize fantasy scores and have a chance to place very highly. If all your players just score close to their averages, you’re not going to cash anyway. If they have off nights, that’s ok, since you are playing tournaments repeatedly and don’t expect to cash most of the time.
    • Perhaps we can find an analogy in dating. If you are trying to win over a really fantastic girl with many, many suitors, doesn’t it make sense to try something a little unusual and perhaps a little more daring? Probably. (Please don’t take this to mean you shouldn’t be yourself!) 
·        Budget – Bankroll management (a topic we will discuss over and over again) is a huge factor in how to approach any daily fantasy game.  Because of the nature of head-to-head play, you will cash more often, allowing you to play higher stakes than your tournament counterparts.  Ensure you track your results so you know how well you are doing.

These are just a couple tips on how to play daily fantasy head-to-head games.  

If you have any other daily fantasy strategy suggestions, post them below or tweet me @rkalwaro on Twitter.  If you’d like to play some tournaments on DraftKings – use https://www.draftkings.com/r/TheConductor44 to sign up!
Always Be Innovating,
Rob

No comments:

Post a Comment