Daily Fantasy Strategy — March 31 — For Draftstreet

Welcome back, degenerates.

We covered a lot of ground last year in our first year covering daily fantasy strategy. We covered everything from big stuff like how big of an impact weather has on run scoring to more minutiae-like topics such as identifying hitters who hit specific pitches well with matchups against pitchers who struggle with that particular pitch. But one thing I don’t believe we ever covered was the value of stacking multiple players from the same team in your lineup.

This offseason I read a great daily fantasy strategy book written by Jonathan Bales, Fantasy Football (and Baseball) for Smart People. It’s 120 pages of well researched, daily strategy goodness, and it only costs seven bucks. It’s highly recommended.  In the book, Bales tackled the issue of how to properly use stacks, and the results of his research might surprise you.

Before we get into Bales’ stacking strategy, let me say that I’m basically ripping all of this straight from the book. But I assume if a few of you buy his book as a result of reading this post, he’ll be cool with it.

Prior to digging into stacking data, Bales theorized that using multiple two to three player stacks from a few teams would be the optimal use of the strategy. But it turns out that’s not the optimal strategy at all. In both double up style contests and GPP contests (top-XX seeds-take-all), stacking more than two or three players from a single team is the way to go.

Below is a chart from the book showing the difference in upside depending on how many players you stack. Note that the upside threshhold in the chart is 150 points which is based on the scoring system of  a different daily fantasy site that shall not be named.

As you can see, your upside goes up by stacking more guys from the same team. This is a really useful piece of information to have when filling out a GPP lineup. You’ve got to hit it big to win a GPP, and this chart makes it plainly obvious that using big stacks makes it more likely that you’ll hit it big. The problem is that Draftstreet will only let you use four players from the same team on a given day. However, the chart below shows that four-man stacks may be the optimal number.

It’s no surprise to see that the more you stack the greater the downside. But it’s important to note that the variance in the amount of downside for stacks ranging from two to seven players is much smaller than the variance in the amount of upside. This means stacking is an obvious value play. If you play on other daily sites that allow you to use as many players from the same team as you want, stack seven.

But maybe the more interesting thing that Bales found is that there’s actually less downside to stacking four players than stacking any other number, even safer than not stacking anyone at all. A Draftstreet salary cap lineup has nine slots for hitters, so it’s probably best to use a couple of four man stacks. Remember when using four man stacks to try and stack guys who all hit next to each other in the lineup like a 1-2-3-4 or 3-4-5-6 run or something.

The Daily Five

Normally we use this space to pick five players we like for the day, but I think I’m going to use it to feature five things: two starters I like for the day, two four man stacks I like, and the best seven man stack of the day.

Starting Pitchers

Let me explain how I determine pitcher value. I start by taking the Steamer projections and breaking them down to an average game projection by dividing all stats used in Draftstreet scoring by the number of games projected. For example, Chris Sale is projected to win 13 games in 30 starts, so he’s projected to win .43 percent of his starts. A win is worth two points on Draftstreet, so Sale is projected to score .867 points per start via the win category. Once I do that math for all categories, I add them up to get a projection for an average game.

Then I adjust for the opponent. For now I’m doing this by increasing or reducing the projection by the percentage by which the opposition’s wOBA was above or below average last year. Opposition data will obviously get much better as the season wears on. Once all that’s done, I take the opposition-adjusted projection and figure out the cost per projected point. Below is a chart showing the cost per projected point for all of today’s starters. It’s probably best to pick your hitter stacks and then just plug in the pitchers with the best value that you can afford.

Best 4-Man Stacks

The chart of pitcher projections above is helpful for more than just selecting your starters when you sort by the adjusted projection. It makes sense to choose hitters from the teams facing the pitcher’s with the lowest projected point totals of the day.

Unfortunately, one of the teams with one of the best matchups is Miami who doesn’t have much of an offense. But the Giants have a good matchup against Brandon McCarthy in a good run scoring environment in Arizona. McCarthy is right-handed so go with lefty Brandon Belt, the switch-hitting Pablo Sandoval who is better from the left side and righties Buster Posey and Hunter Pence who are both comfortably above league average against same-handed pitching.

Let’s also pick on Ricky Nolasco with the White Sox lineup. Their lefties are Adam Eaton, Alejandro de Aza and Adam Dunn, and let’s go ahead and throw Jose Abreu to the fire in his first major league game. That stack will only cost you a quarter of your budget.

Best 7-Man Stack

Justin Masterson struggles mightily with left-handers, and we know Oakland is going to stack their lineup to take advantage of that. Pair their right-handed mashers, Yoenis Cespsedes and Josh Donaldson, with whatever combination of switch hitters and lefties they put in the lineup. The five most likely candidates are Coco Crisp, Brandon Moss, Jed Lowrie, Josh Reddick, and John Jaso.