MLB DFS Advice: Picking Hitters on DraftKings and FanDuel

Welcome back to our DFS strategy series. I've already gone over some tips for picking pitchers in this post, and I covered a lot of the hitting side in this piece where I talked about stacking strategy.

A lot of hitting tips were already covered in that stacking piece, so give that a look after you read this one. For now, let's look at the data and see what we can learn.

First, here are the scoring systems for both of the main DFS websites.

DraftKings Scoring

FanDuel Scoring

The Averages

From here on I'll be using data from the 2019 and 2020 seasons. Here is the breakdown of where all hitter DraftKings points have come from:

The way to read this would be to say there have been 35,223 singles hit in the last two seasons, accounting for 105,669 DraftKings points, which is 24% of the total points scored. The next highest contributor to the total is home runs with a 20% share of the total points scored. To me, those are the two categories you should be focused on if you're looking to safely buy points. A steal is worth 67% more than a single, which can make it enticing to draft a bunch of steals guys day-in and day-out, but it turns out that the rarity of steals makes it a pretty unreliable source of fantasy points.

The most prolific base stealer in the league is Adalberto Mondesi, with 67 steals over the last two seasons. Despite that, only 23% of his DraftKings points scored have come from those steals. He's hit only 15 home runs in the same time frame and has earned 10% of his DraftKings points from those homers. If you're looking for secure points, it makes more sense to look for singles and homers.

The league leader in DraftKings points per plate appearance over the last two seasons has been Yordan Alvarez with 2.46 in his 378 plate appearances. The league average is 1.76. Here is how that changes based on the opposing starting pitcher salary:

You can see the general downtrend here, but it may not be as precipitous as you expected. Much like what we found in the pitching article, there is not much difference in performance against pitchers that are priced between $7,000 and $9,000. Once you get above $10,000 things start getting real tough for offenses, as those pitchers are the most dominant and go the deepest into games, making offensive production hard to find. The general rule here is that you should not view an opposing pitcher priced at $9,000 much differently than one at $7,000. It is possible that the field will favor the hitters against the $7,000 guy, which lowers the ownership on guys facing the $9,000 pitcher.

Lineup Spot

You can see that you lose an average of about 0.1 plate appearances per game as you fall one spot further down in the lineup. That opens up a bit after the four spot because those are pitchers more likely to get pinch-hit for late in the game (this data doesn't account for pinch-hitting). The sweet spots for optimal production, however, are the three and four spots as those spots average the highest points per plate appearance, probably because those are typically the home run hitters that have guys on base for them.

The spot in the lineup doesn't explain all of that though. You typically find the league's best hitters in the three-hole, so the increased ability is what explains a lot of that extra production. I went through the data and looked for hitters that have hit in different lineup spots over the last two seasons and compared their points per plate appearance rates between them. When I found guys with 100 or more plate appearances at different lineup spots, the three-hole and clean-up spot were the clear winners.

Starling Marte has seen 275 plate appearance as the number two hitter and has scored 1.87 points per PA, in his 539 PAs in the three-hole that number is at 2.21. Kyle Schwarber has led off for 258 plate appearances and has scored 1.95 PPA, and that number balloons to 2.31 in 179 PAs as a clean-up hitter. Eric Hosmer ha scored 1.58 PPA in the two spot and 1.96 in the four spot. These are just one-off examples, but the general trend remains. The best lineup spot to hit in for DFS purposes is three or four.

Salary

The name of the game is salary. So much of what you're considering when making your decision is already cooked into the player's salary, which makes the salary the obvious first place to start when making decisions.

For a bunch of different players, over the last seasons, I calculated each player's average DraftKings point output over the last five games and checked to see if their salary rises as that average does as well. Here's what it has looked like for Mookie Betts since 2019 began:

When you look at a ton of these graphs, you notice the upwards trend. Indeed, the DraftKings pricing algorithm does consider recent performance when generating the daily prices. The next question to ask is if that is justified. Do players score more DraftKings points in the game following really high or really low average point outputs over the last five games? The answer turns out to be no. Here's another graph for Mookie's last two seasons:

So on the x-axis there you have the average output of Mookie's last five games over the last two seasons, and on the y-axis, you have how many points he averaged in the next game after putting out that x-axis value the last five. If your last five games were a predictor of your next game, you would see this line be trending upwards. In the case of Mookie, it actually looks to be trending downwards. That turns out to be because he only had averaged 22 points over his last five games one time, and he put up a zero in the next game.

You cannot derive anything from just looking at one player, but I have run this on all players and it is undeniably true that a player's average fantasy points over his last five games has no predictive power on his very next game. Since we know that the last five games do weigh into the salary, that makes it advantageous to play players when they are "cold". This is because the price will be cheaper while the expected production is not.

That is really the best tip I can offer. These pricing algorithms are quite good, except for their bias towards hot and cold streaks. Taking the few hundred dollar discount on players after they have been "cold" is a great way to increase your lineup's expected points.

Salary vs. Production

What you see above is the average DraftKings points per plate appearances at each price point for hitters over the last two seasons.  You can see that the algorithm does a good job from this view. It's a steady incline as you climb up in salaries, with small dips between each $100-$300 interval. Things really start to increase at a faster rate after you reach the $5100 mark, but in general the algorithm does a good job at pricing hitters higher as their production increases.

Handedness

It is pretty common for fantasy players to prefer to play hitters when the opposing SP is of the opposite handedness. I went ahead and checked average DraftKings scores broken down by handedness, and here's how it looks:

For left-handed batters, there's a pretty big difference. Left-handed hitters average 7.15 DraftKings points per game when a southpaw is on the hill against them, and 7.46 when it's a righty. The difference is much smaller for right-handed hitters, and they've actually scored a little bit better against right-handed pitchers than lefties.

When we look at these values by points per plate appearance it looks like this:

So the gap for right-handed hitters completely disappears, but for lefties, it stays the same at a 5% decrease in production when facing a pitcher of the same handedness. This is not me saying that it's no better to play J.D. Martinez against a lefty versus a righty (he scores 2.20 DK pints per PA against lefties and 1.89 against righties), but just pointing out that at the highest level it doesn't look like it makes much difference. That's probably because we're only looking at the handedness of the starter, and typically a hitter will see at least one of their plate appearances against a bullpen pitcher (and that's been more like two on average recently as starters have thrown fewer innings).

You should definitely take a look at each hitter's career splits to see if they are substantially better against one type of pitcher than another, because just assuming that they will do better against a starter of the opposite hand isn't a good assumption to make.

Download Our Free News & Alerts Mobile App

More Fantasy Baseball Advice