If you’ve been a FantasyLabs member playing daily fantasy basketball this season, your process has probably become pretty efficient by this point. You have personalized Player Models that weigh all of the factors you’ve found to be the most predictive. You use the Trends tool to identify all the right plays and fades. You can rattle off Kristaps Porzingis‘ usage differential in games Carmelo Anthony misses thanks to the NBA On/Off tool. And you realize that Isaiah Thomas‘ Defensive Real Plus/Minus score on our Matchups page is not a typo: He really is that bad defensively.
But now NBA is winding down, and it’s time to create and refine an MLB DFS process. In this article, I’ll aim to get you started in the right direction by identifying some key similarities and differences between the two sports.
Projections vs. Ratings
We’ll start with one common misconception about NBA models. Excluding general player information, the two leftmost columns in the NBA Models are labeled “Rating” and “Proj.” Rating is an aggregate value that reflects a player’s score based on the model’s settings. Do you want to find players with the most beneficial positional matchups each night? Drag the “Opponent Rating” setting to the far right, and you’ll see a bunch of players who are facing the Nets and/or Suns leap to the top of the model. Want to find the most consistent players? Increase the “Consistency” setting, and the players at the top will be replaced by a new set of players. Rating will “rate” players based on the criteria prioritized by the model.
One of these criteria is Median Projection, or “Proj”. This is the fantasy point total we’d expect the player to score around 70 percent of the time. (Also, 15 percent of the time we’d expect them to finish closer to the ceiling and 15 percent of the time closer to the floor.) “Proj” is sometimes mistreated as a single be-all, end-all metric, but really it is only one variable among many, all of which contribute to each player’s rating depending on the model’s settings.
There Are No Projections in MLB
If you’re an NBA player exploring the MLB Models for the first time, you may be surprised to learn that “Proj,” “Ceiling,” and “Floor” are missing entirely from the sport. Due to the volatile and somewhat binary nature of MLB scoring, we will not project players to hit, for example, 0.15 home runs in a matchup. Home runs are either hit or not hit and the massive points bonuses that come along with them are either awarded or not awarded.
So what is important in baseball? All of the same salary and Vegas variables you are used to are still available along with a library of metrics designed to find batters and pitchers in the best overall matchups and environments. Want to find pitchers likely to suppress scoring thanks to playing in a deeper ballpark? Want to fade pitchers who risk losing innings due to a weather-related delay? Check out Park Factor. Crank up the Weather slider bar. Batter splits, strikeout predictions, recent performance and luck metrics: Almost anything you could possibly want to consider is available as a variable that you can weigh.
Learn about all of our various MLB metrics in our DFS glossary.
As is the case with NBA, we find out which metrics are important in MLB by evaluating how they impact Plus/Minus, our proprietary salary-adjusted metric. Once you adjust the settings in your MLB Model and hit the “Update” button, the model will be immediately backtested. The Plus/Minus in the top area of the “Model” tab will show you how many fantasy points above or below salary-implied expectations have historically been scored by the top five percent of players based on the model’s settings. The higher the Plus/Minus is, the better.
Learn more about our MLB models with our tutorial videos.
Finding Value
In NBA DFS, players generally become value plays when their projected minutes or usage in a game exceeds the long-term average. Most of the time, this happens when a teammate is out. These situations are easily identified by following our News page, keeping an eye on projected and confirmed starting lineups via our Matchups page, and running On/Off queries to determine which players will be most affected by changes to a team’s active lineup.
In MLB, players don’t automatically become great plays when a teammate is out. In fact, sometimes, they can be hurt by it. There is much more correlation between players in MLB: Batters on base need to be driven in by their teammates to score runs. If a team’s star player is replaced by a scrub, the team may be less likely to score runs, which hurts all offensive players. In NBA, the opposite is true: If a star is out, everybody else on the team is likely to benefit from the extra usage.
Minutes & At-Bats
The closest parallel between value plays in NBA and MLB is the correlation between minutes and at-bats. In NBA, there is no substitution for time on the court. A player who typically plays bench minutes and is priced accordingly will always become a top value play when he is in a position to play starter minutes one night.
More minutes = more opportunity.
In MLB, opportunity comes from at-bats. A player batting leadoff will see the most at-bats, followed by the player batting second, and so on down the line. To find value, we want to discover players who typically bat lower in the order (or who don’t play much) and then find themselves batting in a higher spot one day.
This idea is most obvious when looking at batters priced $3,000 and lower on DraftKings at each spot in the order. There is a difference of close to two fantasy points per game between the leadoff position and ninth spot:
Ownership
As FantasyLabs Co-Founder Jonathan Bales has noted in his most recent book, How to Use Math and Psychology to Win at DFS, one of the main differences between MLB and NBA is the event-driven nature of MLB, which leads to a higher degree of variance. NBA is generally easier to predict than MLB because we can follow the minutes and usage to the most valuable plays. Anthony Davis notwithstanding, more often than not the ‘chalk’ plays in NBA perform well. In tournaments, you can still play chalk while leveraging one or two contrarian plays to differentiate your lineup.
In MLB, one ball can be hit 308 feet for zero fantasy points and another can be hit 308 feet for 14 fantasy points. The event-driven nature of the sport leads to huge and somewhat unpredictable scoring swings on a day-to-day basis. Because there is so much variance involved in baseball, it often makes sense to take better ‘odds’ on lower-owned stacks and players.
For example, let’s say the Cubs and Red Sox are both playing, and the Cubs’ middle-of-the-order is the chalk. Let’s say we expect the Cubs power bats to be 30-35 percent owned and the Red Sox power bats only 10-12 percent. Here, you are ‘getting odds’ that the Red Sox you roster will outscore the Cubs rostered by many others. In NBA, it’s very unlikely that someone like Russell Westbrook will be a complete dud as a chalk play. Sure, he might score 52 rather than 82 points, but he’s probably not going to ruin your lineup completely. In MLB, even the best players regularly go 1-4 or 0-4. Taking shots on power potential with an ownership discount in large tournaments makes a lot of sense.
MLB Ownership Projections & Dashboard
This season, for the first time we are projecting MLB ownership. If you’re coming from NBA, you’re already familiar with the ‘Proj Own’ column in Player Models. As your database grows with results, you’ll be able to utilize our Trends tool to find out how chalky and contrarian stacks have historically fared: When Cubs batters are projected to score more than 4.5 runs with an ownership projection of 20 percent or more, how have they performed? When Mookie Betts is projected for less than 10 percent ownership, how often has he beaten his implied value threshold?
Also new to MLB this year is another concept from our NBA product: the DFS Ownership Dashboard. This tool is arguably even more powerful in MLB than in NBA for one key reason: Late swap. Want to see at a glance what percentage of DFS players are on each pitcher and batter so you can adjust your exposure to players in late-starting games? This is the place to do that, and the data is always available a few minutes after the main slate locks.
All of the same benefits of the NBA Ownership Dashboard still apply: You can match your own plays against the “GPP Grade” column, which grades players based on high- and low-stakes ownership differential:
Typically, you want your exposure to align more closely with the players who have high GPP Grades — those with the highest stakes-based ownership differentials.
Conclusion
There are many other differences between the two sports: For example, in one you are rewarded for causing the ball to leave the playing field, while in the other you could be ejected for doing that. This article doesn’t even explore ballpark or weather factors — two key drivers of MLB value not present in NBA. These, and many other factors, you should research for yourself with our MLB Trends tool.
