Our mission here is to shed some light on why taking a player-based focus when handicapping baseball will lead to increased profit with MLB picks.
Using a Player Based Approach to MLB Handicapping
Advanced baseball statistics have been around for quite some time with the advent of Moneyball and Sabermetrics. This analytics revolution in the MLB put a focus on the availability of data. And with so much MLB data available, the handicapping of an MLB game can be completed using various methods.
Baseball is by far my favorite sport to handicap. Unlike the other major sports, baseball is the only sport with a “start and stop” nature. Additionally, the game is played with several situations with a finite set of outcomes. Because of these situations and the limited number of outcomes, baseball is a game that lends itself to certain types of betting models. With that said, my focus on baseball handicapping is always surrounding the batter/pitcher matchups.
In my approach to MLB, teams are broken down to their true components: the batters/pitchers that are expected to play and their corresponding estimates used to predict their current level of ability. And because of the situational factors that I previously mentioned, my approach will be focused on adjusting these player estimates accordingly.
Tony Larussa was infamous for paying extra special attention to the batter/pitcher matchups. He would not hesitate plugging in a utility player into his starting lineup based off a success against an opposing pitcher. Some hitters tag certain pitchers, and some pitchers dominate certain hitters. The reasons could be swing style, pitch trajectory, or specific tendencies.
While I don’t think many hitters have a statistically significant sample size of matchups against specific pitchers in their careers, I do strongly believe that a game of baseball can be broken down to the series of matchups between each team. More specifically, my MLB handicapping methods aim to dissect each projected matchup between a starting lineup and the opposing starting pitcher. How does the handedness of the pitcher affect each hitter’s projected worth? Does the ballpark affect performance? What about umpires, weather, and fielding?
The Log5 method was a formula created by Bill James, one of the most recognized names from Sabermetrics. The original intention was to predict a team’s probability of winning a game. Additionally, it can be used to project the performance and outcome of a batter/pitcher matchup.
Handicapping the MLB with Player Stat Projections
You'll notice a common theme when I reveal the details of my MLB handicapping model. My predictions are primarily driven by projected player contributions and NOT by team statistics. This should make sense given the fact that team statistics can be useless when a team's most impactful players catch the flu. What about when a manager switches their lineup? When a manager does this frequently, the team statistics will be misleading since they don't account for this. Like I previously discussed, a game of baseball can be broken down to the series of batter/pitcher matchups…two players. By knowing and understanding who is available to play and how much they can contribute can lead to more accurate MLB picks.
The purpose of developing prediction models is to obviously predict a future event. With that said, my approach is to predict team statistics using projected player statistics. Can the past help predict the future? Yes, to a certain extent. But figuring when current/recent statistics become predictive can be its own challenge. Besides, the early parts of the season present the best opportunities to wager in terms of value due to this very fact. MLB odds makers don't have access to a lot of information. They don't know for sure how a particular team is going to be. They rely on as much information as they can, but they rely heavily on last year statistics, offseason changes, etc.
On another note regarding statistics, certain stats may mislead bettors as it pertains to a player/team’s predictive value. The outcomes of sporting events over a small sample size can be full of variance due to luck, randomness, etc. In other words, some teams can perform at a higher level than they are actually worth....and vice versa. This can present valuable wagering opportunities if you have previously made projections that indicate otherwise. For a recent example, the Pittsburgh Pirates started out with a pretty poor record in 2014, but my projections/data were telling me they were a better team than their early performance. As we know, the Pirates went on a successful run late in the season and made the playoffs.
Good luck with you handicapping!