Oct 9, 2018
Oct 9, 2018

How do you make money from MLB predictions?

What are the challenges with MLB predictions?

Why data can be a help and a hindrance

What to do with your MLB predictions

How do you make money from MLB predictions?

There are countless methods for making MLB predictions that will lead to varying degrees of success. If you want to make money from betting on baseball you need to be able to make consistently accurate MLB predictions, but how do you do it? Read on for some expert insight into making MLB predictions.

What are the challenges with MLB predictions?

Making money from betting is hard, a lot harder than people think. While baseball is a completely different sport to soccer, basketball or tennis, it’s just as difficult to beat the bookmaker and make a consistent profit from betting on it. Each of the aforementioned sports (and many more) have differences that present their own unique challenges, but there are also some similarities between them.

If you’re betting on baseball with a bookmaker you are not only pitting your knowledge, model or strategy against that bookmaker, you are up against everyone else that is betting with them. A bookmaker like Pinnacle doesn’t have to make MLB predictions, they simply use the information provided by all the other bettors in that market to make the challenge of finding value a lot more difficult.

The large sample size (there are 2,430 games in the MLB regular season), the countless data providers and the format of the game are all very good reasons to bet on baseball. However, these are also reasons that will encourage a lot more people to bet on the sport - the more crowded the betting market gets, the sharper the bookmaker’s odds will become.

Additionally, baseball is quite unique in the sense that the way the game is played is very dynamic. Tactical shifts and approaches are commonplace in the MLB - the return of Lou Boudreau’s defensive shifts from 1946 in 2014 being one recent example of note. The unpredictability in how the game is played and what aspects of the game a team places value on mean a successful betting strategy can quickly become a redundant one.

MLB predictions: Why data can be a help and a hindrance

As mentioned above, there is an endless ream of data available in baseball - it is probably the most data heavy sport out there. The use of data has became more prevalent after the emergence of sabermetrics with teams being the first to benefit (Billy Beane’s Oakland A’s perhaps the most famous example). Now, the use of data extends to the betting market, fantasy sports and even the average fan.

Regardless of how accurate your MLB predictions are, if you don’t look after your money properly you’ll soon find that you don’t have any left to bet with.

More advanced metrics such as ERA+, FIP, WAR, wOBA and wRC+ (there are many, many more) are now regarded as a basic component of making MLB predictions. However, it is important to note that there is a distinction to be made between using data and using data effectively.

Once you have collected data, the next step is analysing it and building a predictive model. There are so many data inputs and model functions that can be used to make MLB predictions. Run scoring regression analysis could be used to find an edge on Totals betting, a ranking model with all manner of inputs can be used for Money Line matchups, individual team analysis can be used for season win totals projections and so on.

Making considerations for the unique aspects of baseball when using data to make MLB predictions is a crucial part of the process. Failure to account for things like the strength of schedule, the weather, where a game is played and at what point in the season it is played can have a major impact on your predictions and the results you achieve when betting on them.

Despite all the benefits that using data might bring, it’s important to remember that while a pattern may exist or an apparent edge might be found, only after thorough testing will you be able to distinguish whether these findings are legitimate and will hold up against the betting market. This can be summarised by the words of Mark Twain; “It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so.”

There are plenty of useful resources available to help guard against making the common mistakes that come with data analysis and modelling. Nassim Taleb’s book Fooled by Randomness provides a good explanation of (amongst other things) why “the more data we have the more likely we are to drown in it” and while this is true for most sports, it is especially applicable to baseball given the wealth of available data.

Joseph Buchdahl has written at length about the science of probability and uncertainty in Squares & Sharps, Suckers & Sharks and also has plenty of articles on Pinnacle’s Betting Resources covering things like the difference between correlation and causation. Those after literature more specific to baseball data analysis can look to Jim Albert and Jay Bennett’s Curve Ball (a collection of essays on baseball statistics).

What to do once you have your MLB predictions

Producing your own MLB predictions is all well and good but if you want to make money off of them, you need to able to invest in them. A bookmaker will provide odds on an event occurring (which over time will adjust slightly as new information enters the market). If your predictions are consistently more accurate than the one provided the bookmaker, you will be able to make a profit from betting on them.

There are three simple steps to follow if you want start betting on your MLB predictions. Firstly, compare your predictions against the market. The simplest way to do this, depending on what you are trying to predict, is to convert your predictions into percentage chance and compare them against the bookmaker’s odds (in percentage chance) - this will tell you whether you should bet or not.

Failure to account for things like the strength of schedule, the weather, where a game is played and at what point in the season it is played can have a major impact on your predictions.

If after this first step you determine that you should place a bet on your prediction, you need to calculate how much money you should bet - using a staking method to optimise how much money you place on each bet. Finally, you need to measure the success of your approach by analysing results (and this doesn’t just mean over the course of 30, 40 or 50 bets).

Using our examples above, you might have a power rankings model that suggests the Chicago Cubs have a 66% chance of beating the St. Louis Cardinals (after all external factors had been considered). If the bookmaker posts odds of 2.33 (42.92%) for the Cardinals and price the Cubs at 1.694 (59.03%) this would provide you with betting opportunity that offers expected value (you think the Cardinals have a better chance of winning than what they are being given). 

Once you have a means of finding an edge (expected value), using a staking method such as the Kelly Criterion will help you bet amounts relative to that edge and ensure that you don’t exhaust your bankroll. Regardless of how accurate your MLB predictions are, if you don’t look after your money properly you’ll soon find that you don’t have any left to bet with.

In addition to refining your approach, analysing the results is arguably the most important part of the process when betting on your MLB predictions. The question of luck vs. skill in betting is something all bettors should be aware of, but one many fail to recognise.

Whether it’s comparing your bets against the closing line (Pinnacle’s is the most efficient available), using the t-test or Bayesian analysis, ensuring your results are down to your skill (method of prediction) rather than luck (randomness) is essential if you intend to succeed in the long term.

Betting Resources - Empowering your betting

Pinnacle’s Betting Resources is one of the most comprehensive collections of expert betting advice anywhere online. Catering to all experience levels our aim is simply to empower bettors to become more knowledgeable.