Analyzing MLB Stats

Daily Fantasy Baseball Strategy: Analyzing MLB Stats - When is a Sample Big Enough?
by Ray of RotoPicks.com

In a weekly league (season long) you just "set and go". You drafted last year's MVP in the first round, and he is in your starting lineup each week. Only an injury will keep him from your starting lineup, unless you are playing in a league with only one other team, chances are you do not have a better option on your bench. It is not even a thought in your mind to check the matchups against his upcoming opponents, because it simply does not matter. For better or worse, he will be in your lineup for the week, and by the end of the season, the performance will even out.

Even though there are some yearly fantasy leagues where you can set your lineup each day, the same thought process as above likely will prevail and said MVP will be in the lineup regardless of his past performance against that day's starting pitcher. Yes the player is a stud and should only be benched when he or his team is not playing, but is there a better way?

Now we have arrived at one of the benefits of playing in a daily fantasy league. Since each day is a fresh start, when creating your roster, you should remove any preconceived biases. Look at the splits. It is that simple. Anybody can look at the leader boards and select the guys with the most home runs or the players that are performing the best at that current time, but the value lies in the splits from both a positive and negative perspective. Performances against that day's starting pitcher, whether the game is played during the day or at night, and the ballpark where the game is being played should all be weighed when making roster decisions. But you can't blindly let the numbers drive your decision. There are other factors that must be considered.

ONE OF THE ONLY DAILY FANTASY BASEBALL SITES TO OFFER ONLINE "SNAKE DRAFTS"... DRAFTDAY (USE PROMO CODE ROTOPICKS AND GET A SPECIAL BONUS!)

The most commonly used split is batter vs. pitcher and of course pitcher vs. batter. This should ultimately be a determining factor when making your roster decisions, while other factors such as the ballpark the player is playing in as well. Those factors must be analyzed, but ultimately the past history between the batter and pitcher will be the deciding factor. But, they cannot be taken at face value. Also, it might not be the best strategy to solely base your decision on past performance against the opponent. You must put it into context with what is currently happening that season. If you ask most people whether they want a hitter with a .667 average against that night's starting pitcher or a hitter with a .400 average, they would look at you strangely. Of course the .667 hitter is the better option. However, that might not be the right answer. Did you look at the sample size? Were other factors considered? The .667 batting average very well could have been accumulated over 6 at bats, with the hitter producing 4 hits while the .400 hitter had 10 hits over 40 at bats. Obviously you then feel better about the second sample size as it was accumulated over a longer period of time and likely it was naturally adjusted for other variables. It is more likely that the 40 at bats were over the course of a career, or at least a few seasons, while it's possible the 6 at bats came in only one or two games. That is not enough information to make a decision.

Some of the variables that could be hidden by a small sample size, or accounted for in a larger sample size are quite critical factors that seriously impact a player's statistical output. They could be weather, ballpark, health, fielding quality, or luck. Also that .667 average could be misleading if it came 3 years ago back when the pitcher was still developing or before the hitter started showing signs of age. The sample is simply too small that it cannot adjust for those factors, and further research must be done. It is important to know when the statistics were compiled, and how the performance of said players has changed and evolved during that period.

Based on the same logic as above, you are going to always take the pitcher with the lower ERA against a specific team. But this is a picture that is even more cloudy and difficult to analyze. It is very easy to be duped and mislead by a pitcher's performance against a specific team. Aside from the fact that you always would rather have a larger sample size, there are many variables that must be considered. Of course the quality of the pitcher's performance will play a large role. If he was terrible 5 years ago but has since gotten healthy and learned how to throw an off-speed pitch, then the splits are not as valuable. The opposite is also true for a pitcher that has aged. But if said pitcher historically had a better or worse performance against a specific team exclusive of other variables, then maybe that bares some consideration.

It is important to remember that pitchers face hitters, not teams. There is a high probability that only one or two players remain from the period in which the stats were compiled, which therefore make those statistics useless (unless you want to use it as a baseline for that specific ballpark). This means that it is more efficient to study a pitcher the same way you would a hitter, by looking at each individual split. Then the same logic as we mentioned above applies.

Obviously the more information you have, the better. But do not blindly base your decisions on splits. Context is very important. When were these statistics compiled? What has changed for both parties during that time period? What has changed since the batter and pitcher last faced off? If you keep these factors in mind, then you can use splits to make informed and successful roster decisions which will surely increase your chances of winning your daily fantasy baseball league.

Fantasy Baseball Betting

Featured Articles

MISC and FAQ