Someone asked on twitter what statistics I think are most important to consider when taking the under on baseball games. There are several important stats to consider when betting totals, and I’ll try to give a good rundown of the stats I like when considering totals in baseball.
The most important thing to consider when picking any total in the MLB is park factors. Park factors can be pretty confusing, especially when you look up the park factors on ESPN and see some parks be near the top or bottom that have no reputation of being an extreme park. Parks change year to year, and this can be for various reasons. Progressive field where the Cleveland Indians play has suddenly become a much more extreme hitters park, in part because the renovations done at the stadium have opened up a crosswind towards right field that has resulted in many more homeruns. Keeping up with these year to year changes is very hard to do, and the results are even unexpected to the teams. The other problem is that park factors are so dependent on the home team’s strengths and weaknesses. For example, Barry Bonds in his highest home run hitting seasons had huge effects on AT&T Park factors. AT&T park is well known to be hard to hit homeruns towards right field, but almost all of Bonds’ homeruns went to right field this would have greatly skewed their park factor. Generally, it’s probably smart to simply tier the parks, with Coors field, Chase field. Camden Yards, and Miller Park being the most extreme hitter’s parks, and AT&T stadium, Dodger Stadium, Petco Park, Safeco field, and Citi Park being the most extreme pitcher’s parks. The fields in between will change year to year, making it very hard to judge.
The second most important stats that I consider when taking the under is related to contact rate, usually judged by strikeout rate. Pitchers can be great and not strike out a high volume of batters, but it’s still very valuable to prevent even the opportunity to allow a batter to get on base. With strikeout rate, I usually consider walk rate. Again, a pitcher can get away with walking batters to an extent, but allowing baserunners is problematic. I can stomach a high walk rate if it’s paired with a high strikeout rate (eg Carlos Martinez), but I can’t stomach a high walk rate paired with a low strikeout rate. But if a pitcher has both a high strikeout rate, and low walk rate, that means they are limiting runs. Clayton Kershaw has good batted ball numbers, but they aren’t among the best in baseball. What separates Kershaw from the rest of the pitchers in baseball is his ability to limit walks, and strike out a lot of batters. Generally, I will also look at FIP, but you can usually get a good idea of what a pitchers FIP will look like just from knowing their walk rate and strikeout rate.
The final thing that I want to look at for starters is batted ball numbers. People often talk about BABIP as if it’s something completely out of a pitchers control, and to some degree it is. There is very low correlation of BABIP from year to year, but that doesn’t make it completely random. A pitcher who allows a lot of line drives is going to have a higher BABIP than a pitcher who allows an inordinate number of groundballs, because obviously, line drives turn into hits more often than groundballs. One example I always point to is Chris Young. Chris Young has had a long rather ordinary career as a back-end rotation. Strangely however, his ERA has outperformed his FIP every single year of his career, and his BABIP has been above the league average of .300 just one time. Clearly Young is doing something to outperform his FIP so consistently, and it’s not hard to figure out what. Young limits line drives, and hard hit balls. In his career, Young has allowed a line drive percentage of just 18.5%, much better than the league average of around 22-23%. Line drives land for hits over 2/3s of the time making them by far the best hit. He also has a very low hard hit ball rate. Hard hit balls are also very important to keep low, because they often turn into homeruns. Young has a career hard hit ball rate of 16.6%, which once again is much better than the league average that is around 20%. Fangraphs has created a stat SIERA that attempts to take batted ball data into account. It’s a fairly useful stat, but it does not have enough history of success to count on its success quite yet, but in a few years, I would anticipate that SIERA becomes more and more effective to judge pitchers all around.
The previous game can also influence my decision making on betting totals. If in the previous game a starter gets pulled within the first three innings, the manager may be more desperate to have his starter throw as many innings as possible. One of the most important jobs of a manager is to keep his bullpen fresh, but if they were made to pitch 6 innings the previous game, less will be available to pitch the next day. This could lead the starter to be forced to stay in longer than ideal. While managers are generally good at working around this problem, if a team has a very bad week from their starting pitchers it gets much harder to ensure a fresh bullpen.
I did want to briefly touch on offense as well. In the past, there was a lot of value that could be extracted from simple lefty/righty splits, but that is becoming the case less and less. Most teams at this point have some sort of super utility player who can platoon with different positions to ensure that these splits matter less. Teams are also much more likely to platoon in positions where they lack a great player. The Indians for example platoon their right field with Chisenhall and Guyer who will start depending on what side the starter pitches from. You can usually get a good idea of how strong an offense is by simply looking at the team wRC+, or WOBA. Both of these stats look at the true value of each potential play that a hitter can produce. While offense of course affects how many runs will be scored, run prevention is usually the most important factor to consider when betting on totals in the MLB.