Part of Major League Baseball’s draw in the last decade has been the showdowns between sluggers and top-tier pitching. Flamethrowers versus pure power, long seasons with storylines spanning months and rivalries revisited from the last days of March to frostbitten early November. Not so for the 2020 campaign.
There are already new strategies in place for lineups on the offensive side, now that both leagues have a designated hitter (for good reason, because runs win games). As the adage goes, “get ‘em on, get ‘em over, get ‘em in.” Create runs wherever possible, save runs wherever possible. This has been the strategy for eons, but when the season becomes compressed like it is, the statistical likelihood of teams putting up extraordinary results skyrockets.
On the defensive side, though, the strategy is less clear beyond looking for the fielder with the fewest errors in order to keep runs to a minimum. So how do teams know who’s got it?
Since the 70s and Bill James’ Baseball Abstracts publications, the study of sabermetrics has grown to be instrumental in every aspect of the game and used everywhere, from coaching to managing, hiring practices, team spending, drafting, trading, and more. James, along with his Society for American Baseball Research (SABR) colleagues, ushered in the modern era of “moneyball” and of finding competitive advantages using new statistical formulas in player and organisational decisions.
The basic measures for today’s game probably look familiar: batting average, earned run average, and fielding percentage. How much, how many, how often, and percentage of success. What sabermetrics can do with these base numbers is account for variables and create a uniform measure of a player’s performance independent of outside influences like teammate error, altitude, or ballpark dimensions.
For teams looking to capitalise on players’ strengths over a short season, sabermetric analysis has the power to locate specific situations to determine how a player is statistically likely to perform, based on existing data. Now more than ever, the gamble is in trying to design the roster to win a 60-game sprint rather than a 162-game marathon.
Let’s see what sabermetric tools could impact the 2020 MLB season strategy on defence.
The Ringer’s Michael Baumann gets it right: “[M]anaging a pitching staff may well be the most complicated part” of the 2020 season. Minnesota Twins pitching coach Wes Johnson says the most difficult adjustment is that “the body doesn’t know that we’re in a sixty-game season.”
Unconventional pitching strategy might come into play here since teams start with thirty roster spots (potentially up to eighteen or even nineteen pitchers on the staff). As pitchers work up to midseason form, teams can ramp up the intensity of outings as the season goes along with little fear of overextending the bullpen.
Once a rarity, strategies like openers (a reliever throwing the first few innings) and piggybacking (back-to-back short outings in the same game by starters) are now part of the conversation in almost every major league clubhouse because they’re the most direct method of putting the best pitcher on the mound in any given situation. But what makes a pitcher the right one?
Two of the most insightful sabermetric measures for a pitcher are BABIP and SIERA. Both use large volumes of data to compile, making them more precise than other statistics, and their results provide insight into intangible aspects of the game. The former is the more straightforward of the two but has implications on the style of a pitcher’s game. The latter performs well as a complete measurement of skills, taking enormous computing power to produce and demonstrating just how good a pitcher is without certain statistical advantages.
BABIP refers to batting average on balls in play: essentially, what percentage of fair territory contact results in a hit. In the context of choosing the right pitcher, there’s already conversation about those more likely to register a strikeout, a ground ball, or a fly ball. BABIP is able to drill down to the quality of each batted ball to see the quality of both the pitches and the contact. Pitchers with exceptionally low BABIP (like Justin Verlander’s .218 and Jeff Samardzjia’s .240 in 2019) are likely to induce more soft than hard contact, since so many balls in play don’t result in a hit. It also points to a higher swing and miss rate, given that batters hit fewer balls squarely.
Generally, BABIP is fairly insightful for pitchers’ ability to record outs, preserve leads, and can suggest that certain players may be better suited for situational pitching than others. In terms of a holistic evaluation of a pitcher’s abilities, it falls short.
Enter: SIERA. Using value-based statistical weighting where certain outcomes are preferred over others, SIERA (skill-interactive earned run average) can paint a clearer picture of what pitchers are capable of in all aspects of their game.
SIERA is the measure that can determine the true skill level of a pitcher and is able to relate the value, relevance, and potential bias of a given metric will be, whether it’s strikeout to walk ratio, BABIP, or home runs from fly balls. FanGraphs breaks down what SIERA tells us and why it’s a groundbreaking concept:
“Strikeouts are good…even better than [fielding independent pitching] suggests. High strikeout pitchers generate weaker contact, which means they allow fewer hits (AKA have lower BABIPs) and have lower homerun rates. The same can be said of relievers, as they enter the game for a short period of time and pitch with more intensity.
Also, high strikeout pitchers can increase their groundball rate in double play situations. Situational pitching is a skill for pitchers with dominant stuff.
Walks are bad…but not that bad if you don’t allow many of them. Walks don’t hurt low-walk pitcher nearly as much as they hurt other pitchers, since low-walk pitchers can limit further baserunners. Similarly, if a pitcher allows a large [number] of baserunners, they are more likely to allow a high percentage of those baserunners to score.
Balls in play are complicated. In general, groundballs go for hits more often than flyballs (although they don’t result in extra base hits as often). But the higher a pitcher’s groundball rate, the easier it is for their defense to turn those ground balls into outs. In other words, a pitcher with a 55% groundball rate will have a lower BABIP on grounders than a pitcher with a 45% groundball rate. And if a pitcher walks a large number of batters and also has a high groundball rate, their double-play rate will be higher as well.
As for flyballs, pitchers with a high flyball rate will have a lower Homerun Per Flyball rate than other pitchers.”
Basically, SIERA takes into account that different at-bat outcomes mean different things for different pitching styles. For example, pitchers who often play in ballparks that favour hitters are compensated for their disadvantage, while those who often play in pitchers’ ballparks receive equal adjustments in the opposite direction. Coors Field in Colorado, notorious for home runs because of altitude-thinned air, comes to mind. Ground-ball pitchers who rely more on their defence for outs and a strikeout pitcher are assessed differently for their disparity in BABIP. These sorts of equalising elements focus on pitchers’ individual contributions to various elements of play.
Coaching staffs now have the ability to project how their starters and bullpen arms will perform, but on a scale where their tendencies and situational roles are accounted for. It’s the ultimate unifier for determining where a pitcher’s skills are at, and a tool to decide who’s likely to perform.
Despite its complexity, it all boils down to a regular old earned run average number, just run through the wringer a few times. Unsurprisingly, the pitchers with the top SIERA in 2019 belonged to Gerrit Cole (2.62), Max Scherzer (2.93), and Justin Verlander (2.95).
Forget fielding percentage. The newest innovation in calculating fielding ability is found in ultimate zone rating (UZR) and defensive runs saved (DRS). The two have different expressions but very similar concept: how did a player do in the pursuit of preventing as many runs as possible?
DRS and UZR serve as excellent measures of which defensive players will make extraordinary plays in the field, and which of them can even attempt extraordinary plays. Both are measured as a number relative to zero, with zero representing the average replacement player.
Angels shortstop Andrelton Simmons holds the current record for most runs saved in a season, at a DRS of +40 in 2017. The all-time worst player in the DRS category is, shockingly, Hall of Fame shortstop Derek Jeter, who recorded a -162 DRS between 2003 and his retirement in 2014.
The difference between DRS and fielding percentage is that DRS measures the range of a defender in addition to rewarding or penalising them for the outcome. Fielding percentage is whether or not a player made an error on a putout or assist opportunity, but DRS accounts for how difficult the play is to make based on the statistical likelihood of success and adds or retracts points accordingly.
Let’s say there’s a 20% chance that Andrelton Simmons can make the out at first on a sharp grounder deep in the hole at short. Fielding percentage takes the success/failure as binary, either 100% or 0% with nothing in between. If Simmons can’t get to the ball, it’s neither a putout nor an assist opportunity and he doesn’t receive a fielding percentage score. DRS takes that 20% likelihood of making the play and awards him .8 points for success and only penalises .2 points for failure. Defenders who make statistically more difficult plays are rewarded more than on easier ones.
Where UZR comes in handy is where, similar to SIERA, more equalisation elements come into play. DRS compares based on a positional average and doesn’t serve statisticians who want to compare DRS over different positions. With UZR, the differential is built in and every position can be compared equally.
For the UZR crowd, it’s a little more complicated but reads the same. A UZR score takes the DRS factors into account (likelihood of making the play) and adds the average run value of the batted ball. An extraordinary play on a difficult ball that would lead to a run, perhaps a long fly ball caught over the centre field fence, would be worth more on the run prevention side as well as the positional difficulty.
DRS is a great tool for managers in roster-building mode. It’s a clear method to decide on fielders at any given position because every right fielder is stacked up against their cohort, as with shortstops, catchers, and so on. UZR is more a method to set the best possible lineups and assess overall talent in throwing accuracy, range, and more.
The sabermetric world is expanding all the time. Every major league club employs at least one person dedicated to the science and they use it in just about every team situation. The stakes are just as high as sabermetrics’ potential to affect the outcome of games and even seasons. Now that every single series has the potential to be pivotal in 2020 given the sheer number of divisional games (40), calculating the matchups looks to be a matter of millimetres and the slimmest percentage points.