Impact of the new shift rules

Blame Cy Williams.

Williams was the National League’s career home run leader through 1929, before ceding the top spot to Rogers Hornsby. Opposing managers would position two outfielders in deep right in an effort to catch anything that Williams didn’t hit over the shallow right field wall in the Baker Bowl.

Or maybe the fault belongs to Ted Williams. In 1941, White Sox manager Jimmy Dykes instructed his second baseman to play in shallow right field, directing his shortstop to move to the right side of the keystone while positioning his third baseman where the shortstop usually plays. Sound familiar? The famous “Williams shift” is the precursor to many of the defenses currently deployed by MLB managers.

Well, “currently” is no longer true as MLB will legislate the shift, beginning in the 2023 season. What Joe Maddon and the Tampa Bay Rays popularized around 12 years ago grew to the point where many felt that defensive shifts were hurting the game. Analytics evolved such that the shift was used on 30.9% of plate appearances last year. It was only 13.7% when Statcast first started tracking in 2016.

In 2022, left-handed batters faced the shift a whopping 55% of the time. Their right-handed brethren were only burdened at a 19.6% clip. Base hits have been down in recent seasons and MLB is theorizing that the use of shifts is at least partly to blame. So, instead of rewarding teams for advanced scouting and optimally positioning their players, they will now need to adhere to new rules governing where infielders can be positioned.

Starting this season, two infielders need to set up on either side of the keystone, with their cleats on the dirt or infield grass. Something to keep in mind is that teams can still situate their defenders in a “non-conventional” manner. The third baseman can still slide to the right while the shortstop can continue to guard up-the-middle, so long as he remains just left of the second base bag. The second baseman can’t venture into a rover position, but he can still play to the edge of the outfield grass.

However, there are no restrictions on outfielders. There has been some scuttlebutt regarding whether teams will use some extreme alignments with their outfielders. For instance, the left fielder could assume the rover spot in right field with the other two covering the rest of the acreage.

Shifting your fantasy rankings

From a fantasy baseball perspective, ascertaining which batters will benefit the most from the elimination of the shift is a great way to gain an edge. Of course, everyone knows this — so it’s a matter of the accuracy of the various methods used to identify these batters and how to best quantify the effect.

On paper, this seems like a plausible puzzle to solve as there is a bevy of shift-related data available. Unfortunately, there is a statistical elephant in the room. The operative question is, “What would have happened if there were no shift?” Again, there is accessible data to address the query. The problem is, to do so, the adjusted defensive alignment needs to be presumed.

The simple truth is no one knows what the alternate positioning facing these hitters in 2023 will be. As I mentioned, teams do not need to place their infielders in conventional spots across the diamond. Furthermore, the approach of both pitchers and batters may be different with the fielders in different spots. However, to quantify the effect — which is what everyone wants — the derived outcome is based on where they expect the fielders would have been placed, with the same pitch and batted ball event. As disheartening as this may seem, all it means is a numerical adjustment is no more than a “guesstimate.”

Now, it’s still logical to pinpoint certain hitters as being likely to benefit more than others. In scientific terms, the exercise can be accurate, but not precise. This group of players stands to gain the most, but the extent is unclear. The caveat is those of us required to project and rank players need to assign a static adjustment. That said, we need to do the same when regressing a hitter’s or pitcher’s BABIP and HR/FB and all the other metrics impacted by this new change, such as LOB%. What follows is a description of how the 2023 ESPN Fantasy Baseball projections account for shift legislation.

Shifting your focus

The first step is discerning what subset of batters are most likely to aided by the new rules. Statcast has been tracking shifts since 2015. Looking for BABIP trends on various batted-ball types is a good means to gauge shift effectiveness. By means of reminder, BABIP stands for batting average on balls in play. It’s the same as the standard batting average formula, except that homers are excluded from the numerator while homers and strikeouts are omitted from the denominator (although sacrifice flies are included).

Ignoring the element of luck intrinsic to BABIP, a lower BABIP indicates better defense. Since data from an entire MLB season will be incorporated, the luck element should be minimal, expect perhaps in the truncated 2020 campaign where the sample size remains unavoidable small by comparison to other seasons.

Here is the league-wide BABIP on ground balls since 2015, broken down by batter handedness:

It’s always nice when intuition is supported by data. It makes sense that left-handed hitters are more affected by shifting, since most of these now-banned defensive alignments are designed to combat the tendencies of left-handed hitters. This suggests that left-handed pull hitters, especially those hitting a higher percentage of ground balls, stand to be helped the most by the new rules, assuming they frequently faced the shift in the first place.

The data for outfield line drives and fly balls isn’t as definitive. In fact, the respective 2022 BABIP calculations were higher than 2015. The yearly numbers jump around. That is, there isn’t a linear trend. The shift may in fact be hurting batters with respect to balls lofted into the outfield, but the nature of the BABIP formula may be making the trend. Specifically, home runs are left out of the equation, so the nature of the actual baseball used each season influences outfield line drive and fly ball BABIP.

For example, in 2019, the baseball traveled further, increasing homers. A batted-ball event in 2019 may have resulted in a home run, but it would have been an out or hit in another season, affecting the BABIP. As such, making individual player adjustments based on their outfield line drive and fly ball BABIP is sketchy. The sage approach is to focus on estimating how many more ground ball hits a batter would have had if defenses were not allowed to shift.

To be honest, everybody comes to a similar conclusion, sometimes via intuition. Still, it’s reassuring when the data corroborates the assertion. However, here is where different approaches veer off. Some overlay a batter’s spray chart over a conventional defense and determine how many of the grounders would have been fielded successfully. Others compare results both with and without the shift and simply well, assume the outcomes of the former will just “become” the latter.

Shifting your expectations

Look, there are going to be flaws with any approach. Yet, some means of quantifying the effect is necessary, so calling any of the above “wrong” is too harsh. The key is to treat those numbers like any expected numbers, and regressing towards them, often splitting the difference. As an example, there are various expected HR metrics. Some utilize a park overlay, others use trajectory and distance to judge how many venues in which each batted ball would have resulted in a homer.

The idea is that the hitter just happened to strike the ball in a particular manner in a specific venue. In another park, the same exact flight may have had a different outcome, quantified by an expected stat. A wise projection process regresses actual stats to their respective expected measure. In other words, if a batter hit 25 homers, but the expected number was 29, a fair way to handle the difference would be to average the two and project the upcoming campaign as if the hitter had in fact clubbed 27 homers.

Circling back to the shift data, what would have happened without the shift is akin to expected hits (xHits). The best way to quantify future expectations is to regress the actual results halfway towards the expected results — in other words, to average the two numbers. There are some xHits determinations publicly available (Statcast) and other proprietary sources that account for the shift (among other things).

The system used to generate ESPN’s fantasy baseball projections regresses these xHit measures, so they are accounting for the shift. Keep in mind that the adjustment is an average of actual and expected, which will be lower than other systems projecting 2023 numbers based strictly on xHits.

Ultimately, it is up to each fantasy baseball manager to decide if a particular player will outperform his hit projection due to shift legislation. This is no different than deciding how much Kevin Gausman‘s .363 BABIP from last season needs to be regressed or how much of Tony Gonsolin‘s 83.8% LOB mark was merely good fortune.

Let’s conclude with a table displaying the top-50 players whose 2023 projected average is most improved from their 2022 mark. To make the list, the hitter needed to face the shift at least 50% of the time, with switch-hitters qualifying if that percentage was reached from either side of the plate.

Clearly, there is more to a hitter’s improved projection than just a shift adjustment. That said, It is not a coincidence that none of the first 23 hitters on this list are right-handed and only seven righty swingers made the cut at all.

While it isn’t reasonable to precisely pinpoint how any individual player will be influenced by the new rules, there are characteristics that do help us to identify those most likely to notch more base hits in the upcoming season. Most are left-handed (or switch-hitters) with an above-average ground ball rate. Still, in the end, what we don’t know will always outweigh what we do know.

Arm yourself with the best information you can, trust your gut, and when the evidence starts to point towards the wrong direction, shift your conclusions and embrace the new reality.