New MLB Rule Changes - How Fantasy Baseball Experts are Handling their 2023 Projections

Earlier this September, Commissioner of Baseball Robert D. Manfred Jr. announced a number of rule changes that Major League Baseball would begin to employ starting in 2023. The rule changes addressed a few general areas of the game of baseball from the pace of play – to the physical dimensions of the field – to defensive limitations.

Without going through all of its specifics, the highlights of the new MLB rules are as follows:

• MLB will incorporate a pitch clock/pitch timer.
• Pitchers will be limited in the number of pickoff attempts.
• There must be four defensive infielders to start each play, two on either side of second base.
• The bases will be enlarged from 15 inches to 18 inches.

In addition to the above rule changes, MLB will also now implement a different team schedule blueprint for 2023. The new schedule will be more of a “balanced” schedule. Each team will now play every other team at least once in the season. Teams will only face division rivals 14 times instead of the previous 19 times.

Be sure to check all of our fantasy baseball lineup tools and weekly lineup resources:

• Fantasy baseball injury reports
• Fantasy baseball trade analyzer
• Daily MLB starting lineups for fantasy baseball
• Fantasy baseball BvP matchups data (Batter vs. Pitcher)
• Fantasy baseball PvB matchups data (Pitcher vs. Batter)
• Who should I start? Fantasy baseball player comparisons
• Fantasy baseball closer depth charts, bullpens, saves
• Fantasy Baseball live scoreboard, daily leaderboards

 

How Will These Changes Affect The Game?

We can (and should) debate all of the positives and negatives that will result from the new MLB rules and schedule. Fans across the country will now be able to see all Major League teams at least one time every other year. The average game time will undoubtedly reduce as evidenced by pitch clock trials in the minor leagues. In addition, there will be more balls in play for 2023, as well as an increase in the number of stolen bases.

Perhaps this will all lead to a better product for the fans. Hopefully so. As a fantasy baseball analyst, it is part of my task to try and understand how the changes will affect various accumulated player statistics – and to convey them to my readers and listeners.

Today’s article will start that process. We will discuss how/what a number of expert projection artists are incorporating into their projections for the coming season. Even if we cannot at this time zero in on all of the potential quantitative effects of all the new league changes – determining the approximate magnitude, direction, and difficulty associated with the tasks will be of great benefit.

 

Projections

Let’s talk a bit about projections. For many (including myself), player projections are the backbone that forms the strategies and planning for the upcoming fantasy baseball season. Understanding how player statistics are forecasted for the coming season is the essential part of fantasy preparation.

In its simplest form, a projection is a forecast of the future. A baseball projection is the best estimate of a player’s true talent for a given period of time. Well-designed baseball projections always start by studying historical data – whether it be the pure top-line statistics or underlying sabermetric components. Projectors develop predictive models laced with mathematic formulae to capture their inherent volatility.

Statistically, the year 2020 was a significant disruptor for projections. The 60-game short season at the height of COVID provided issues for projections. With a partial season as the most recent historical data point – 2021 models had to quickly recalibrate.

In the past few seasons, we have also dealt with some other abrupt changes to the game context. In 2021, we saw changes due to the sticky ball, and in 2022, the deadened ball era continued and compounded. These soft issues are even harder to tackle, as we simply do not know what the short-term future baseball environment will consist of.

Today’s article will discuss how the experts intend to handle more transparent issues. Pinpointing the precise mathematics is an unattainable result - but learning what/how the projection experts are adjusting is a necessary first step.

Finally, prior to getting down to the nitty gritty – a fair warning. The accuracy of 2023 projections will undoubtedly be worse than the accuracy of some other recent season projections. Accurate projections rely on a fairly stable environment, to which these rule and schedule changes provide disorder. The errors surrounding statistical deviations will be larger in 2023. Our goal – simply put – will be to do the best that we can.

 

The Experts

As always, I believe in the wisdom of crowds. I believe that listening to a wide range of smart people is the optimal way to learn and grow. Today’s article is a public service for all those who use projections. The most important piece of information that you need to know prior to using 2023 projections will be whether or not published figures have been adjusted to account for the changes.

I have surveyed seven fantasy baseball industry experts. These individuals have created excellent projection forecasting systems. These are not the same panel of projections that are included in the ATC Projections. Rather, these are all industry friends that I personally know and more importantly – whose opinions I greatly trust.

I sent out a questionnaire/survey to these experts a few weeks ago with pointed questions as to whether or not (and possibly how) they will be handling 2023 projections [with regards to new rule changes and to the new balanced schedule. In this article, I will share with you many of the questions and their corresponding responses. I will also include my own take where I feel that I can add some additional value, as well as let you know what I plan on doing for the upcoming 2023 ATC Projections based on their feedback.

Below are the seven experts surveyed, along with their affiliation and/or projection system model (in alphabetical order):

• Derek Carty – RotoGrinders / THE BAT
• Jared Cross – Steamer
• Rudy Gamble – Razzball
• Ray Murphy – BaseballHQ.com / Ron Shandler’s Baseball Forecaster
• Mike Podhorzer – RotoGraphs / ProjectingX.com / Pod Projections
• Dan Szymborski – FanGraphs / ZiPS
• Todd Zola - Mastersball / The Zystem

Now, let’s go through the survey questions and the panel’s responses, along with my additional commentary.

 

General Projection Adjustments

Question 1 - In general, do you plan on attempting to adjust your projections to account for the effect of the new rule changes?

Question 2 - In general, do you plan on attempting to adjust your projections to account for the effect of the change in schedule?

Question 3 - As many users of your projections are likely to make their own assumptions on the changes – so as not to have them double count effects … Do you plan on providing just one set of 2023 projections, or do you plan on providing two sets of projections (with and without changes)?

• Ray Murphy (BaseballHQ) –
  • Rule Changes - Yes.
  • Schedule Changes - Yes.
  • Multiple Projection Sets - We will provide one set of projections.

• Jared Cross (Steamer) –
  • Rule Changes - Yes.
  • Schedule Changes - Anticipated average park factors will be affected.
  • Multiple Projection Sets – Just one set.

• Rudy Gamble (Razzball) –
  • Rule Changes - Yes.
  • Schedule Changes - Our preseason/rest of season projection already calculate schedule impact.
  • Multiple Projection Sets – One set.

• Todd Zola (Mastersball) –
  • Rule Changes – I’ll adjust where I see fit - mostly stolen bases and those likely to be most affected via shift legislation. I need to think about how pitchers may be affected.
  • Schedule Changes - Yes, this is where I think pitching will be more influenced than batting.
  • Multiple Projection Sets – I’ll have one set with a clear explanation of the approach.

• Mike Podhorzer (Pod Projections) –
  • Rule Changes - Yes, as some, if not all, of these rule changes will very clearly impact performance. The challenge is figuring out what each rule change’s effect is going to be! ‘Attempting’ is a good description, as I highly doubt I or any forecaster will get the exact effects right.
  • Schedule Changes - No, there are far too many variables already. Having to now account for different opponents doesn’t make sense. The effects would be washed out from everything else going on anyway.
  • Multiple Projection Sets – Just one set, I never thought two sets would even be a thing. Why would someone want a set of projections that ignore rule changes, and are therefore likely to be less accurate? That doesn’t make any sense!

• Derek Carty (THE BAT) –
  • Rule Changes - Yes, at least to the degree it is feasible/that I feel comfortable.  I’m of the belief that, even if we can make some general assumptions (especially for cases where we have some data to work off of), we’ll improve accuracy by doing so rather than doing nothing. There will always be variance and unintended consequences we can’t account for, but on the whole, we ought to add some accuracy by at least being directionally correct on the intended effect.  That said, it’s still very early in the off-season and I haven’t dug too deep into any particular rule change yet.
  • Schedule Changes - THE BAT X will do this automatically. The way it’s designed, it starts by looking at every plate appearance every player has ever taken and the associated context (opposition, park, weather, umpire, defense, framing, etc.).  It then adjusts every player’s underlying talent level to be reflective of a neutral context. Then it applies the context associated with the upcoming schedule, so a player’s final projection for any given season will be higher or lower than their underlying context-neutral talent projection based on their own specific strength of schedule. For players with easy past schedules (AL Central pitchers, for instance), that gets stripped out automatically. The new balanced schedule will get applied the same way it always has.
  • Multiple Projection Sets – Just one set, but I’ll let people know what I have and haven’t tried to account for.

• Dan Szymborski (ZiPS) –
  • Rule Changes - In general, no. I don't like introducing things that are wild guesses; I prefer that users of my projections simply know something is out there rather than some unknown blend of adjustment mixing in somehow. For example, even things like shifts aren't straightforward. Teams that normally shift aren't necessarily going to go back to bog standard 1970 infield positioning; we may see shortstops right at the edge of the base and maybe even a defender in motion like a WR in football if teams decide to get creative. 
  • Schedule Changes - Yes. ZiPS uses a simple measure for predicting the strength of the schedule for the upcoming season and the projections reflect those changes.
  • Multiple Projection Sets – Just one. The possible increased accuracy isn't worth the confusion caused by having multiple sets of numbers. Accuracy isn't always the crucial thing.

Change In Schedule

Let’s start with question number two. Almost every expert stated that they will indeed adjust projections to account for the new balanced schedule. In fact, it seems that this particular adjustment would almost be an automatic one.

Most of the more automated-based projections either calculate the strength of the schedule or calculate park factors for each player. These are often updated annually and are a routine part of the projection process.

One of the first steps of generating projections from historical data involves adjusting all historical player stats to a neutral context (both environment and park). Next, the years are combined by some weighted average (ex. 50%/30%/20% for the past three seasons) and are then adjusted based on the projection’s own proprietary methodology. Lastly, the new park factors/strength of schedule factors are reapplied to generate the stats in a 2023 context.

Certainly, I have glossed over many details, but to repeat/summarize – the general projection process incorporates:

1. Strip park factors from historical data to a neutral context
2. Combine historical years by the weighted average
3. Apply proprietary methodology
4. Apply park factors to generate the 2023 context

The above process is especially important for players who change teams in the offseason. A hitter who previously played his home games at pitching friendly loanDepot park – and then signs with offensive-heavy Coors Field for the following year – will have a larger expected home run total in the coming season. This should be fairly intuitive.

Taken a step further – what if a player played in 2021 with the Marlins, then played in 2022 with the Rockies ... and signed with the New York Mets in 2023? Without converting to a neutral baseball context, it is difficult to combine the previous two seasons to analyze and determine his true level of talent.

The more automated projections use the process outlined about (park factors), and since they do, incorporating the balanced schedule is simply a part of the forecasting process. The 2023 context will simply adjust with different factors, which reflect different away ballparks.

Mike Podhorzer's Pod projections and other more manually based projection sets will not make a specific adjustment for the new balanced schedule. Instead, manual projection artists will look at skills/rates from a number of previous seasons. They then (based on experience) make an educated, but manual selection for the coming year. For example, Mike might see strikeout rates of 25%, 22%, and 27% in the past three seasons, and then choose 25% for next year, etc.

One large benefit of projecting rates manually is that soft information can more easily be incorporated. Suppose that the past four seasons of walk rates for a player were: 12%, 12%, 6%, and 12%. Mike might also know that the player was injured for the 6% season. He may choose to ignore the down year and stick with a 12% selection for the future.

Or say another player has exhibited groundball rates of 50%, 50%, and then 40%. Mike might also determine that the latest year of 40% was the result of a change in swing mechanics. In that case, Mike might select 40% for the following year, recognizing that the new approach is more indicative of the future than any of his prior seasons.

However, the manual process does not explicitly enumerate park factors. Shifting figures to the new balanced schedule may not be specifically addressed. For a player who played on the Marlins in 2022 and will head to the Rockies in 2023, the manual projector would make a manual adjustment anyways – it just won’t involve removing and reinserting park factors and the like.

By and large, automated projections do a better job at player skills than manual ones do. For that reason, ATC Projections incorporate the vast majority of its weight for player rates from calculation-based projections. Since the more automated projections do not need to make an adjustment for the new balanced schedule – ATC will not need to either. ATC Projections should already be correctly accounting for the new schedule upon its release in 2023.

 

New Rules Changes

As for the first question (adjusting for the new game rules) – the general consensus from the experts is that they should and will make an attempt to address the changes. It is evidently clear that without being able to over-shift on the infield, Joey Gallo will pick up a few extra hits over the course of a season. With the bases coming closer together, the stolen base rate will indeed increase from where it was the previous season.

The one expert above who stated that he won’t make explicit adjustments for rule changes is Dan Szymborski and his ZiPS projections. Dan has always been of the opinion that for projection purposes - for a highly unclear and hard-to-calculate changing situation – one should not inject “guesswork” into the figures. Dan deems these new rule changes to be squarely in this category and will thus wait until 2024 in order to employ them. Since ZiPS is largely based on prior player historical comparisons, it makes sense for him not to make guesses. Instead, the user should apply any topside adjustments on their own to their liking.

There may be other public projection sets that will not make an attempt to account for the MLB rule changes, but for the most part, you may assume that the new landscape will be addressed in some capacity.

Although the surveyed experts did not provide specifics as to how they will make adjustments (many are still in the process of working out the details), later in this article they will describe the magnitude of the anticipated changes, and to what fraction of the player pool the changes will affect.

As for how the ATC Projections will react for the 2023 season – for the most part, I intend to leave the process alone. That is, I will not change the weighting of the underlying contributing projections, nor will I make topside adjustments on my own. I will assume that the projections have properly adjusted for the 2023 season on their own.

Sure, there may be some projector that either neglected or mishandled the rules changes. However, the beauty of the wisdom of the crowds' methodology is that if any one projection artist did not make the proper adjustments, the overall ATC average will not be dramatically affected.

In some cases, if I can clearly determine that one projection system has not made an adjustment for a particular player – I will reduce that projection’s weight for the one-off player. If ZiPS clearly has an outsized batting average projection for Anthony Rizzo, I may manually reduce its share for the Yankee first baseman. ATC is not fully automated; I do comb projections prior to its release for detectable outliers, regardless. 2023 projections will be no different – it will just be a bit more work on my end for this upcoming season.

 

Multiple Projection Sets

We have an agreement across the board here. All makers of projections will only provide one set to the public. I was curious here to see if anyone would provide the pre and post-adjustment figures, but alas – they will not.

What will be most helpful to us are the explanations that projectors will provide to go along with how they addressed the changes. Todd Zola and Derek Carty explicitly mention that they will provide users with some background of their attempts, and I am sure that others will do as well.

 

Projection Release Dates

Fantasy baseball players often ask me about the release dates for particular projection sets. For your information, I have asked the experts to let us know when they intend to release their 2023 projections. Here is what they had to say:

Question 4 - Approximately when do you plan on releasing your initial projections?

• Jared Cross (Steamer) - Initial projections have been released but adjustments for rule changes should happen in January.
• Rudy Gamble (Razzball) – December 2022 but all the changes likely not accounted for until late February.
• Ray Murphy (BaseballHQ) – Our projections are first available in the 2023 Baseball Forecaster, shipping in early December. The first online projections at BaseballHQ.com will be available in mid-December.
• Mike Podhorzer (Pod Projections) – Early February. 
• Todd Zola (Mastersball) – I launched November 1, but without any overriding of the standard engine. I plan on making the stolen base and schedule adjustments by mid-November, then will roll in shift adjustments as more research (mine and others) unfolds.
• Dan Szymborski (ZiPS) – Team-by-team over late November and December.
• Derek Carty (THE BAT) – Usually late December/early January, but I may try to put a preliminary run-up earlier this year.

The majority of the projections have already been released or will be released in the next few weeks. However, it appears that the adjustments due to the new rule changes may not hit projections until late January or even February. You may want to download an initial projection set from your favorite provider upon release and monitor changes as you get closer to your draft day.

The initial ATC Projections will be released during the third week of January. I am hoping by then that most of the underlying projections will have already incorporated their adjustments. However, ATC may shift during the month of February as projections start to incorporate more math behind the upcoming rule changes. In 2023, ATC may move more than usual (for some players) from its initial release. Try to download ATC projections as close to your draft day to get the most up-to-date projections available this season.

 

Spring Training Data

Question 5 - Will anything that transpires in Spring Training affect your projections as far as rule changes? (i.e., will your correct projections be based on Spring Training data?)

• Jared Cross (Steamer) - No.

• Rudy Gamble (Razzball) – Playing time estimates and batting order estimates will change so yes. But not necessarily the projected rates.

• Ray Murphy (BaseballHQ) – Possibly. Prior research has shown that team-level SB data in spring training can be predictive, so that’s one example of where we might react to spring data. But more generally, since spring training will be an opportunity for players to be adjusting to pitch clock, etc., we will likely default to treating spring data as skeptically as we usually do. (It’s also a WBC year, which tends to even further mess with spring data).

• Mike Podhorzer (Pod Projections) – Yes, I am open to making adjustments based on spring training data, but it will have to be overwhelming evidence that a specific rule change is causing B and not A to make that adjustment. 

• Todd Zola (Mastersball) – I never say never, so it would have to be something way extreme – and I can’t envision what that might be.

• Dan Szymborski (ZiPS) – Nope. Again, it might be helpful from a numbers perspective, but I'm against things that make the data more of a challenge to use.

• Derek Carty (THE BAT) – Nope, very unlikely.  Not sure we’ll be able to learn much by spring training.  But I guess we’ll have to see.

I thought that this was a critical question for projections experts – to see if they were thinking about using Spring Training data in any way for 2023 and if they would quickly incorporate any learnings into their projection sets. By and large, the answer was no. Most projection artists wouldn’t consider it. A few will only entertain doing so if they happen to spot something significant.

Ray Murphy’s response is interesting in that Spring Training team-level stolen base data can be predictive. It is also possible that we can compare stolen base rates from spring to spring, and then compare them to pertinent minor league data. Ray also points out that this is a World Baseball Classic year, and Spring Training data will be wonkier, regardless.

Let’s see. Time is short during the month of March for us fantasy baseball players. If Spring Training data would be used in any way, we may not realistically have the chance to properly incorporate it until the season has already begun anyways.

 

Impact, Difficulty, and Depth

As described above, determining exactly how to adjust projections is still in the works for most projection systems. While there is minor league data to observe, much of the anticipated changes require a lot of guesswork.

In addition, we aren’t clear on how major league ballclubs will react to counteract the effects of some of the changes. For example, outfields may now employ their own shift. Pitching coaches may tweak a groundball pitcher’s pitch mix in order to induce more strikeouts, etc.

Rather than pinpoint an exact methodology to handle each rule change effect, I have asked our expert panel to classify each rule change in terms of:

1. Impact
2. Difficulty
3. Player Depth
4. Stats Affected

Impact – The level of impact that the change will produce.

Difficulty – The level of difficulty in order to implement the change in projections.

Player Depth – The number of players in the player pool that will be affected by the change.

Stats Affected – Which player statistics would be most affected by the change.

Rather than detailing the expert-by-expert responses for all changes and classifications – here is a summary chart of the consensus responses, followed by some commentary.

Banning of the Shift

• High groundball hitters are the batters most affected, and left-handed hitters are more affected than right-handed hitters. For a few hitters, the changes may be dramatic.
• This will be a medium-level difficulty item for projection systems, with a few experts devoting a large amount of time to implement.
• It is unclear if projections will be adjusting their pitcher stats; however, groundball pitchers on poor fielding teams would be most affected.

Larger Bases

• Surprisingly, most of the experts have placed this as a small impact change.
• My personal belief is that the impact will generally be as follows:
  • No effect on players who steal 0-5 bases per season.
  • A small effect on players who already attempt a large number of steals when they have the opportunity, and on those who already have a high SB success rate.
  • A large effect on players who have a 60-75% stolen base success rate, and who could stand to rise in SB attempts. Very roughly, for players who currently steal 10-15 bases – they may add in another 5-10 swipes.
• If projections do not capture these possible impacts, fantasy drafters should have in mind that:
  • The mid-range stolen base players may be undervalued.
  • Large quantity stolen base players may be overvalued.

Pickoff Rules

• Most projection sets will not separate out the effect of the pickoff rules from the effect of the larger bases. There will be one sweeping adjustment where applicable.
• Pitchers who have difficulty holding runners will have their situation exacerbated.

Pitch Clock

• Most projection sets will not make any adjustments. Those that do will target a small list of pitchers who currently have a high average time per pitch.
• Although the experts did not discuss this in their responses, this change should result in a few more walks and a few more hits for applicable pitchers (a few pitchers will see a slight increase in their WHIP). They will also see modestly lower strikeout rates.

Balanced Schedule

• We discussed this in detail earlier in the article; you can trust that projections will have incorporated the new balanced schedule properly in their output.
• Most projection systems will not find this change to be difficult, other than to modify the applicable park factors.
• The change will affect all players, with pitchers affected more than hitters.

Download Our Free News & Alerts Mobile App

Like what you see? Download our updated fantasy baseball app for iPhone and Android with 24x7 player news, injury alerts, sleepers, prospects & more. All free!

More Fantasy Baseball Advice