
How do we recover what progress leaves behind without giving up the gains?
There is a particular kind of baseball game that has almost disappeared. The starter is still on the mound in the eighth inning. He has faced the same hitters three times. They know what he has, and he knows what they know. His fastball may be a little slower than it was in the second. His command may not be quite as sharp. The opposing lineup has had hours to study him, and his own manager has to decide whether the person who got them this far is still the right person to finish the job.
That used to be a fairly ordinary part of baseball. Now it feels almost exotic.
Modern pitching strategy evolved for good reasons. Teams learned more about fatigue, injury, velocity, matchups, pitch counts, and the diminishing advantage of letting the same pitcher face a lineup repeatedly. Relief pitching became increasingly specialized, and the complete game moved from expectation to exception. SABR’s history of the modern bullpen traces that development over more than a century and shows how relief pitching evolved from emergency replacement into a sophisticated strategic system.
It would be silly to pretend none of that knowledge exists. It would be even sillier to conclude that because some people miss seeing starters work deep into games, what they really want is the return of exhausted pitchers throwing until something in the elbow starts making an expensive sound.
That is where I think we often misunderstand nostalgia. We assume that when people say they miss an older version of something, they want the whole old system back. Usually they do not. They miss a quality the old system happened to produce.
In the case of the starting pitcher, maybe the thing people miss is not the workload at all. Maybe they miss continuity. They miss the story of one person solving the same problem over time. They miss adjustment, endurance, tension, and the sense that the person who started the game still owns what happens near the end of it.
If that is the real loss, the useful question is not, “How do we make baseball look like 1975 again?” It is, “How do we preserve what we have learned while recovering more of what made that experience compelling?”
That is the idea behind Constructive Reversion: the selective recovery of a valuable quality lost during change without restoring the obsolete or harmful system that once surrounded it.
The distinction matters because progress is rarely as clean as we like to imagine. We talk about change as though the old system simply disappears and the better one takes its place. In reality, a successful change can solve exactly the problem it was designed to solve and still remove something else that nobody had thought to measure.
Remote work is a good example. The gains are real. People gained flexibility. Commutes disappeared for some workers. Organizations gained access to talent outside a commuting radius. Employees gained more control over where, and often how, they worked. Those are meaningful improvements, and there is no reason to apologize for them simply because someone remembers a particularly productive conference-room whiteboard session from 2007.
But offices were also doing things that were difficult to see precisely because nobody had designed them to happen. People overheard conversations. Junior employees watched senior employees deal with difficult situations. Colleagues from different parts of a company encountered one another without first deciding whether the interaction deserved thirty minutes on a calendar. A manager noticed when someone was unusually quiet. A new employee learned what the organization actually valued by watching how people behaved when nobody was presenting a slide about values.
Research following Microsoft’s shift to remote work found that collaboration networks became more siloed and more static, with fewer bridges between groups. That does not mean remote work is inherently bad or that everybody belongs back in an office five days a week. It suggests that proximity had been providing some useful functions that were easy to overlook until they disappeared.
Once we understand that, the question changes. We no longer have to choose between declaring the office obsolete and treating attendance as proof of culture. Instead, we can ask what physical proximity was actually giving us. If it was mentorship, then mentorship should be built deliberately. If it was cross-functional learning, create more opportunities for it. If it was trust, disagreement, creativity, or simply the ability to know the people you work with as actual people, then bring teams together when those things matter.
The commute does not need a comeback. Presence might.
This same distinction appears in ordinary life. People say they miss family dinner, but “family dinner” can carry a lot of mythology with it. The old version may have depended on schedules, household roles, and expectations many people have no interest in recreating. Perhaps the table itself was never the important part. Perhaps what people miss is shared attention: some period of time when everyone is in one place and expected to pay attention to one another.
If that is the value, there is no reason to rebuild an idealized household from sixty years ago. You can keep the phones, the careers, the complicated schedules, and the rest of modern life while deciding that occasionally, for half an hour, everyone is actually there. The old form is optional. The human function may still matter.
That may be the most important discipline inside Constructive Reversion: learning to separate the value from the package it arrived in. We are not always very good at that. Sometimes we romanticize the whole package because one part of it mattered. Other times we discard the whole thing because some parts deserved to disappear.
Businesses make the second mistake particularly well when they modernize physical spaces and brands. Walk through enough airports, office lobbies, hotels, retail locations, and newly renovated venues and eventually you begin to wonder whether one very busy designer has been working on all of them. There is the tasteful wood panel. There is the gray chair. There is the determined little plant beside a charging outlet.
None of those things is wrong. Standardization solves real problems. It creates consistency, reduces cost, protects quality, and can make spaces safer, easier to navigate, and easier to maintain. But there is a point at which consistency becomes anonymity.
Baseball parks make that easy to see. The great ones are not loved because they reject modern standards. They are loved because they remain attached to a place. PNC Park does not hide Pittsburgh from the experience. The Allegheny River, the Roberto Clemente Bridge, and the skyline are part of the stadium’s identity. The building works as a modern ballpark while still telling you where you are.
That is not nostalgia. It is character.
The same principle applies to organizations. A company can have strong brand standards without making every office, customer interaction, internal message, and local market feel as though it arrived shrink-wrapped from headquarters. It can centralize important decisions without eliminating every trace of local judgment. It can protect consistency without sanding away personality.
The thing worth recovering is not the old carpet, the old office, or the old building. It is the sense that this place belongs somewhere and to someone.
Technology raises a more serious version of the same problem because what can disappear is not merely character. It can be responsibility.
Technology is very good at removing friction, and much of that friction should be removed. Let software schedule the appointment, process the receipt, route the form, flag unusual patterns, summarize large volumes of information, and remind me that I did not attach the attachment I specifically said was attached. There is no virtue in preserving boring work simply because people have always performed it manually.
The problem begins when automation moves from assisting judgment to absorbing responsibility.
Baseball offers a useful version of this distinction through instant replay. Major League Baseball expanded replay because cameras could reveal things an umpire missed. Technology improved the evidence available to the people making the call. It did not make the decision ownerless. Human officials still had to review the play, interpret what happened, administer the ruling, and stand behind it.
That distinction matters far beyond sports. We are entering a period in which organizations will be able to automate increasingly consequential decisions, and some of those systems will be better than the human processes they replace. That still does not answer the question of responsibility.
NIST’s Artificial Intelligence Risk Management Framework makes this point in more formal language by emphasizing clearly defined human roles, responsibilities, and accountability around AI systems. The technical system can assist with a decision. The organization still needs to know who owns the result.
In ordinary language, “the system made the decision” is not good enough.
Use AI to summarize three hundred employee comments, but a leader should still decide what those comments mean. Use automation to flag a financial transaction, but someone should own the decision to freeze an account. Use a chatbot for routine questions, but if the situation becomes emotional, unusual, or consequential, somebody should be able to step out from behind the workflow and take responsibility.
The more serious the decision becomes, the less acceptable it is for accountability to disappear into software. That is not an argument against technology. It is an argument against hiding behind it.
And that is where Constructive Reversion becomes more than nostalgia dressed up in better language. It becomes a way of correcting the excesses that sometimes follow successful change.
Every useful revision contains a temptation to overlearn its lesson. We discover flexibility and begin treating presence as obsolete. We discover standardization and begin treating local character as inefficiency. We discover automation and begin treating human judgment as friction. We discover measurement and begin treating anything difficult to quantify as unimportant.
The original improvement may be completely legitimate. The excess comes later.
That is why Reversion should not be understood as the enemy of Revision. In many cases, it is Revision’s second act. First, we improve the system. Then we live with the new system long enough to discover what the improvement cost. Only then can we ask the more mature question: what should remain gone, and what deserves to return under better conditions?
Sometimes the answer will be nothing. Some practices were inefficient, unsafe, exclusionary, or dependent on burdens we should have abandoned long ago. Constructive Reversion has no obligation to rescue them. The past is not automatically wise because it is past, just as the present is not automatically wise because it arrived later.
The work is separating the value from the package.
That brings us back to the starting pitcher.
He does not need to throw 140 pitches to make the game feel like his. A manager does not need to ignore everything we have learned about fatigue and injury. Baseball does not need to become what it was. But if something worthwhile disappeared while the game got smarter—continuity, drama, adjustment, endurance, visible ownership of the outcome—there is nothing backward about trying to recover more of it.
The same is true in a business. A company does not need to undo remote work to rebuild mentorship, abandon automation to preserve accountability, or reject standards to retain local character. It needs to understand which gains are genuine, which human qualities were lost along the way, and whether those qualities can be designed back into the new system without recreating the flaws of the old one.
The same is true in our personal lives. A better schedule can still leave us missing unstructured time. A useful device can still consume too much attention. A healthy boundary can still require us to rebuild a relationship in a different form. Progress does not become a mistake merely because it created a tradeoff. It simply gives us another decision to make.
That, to me, is the point of Constructive Reversion. It is not a demand to go backward. It is permission to admit that progress can be incomplete. We can preserve what genuinely improved while recovering qualities such as continuity, presence, character, judgment, attention, and ownership under better conditions than the ones that produced them before.
The rule is simple enough to remember, but difficult enough to require judgment: recover the function, not the defect.
Keep what the new system taught us. Keep what it made safer, faster, fairer, or more capable. Then look carefully at what disappeared on the way there and decide whether any of it still matters.
That is not retreat.
It is a more complete version of progress.
Keep the gain. Recover the value.
References
Hooley, David, John Daniels, and Sara Andrasik. “The Specialized Bullpen: History, Analysis, and Strategic Models for Success.” Baseball Research Journal, Society for American Baseball Research, 2018.
Yang, Longqi, et al. “The Effects of Remote Work on Collaboration Among Information Workers.” Nature Human Behaviour, 2022.
Major League Baseball. “PNC Park Guide: Capacity, Seating Chart, Parking, and More.” MLB.com.
Major League Baseball. “MLB Clubs Unanimously Approve Expansion of Instant Replay.” MLB.com, 2014.
National Institute of Standards and Technology. Artificial Intelligence Risk Management Framework (AI RMF 1.0). NIST AI 100-1, 2023.