So, I’m running a Star Trek RPG campaign with some friends of mine that is also a podcast/audio drama. It’s set in the Lost Era about 25 years before the start of TNG, so to get a sense of what the right registry numbers would be for the time period, I created this spreadsheet and plotted out the registries over time from the founding of the Federation up through the 2380s, making sure they conformed to all known canon instances and - as much as possible - licensed non-canon instances as well.

You can play around with it yourself by adjusting the different factors: Addition (how many ships were ordered each year), the fleet’s overall attrition rate (which would include all types of ships removed from active service - destroyed, lost, retired, decommissioned, etc.), and the number of UFP member worlds and sectors of space (I assume that the UFP adds about one member or protectorate every year and one sector of territory every other year).

In the process of putting this together, I developed a theory (actually two) about why the registry numbers skyrocket between the TOS and TNG eras. First, some assumptions:

1. Each starship receives a unique number with one of several predetermined prefixes (NX, NCC, NAR, NST, etc.), which combine to form that starship’s registry.
2. No starships are ever assigned the same number (with one very notable exception and that one gets a letter suffix). For example, there would never be one ship with the registry NCC-74656 and another with NST-74656.
3. Registry numbers are assigned sequentially (we’ll refer to this as the “registry number stream” moving forward).
4. Registry numbers are assigned when a starship is ordered by Starfleet Command, not when its construction starts, finishes, when it's launched, officially commissioned, or any other event.
5. The time between when a starship is ordered and when it’s launched (if it’s ever launched at all) can vary greatly depending on several factors, including but not limited to:
   1. A prototype versus a standard design
   2. The number of ships ahead of it in the ordering queue
   3. Other logistical factors, etc. For example, the TNG Tech Manual mentions six Galaxy space frames that were started but intentionally left half completed for an indeterminate period of time. These space frames would’ve received registry numbers at the time they were ordered, even though the ships may not have entered service until several years after the fact.
6. The rate of registry number increases starts to rise dramatically sometime around 2320 (based on this analysis by Ex Astris Scientia).

My Theory:

In the early 2320s, Starfleet realized something fairly simple, though not necessarily immediately obvious: its main bottleneck in the maximum fleet size it could field didn’t have anything to do with how quickly or how many ships it could build. Not by the amount of raw materials available for construction, by construction time needed, or the throughput capacity of its shipyards. In fact, the upper limit on Starfleet’s capacity for fielding ships was not related to ship production at all. Instead, it was how quickly it could train new officers and crew to staff those starships. Since the Federation had already unequivocally turned away from genetic and widespread/routine cybernetic enhancement, the lowest logistical ceiling was the four years needed to train an officer to a minimum proficiency. And then another 10-15 years at least to train that officer to a command ability.

At the same time, Starfleet also knew that someday the unthinkable was bound to happen - someday, somebody would launch a large scale invasion of Federation space. Whether it would be the Klingons, Romulans, the recently encountered Cardassians, or some as yet unknown species, Starfleet couldn’t guess. But it knew that this was an eventuality it had to be prepared for.

So in the early 2320s, they decided to simultaneously both prepare for and create a deterrence for that eventuality by building a surplus of starships that would essentially go straight into mothballs, but which could be reactivated and pressed into service in as little as a few days. Starfleet would start steadily building a tactical advantage that would only grow stronger over time. From that point forward, Starfleet started running all of its shipyards at 100% capacity at all times, despite knowing full well that this would create ten times or more as many ships as they had officers to crew them. They were mostly tried and true designs: Mirandas, Excelsiors, Oberths and most of them, upon completion, were quietly flown to any number of a hundred different surplus yards spread across the entirety of Federation space. Whereupon they were powered down and stripped of their dilithium crystals, torpedo inventories, and phaser coils.

When Shelby says in Best of Both Worlds that they’ll have the fleet back up and running in less than a year, she doesn’t mean they’re going to build 40 ships from scratch and launch them all within that time. She means that they’re going to dip into the Reserve Surplus - reactivate them, reassign officers to staff them, and have them filling the holes in the deployment profile that Wolf 359 left behind.

Further, when the Dominion War breaks out, Starfleet suffers continuous heavy casualties for months - hundreds of ships are damaged or destroyed in the war’s early days. And that’s even before we ever see the 1000 ship battles that the Defiant later participates in. We can reconcile this with a “small” Starfleet by supposing that the bulk of those ships were being reactivated from surplus yards and staffed as quickly as possible, but that they were not on active operational duty prior to being pressed into combat. It also explains why there are so many Miranda-class in the Dominion fleets. It's not because the bulk of Starfleet is Miranda-class, it's because the bulk of the Reserve Fleet is Miranda-class, because that's what was politically viable when the Reserve Fleet first started taking shape.

This can also help explain the Enterprise-E’s absence for the visible parts of the Dominion War - if the DS9 fleets were largely made up of Reserve Surplus ships, then that lessens the need for otherwise “regular” starships to be present in those fleets. It makes more sense why a ship like the Enterprise-E would continue in her normal operations while there are these great assemblages of fleets otherwise fighting in coordinated campaigns.

The Galaxy-class ships we see in DS9’s fleet battles are probably an exception to this rule and are likely active-duty ships that serve as the nuclei to groups of Reserve Fleet ships (i.e. Galaxy wings 9-1 and 9-3), much the way aircraft carriers serve as the nuclei of modern day naval carrier groups.

Objection 1: But then why doesn’t Shelby explicitly mention the Reserve Surplus when she talks about rebuilding the fleet? How does it never come up during all the conversations between characters during the Dominion War?

It’s true that the general existence of the policy would be hard to keep a secret, especially once war time actually came. To limit that liability, Starfleet officers would be explicitly discouraged from discussing the details and specifics of it - something which only admirals, their adjutants, and maybe some captains would do only when absolutely necessary. For your run of the mill officer that ends up assigned to that ship, they would only really be aware of the fact that they’ve been assigned to the USS Lee (NCC-31450). If it ever even occurred to them to be curious about the ship’s service history, official records might only say something terse and generic like “Reactivated from Reserve Status on Stardate 50534.1.”

Objection 2: But how can it be a secret at all when there are hundreds of starships flying around with 7xxxx registry numbers painted on their hulls, plain as day? Why wouldn’t they use some separate registry stream or no registry at all to hide the reserve fleet’s existence?

This is where I think this theory goes beyond a nerdy fan theory and starts to suggest some really interesting political world building and provide ground for actual dramatic premises. And I imagine that when the Reserve Surplus policy was first proposed and then committed to, more than one admiral asked this exact question. And to avoid the problem, the admirals seriously considered giving the surplus ships a different registry code format, which would limit the rise of registry numbers in the regular service ships the other powers were likely to encounter. But ultimately, they decided against it to avoid some thorny diplomatic issues that would result:

Starfleet prides itself on its multi-role mission. It doesn’t make warships and science ships. It makes starships, which are supposed to be able to serve either function. Giving the Reserve Fleet a distinctive registry code format would’ve been a tacit admission of its more cynical purpose - that those ships really were built and stored solely for the purpose of waging war. Even though the Federation would never use such a resource for a preemptive invasion, its potential enemies wouldn’t be so trusting, and they might then have used the special registry code as evidence of Starfleet’s hostility (the term “secret invasion fleet” would undoubtedly be thrown around). So, in the end and to avoid diplomatic objections, Starfleet kept its deceptively straightforward chronological number scheme and let starships fly around with registry numbers that gave the game away.

Moreover, even though it may have been the policy’s original impetus, these ships weren’t meant for just defense against invasion. They were ultimately meant to be drawn upon for any purpose and therefore, they had to be as capable of fulfilling all of a starship’s potential mission profiles as any other ship. And their registries, the Admiralty decided, should reflect that.

All of which combined to create a situation where the Reserve Surplus policy became paradoxically both one of Starfleet’s most closely guarded secrets and an open secret among all of the great powers of the Alpha and Beta Quadrants.

Sure enough, by the 2330’s, the other powers started noticing the rapid rise in starship registry numbers. And when they started extrapolating those numbers, they quickly realized there were hundreds and then thousands of ships they weren’t seeing. At first, they weren’t sure where they were or what they were doing. Space is, after all, quite a big place and no spacefaring power has the ability to effectively surveil the entirety of another power’s territory. When accused, Federation diplomats feigned ignorance. When confronted with evidence, they neither confirmed nor denied anything and the consistent registry format kept that denial credible, if not believable. But in all that was unspoken, a very clear message was sent: the Federation will never seek war. But if pressed, it was prepared for it.

The Klingons provide an illustrative and perhaps the most important example of how this deterrence worked in practice. At the policy’s inception, peace with the Klingons was still a relatively new situation. The Federation wanted to maintain that peace and build upon it, but at the same time, it also needed to be prepared in case that peace failed. The Reserve Surplus policy accomplished the second goal at the cost of the first. The Admiralty feared that the Klingons might even decide it constituted an invasion fleet and use that as justification in attacking first. In the wake of Praxis, had the Klingons started that war, they would’ve lost it. But the Federation avoids wars, even when it can win them. But even so the Federation couldn’t fully trust the Klingons to honor that peace, the bad blood running too deep for too long. By building the Reserve Surplus and using familiar designs like the Miranda and Oberth, as well as keeping them within the same registry scheme as operational starships, Starfleet was able to walk a fine line of being prepared for war without outright courting it. And, in part because of the deterrence created by the Reserve Surplus policy, war with the Klingons never came (though of course there were other factors, the most significant being the sacrifice of the Enterprise-C, which began to genuinely sway Klingon public opinion towards the Federation in a way that no other outreach had been able to do).

On the other hand, even despite their general knowledge of it, the Reserve Surplus was not enough to dissuade the Cardassians from going to war. They poured the majority of their economy into their war efforts once the conflict kicked into high gear in the 2350s. And for ten years, the Cardassians produced as many ships as they could as fast as they could, strip mining both their own planet and those of annexed territories like Bajor in the process. But despite a single minded focus on military superiority, they simply could not overcome the seemingly inexhaustible supply of replacement ships that Starfleet was able to field. Whenever the Cardassians managed to destroy a starship, it was replaced within weeks, sometimes days. Nonetheless, the Cardassians were relentless in their efforts. Where Klingons on the battlefield might be summed up with the word “ferocity” or Romulans with the word “cunning”, Cardassians could be summed up with the word “tenacity.” For ten years and with an uncommon single-mindedness, the Cardassians threw their focused might on defeating Starfleet and acquiring Federation territory. And though they made some little gains here and there, the Reserve Surplus policy created a logistical wall that they simply could not breach, to the point where Starfleet was able to demobilize from a war footing a full three years before the Cardassians did. Eventually, even they realized that the pursuit of such gains at Starfleet’s expense was an ultimately futile effort, paving the way for an eventual armistice.

Objection 3: But if the bottleneck is trained officers, that still doesn’t explain how Starfleet staffs those ships when war does come.

Yes, this is true. So let’s stretch this theory a bit further and also use it to explain the officer/enlisted situation. We know that enlisted crew ranks exist, but seem to be extraordinarily rare compared against present day militaries. Gene Roddenberry may at one point have had a vision of a Starfleet with only officers, but that’s now patently contradicted by on screen evidence, Miles O’Brien being the most substantial example (but others existing as well, like Crewman Tarses).

So how do we reconcile this? This is how:

In peace time, Starfleet maintains an officer corps with at least four years of training, usually from Starfleet Academy, and enough to fully staff a “small” fleet for exploratory, humanitarian, and defensive patrol missions. In peacetime, this mostly suffices. Starfleet also maintains a small corps of enlisted personnel, but this is mostly just to retain the institutional memory of how to train and field enlisted personnel at all. In peace time, enlisted ranks make up a very small percentage of overall Starfleet personnel.

But should Starfleet have to transition into a wartime footing, it can use that institutional memory to rapidly train tens of thousands of enlisted crewmembers in the bare essentials of starship operations within the span of a few months. As Starfleet mobilizes for war, it activates its Reserve Surplus fleet and spreads out its Academy trained officer corps over the entirety of Starfleet, with enlisted personnel filling up the bottom ranks of operations. The enlisted crewmember takes the station that would normally go to a newly commissioned ensign (or even a Starfleet Academy cadet on field training).

(we'll leave aside the question of where these extra bodies come from, whether the Federation would institute a draft or in some other way compel a larger percent of the population to participate in war making. I think it's an interesting question and one I'd like to explore at some point, but I don't have any formed thoughts at the moment)

In peacetime, that ensign might be the low-person on the operations department totem pole, scrubbing plasma conduits on a long serving starship without anyone under their supervision. But in wartime, that same ensign might find their first assignment is onboard an activated Reserve Surplus ship, overseeing one or more enlisted personnel with three to six months of department-specific training, much more similar to how current-day militaries function. Starfleet doesn’t consider it an ideal situation, preferring its personnel to be more well rounded. But it’s an acceptable compromise to make in order to wield the most effective fighting force possible, as long as that compromise is a temporary one.

(we can also use this theory to put one more notch into Ensign Kim’s butt monkey status - had he commissioned just a bit later, he would’ve done so into a Starfleet mobilized for the Dominion War. Whereupon he might’ve been immediately in charge of some amount of enlisted personnel. Even if he was still just an ensign, at least he wouldn’t have been at the absolute bottom of the ladder. But alas, poor Harry missed it by that much!)

Once the conflict concludes and Starfleet is able to demobilize from a wartime footing, most of those enlisted personnel are deactivated along with the Reserve Surplus ships they staffed. Some few may choose to stay in the service and almost all of those go through further training to become commissioned officers. Some exceedingly few may choose to stay in the service but not go through officer training and become non-commissioned officers. This is how we explain Chief O’Brien: he originally enlisted as part of the Cardassian Wars and was able to rise at least as far as Tactical Officer on a New Orleans-class ship. After Starfleet demobilized (though before the official armistice was signed), he chose to stay on but also chose not to transition into a commissioned officer, perhaps one of only a very few to do so. From what we know of him, we can surmise he liked the opportunities to work on interesting machinery, but never had any interest in commanding a starship or one of its departments, and thus never applied to the Academy before enlisting nor pursued an officer’s commission following the war. To remain a Tactical Officer/Chief of Security would've required becoming a commissioned officer and department head. He didn't care about any of that, so he moved to transporters and was perfectly content there until an opportunity came along to work on an even more interesting engineering challenge - cobbling together a monstrosity of Starfleet and Cardassian technology. The attraction of that challenge might've overwhelmed his aversion to becoming a department head with people under his command.

So my supposition is that enlisted crewmen are uncommon but not rare, but if one stays in Starfleet long enough, the natural course is to go through officer training and receive a commission. One has to make an active choice to stay in Starfleet for an extended period of time but remain an NCO rather than become an officer, making NCOs legitimately rare. This is based on the observation that we have only seen two Chief Petty Officers in all of Star Trek: Miles O'Brien and Sergey Rozchenko, both of whom expressed an explicit preference to remain enlisted rather than become officers.

Special Bonus Theory: The Starship Registration Reorganization Act

Even with this theory, we’re still talking about an awful lot of ships - between one and two thousand every year for decades upon decades. So, I actually have a second theory on registry numbers that I use in tandem with the Reserve Surplus theory. Though it’s much duller and doesn’t offer any of the diplomatic intrigue that the Reserve Surplus theory does, it does do yeoman’s work in lifting the yearly average. And it’s not really linked to the Reserve Surplus theory by necessity, so you can feel free to take either one and leave the other if you choose.

Anyway, it goes like this:

Also in the early 2320s, the Federation Council passed into law the Starship Registry Reorganization Act, significantly altering how registry numbers were managed and folding the administration of vessel registry, civilian and Starfleet, under one common office and therefore, into one stream of registry numbers. Starfleet, as we all know, serves a broader role than just military or defense and in practice probably handles many of the duties that a present day government’s “alphabet soup” of offices does. To draw a parallel to present day organizations, Starfleet is both the Navy and NASA. But beyond that, it’s also the Coast Guard, NOAA, and probably the FAA, NTSB, CTB, ATF, DEA, and several others.

In the 20th/21st century United States, the Navy assigns hull numbers to Navy vessels while the Coast Guard maintains registries for civilian vessels. Similarly, the various service branches assign tail codes to their own aircraft while the FAA assigns and manages tail numbers for all civilian aircraft. But if all of the future versions of these disparate offices fall under the general umbrella of Starfleet, then it makes sense that those individual functions would also be consolidated into a single office.

(in real life, you apply for the tail number you want and the FAA gives it to you if available. But we’ll dispense with that part in this analogy, both to keep the sequential registry framework intact and to avoid the prospect of starship vanity plates. Instead, we’ll say that if you’re building a starship for non-Starfleet purposes, Starfleet just hands you the next number in the stream)

So in this construct, prior to the 2320s, Starfleet maintained a simple stream of numbers that they applied only to their own ship registries. And since each vessel was supposed to be all purpose - in theory capable of both exploration and defense - they dispensed with type-specific registry codes and kept only a distinction between prototype and operational vessels. In other words, instead of the BB-62, CVN-65, SSN-688, etc. of the US Navy, Starfleet simplified its system to just NX and NCC, with the caveat that NX prototypes could be re-designated as NCC if and when the ship entered operational service (a la Excelsior).

As a result, between 2161 and the early 2320s, only NCC and NX numbers were counted towards Starfleet’s official registry number stream, which started at NX-01 and counted up to about NCC-10000 over the course of about 160 years (which averages out to about 60 ships per year). Meanwhile, a civilian Federation agency, akin to today’s FAA, maintained a parallel but separate number stream for registering non-Starfleet vessels, usually with registries like NAR and NST. Then, in the 2320s, that civilian agency was folded into Starfleet’s registry office and from that point on, all Federation vessels - civilian and Starfleet alike - drew their registries from a single stream of numbers, with Starfleet applying the appropriate prefix code based on the ship’s intended purpose. NX-25001 might be Starfleet’s newest and most advanced prototype starship, while NAR-25002 might be a run of the mill automated freighter making the run from Earth to Vulcan and back.

This is a far more pedestrian theory to explain prolific registry numbers than the Reserve Surplus policy, and with none of the diplomatic intrigue. But it offers a nice way to explain the rise in a way that will never come up in conversation between two Starfleet officers in the middle of an hour-long drama. It can also be used to illuminate the evolving nature of the Federation itself over time. In other words, the Federation starts as a self-defense focused military pact among otherwise sovereign nations, a la NATO. They maintain a common defense force - Starfleet - but most other operations, including the mundanities of civilian vessel registration, stay with the individual planetary governments. But over time, the Federation evolves from a NATO-esque setup into a more federalized government with more centralized operations. And part of that evolution is things like civilian vessel registration eventually being folded into an appropriate federal agency.

And depending on where your vision of the Federation/Starfleet falls on the idealistic/cynical spectrum, you can tie Reserve Surplus and SRRA together to varying degrees. You can say that the two policies were unrelated and only coincidentally happened to occur at about the same time. Or you can say that Starfleet used the latter as cover for the former. Personally, my head canon is that they had separate origins and the timing was a coincidence, but that advocates of Reserve Surplus used the timing as part of their argument on how “hiding” the extra registry numbers in plain sight was feasible.

Extra Special Bonus Theory: The 2350s Fleet Modernization Program

Because even when used in tandem, both theories can't explain an inconsistency with the ships first seen in First Contact (Akira, Norway, Steamrunner) and chronological registries. The First Contact ships have 5xxxx and 6xxxx registry numbers, meaning that under a chronological system, they must precede ships that we have seen long before them (Galaxy, Nebula, Defiant, Intrepid). The conventional wisdom that these ships were started in response to the Borg threat in 2366 (which we know canonically does include the Defiant) cannot be true if they have 5/6xxxx numbers and the Defiant has a number of 74205. So, this is my workaround:

By the reckoning of my spreadsheet, 5/6xxx registry numbers would be ordered in the late 2340's and into the 2350's, which puts them solidly post-Narendra III. So in the late 40's into the 50's, the Admiralty probably realized that after a certain point, no matter how many were on tap, 100 year old Miranda and Oberth designs would only be viable assets for so long. And with peace with the Klingons solidifying into something truly stable, Starfleet decided it could start populating the fleet with newer, more combat capable designs without risking a war. Thus, the ships intended to replace the earliest reserve ships started being designed in the 2350s and started populating the reserve fleet by the 2360s. However, they remained intended as assets mostly tied to the Reserve Surplus fleet and therefore - at least to some extent - were Starfleet's ace in the hole for military action. For this reason, they remained largely unseen until called upon for that purpose - namely the Battle of Sector 001 and the Dominion War. They weren't used in the Cardassian Wars because Starfleet didn't need them. They could fend off the Cardassians without revealing their advanced new ships. And they didn't pull them for Wolf 359 because there was only a few days notice, not enough time to reactivate a significant number of mothballed ships.

Caveats:

Unfortunately, with even all of these theories worked together, there's still some data points we have to disregard, but I think they're mostly production mistakes and things that are clearly retconned by other sources:

• USS Yamato NCC-1305-E: Unequivocally a production mistake as confirmed by Okuda.
• Excelsior/Nebula USS Melbourne (NCC-62043): Another clear production mistake, the result of miscommunication between "Best of Both Worlds" and "Emissary."
• The 8xxxx "Conspiracy" registries: Clearly these were never meant to be readable during broadcast. I'm surprised they weren't redone for TNG-R, but alas they weren't. Ultimately, I don't see how we can reconcile these registries when we know for a fact there is a ship class that started being developed after this point but still has a lower registry (Defiant NX-72405) and thus, just have to toss them.
• USS Prometheus NX-59650/NX-74915: Another production mistake, the result of lack of coordination between VFX and set design teams. Here maybe we can say the multi-vector assault mode nature of the prototype has something to do with the presence of two space frame numbers.
• Constitution-class ships below NCC-1700: Without USS Constellation (NCC-1017), we could write off the others as non-canon. But Constellation nixes that answer. My solution here is that on rare occasion, a ship may be ordered as one class, but later held over to be built as another class. So in this case, the Constellation (and maybe Eagle, Republic, etc.) were first ordered as some other class, but then held over and reassigned as Constitution. I use Asia-class to fit with Masao Okazaki's and Bernd Schneider's work, but you can use whatever you like.

So, there you have it. In what I hope is true Daystrom fashion, I present these absurdly detailed conjectures to explain exceptionally minor aspects of a TV show’s production. The Reserve Surplus theory I do intend to eventually use as the premise for an eventual storyline on our show, probably something like:

"When ships in Starfleet’s secret surplus fleet begin mysteriously self-destructing, the Tempest must discover the culprit and stop any further destruction before the important strategic resource is lost entirely."

Though that's still a fair way down the road from now.