How (My) Stardates Work

It’s only fair to point out that there is no “real” system for stardates. In fact there were quite deliberate attempts to ensure that there was no rational system to stardates at all, just so that no-one could accuse the show of getting things wrong. My suggestions here are just suggestions, and no more. They require an awful lot of “tweaking,” selective interpretation of the evidence and a charitable disposition on the part of the reader. In the end, I stuck with stardates, mainly because it seemed really strange not to use the chronological system used in the show in a chronology.

For people who like long explanations, I have written a much longer introduction. This is just a quick “how to.”

Alas, nifty Java calculators are quite beyond me, but if you want to play around with stardates (and other dates, too), there’s a spreadsheet on the Calendars page.

Stardates, as I See Them

The base unit of the stardate is a standard 24-hour day.

Once you get beyond that, things start to get a bit complicated. Although there is no definitive and consistent value for “1,000” stardate units (the gap between S.D. 41XXX and 42XXX, for example), what evidence there is suggests that it is the same as an Earth year.

After struggling with this for a long time, I have eventually reached the conclusion that stardates operate according to a sort of “decimal calendar.” Each year is divided into ten “months,” each of which are either 40 or 30 days long:
“Month” Length in days
0 40
1 30 (40 in a “long” year)
2 40
3 30
4 40
5 30
6 40
7 30
8 40
9 40

By constructing a “short” year of 360 days and a “long” year of 370 days, alternating them will provide a 365-day year. Since the idea would then be to keep a fairly direct link between stardates and the conventional calendar, there is an extra “long” year every 40 years to take account of leap-years.

It’s not very pretty, and it has a big problem: the “months” obviously each have 100 days in them, not 30 or 40. This is not something that has any obvious or logical solution. Stardates really were set up so that “1,000” days is equal to 365, and there are no “impossible numbers” in the dates; you can have any from 00 to 99 in any given “month.”

My solution (explored in slightly more detail in the longer version) is that there is an “extra component” to stardates, something that the various “official” explanations for them have specifically mentioned.

To start off with, I envisaged it as a “time-space correction factor” added to the basic date. That makes it very difficult to see why 30-day “months” should have as many days in them as the 40-day ones. Eventually, I came to see things more like this:

Calendar1a.png

How closely this illustrates what’s “really going on” I’m not sure. I think it’s a very vague approximation. The idea is that each month is more of a “time-space grid” and each stardate reflects a discrete time-space location, not just the simple passage of time. The example I’ve given is a “short” month followed by a “long” one, indicating the whole range of possible stardates, and the ones actually used in the show.

What this represents in detail I don’t know. It could be a “world line” tracing the path of the ship through space and time in relation to referents that are currently beyond known science. I’m pretty sure that they can’t be “world lines” in the currently accepted sense, because the numbers are pretty much random. The spreadsheet I’ve made to convert from conventional dates just uses a random-number generator for the value, and certainly the stardates used in the shows follow no discernible pattern.

Although it isn’t immediately self-evident, there are a few occasions where it is quite plain that when it is S.D. 44723.4 in one part of the Federation, it’s S.D. 44703.4 somewhere else. How else can the launch of the São Paulo be on S.D. 52889.3, and yet it’s S.D. 52861.3 when Captain Sisko takes command?

This is Really It?

I’m afraid so. It's based directly on the on-screen evidence, and I haven’t come up with anything else that works even a fraction as well. Every “official” explanation of stardates ever published is very clear that they reflect location as well as time, so how else is that requirement to be addressed? “Star Trek” is, after all, a science fiction story: not everything can be explained in full detail, and sometimes you just have to accept things.
Stardate Date ~Reset
(0)0000.0 00:00:00 on Saturday 26th February, 2225 Zero for stardates
(4)0000.0 00:00:00 on Monday 30th November, 2263 “Star Trek: Discovery” re-zeroing, for reasons unknown
00000.0 00:00:00 on Sunday 17th June, 2323 Shift from “four figure” to “five figure” stardates

The table above shows the three “re-zeroings” that I just can’t seem to avoid. The 2323 one is vital to get the gaps between events in the 23rd and 24th centuries to be in the right places. The 2263 shift is forced on me by the new data from “Star Trek: Discovery”. I have no idea what might make it necessary, it just is.

The numbers in brackets preceding the “four digit” stardates are there for my convenience, and may or may not reflect the way they operated in “Star Trek”. They allow me to distinguish between S.D. (3)2136 in 2257 and (4)2124 in 2266. My theory is that the leading digit is a “later addition” that wasn’t part of the original system.

Any given stardate will have between one and four possible “normal date” values. For example, S.D. 44392 has to be Tuesday 24th October 2367, because the “space location correction factor” has to be at its maximum value of 60, whereas S.D. 44372 could be Friday 6th, Sunday 15th or Tuesday 24th October, depending on the “time-space correction factor.” This is a vital flexibility for me, and I have to assume that the people “really” using stardates in “Star Trek” find the system useful and informative about things I just can’t guess at. Of course, since my stardate units don’t match exactly with days, all of these date conversions could stretch over into adjacent days. I’ve picked the date at 44392.5 and 44372.5; all this is complicated enough already.

How Do People Measure the Time on Starships?

Initially, I assumed that stardates were the standard way of keeping time aboard ship. After all, when the fake Lincoln asks “Do you still measure time in minutes?” in TOS “The Savage Curtain”, Captain Kirk replies “We can convert to it, sir.” (Thanks as usual to Chrissie’s Transcripts website for this quote.) That’s one of the reasons I’m so attached to the idea of a stardate unit being a 24-hour day. Deep Space 9 keeps Bajoran time, and Starfleet ships keep Earth time. Why? Because people know what day of the week it is, when festivals will fall, even what the year is (although all these things are subject to weird lapses). They also use the standard 24-hour clock to measure the time.

It might be possible to “tweak” the relationship between the UTC time and stardates to get things to work slightly better. My own feeling is that every “solution” will just raise a different crop of problems, so it’s better to keep things simple. Of course, I reserve the right to change my mind.

This also addresses a question that I’d kind of dismissed before; isn’t is awfully confusing if the date keeps flipping around? Potentially, you could get up on S.D. 45213, eat breakfast on S.D. 45203, report for duty at S.D. 45233. You can see the problem. Although the stardates as used don’t seem quite that volatile (except “off-screen”) there’s nothing in what I’m suggesting to rule that out. If the stardate is simply used to fix things like log entries and space phenomena to particular time, then this problem is going to be much less pronounced.

Since When is It Up to You to Change Stardates?

Put simply: it isn’t. Some stardates can be made to agree with my ideas, some just can’t. All of the stardates quoted in the timeline reflect the actual dates used. Sometimes they agree with the place I’ve put them, often they don’t.


Welcome
Introduction to Stardates

by StrauchiusStrauchius on 05 Dec 2010 21:08, last updated on 20 Nov 2017 08:23