Onyx Storm (Spotlight on Scale)
Hello, all. Welcome to this Saturday review.
Originally, this spotlight analysis was going to be part of yesterday’s review. However, upon reflection, that didn’t seem the best choice. My analytics indicate that most of you read my reviews from your phones. Incredibly long posts aren’t conducive to that reader experience.
So, in the interest of smoothing things out for you all, I’m going to be separating the Spotlight analyses from the main review posts going forward. If the Sunday after the relevant portion of the review is open, I’ll be scheduling the Spotlights for then. If something is already scheduled (as was the case this week, with Part 2 of the Dot Monster Re:Volution review), I’ll slot it into Tuesday, Saturday, Monday, or whatever else the next available day happens to be.
Thank you for bearing with me on this slight format adjustment. I do hope it provides everyone with a more pleasant reading experience.
STATS
Title: Onyx Storm
Series: The Empyrean (Book 3)
Author(s): Rebecca Yarros
Genre: Fantasy (Epic)
First Printing: January 2025
Publisher: Red Tower Books
Rating: 1.5 / 10
SPOILER WARNING
Heavy spoilers will be provided for the entirety of The Empyrean up through the end of Chapter 44 of Onyx Storm. Mild spoilers for elements later in Onyx Storm may be provided, but I will keep the first paragraph of each section as spoiler-free as possible. Heavy spoilers from later in Onyx Storm will be confined to clearly labelled sections.
THE IMPORTANCE OF SCALE IN WOLRDBUILDING
When it comes to writing fiction, scale matters.
In most genres of fiction, this is an issue that resolves itself naturally. Stories set in the real world (especially if they are set in the modern era and lack any speculative elements) operate by rules we are familiar with. Hard numbers of distances, travel times, sizes, weights, costs, and so much more often don’t need to be provided because most members of the audience will have a general idea of how things scale off each other. When numbers are used, it’s also relatively easy to research the correct ones, unless the subject matter is extremely technical.
For example, in The Living Daylights, James Bond covers a lot of ground in only a few days. He goes from Czechoslovakia to England, back to Czechoslovakia, onward to Austria, and then to Tangiers. At least a portion of this travel was done via his car and other motor vehicles rather than by airplane. The third act then takes place in Afghanistan, with us being shown only a single airplane flight to there from Tangiers. The audience doesn’t need to understand the precise distances and travel times to find this believable. We have a general understanding that Europe and the Mediterranean region are fairly close together (both in terms of geography and interconnectivity) and that a large plane can reasonably be expected to travel from North Africa to Southwest Asia in one flight.
As a setting moves farther away from our modern reality, the situation changes. The audience cannot rely on basic assumptions and common knowledge anymore. This isn’t a huge deal with historical fiction, as there are resources an author can use to make things reasonable, but speculative elements (whether Fantasy or Science Fiction) can really muddle the waters, especially if these elements have a direct impact on the narrative.
When using a “secondary” or “created” world, this is at its most extreme. The audience has no reference points to work with unless the author provides them. Some things can be addressed by common sense or historical reference - for example, we don’t need the properties of steel explained to us in The Lord of the Rings - but for the most part, we’re reliant on the author to give us a sense of scale.
We don’t need hard numbers for this. In fact, hard numbers are incredibly risky. Unless an author thinks through things extremely carefully, there’s a significant danger of overlooking some small factor and opening ludicrous plot holes. All the online speculation about just how the economy of the Harry Potter universe works is a good example of this. Another example is the absurdly small size of the Grand Army of the Republic in Star Wars, where a few million personnel are expected to fight a war across several thousand planets. (Yes, I know the latter example was retroactively explained, but the fact it needed to be retroactively explained highlights the scale issue.)
Using numbers that don’t make sense damages audience immersion and makes the world feel artificial. Because of this problem, many Fantasy and Science Fiction writers who use hard numbers in their settings will numbers will obsess over their figures to ensure that everything is internally consistent enough to preserve audience immersion.
I’m sure you can see where I’m going with this.
Throughout this series, Yarros has given us hard numbers a handful of times. We know that Conscription Day brings “thousands of twenty-year-olds” to Basgiath. At various points, we have been provided with travel times. Other times, she has either provided measurements for the dragons or measurements for something the dragons scale off of, like the wall the dragons perch upon at the Riders Quadrant.
Prior to Onyx Storm, most of these numbers weren’t a problem The glaring exceptions were Yarros’s use of the death toll figures in Fourth Wing and how she inflated the scale of the wyvern horde at the Cliffs of Dralor and the forces advancing on Basgiath in Iron Flame. In these cases, there were two important factors at play.
Yarros made the statistics narratively relevant. She was using them to force a sense of stakes and tension. In the case of the death toll, she used it as a countdown element; for the scale of the wyvern and venin forces, she wanted us to be intimidated by large numbers.
Yarros provided enough information to create a contradiction that wasn’t acknowledged by the story, thereby making the world fell artificial. Her death toll figures for Violet’s first-year class hit the average total for an entire year by the time Threshing occurred, yet the narrative doesn’t acknowledge that their losses are out of the ordinary, despite the fact that many more were certain to die over the following nine months. For the venin and wyvern, she made the numbers so cartoonishly huge that she’d need to hand the heroes additional power for their survival to make any sense, only to overcompensate to the point that the threat could no longer be taken seriously.
What makes Onyx Storm special is that Yarros makes both of these mistakes again.
DISTANCE AND DIAMETER
Yarros has made a big deal about the ticking clock to defeat the venin. Violet and her squad need to travel the world, find the rainbow dragons, and bring them back to Navarre as fast as possible, since otherwise, the venin will overrun Navarre. A big deal is also made about the fact that so little of the world is explored and how hard it will be to fly to all the islands.
Rather than actually writing about a lengthy, globetrotting journey, Yarros chose to compress all the travel, and in doing so, she tells her audience that her apparently Earth-like planet is about the size of Mercury.
This revelation hit me in Chapter 33. Upon the squad’s arrival in Hedotis, we get this from Violet.
The weather is slightly colder this far south, and I strip off my gloves before dismounting.
In Chapter 21, Yarros told us that the journey from Teclis’s palace in Poromiel to Deverelli was “a twelve-hour flight due south” (“sixteen with the gryphons and what [his] texts have provided about historical wind patterns”). This represented such a dramatic shift in latitude that Violet not only had to start worrying about sunburn but also reacted to the change in temperature with, “Holy shit it’s hot here, and by estimates it must only be around nine o’clock in the morning”.
The very clear implication here is that, as warm as the southern coast of the Continent might be (remember, Navarre has access to oranges year-round), it is not a tropical climate, while Deverelli is. This gives us a sense of the relative position of the Continent and Deverelli on a globe (assuming this world is round, which is reasonable, since Yarros has not told us otherwise). The Continent is in a temperate zone, while Deverelli is in the tropics.
The “total flight time” from Deverelli to Unnbriel was another “twelve” hours. Judging by the last reference to Deverelli being its “southwestern” coast, they were not flying due south (but I will calculate things as if they were, for the fact of tipping the scales in Yarros’s favor). Upon arriving in Unnbriel, we are told of “suffocating heat and humidity that rivals Deverelli”. This implies that Unnbriel is also in the tropics.
Now, after another “a thirteen-hour flight” (which was also not specified to be due south, but again, I’ll help Yarros out here), we are told that they are “so far south” and the weather is “colder”. Not “cooler”. Not “less stifling”. “Colder”. There’s also no mention of heat or humidity here. Hedotis is so far south that it is outside of the tropics, back in a temperate zone where there are four seasons.
(Also, while this isn’t necessarily a flaw, the fact it is “colder” tells us that it is winter in Hedotis. It’s currently winter on the Continent. This means that both hemispheres have the same seasons at the same time, implying that this planet has no axial tilt relative to its sun. Where are seasons coming from, then? Does this world have a highly elliptical orbit? Or is there some magical effect at play that works despite magic being confined to specific areas?)
How tiny is this planet if the dragons and gryphons can across the entirety of the tropics in … 41 hours?!
Even if the gryphons (remember, they set the travel pace) can maintain an average cruising speed of 60 miles per hour, that total distance from the temperate southern edge of the Continent to this temperate island cannot be more than 2,460 miles. That’s about three-quarters of the total width of our real-world tropics. Imagine the southern coast of the Continent as the Florida Keys, while Hedotis is Rio de Janeiro … then cut that by 25%.
Except … no … the gryphons can’t be maintaining a flight speed of 60 mph. Their flight speed must be way lower than that, due to this line in Chapter 30.
Eistol, the capital of Unnbriel, is less than a twenty-minute flight inland, but it takes two hours for the calvary to wind their way through the steep terrain and over the ridgeline to the heavily fortified city.
Okay, so a gryphons can cover a straight-line distance about six times as fast as horses that need to deal with a winding route and elevation changes. Even if we assume these horses could maintain a steady canter (up to 12 mph), that “winding” bit and the fact they need to go “over the ridgeline” means that they could easily be covering three times the straight-line distance (since “winding” over a ridgeline implies a path that doubles back over itself). Let’s generously assume that horses are advancing over the straight-line distance at an average rate of 5 mph. That caps the gryphons out at 30 miles per hour. Bear in mind, this is for a short journey, not a sustained flight. It’s a casual jaunt, not a ultramarathon.
Is Yarros implying that this planet is so small that the tropics are only … 1,230 miles across? Again, assuming this planet is round, we can use this arc to figure out this planet’s diameter relative to Earth’s. That makes this planet roughly 3,000 miles in diameter - in other words, the size of Mercury.
Obviously, this is all armchair math. It also doesn’t really matter if this Earthlike world is the size of Earth, Mercury, Jupiter, or one of those tiny planets from Super Mario Galaxy. You can get away with that sort of thing in speculative fiction, especially Fantasy. I don’t consider this to be a flaw … unless Yarros actually intends for this world to bigger than Mercury.
MILITARY MIGHT
It’s time to talk about those 40 thousand troops from Zehyllna.
Prior to Chapter 44
When I first heard this figure, I found it puzzling. Yes, 40 thousand is a big number (far larger than most medieval or Renaissance armies in Europe), but Zehyllna is an island that is not actively at war with anyone, while Navarre and Poromiel represent an entire continent and have been in an active state of war, against either each other or the venin, for six centuries. Surely they have such vast armies already that 40 thousand people won’t make much of a difference.
Then there’s what this tells us about Zehyllna.
While the exact rate of military personnel to civilians varies by both location, and historical era, I’ve heard 1% of the adult population cited as a common benchmark. Larger sizes are certainly possible, especially in a crisis (in 1945, somewhere around 6% of the total US population, which we can extrapolate to about 8% or 9% of the adult population based on life expectancy for the era, was employed by the military), but having a large rate of military employment is incredibly difficult to sustain for long periods. Every person employed by the military is one less person engaging in economically productive work. The 1% figure is a number that is relatively sustainable.
All this is to say that we can reasonably assume the population of Zehyllna is about 4 million people. Nothing wrong there. We don't actually know how big the island kingdoms are, so even after accounting for the fact plant life is less vibrant, it’s not impossible that they have the space and agriculture needed to sustain such a population.
Now, there are still some outstanding issues here, but there’s enough information available handwave those issues:
This can’t possibly be Zehyllna’s full army. Why would the queen of this island agree to throw her entire army at a foreign conflict, leaving her people exposed? The standing army must be substantially larger than this, and thus, the population should be even higher. However, given the way this culture shows their devotion to the God of Luck, it would be consistent for them to take a wild gabble like throwing their entire army at the problem, so this is excusable if everything else makes sense.
Logistics rears its ugly head again. Even if this entire army is ready to deploy tomorrow, we’re really supposed to believe that Zehyllna has the naval assets to move all those people right away? However, we were told in Chapter 39 that the troops weren’t all going to be sent at once, so we can assume that the actually deployment will be scaled to whatever assets Zehyllna has on hand.
All right - that’s Zehyllna. If things had stopped there, I wouldn’t have been inspired to write this spotlight.
Unfortunately, Yarros chooses to throw more numbers at us in this Battle Brief.
Chapter 44
When discussing the Battle of Suniva, we are told a death toll and get Violet’s reaction to it.
Twenty-five thousand people. We’ve never studied a modern battle with so many casualties before. How is Amari’s name are we supposed to dissect one that not only killed some of our classmate’s families but took the life of their queen?
Hold up. How exactly have Navarre and Poromiel not suffered casualties like this?
Let’s really think about this. In most wars, the rate of casualties relative to the number of solders is typically low, especially for wars happening in a pre-Industrial era. Between self-preservation instinct and the impact that losing comrades has on morale, an army will retreat or surrender after a relatively small number of deaths. There’s a reason why the practice of “decimation” was used by the Romans to force disobedient soldiers to fall into one. Killing one in ten men was enough to break rebellious spirits.
However, neither retreat nor surrender are options in this war. Yes, riders and fliers can flee from a conflict on their mounts, but between wyverns and venin death wave attacks, large groups of people will not escape battlefields unscathed. Surrender is likewise not an option, unless the surrendering soldier transforms into a venin. This is a war of extinction. Therefore, when we are told casualties, we can quite reasonably assume that the overwhelming majority of people at that location died.
Add to this what we are told about the battle at Suniva.
The city had “fifty-foot walls”.
We are explicitly told “most everyone burned to death”.
Four venin, mounted on wyverns, were there to “hold the perimeter” while the remaining eight targeted the daggers and the Queen.
The only reason Basgiath has any information at all about the battle is that a couple of riders escaped after the Queen’s death.
Twenty-five thousand people isn’t a small portion of an army dying before everyone turned and ran. 25 thousand people was the entire population of the “fourth-largest city” in Poromiel.
Something Didn’t Feel Right
When I read that Zehyllna had sent an army that was larger than the fourth-largest city in Poromiel, something felt very wrong to me. I crunched numbers to figure out why.
And this is where things got really stupid.
Much like the 1% rule I’ve seen quoted for armies, the rule for urbanized populations in a pre-Industrial society is 5%. This is because agriculture without modern fertilizers and machinery demands so much labor that the overwhelming majority of the population needs to be devoted to feeding everyone. This means that the population of the region around Suniva would come out to … 500,000 people.
Less than 12.5% that of the island of Zehyllna.
Well, that’s only one city. Maybe the rest of Poromiel is different?
Unfortunately, Yarros does not give us population numbers for any cities other than Draithus, nor even any relative sizes beyond a ranking. In the interest of being as fair to her as possible, I will extrapolate populations for the rest of Poromiel based upon estimates for the population of Great Britain’s cities in 1750. This would make Suniva analogous to Liverpool. Based on this data, let’s assume that:
Cordyn, the third-largest city (analogous to Birmingham), has a population roughly 1.1 times that of Suniva. That produces an the urban population of about 27,270 people, for a total of 545,500 people for the city plus the surrounding area.
We are told that Draithus has a population of 30,000 people. This yields a total population of 600,000 people for the city plus the surrounding area.
The largest city would have a population 30.7 times that of Suniva. I have to draw a line here, because Yarros has not established the existence of a massive city to parallel London within Poromiel’s borders. Let’s instead assume that the largest city is 10 times the size of Suniva. That’s still absurdly massive for something that hasn’t been mentioned previously, but it restricts the population of the city and its supporting agricultural lands to 5,000,000 people.
To be exceptionally generous, I will assume Poromiel has or had three additional major population centers that are analogous to Manchester, Leeds, and Sheffield (also on this table of figures from 1750). These cities and the rural populations supporting them would collectively represent another 1,045, 500 people.
That puts the total population of Poromiel at about 7.7 million people.
Poromiel - the kingdom that covers half a continent, home to “endless fields of grain” to feed its population - has less than double the population of this island kingdom that does not have established agricultural abundance and has far less magic to sustain its vegetation even if it could grow an equivalent acreage of food.
And while that may seem strange enough by itself, we then need to consider the military.
While Zehyllna having a 1% rate of military employment makes sense as a sustainable size for a nation not actively at war, Poromiel is in an existential conflict for survival, one in which surrender is not an option. They have also been hardened by centuries of war, so they should understand the risks. If this military isn’t swollen like the US’s during World War II, they are actively courting extinction. This means Poromiel’s army would be around 462,000 people.
Suddenly, those 40 thousand extra troops don’t seem like such a game-changer, especially since magic, alloy, gryphons mean more to this conflict than raw manpower.
Things get even worse, in both directions, when we then remember that Navarre is also part of this conflict. We don't have specific numbers for Navarre’s population, but we do get this line regarding the influx of refugees in Chapter 46.
“You let them in, and there will be no kingdom to serve.” Calldyr lifts his nose. “We have already weakened the outposts by stripping them of all but the necessary alloy, and look what that got us in Suniva. We have sent riders. Lost riders. What more would you have us do? Starve when we cannot feed double our current population?”
The clear implication here is that Poromiel’s population is equal to that if Navarre. This puts the population of the Continent and the total armed forces against the venin at 15.4 million and 924,000 people, respectively.
The entire continent, with all of the abundance supplied by magic, has less than 4 times the population of an area that has only been described to us an as “island” and explicitly has less magic to support abundant food supplies.
Even if we accept that, this leaves the issue of how Zehyllna’s contribution is being treated as a game-changer, despite:
Those 40 thousand people are going into a conflict that likely already has 23 times as many soldiers.
Those 40 thousand people are not any sort of specialists, nor are they bringing magic, alloy, dragons, or gryphons.
Look, I get it - in both world wars (and, indeed, many other wars throughout history), many countries contributed relatively small numbers of soldiers to the conflicts. My issue here is how Yarros frames this. If these were the warriors of Unnbrielf (and if Yarros hasn’t destroyed the credibility of Unnbriel’s warriors), or if Zehyllna has its own magic that it was bringing to the battlefield, I would have no trouble believing that 40 thousand people could tip the scales. However, these soldiers are the equivalent to Navarre’s infantry, a faction Yarros has gone out of her way to punch down to make her special OC factions look better. No amount of backtracking to talk about manning weapons emplacements that may not even exist or goring wyverns that may not even get driven down to ground level is going to correct that. Yarros isn’t dropping Spartans or Space Marines or even the Death Korps of Krieg into this theater of war. She’s delivered the equivalent of Grunts or low-quality Planetary Defense Force personnel, and they simply can’t make a difference in this kind of conflict with these sorts of numbers.
Getting to the Point
Yes, much like the size of the planet, everything here was armchair math, and it leans on a lot of assumptions.
However, I should never have felt a need to check Yarros’s work in the first place.
That’s the problem with hard numbers: even if people aren’t doing hard math, those numbers can feel wrong. The moment we were told that the military force sent be a distant island nation was 1.6 times the size of one of the total population of one of the largest cities on the Continent, the scale of the setting stopped making sense. This is a problem when Yarros is relying on the big numbers to tell us that this is an epic conflict with dire stakes and horrific losses. She is relying on hard numbers to drive her story forward, and those numbers simply don’t work.
Literally all Yarros had to do to avoid this mess was not use hard numbers. If Zehnyllna just sent “thousands” of soldiers, and if all we knew about Suniva was that the entire population of the foruth-largest citiy on Poromiel was wiped out, that would keep things flexible. If Yarros really wanted to give us a number to work with, she should have only have told us the population of Suniva, since she is relying on those numbers specifically to give a sense of just how much death is occurring.
FINAL THOUGHTS ON YARROS AND SCALE
I am constantly encountering amateur writers who are deeply passionate about their worldbuilding. Some do use hard numbers in their setting, but when they do, they agonize over these figures. They consider every possible detail and implication to ensure their settings hold together under any reasonable amount of scrutiny. They do this because they care about Fantasy and respect the intelligence of their audience. These writers aren’t content with shoddy craftsmanship, nor do they demand that the audience turn their brains off.
Yarros’s careless use of numbers demonstrates that she doesn’t care. She throws distances and travel times and body counts and relative populations onto a page because they sound good in the moment. She relies on the audience not comparing these numbers, even if she puts a set of contradicting numbers into focus in the same scene.
The lesson to be learned here is pretty simple: don’t use hard numbers in your book unless you care enough about your setting and your audience to make those numbers make sense.
