Accept our Terms of Service
Our Terms of Service have recently changed! Please read and agree to the Terms of Service and the Privacy Policy
Traceless is not the only esper looking at the big picture and occasionally writing about it
Next Post »
« Previous Post
Permalink

(Posted to Causal Web's public blog, Untangling The Webs, on October 10th, 2036)

A common sentiment among some dungeon espers (and internet edgelords) is that dungeon espers don't live to be fifty and that the scary dungeons get scarier each time they get away and nothing good ever happens, so probably this is getting worse. This is mostly incorrect on the merits (dungeon esper lifespan statistics sample heavily from the bad old days and if you do reasonable forward projections you get much more optimistic results, good things happen), but the case for optimism is actually much stronger than that! 

We're Getting Better At This Every Year

Controlling for confluences, fewer espers in dungeons die every year. Note that that isn't as a fraction of living or active combat espers, it's an absolute number. Less deaths in 2035 than in 2034, and so on, in a trend that is noisy due to confluences and was much less clear thirty years ago but is very obviously a thing now. And if you plot out a trendline, the improvement is better than linear!

Why is this happening? One big reason is momentum. Each new esper who doesn't die tragically in their first few dungeons could become the next Skybreaker or Kang Jaeha, people whose very existence is a major game-changer to the landscape. But more practically, every esper who survives and stays in the business is around to make next year's load lighter for the incoming espers, with a year's worth of experience and connections to make them more effective, too. 

There's also been a lot of procedural improvements - the DRT and their international equivalents are better prepared and better equipped than they were ten or twenty or forty years ago, in some cases dramatically so. (This assortment of notable historical DRT failures is depressing to read through, but it's a very useful reminder that lots of things that we all know now used to get people killed!). I personally suspect we're seeing diminishing returns from improvements here, though there's still significant gains to be made. 

And then there's the other big reason for optimism, which is that

Every Dungeon Is A Lootbox

I have two graphs I would like to show everyone.

This first one is a graph of A-and-S class dungeons fatality rates against the number of espers that were able to be teleported in to handle it. You may notice that even when a single esper (usually an A or S class with a relevant power) is brought in, the rate drops dramatically. And then it does again at the second, and the third, and so on, because having the right sensor and the right type of shielding, mobility, and utility powers instead of just what's nearby on-hand matters a lot. Pushing back or killing the really scary dungeons safely is usually a matter of getting the right combination of powers there.

The second graph tracks the availability of long-range, unrestricted (no shade intended to Vera Ablinger, who's done incredible things for the world of logistics)  teleports over time, using a weighted average of the public and  DRT-internal prices of such teleports (this gives a clearer picture than using either alone, but if you're skeptical, scroll down on that link and you can see the individual numbers and also rough estimates of actual teleports used per unit time). 

Looking at that second graph, you might notice a huge permanent drop in price in 2030, bigger than any other individual change, and wonder "Wait. What happened in 2030?" The answer is that Korea's Quasar Guild recruited the best teleporter the world has ever seen (by a huge margin) in late 2029, and then discovered they had a dungeon material that acts as a battery for his power. (Veterans of the scene already know this, if they have eyes. The changes we've seen in the last 6 years have been nothing short of incredible, in terms of international coordination and utilization of esper strength.)

And the thing is - 20000 new dungeons appear every day. Any one of those could have something as useful as those batteries, or even moreso. Most of them won't! But we're getting better at finding out when they do and getting better at extracting dungeon materials proactively all the time. 

We all know there's a confluence coming. With the length of this lull, it's probably going to be a pretty bad one, too. But we're getting better at this much faster than they are. Don't lose sight of that, and don't give up on tomorrow.


(Causal Web moderates her blog comments with her spare time and is not interested in being a platform for people who aren't interested in actually reading what she wrote, are not interested in engaging intellectually with the topic at hand, are statistically illiterate and argue statistics with her, have a mysterious inability to search the web for themselves, or are rude, unpleasant, or not contributing to the discussion at hand. If your commenting indicates that you fall into one of those categories, you will be IP-banned from this site and the reason why will be publicly edited into your comment as a warning for those to come. If you understand this, please type "I Understand That I Will Be Judged And Publicly Humiliated If Found Lacking" into the box below and press enter, at which point you will be able to compose your comments.)

Total: 11
Posts Per Page:
Permalink

NOTE: USER WAS BANNED FOR THIS POST (reasons: not engaging intellectually with the topic at hand, general rudeness and unpleasantness, depressing lack of imagination, disappointing use of a funny username in this context (said username is now up for grabs))

Sorry but this post is kinda stupid and tone-deaf?? Espers are dying every day in dungeons and when people are appropriately worried about this now there's assholes linking this post. Who tf cares if teleports are cheaper? It doesn't fucking matter who you cart over to Volcanic Range, we're never finding the core and that shit  isn't going to fucking care who you bring in, we're still never finding that fucking core and it'll keep taking chunks out of cities every time it pops up.

PS it's fucking unreal that you'd write this fucking middle schooler's optimistic drivel and then make me type out that sentence before commenting. Ego much?

Permalink

The commenter above me isn't wrong that everyone Hellscape kills is because we weren't good enough to stop it yet. But I went and compared the data on it from the last three confluences, on a hunch, and

[a number of citations and stat comparisons follow]

There are a lot of reasons the deaths might be down last time, of course. But you could be right, that we're getting better at this. As long as none of those key espers die on the job, like Red Light.

Permalink

I want to avoid over-updating on improvements that are traceable to finite supplies of dungeon mats that we cannot reverse-engineer, or single exceptional espers. We can accomplish more things with these in play, but relying on them makes a lot of systems (and all things that depend upon them) more fragile. If Min Woo-Young gets deployed in a dungeon and dies, or even if they stop sending him into dungeons and an earthquake drops his ceiling on him in his sleep, there's no reserve capacity ready to take over there. If he retires in sixty years (which is already presuming a lot about how long he wants to keep up with this - I haven't been able to find any information about what he's doing with his presumable fortune so I don't know how to project into the next decades whether any of this will motivate him), and nobody quite as good pops up in those sixty years just like nobody quite as good popped up in the last sixty years, then sixty years' worth of infrastructure that depended on routine use of or emergency availability of his teleports will all fall back to worse alternatives simultaneously. Not on the instant, I know they have a strategic reserve, but the strategic reserve can only get so big (there's only so much battery capacity) and just like the stock market responds to news before any actual container ships are delayed hitting port, the comedown won't be pretty.

We can't make more espers (except by growing the population and getting better at managing awakenings), and we can't control what kind we get, and we can't expect any particular correlation between power value and personal cooperativeness; there's no Esper Moore's Law and the market will fundamentally not be able to work its magic to respond to the immense value being generated here.

(Someone is going to say I'm just saying all this to encourage more monster capture. I'm not, particularly, but to summarize the argument that someone might be trying to have: there are eight monsters currently routinely doing dungeons, between usefulness and amenability limitations, and this is an absolutely terrible conversion rate from the number of live captures that have been performed (it's actually better than "increase the human population to get more espers" if you instead compare to the number of ongoingly maintained monsters that survived their dungeons, though). It is, however, an absolutely terrible conversion rate from an input we control and that the market is theoretically capable of scaling. I'm good at catching monsters but I'm not the only person who can.)

Permalink

[responding directly to Traceless's comment] 

fwiw Woo-young is mostly donating his money to various charities and doesn't really do this for the money, he's really relatively frugal, he mostly does it because he thinks it's the right thing to do

sauce: he's a personal friend

Permalink

There are a lot of reasons the deaths might be down last time, of course. But you could be right, that we're getting better at this. As long as none of those key espers die on the job, like Red Light.

I do think that what we're seeing is us getting better but I do understand the "it's too early to tell" perspective. And you and Traceless are both right that an important part of defending our survival gains is keeping the key players protected. Sometimes I spend too long reading deployment logs for some espers (not naming any names) and have to go lie down for a bit.    

(Red Light's death was a horrible tragedy. I only did a few hours of research, at the time, but - IMO his agent should have told him to be taking slightly fewer dungeons. There wasn't enough time in his schedule for reliable sleep and hadn't been for days.)

Permalink

We can accomplish more things with these in play, but relying on them makes a lot of systems (and all things that depend upon them) more fragile.

Agreed. One of my pet charities is an Single point of failure analyzing think tank that looks into this stuff - if you're interested, they just released a deep dive on the state of teleportation and what we'd need to start doing immediately if MWY died tomorrow. (It is so stressful hearing about him getting deployed into dungeons. Quasar does invest appropriately in protecting him, as far as I can tell, but it's unnerving having so much of our logistical capacity on one pair of shoulders. We're very very lucky he's as committed as he is to keeping those batteries filled up, and that Quasar has as many of those as they do.)

Not on the instant, I know they have a strategic reserve, but the strategic reserve can only get so big (there's only so much battery capacity) and just like the stock market responds to news before any actual container ships are delayed hitting port, the comedown won't be pretty.

Hmmm, the case where he retires I think it'll be... not fine, obviously, but very manageable. We just need to start planning the transition in earnest a decade in advance, and we can do that. I expect we'll start swapping out the lower priority uses while he's still active and work out the pain points as we go. And the batteries retain years worth of his power; I expect we'll stop selling it for public use and keep a very comfortable reserve for the true emergencies (which, ideally, will get rarer and rarer.)

In terms of SPoF espers, I worry a lot about Skybreaker. She's repeatedly said she won't stop punching hurricanes until she's dead, for which she has my utmost respect and gratitude. She also deploys a lot in high class dungeons. (In which she does amazing and important work, to be clear!)

We don't have anyone else who can do what she does and a lot of those storms would have destroyed cities, if left unchecked. This paper looks at what would probably would have happened in New Orleans and it is. Not a fun read.

(Always nice to see you in the comments! Say hi to Cricket for me.)

Permalink

It might be a little pointless to engage with somebody who's already IP banned but there might be people reading who are thinking the same thing and just not saying it, so as for 'we're never finding the core' they said that about Omen. They said that about Escher. Those dungeons are dead now. You can't say 'we'll never find the core' about any dungeon because it only takes the right person or combination of people in the right place, and we're getting better and better at that.

Permalink

(Red Light's death was a horrible tragedy. I only did a few hours of research, at the time, but - IMO his agent should have told him to be taking slightly fewer dungeons. There wasn't enough time in his schedule for reliable sleep and hadn't been for days.)

I was gathering citations to argue about this (I'll come back to that) but now I am dealing with the number of people in different serious reports who think lack of sleep contributes to a lot of esper deaths. Why is this a problem.

Permalink

now I am dealing with the number of people in different serious reports who think lack of sleep contributes to a lot of esper deaths. Why is this a problem.

People can often function tolerably for years and even decades while sleep deprived but it reliably impairs reflexes and decisionmaking, even in people using stimulants to shore up alertness. And espers are no less immune to "I can just miss a few hours of sleep, it'll be fiiiiine, trust" disease than busy humans as a class have been historically, which is to say, distressingly susceptible. (I helped fund a PSA about sleep deprivation for espers, once. It was well-received in focus groups and got good penetration and didn't seem to move the needle at all. People want more time than they get.)

Permalink

Also, terrified families absolutely hate it if they find out that you won't even try waking up an esper who might be able to solve the dungeon faster and might do it if you called them, because you have your threshold for that set to when it'll swing forty lives instead of only twenty (because interrupted sleep is also worse than uninterrupted, on top of less being worse than plenty), and there are multiple different links in the deployment system besides the esper themselves which have to be robust to that pressure, and then the media has to know not to press on that too hard.

Permalink

(Note: the timestamp for this comment is three days after the Confluence starts.)

I don't think you meant for this to be used as comfort reading, but I must say it has helped me a lot! I am staying with some German relatives and decided to weather the storm here rather than go back to the UK. I am mostly worried about my Esper cousin being overworked.

Total: 11
Posts Per Page: