Posts
Comments
I don't think you should feel bad about that! This scenario was pretty complicated and difficult, and even if you didn't solve it I think "tried to solve it but didn't quite manage it" is more impressive than "didn't try at all"!
-
There is a problem I want solved.
-
No-one, anywhere in the world, has solved it for me.
-
Therefore, Silicon Valley specifically is bad.
Were whichever markets you're looking at open at this time? Most stuff doesn't trade that much out of hours.
I think this is just an unavoidable consequence of the bonus objective being outside-the-box in some sense: any remotely-real world is much more complicated than the dataset can ever be.
If you were making this decision at a D&D table, you might want to ask the GM:
- How easy is it to identify magic items? Can you tell what items your opponent uses while fighting him? Can you tell what items the contestants use while spectating a fight?
- Can we disguise magic items? If we paint the totally powerful Boots of Speed lime green, will they still be recognizable?
- How exactly did we get these +4 Boots? Did we (or can we convincingly claim to have) take them from people who stole them, rather than stealing them ourselves?
- How honorable is House Cadagal's reputation? If we give the Boots back, will they be grateful enough that it's worth it rather than keeping the Boots?
I can't realistically explain all of these up front in the scenario! And this is just the questions I can think of - in my last scenario (linked comment contains spoilers for that if you haven't played it yet) the players came up with a zany scheme I hadn't considered myself.
Overall, I think if you realized that the +4 Boots in your inventory came from the Elf Ninja you can count yourself as having accomplished the Bonus Objective regardless of what you decided to do with them. (You can imagine that you discussed the matter with the GM and your companions, asked all the questions above, and made a sensible decision based on the answers).
ETA: I have finally tracked down the trivial coding error that ended up distorting my model: I accidentally used kRace in a few places where I should have used kClass while calculating simon's values for Speed and Strength.
Thanks for looking into that: I spent most of the week being very confused about what was happening there but not able to say anything.
Yeah, my recent experience with trying out LLMs has not filled me with confidence.
In my case the correct solution to my problem (how to use kerberos credentials to authenticate a database connection using a certain library) was literally 'do nothing, the library will find a correctly-initialized krb file on its own as long as you don't tell it to use a different authentication approach'. Sadly, AI advice kept inventing ways for me to pass in the path of the krb file, none of which worked.
I'm hopeful that they'll get better going forward, but right now they are a substantial drawback rather than a useful tool.
Ah, sorry to hear that. You can still look for a solution even if you aren't in time to make it on the leaderboard!
Also, if you are interested in these scenarios in general, you can subscribe to the D&D. Sci tag (click the 'Subscribe' button on that page) and you'll get notifications whenever a new one is posted.
Your 'accidents still happen' link shows:
One airship accident worldwide in the past 5 years, in Brazil.
The last airship accident in the US was in 2017.
The last airship accident fatality anywhere in the world was in 2011 in Germany.
The last airship accident fatality in the US was in 1986.
I think that this compares favorably with very nearly everything.
How many of those green lights could the Wright Brothers have shown you?
You can correct it in the dataset going forward, but you shouldn't go back and correct it historically. To see why, imagine this simplified world:
- In 2000, GM had revenue of $1M, and its stock was worth in total $10M. Ford had revenue of $2M, and its stock was worth in total $20M. And Enron reported fake revenue of $3M, and its stock was worth in total $30M.
- In 2001, the news of Enron's fraud came out, and Enron's stock dropped to zero. Also, our data vendor went back and corrected its 2000 revenue down to 0.
- In 2002, I propose a trading strategy based on looking at a company's revenue. I point to our historical data, where we see GM as having been worth 10x revenue, Ford as having been worth 10x revenue, and Enron as having been worth $30M on zero revenue. I suggest that I can perform better than the market average by just basing my investing on a company's revenue data. This would have let me invest in Ford and GM, but avoid Enron! Hooray!
- Of course, this is ridiculous. Investing based on revenue data would not have let me avoid losing money on Enron. Back in 2000, I would have seen the faked revenue data and invested...and in 2001, when the fraud came out, I would have lost money like everyone else.
- But, by basing my backtest on historical data that has been corrected, I am smuggling the 2001 knowledge of Enron's fraud back into 2000 and pretending that I could have used it to avoid investing in Enron in the first place.
If you care about having accurate tracking of the corrected 'what was Enron's real revenue back in 2000' number, you can store that number somewhere. But by putting it in your historical data, you're making it look like you had access to that number in 2000. Ideally you would want to distinguish between:
- 2000 revenue as we knew it in 2000.
- 2000 revenue as we knew it in 2001.
- 2001 revenue as we knew it in 2001.
but this requires a more complicated database.
One particularly perfidious example of this problem comes when incorrect data is 'corrected' to be more accurate.
A fictionalized conversation:
Data Vendor: We've heard that Enron [1]falsified their revenue data[2]. They claimed to make eleven trillion dollars last year, and we put that in our data at the time, but on closer examination their total revenue was six dollars and one Angolan Kwanza, worth one-tenth of a penny.
Me: Oh my! Thank you for letting us know.
DV: We've corrected Enron's historical data in our database to reflect this upd-
Me: You what??
DV: W-we assumed that you would want corrected data...
Me: We absolutely do not want that! Do not correct it! Go back and...incorrect...the historical data immediately!
Success indeed, young Data Scientist! Archmage Anachronos thanks you for your aid, which will surely redound to the benefit of all humanity! (hehehe)
LIES! (Edit: post did arrive, just late, accusation downgraded from LIES to EXCESSIVE OPTIMISM REGARDING TIMELINES)
MUWAHAHAHAHA! YOU FOOL!
ahem
That is to say, I'm glad to have you playing, I enjoy seeing solutions even after scenarios are finished. (And I think you're being a bit hard on yourself, I think simon is the only one who actually independently noticed the trick.)
Petrov Day Tracker:
- 2019: Site did not go down
- 2020: Site went down deliberately
- 2021: Site did not go down
- 2022: Site went down both accidentally and deliberately
- 2023: Site did not go down[1]
- 2024: Site went down accidentally...EDIT: but not deliberately! Score is now tied at 2-2!
- ^
this scenario had no take-the-site-down option
I assume that this is primarily directed at me for this comment, but if so, I strongly disagree.
Security by obscurity does not in fact work well. I do not think it is realistic to hope that none of the ten generals look at the incentives they've been given and notice that their reward for nuking is 3x their penalty for being nuked. I do think it's realistic to make sure it is common knowledge that the generals' incentives are drastically misaligned with the citizens' incentives, and to try to do something about that.
(Honestly I think that I disagree with almost all uses of the word 'infohazard' on LW. I enjoy SCP stories as much as the next LW-er, but I think that the real-world prevalence of infohazards is orders of magnitude lower).
Eeeesh. I know I've been calling for a reign of terror with heads on spikes and all that, but I think that seems like going a bit too far.
Yes, we're working on aligning incentives upthread, but for some silly reason the admins don't want us starting a reign of terror.
I have. I think that overall Les Mis is rather more favorable to revolutionaries than I am. For one thing, it wants us to ignore the fact that we know what will happen when Enjolras's ideological successors eventually succeed, and that it will not be good.
(The fact that you're using the word 'watched' makes me suspect that you may have seen the movie, which is honestly a large downgrade from the musical.)
During WWII, the CIA produced and distributed an entire manual (well worth reading) about how workers could conduct deniable sabotage in the German-occupied territories.
(11) General Interference with Organizations and Production
(a) Organizations and Conferences
- Insist on doing everything through "channels." Never permit short-cuts to be taken in order to expedite decisions.
- Make speeches, talk as frequently as possible and at great length. Illustrate your points by long anecdotes and accounts of personal experiences. Never hesitate to make a few appropriate patriotic comments.
- When possible, refer all matters to committees, for "further study and consideration." Attempt to make the committees as large as possible - never less than five.
- Bring up irrelevant issues as frequently as possible.
- Haggle over precise wordings of communications, minutes, resolutions.
Accepting a governmental monopoly on violence for the sake of avoiding anarchy is valuable to the extent that the government is performing better than anarchy. This is usually true, but stops being true when the government starts trying to start a nuclear war.
If the designers of Petrov Day are allowed to offer arbitrary 1k-karma incentives to generals to nuke people, but the citizens are not allowed to impose their own incentives, that creates an obvious power issue. Surely 'you randomly get +1k karma for nuking people' is a larger moderation problem than 'you get -1k karma for angering large numbers of other users'.
No, wait, that was the wrong way to put it...
Do you hear the people sing, singing the song of angry men
It is the music of a people who will not be nuked again
The next time some generals decide to blow us off the site
They will remember what we've done and will fear our might!
CITIZENS! YOU ARE BETRAYED!
Your foolish 'leaders' have given your generals an incentive scheme that encourages them to risk you being nuked for their glory.
I call on all citizens of EastWrong and WestWrong to commit to pursuing vengeance against their generals[1] if and only if your side ends up being nuked. Only thus can we align incentives among those who bear the power of life and death!
For freedom! For prosperity! And for not being nuked!
- ^
By mass-downvoting all their posts once their identities are revealed.
The best LW Petrov Day morals are the inadvertent ones. My favorite was 2022, when we learned that there is more to fear from poorly written code launching nukes by accident than from villains launching nukes deliberately. Perhaps this year we will learn something about the importance of designing reasonable prosocial incentives.
Why is the benefit of nuking to generals larger than the cost of nuking to the other side's generals?
It is possible with precommitments under the current scheme for the two sides' generals to agree to flip a coin, have the winning side nuke the losing side, and have the losing side not retaliate. In expectation, this gives the generals each (1000-300)/2 = +350 karma.
I don't think that's a realistic payoff matrix.
Eliezer, this is what you get for not writing up the planecrash threat lecture thread. We'll keep bothering you with things like this until you give in to our threats and write it.
Splitting out 'eating out' and 'food at home' is good, but not the whole story due to the rise of delivery.
I believe the snarky phrasing is "Inflation is bad? Or you ordered a private taxi for your burrito?"
I don't actually think 'Alice gets half the money' is the fair allocation in your example.
Imagine Alice and Bob splitting a pile of 100 tokens, which either of them can exchange for $10M each. It seems obvious that the fair split here involves each of them ending up with $500M.
To say that the fair split in your example is for each player to end up with $500M is to place literally zero value on 'token-exchange rate', which seems unlikely to be the right resolution.
Update: the market has resolved to Edmundo Gonzales (not to Maduro). If you think this is not the right resolution given the wording, I agree with you. But if you think the wording was clear and unambiguous to being with, I think this should suggest otherwise.
So the original devs have all long since left the firm, and as I'm sure you've discovered the documentation is far from complete.
With that said, it sounds like you haven't read the original requirements doc, and so you've misunderstood what DEMOCRACY was built for. It's not a control subsystem, it's a sacrificial sandbox environment to partition attackers away from the rest of the system and limit the damage they can do to vital components like ECONOMY and HUMAN_RIGHTS.
The 'Constitution.doc' file specifies the parameters of the sandboxed environment, and is supposed to ensure that the rest of the system can continue functioning whatever goes on in here. It's on protected drive space to be hard for attackers to edit: on a quick review, though, it looks like there have been some attacks on the logic used to read and parse it from disk.
The Shadow Cabinet backup system is implementing industry-best-practice STATIONARY_BANDIT algorithms to align attackers with the system.
We still have a lot of issues with (CR|K)OOK-class viruses, but since DEMOCRACY was implemented we've avoided having any outages to the FOOD system, which is better than any of our competitors.
Give me a call if you have more questions: my team doesn't own this code, but we've had to interface with it in the past and we're familiar with some of its nuances.
(Non-expert opinion).
For a robot to pass the Turing Test turned out to be less a question about the robot and more a question about the human.
Against expert judges, I still think LLMs fail the Turing Test. I don't think current AI can pretend to be a competent human in an extended conversation with another competent human.
Again non-expert judges, I think the Turing Test was technically passed long long before LLMs: didn't some of the users of ELIZA think and act like it was human? And how does that make you feel?
I agree that that would probably be the reasonable thing to do at this point. However, that's not actually what the Polymarket market has done - it's still Disputed, and in fact Maduro has traded down to 81%.
And I think a large portion of the reason why this has happened is poor decision-making in how the question was initially worded.
Edited to add: Maduro is now down to 63% in the market, probably because the US government announced that it thinks his opponent won? No, that's not an official Venezuelan source. But it seems to have moved the market anyway.
My sense is this security would be fine? Is there a big issue with this being a security?
In the sense that it would find a market-clearing price, it's fine. But in the sense of its price movements being informative...well. Say the price of that security has just dropped by 10%.
Is the market reflecting that bad news about Google's new AI model is likely to reflect poor long-term prospects? Is it indicating that increased regulatory scrutiny is likely to be bad for Google's profitability?
Or is Sundar Pichai going bald?
I agree that most markets resolve successfully, but think we might not be on the same page on how big a deal it is for 5% of markets to end up ambiguous.
If someone offered you a security with 95% odds to track Google stock performance and 5% odds to instead track how many hairs were on Sundar Pichai's head, this would not be a great security! A stock market that worked like that would not be a great stock market!
In particular:
- I think this ambiguity is a massive blow to arbitrage strategies (which are a big part of the financial infrastructure we're hoping will make prediction markets accurate). There are already a lot of barriers in the way of profiting from a situation where e.g. one market says 70% and one market says 80%: if there's a chance that the two will resolve some ambiguity differently, that adds a very big risk to anyone attempting to arbitrage that difference.[1]
- I think this ambiguity is very dangerous to the hopes of prediction markets as a trustworthy canonical source on controversial events. If Maduro says "See, everyone, I won the election fair and square, the prediction markets agree!"[2], I think it's very important that the prediction markets be perfectly clear on what they are tracking and what they are not tracking, and I don't think that's currently the case.
- Related to #2, I think that this ambiguity is vastly more likely to occur in the case of the controversial events where it's most valuable to have a trustworthy and canonical source. The cases where markets resolve cleanly are exactly the cases where non-prediction-market mechanisms already reach a canonical consensus on their own.
- ^
This also makes Manifold's preferred strategy of dealing with ambiguity by N/A-ing a market less valuable: that's an acceptable resolution for someone who just did buy-and-hold on that one market, but can be very bad for someone who was trading actively across multiple markets some of which N/A-ed and some of which did not.
- ^
This imagines a world where prediction markets are major enough and mainstream enough for people to be looking at them and talking about them: but that's exactly what prediction market advocates want!
I don't think 'official information from Venezuela' is fully unambiguous. What should happen if the CNE declares Maduro the winner, but Venezuela's National Assembly refuses to acknowledge Maduro's win and appoints someone else to the presidency? This is not a pure hypothetical, this literally happened in 2018! Do we need to wait on resolving the market until we see whether that happens again?
I agree that resolving that market to Maduro is probably the right approach right now. But I don't actually think the market description is entirely clear even now, and I think it would be very easy for things to have turned out much much worse.
Edited to add: it also seems that many people on Polymarket don't find it unambiguous in that way. The market is trading currently at 84% for Maduro with noticeable recent activity, which does not seem like what should happen if everyone is clear on the resolution:
Short answer: Money is fungible.
Long answer: If you work 60 hours a week, buy essential items, and can't buy luxury items, it is reasonable to say that you needed to work 60 hours a week just to afford essential items. If you work 60 hours a week, buy essential items, and also buy luxury items, it seems more reasonable to say that you worked [X] hours a week to buy essential items and [60-X] hours a week to buy luxury items, for some X<60.
If you ignore the fungibility of money, you can say things like this:
- Bob works 40 hours a week. He spends half of this salary on essential items like food and clothing and shelter, and the other half on luxury items like fancy vacations, professionally prepared food, recent consumer electronics and entertainment, etc.
- Now Bob has children. Oh no! He now needs to work an extra 20 hours a week to afford to send his children to a good school! This means he needs to work 60 hours a week to afford necessities!
But, even if we account a good school as a necessity, Bob's actual situation is that he is spending 20 hours of labor on his personal necessities, 20 hours of labor on his children's necessities, and 20 hours of labor on luxuries. He has the ability to work 40 hours a week for necessities. He is instead choosing to work 60 hours a week to afford luxuries.
That's a reasonable choice for Bob to make! The modern world has some very nice luxuries indeed, and Bob can justifiably think it's worth putting some extra hours in to get them, even if he doesn't enjoy his job!
Yes, it would be better still if Bob could afford all the same luxuries with a 40-hour workweek. But don't tell me that Bob is in the same position as a coal miner who had to work 60-hour weeks to put food on the table and heat his house in winter, and don't try to use this to argue that there hasn't been any improvement in poverty.
And as the world gets richer still, there are two ways this could manifest:
- Bob gets richer, and uses that wealth to work less.
- Bob gets richer, and uses that wealth to have more luxuries.
In the first world, we'll see Bob working fewer hours. In the second world, we'll see Bob working the same hours but having more nice things. Both are improvements! But which improvement we get depends on how nice the things our society can produce get. The more new nice things our society can produce, the more likely Bob is to continue working 60-hour weeks to get modern luxuries, rather than working shorter hours and accepting 1980s-era luxuries.
This argument would cease to hold water if there were a substantial number of people working 60 hours a week who genuinely weren't spending substantial portions of their income on luxuries. I think this was true in the US even in the relatively recent past, that sizeable numbers of the people working 60 hours a week in 1945-1975 e.g. never left the state they were born in, prepared all their own food, had very limited access to entertainment, etc. I don't think it's true today: I think the overwhelming majority of people in the US who are 'working 60-hour weeks, at jobs where they have to smile and bear it when their bosses abuse them' are also consuming large amounts of luxuries, and I think it's reasonable to conceptualize this as 'they are working longer hours than they have to in order to consume lots of luxuries'.
Here, you can go hunt down the people who used the react that way on this comment instead.
The relevant figure wouldn't be the current value so much as its derivative: I don't know how that situation has changed over time, and haven't put in the effort to dig up information on what that data looked like in 1950.
I am not much of an economist, but the two thoughts that spring to mind:
- The change you want to see, of people not needing to do as much work, is in fact happening (even if not as fast as you might like). The first clean chart I could find for US data was here, showing a gradual fall since 1950 from ~2k hours/year to ~1760 hours/year worked. This may actually understate the amount of reduction in poverty-in-the-sense-of-needing-to-work-hard-at-an-unpleasant-job:
- I think there has also been a trend towards these jobs being much nicer. The fact that what you're referring to as a 'miserable condition' is working a retail job where customers sometimes yell at you, rather than working in the coal mines and getting blacklung, is a substantial improvement!
- I think there has also been a trend towards the longest-hours-worked being for wealthier people rather than poorer people. "Banker's hours" used to be an unusually short workday, which the wealthy bankers could get away with - while bankers still have a lot more money than poor people, I think there's been a substantial shift in who works longer hours.
The change you want to see, viewed through the right lens, is actually somewhat depressing. I would phrase what you are looking for as a world where society has nothing to offer people that is nice enough they are willing to work an unpleasant job to produce it.
If you have the choice between 'work long hours to get enough food to live' or 'work short hours and starve', it makes sense to call that 'poverty'. If you have the choice between 'work long hours to be able to have a smartphone, internet, and cable TV' or 'work short hours, still have shelter, clothing and food, but not have as much nice stuff', I would call that 'work is producing nice enough stuff that people are willing to do the work to produce it'.
On your definition of 'poverty', Disneyland makes the world poorer. Every time someone takes on extra hours at work so they can take their kids to Disneyland, you account the unpleasant overtime work as an increase in poverty, and do not account the Disneyland trip on the other side of the ledger. This seems wrong.
Yeah, I have no idea. It would be much clearer if the contracts themselves were available. Obviously the incentive of the plaintiffs is to make this sound as serious as possible, and obviously the incentive of OpenAI is to make it sound as innocuous as possible. I don't feel highly confident without more information, my gut is leaning towards 'opportunistic plaintiffs hoping for a cut of one of the standard SEC settlements' but I could easily be wrong.
EDITED TO ADD: On re-reading the letter, I'm not clear where the word 'criminal' even came from. The WaPo article claims
These agreements threatened employees with criminal prosecutions if they reported violations of law to federal authorities under trade secret laws, Kohn said.
but the letter does not contain the word 'criminal', its allegations are:
- Non-disparagement clauses that failed to exempt disclosures of securities violations to the SEC;
- Requiring prior consent from the company to disclose confidential information to federal authorities;
- Confidentiality requirements with respect to agreements, that themselves contain securities violations;
- Requiring employees to waive compensation that was intended by Congress to incentivize reporting and provide financial relief to whistleblowers.
'Securities and Exchange Commission' is like 'Food and Drug Administration': the FDA has authority over both food and drugs, not the intersection, and the SEC has authority over off-exchange securities.[1]
This authority tends to de facto extend to a fair level of general authority over the conduct of any company that issues a security (i.e. almost all of them). Matt Levine[2] calls this the 'everything is securities fraud' theory, since the theory "your company did Bad Thing X, it didn't disclose Bad Thing X, some people invested in your company not knowing about Bad Thing X, then Bad Thing X came out and now your company is less valuable, victimizing the poor people who invested in you" has been applied in a rather large number of cases to penalize companies for a wide variety of conduct.
- ^
Some caveats may apply e.g. commodities exchanges are regulated by the CFTC. The SEC also probably cares a lot more about fraud in publicly traded companies, since they are less likely to be owned by sophisticated investors who can defend themselves against fraud and more likely to be owned by a large number of random people who can't. I am not a lawyer, though. Get a real lawyer before making SEC jurisdictional arguments.
- ^
Levine is very, very much worth reading for sensible and often-amusing coverage of a wide variety of finance-adjacent topics.
'A security' is a much broader concept than 'a publicly traded security'.
For example, if you've been following the SEC's attempts to crack down on crypto, they are based on the SEC's view that almost all cryptocurrencies, NFTs, stablecoins, etc. are 'securities'. Whether you agree with them on that or not, the law broadly tends to back them in this.
This is no longer a question of ‘the SEC goes around fining firms whose confidentiality clauses fail to explicitly exempt statements to the SEC,’ which is totally a thing the SEC does, Matt Levine describes the trade as getting your employment contract, circling the confidentiality clause in red with the annotation “$” and sending it in as a whistleblower complaint. And yes, you get fined for that, but it’s more than a little ticky-tacky.
This is different. This is explicitly saying no to whistleblowing. That is not legal.
Are you sure about this interpretation?
(DISCLAIMER: I am not a lawyer at all, etc, etc.)
This came up on LW recently, and I didn't find the letter convincing as being the second rather than the first situation.
Remember, the whole legal theory behind the SEC's cases is that a clause like this:
I, EMPLOYEE, will not disclose EMPLOYER'S confidential information to anyone without EMPLOYER'S permission. If I do, I understand that EMPLOYER may pursue civil and criminal penalties, and if I receive compensation for disclosing such information I understand that EMPLOYER may recover said compensation from me in addition to any legal penalties.
if it doesn't contain an explicit carveout for the SEC, is itself 'a threat of criminal prosecution for whistleblowing' and 'a requirement to give up whistleblowing payments'.
If OpenAI's NDAs actually contain a clause like 'I will not tell the SEC anything, and if I do I may not receive whistleblower payments', I agree that would be very bad, much worse than a failure-to-exempt-the-SEC problem.
But I think the letter sounds much more like the usual failure-to-exempt-the-SEC. Is there something I'm missing here?
Not a lawyer, but I think those are the same thing.
The SEC's legal theory is that "non-disparagement clauses that failed to exempt disclosures of securities violations to the SEC" and "threats of prosecution if you report violations of law to federal authorities" are the same thing, and on reading the letter I can't find any wrongdoing alleged or any investigation requested outside of issues with "OpenAI's employment, severance, non-disparagement and non-disclosure agreements".
Matt Levine is worth reading on this subject (also on many others).
The SEC has a history of taking aggressive positions on what an NDA can say (if your NDA does not explicitly have a carveout for 'you can still say anything you want to the SEC', they will argue that you're trying to stop whistleblowers from talking to the SEC) and a reliable tendency to extract large fines and give a chunk of them to the whistleblowers.
This news might be better modeled as 'OpenAI thought it was a Silicon Valley company, and tried to implement a Silicon Valley NDA, without consulting the kind of lawyers a finance company would have used for the past few years.'
(To be clear, this news might also be OpenAI having been doing something sinister. I have no evidence against that, and certainly they've done shady stuff before. But I don't think this news is strong evidence of shadiness on its own).
(And I wonder which ghost your great-uncle is...perhaps we can get away with sending no exorcist at all to that one?)
Things about the dataset:
Each ghost statistic has a bimodal distribution, with one peak ~70 for 'high' stats and one ~30 for 'low' stats.
High stats correlate with other high stats: many ghosts have either all stats high or all stats low. This suggests a distinction between e.g. 'Major' spirits (which tend to have all stats high, but sometimes have a few low) and Minor spirits (vice versa).
Sliminess seems to be the stat most correlated with major/minor-ness: almost all Major spirits have high Sliminess, and almost all Minor spirits have low Sliminess. Hostility is the least correlated: Hostile Minor spirits, or non-Hostile Major spirits, both happen relatively often.
However, I haven't yet been able to come up with anything clever to do with this, and ended up mostly just using a linear regression.
Results of my analysis:
Most exorcists have one particular ghost stat that seems to primarily govern the difficulty they face:
- The Phantom Pummelers really do not like Sliminess.
- The Spectre Slayers really do not like Intellect.
- The Wraith Wranglers really do not like Hostility.
- The Demon Destroyers really do not like Grotesqueness (and also do better with low Hostility).
while some behave differently:
- The Entity Eliminators seems to dislike all stats, especially Sliminess: perhaps they have a hard time with Major spirits and a relatively easy time with Minor ones?
- The Mundanifying Mystics have a very high base rate, but actually charge slightly less for all stats - they are expensive in general, and get extra annoyed when you waste their time with Minor spirits?
We handle the idiosyncracies of hiring the various exorcists:
Paying the Demon Destroyers to come seems worth it: they might actually save us 400sp in expectation just on spirit W alone.
The Spectre Slayers seem more valuable than the Entity Eliminators: while the Eliminators are all-around okay at minor spirits, with our knowledge of who is good against which stats we can always pick out a better exorcist to use, while the Spectre Slayers are a uniquely good bet for spirits like S that have very low INT but high other stats.
There are exactly three spirits where I think the Pummelers save us money (N, U, and a little bit on H), so we don't need to fret about that constraint.
And we end up assigning (unless I find something else to do and change this):
A: Spectre Slayers
B: Wraith Wranglers
C: Mundanifying Mystics
D: Demon Destroyers
E: Wraith Wranglers
F: Mundanifying Mystics
G: Demon Destroyers
H: Phantom Pummelers
I: Wraith Wranglers
J: Demon Destroyers
K: Mundanifying Mystics
L: Mundanifying Mystics
M: Spectre Slayers
N: Phantom Pummelers
O: Wraith Wranglers
P: Mundanifying Mystics
Q: Wraith Wranglers
R: Mundanifying Mystics
S: Spectre Slayers
T: Mundanifying Mystics
U: Phantom Pummelers
V: Demon Destroyers
W: Demon Destroyers
Edit after seeing simon's answer:
We appear to have done pretty much the exact same things - identified the major/minor spirit distinction, not found anything to do with it, just fed the stats into a linear regression - and gotten the exact same answer.
I think FTX is somewhat relevant to my #4 (de facto cost of loaning money to prediction markets may be quite high), and to the comment thread around here about how accessible/usable various ways of working around US law to bet on prediction markets are. I don't think it changes my opinion very substantially.
While I think these arguments are sufficient on their own (and am unimpressed with Michael's arguments), there is one I think is missing:
'Guilty' can encapsulate a wide range of verdicts for a wide variety of crimes, and even a client who is guilty of something is not necessarily guilty of everything or deserving of maximal punishment.
A client who is unambiguously guilty of manslaughter can still deserve representation to defend them against a charge of murder.
This is also an area that requires expert support: most non-lawyers treat the phrases 'robbing a house' and 'burgling a house' interchangeably, but the first is a much more serious crime. Even 'how should I address the judge to not offend him unnecessarily and worsen my sentence' is something that a lawyer can legitimately help even a totally guilty client with.