A minor complaint about randomized rewards

Fallen London
1

Well, the rewards for the POSI ‘Unfinished Business’ actions have been randomized for a while now. That’s fine - it does serve to make things more realistic. You’re not going to come away with an even 100 rats every time you pilfer a stall; you’re not going to find the exact same amount of prisoner’s honey among abandoned stage props. But the ranges are so wide they warp back into unrealistic again, I’d argue. Your experienced thief, stealing one rat from a stall and then calling it a day? To say nothing of the effect it has on the player.

For instance, take this, from this morning;

757×510 56 KB

A bit of a ‘Well, then’ moment. Sadly, not the first time this has happened to me. We’ve all been there.

I don’t know about everyone else, but I’d personally be happier if the range for something - like the aforementioned - was 50-250, or even 100-200, than the apparent 1-300. Getting high values feels nice, certainly, but when you happen across the lowest possible values, you feel downright cheated for your actions. Even if the average comes out to be the same (or it should, generally speaking), you still get a pretty deep feeling of disappointment when you get a 1 PPA yield from something that used to be a flat 100. Note the following, which happened to NiteBrite (and keep in mind that this is a 5-action option);

I felt annoyed when I only got maybe 80 silk scraps and 10 surface silk out of this one. But this? Eesh.

Anyway… your thoughts? Agree, disagree? Suggestions?[li]

2

Agreed, When I get two souls for 5 actions in unfinished business in spite i feel like breaking the monitor.

3

I have to say I agree that this change has been an unpleasant one. Granted I’m never a fan of combining luck checks with grinding, and preferred the consistent (save rare successes) results anyway. Perhaps that is an uncommonly curmudgeonly view, but in any case I do think the breadth of the range of results is unnecessarily painful, especially for the multiple-action options.

4

I have to agree that it does seem like too broad a range. Getting such a pitiful return on an investment feels like being cheated, and pulls you out of the game.

5

I know people have said before that the max that they’ve been getting appears to be determied on thier max in the applicable stat, so maybe the mimimum should be as well. Maybe something like (1/4 of stat) to (1.25 of stat) with a weight towards the center (give more of a bell curve)

6

I guess I have the opposing viewpoint. Highly variable results allow math-savvy people to exploit the law of large numbers, while driving people to also explore other pathways or stories. If the most optimal path was always the easiest and most attractive, no one would do anything but.

7

Giving a weight on these actions might be more than StoryNexus can easily handle.

I don’t think these particular things are based on your stats, although some other options are (the one thing where you melt a hole through the floor, for instance. The name eludes me right now.)

8

I don’t mind variable, but it literally makes no sense to break an Appalling Secret down into a single Whispered Secret (and so forth).

9

[li]
There’s a reason Unfinished Business is only available to PoSI; it’s intended for high-level grinding. It serves no other purpose. The fact that it’s extremely unpredictable is counter-productive.
You might also want to look up optimal, because by its definition the most optimal path should be the easiest and most attractive–otherwise it’s not the optimal path.

To suggest a fix rather than poke holes in someone else’s: maybe the chances should be based on the value of the item in question, rather than your level. More valuable items should net you more cash if you get lucky, but the chances should be higher of getting screwed over; lower-value items could be a more stable source of income, but less potentially lucrative in the short run. The range for, say, rats could be 100-300; souls could be 1-200. You’re making more on average with Souls, (Snowskeeper’s A Nice Guy Math Corner: 1/3 chance of netting 200 pence for rats, 1/2 chance of netting 200 pence for Souls) but there’s a bigger chance of getting screwed over.

10

I’ve given this some thought. I’ll try to post something substantive tomorrow, when (I devoutly hope) I won’t be stuck in an airport. Just two things for now. First, going by past experience, suggestions which require significant modifications to the underlying tech, particularly in the direction of mathematical complexity, are unlikely to be enthusiastically received. So I don’t think proposing that rewards should be normally distributed (or weighted in some more ad hoc way) will do much good. Second, I think some of the existing mechanics which generate random rewards are much more successful than others: Seeking Curios, for instance. Given that randomised rewards have become pretty much standard in new content, it might be worth talking about the ones we like as well, and why.

11

Well, I like the randomness of Seeking Curios because it’s thematic. You’re scrounging for treasure in a ruined part of the city - getting consistent rewards doesn’t make an awful lot of sense. The airs-changing options also help lend that air of authenticity to your trawls.

Having said that, coming out of ‘Explore the darker corners’ with one Whispered Secret was always a bit of a downer. :x But at least I wasn’t losing an Appalling Secret for it. Or when you spend three actions to talk to the Dean, and you get single digits of everything as a return. I don’t think anyone particularly likes getting the worst end of the RNG. It’s just that the worst end is much more severe when your ranges span from 0(median) to 2(median), as opposed to maybe .5(median) to 1.5(median), or even .3(median) to 1.7(median). And the wider the range is (1-20 versus 1-300), the more cheated you feel when you get a very low result.

I think a good example of random results is the Primordial Shriek bundle of oddities. The range there is 20-80. The median here is, of course, 50. Even when you hit 20, though, it’s not too bad; it’s the lowest result, sure, and you’re quite aware you could have done better, but you still got 40 pence worth of Shrieks, and you can do something with that. And you’re very happy when you get 80 Shrieks of out if, of course, because whoo, 80 Shrieks! Now, this range is .4(median) to 1.6(median). If it was ~0(median) to 2(median), or 1-100 Shrieks - well, I’d be annoyed if I got a single Shriek fron a bundle. I think getting 100 would really have the same impact on the player as getting 80; maybe a little better, but not as stark as the difference between getting 1 and 20.

Also, I’m going to repeat what Snowskeeper said and point out that Unfinished Business was created for POSI item grinding. Seeking Curios is not.
edited by Laluzi on 1/8/2014

12

[quote=Theus]I guess I have the opposing viewpoint. Highly variable results allow math-savvy people to exploit the law of large numbers, while driving people to also explore other pathways or stories. If the most optimal path was always the easiest and most attractive, no one would do anything but.[/quote]The problem with this is that Unfinished Business was rarely the optimal path, instead the most reliable. The only thing it was really good for was expansive grinds of particular materials. The problem with making it random, though, is that chance is the antithesis of long-term or high-stakes grinding. This is why high-tier Relicker items are weighed beyond their echo value, why Yachts/Clippers/Zubmarines are so prized, and why Tears of the Bazaar/Master’s Blood is so infuriatingly hard to obtain. UB isn’t even that good for echo grinding, either, because at that level you could just go with Pygmalion or Boxfuls of Intrigue for a better E/A profit.

When I get 2 foxfire candles from Veilgarden I strangle one of my housepets. When I get 187 candles from Veilgarden I don’t care because I’m only making up for a loss I either have or will suffer.
edited by OPG on 1/8/2014

13

This is incorrect. Over long time horizons, randomness doesn’t matter. Buy-and-hold, etc.

If you take a look at the responses, it seems clear to me that the reliability of Unfinished Business has been intentionally diminished. The fitness for some purposes remains, but now people have incentive to look elsewhere for reliability. It’s like Fidgeting Writer. It’s very good over the long term, but some people choose to avoid it in exchange for a more consistent return over the short term.

Instead of saying &quotThis is less desirable, we should inform the designers so they can change it back.&quot we could be saying &quotThis is less desirable, the developers made it that way intentionally, we should explore why and adjust as needed.&quot

14

Without developer input we have no idea why they changed the UB rewards. We can feedback and say ‘we don’t like it because low rewards are more likely to cause aggravation than high rewards exhilaration’ but in the end, it’s up to them.

I presume that the devs have access to the number of times people use the UB options and can tell if the number has changed. Presumably they don’t want it to be too popular, but want it to be considered a viable option. Their call.

The danger, as I see it, is that it will cause frustration in the non-power users. The people who don’t read the forum or scour the wikis. They want something to achieve a goal and will go to UB rather than try and find another source. They could well get frustrated by a short series of bad rolls and look elsewhere for entertainment. No idea how valid that worry is.

15

[quote=RandomWalker]Without developer input we have no idea why they changed the UB rewards. We can feedback and say ‘we don’t like it because low rewards are more likely to cause aggravation than high rewards exhilaration’ but in the end, it’s up to them.

I presume that the devs have access to the number of times people use the UB options and can tell if the number has changed. Presumably they don’t want it to be too popular, but want it to be considered a viable option. Their call.

The danger, as I see it, is that it will cause frustration in the non-power users. The people who don’t read the forum or scour the wikis. They want something to achieve a goal and will go to UB rather than try and find another source. They could well get frustrated by a short series of bad rolls and look elsewhere for entertainment. No idea how valid that worry is.[/quote]

I don’t think the non-power users will care that much about how reliable the random rewards are, because they are less likely to be interested in grinding large quantities of commodities for particular rewards any way. They will check in periodically for new content when they run out of storylets they haven’t seen, and that’s about it. (Of course, I could be wrong…)

16

[quote=Catherine Raymond][quote=RandomWalker]Without developer input we have no idea why they changed the UB rewards. We can feedback and say ‘we don’t like it because low rewards are more likely to cause aggravation than high rewards exhilaration’ but in the end, it’s up to them.

I presume that the devs have access to the number of times people use the UB options and can tell if the number has changed. Presumably they don’t want it to be too popular, but want it to be considered a viable option. Their call.

The danger, as I see it, is that it will cause frustration in the non-power users. The people who don’t read the forum or scour the wikis. They want something to achieve a goal and will go to UB rather than try and find another source. They could well get frustrated by a short series of bad rolls and look elsewhere for entertainment. No idea how valid that worry is.[/quote]

I don’t think the non-power users will care that much about how reliable the random rewards are, because they are less likely to be interested in grinding large quantities of commodities for particular rewards any way. They will check in periodically for new content when they run out of storylets they haven’t seen, and that’s about it. (Of course, I could be wrong…)[/quote]

But what about when their ambition updates and they end up having to grind for some sort of POSI item and now all the ways of getting things have changed? That’s gonna be a hell of a kick to the head when they get there.

17

[li]If users are unhappy with something, there’s no harm in telling the developers that, as long as it’s done politely and rationally. If they come to us and say ‘yeah, we did this on purpose, here’s why’, then that’s that. But there’s also a chance that they didn’t mean for it to turn out quite like this - and Failbetter Games has always shown a very courteous reception to player opinion in the past.

Besides, I’m not saying things should be changed back - my beef is that the ranges are a little too wide. If they were narrower, I personally would be less frustrated. I made this thread to see if other people agreed or not, or what they’d like to see out of it. I see absolutely no harm in this; I’m rather puzzled why you think that we have no reason to bring it up.

18

This is incorrect. Over long time horizons, randomness doesn’t matter. Buy-and-hold, etc.

If you take a look at the responses, it seems clear to me that the reliability of Unfinished Business has been intentionally diminished. The fitness for some purposes remains, but now people have incentive to look elsewhere for reliability. It’s like Fidgeting Writer. It’s very good over the long term, but some people choose to avoid it in exchange for a more consistent return over the short term.

Instead of saying &quotThis is less desirable, we should inform the designers so they can change it back.&quot we could be saying &quotThis is less desirable, the developers made it that way intentionally, we should explore why and adjust as needed.&quot[/quote]
Several problems here. For one, you’re either misapplying the Law of Large Numbers or mistaken as to the values involved in Unfinished Business; your earlier post suggests that you seem to think that in the long run these would be more profitable, but they aren’t, as the mean (and what the Law of Large Numbers pushes towards) is actually an echo per action. That’s below the short and long-term average for other places. Unless you need a small amount of something and get lucky, Unfinished Business (the randomized ones, at least) is not the way to go now.

Secondly, you’re avoiding the main point; the main complaint isn’t that there’s a spread, it’s that the spread is far too wide, with a low end so low that it can achieve results too terrible to get anywhere else. Unfinished Business isn’t the only place this happens either; the Rooftop Shack’s option for Making Waves can (and has, frustratingly) give a single point of MW in exchange for a moderately expensive item and quite a bit of Connected: Society, and unlike Unfinished Business you can’t even trust in the LLN because you get the card infrequently and are unlikely to have many Favours in High Places unless you have the right Profession. The spread can be reduced to a point where it’s still arguably worthwhile, and still have the same mean, without being scrapped entirely.

Thirdly, just because something’s the developers’ intention doesn’t mean it’s working that way. Notability, for example, was something you could try for whenever you got the Amanuensis to show up, resulting in someone abusing probability and making numerous attempts at extremely low chances and hitting the cap in less than a week; probability-wise, it was never favorable to grind towards higher percentages. The dev response to this was to institute a minimum MW requirement for Notability attempts, to prevent this kind of behavior. However, it did not address the problem; making minimum-MW attempts was still the most favorable course of action, it just required more grinding to attempt, and it still had the illusion that trying for higher percentages was better when in fact it would still likely require many, many more actions than the low-MW attempts. This ended up being changed to a flat MW requirement to advance Notability, addressing the problems and making a much less painful game experience, thanks in part to user feedback.

This is the same thing. Users are responding to a frustrating mechanic (the possibility of getting effectively nothing for what is given up, be it items or actions, on what is ostensibly a successful attempt at something). The devs may end up agreeing, as with the Notability incident, or they may not. Either way, this sort of discourse is valuable, and while you might disagree with the point saying &quotwell the devs want it so it should be&quot is attempting to shut down important discussion.[li]
edited by Roland Jones on 1/8/2014

19

Just chiming in here to agree that the range is far too wide now. I don’t bother with the “unfinished business” storylets anymore.

20

But what about when their ambition updates and they end up having to grind for some sort of POSI item and now all the ways of getting things have changed? That’s gonna be a hell of a kick to the head when they get there.[/quote]Hello! Non-power-user here.

And, honestly, we’re fine. I think I’m up to date with my ambition, and I’ve never ground anything to progress it, except perhaps in the very early days. My character is hilariously impoverished by the standards of this forum - he doesn’t even have a four-card lodging - but you don’t actually need much expensive stuff to get on with the principal storylines.

Also, bling gets more affordable at a startling rate. Fallen London has a weird kind of runaway inflation, where rewards spiral upwards but the price of anything that already exists stays exactly the same. So if you want to be among the first to have something, you need to grind for it - but if you’re not that fussed, you can just bumble around and it falls into your lap after a while.

Cheers
Richard