Powered by Jitbit .Net Forum free trial version.

HomeFallen London » The Bazaar

This is the place to discuss playing the game. Find tips, debate the best places to find certain items and share advice.

EOR: Towards a new measure of profitability Messages in this topic - RSS

Ben
Ben
Posts: 614

5 days ago
... "Goat Demon grind" ?

Somebody explain that for me?

I know affair of the box. I went the OTHER way on sprifterage *shows his watch* but knowing there might be another way still...

--
The wind has no destination.
http://fallenlondon.storynexus.com/Profile/IcountFrom0
0 link
Kaijyuu
Kaijyuu
Posts: 332

5 days ago
The goat demon grind was an old grind that's no longer available (or at least in its old form). Previously you could not advance past the 3rd coil in Labyrinth of Tigers, use Dealing with Mirror Smugglers to gain The Hunt is On, and capture a goat demon in the flit for ~1.80 EPA. Nowadays you can't get above 7 Hunt from the labyrinth storylets, so the best option is to kill marsh wolves for ~1.73 EPA.

A similar grind is available in the university for those who haven't gotten kicked out (grinding up investigating and turning in in the flit, iirc).
0 link
Kaigen
Kaigen
Posts: 423

4 days ago
Okay, let me see if I'm getting this. If I want to calculate the value of 1 CP of Someone is Coming with the EOR method, I look at my expected payout from robbing the rat, (let's assume 246 Shadowy and a 1% chance of getting the watch) 5.95 echoes. I plug that into the equation with my replacement value of 1.64 for the box grind. Since we're only looking at the cash in, the action value is 1, so the resulting EOR is 4.31 for 6 CP of SiC, reducing to ~0.718 per CP.

I then want to see how much EOR I get from seducing the group's leader on "The kaleidoscopic church" card so I can judge whether to play that or find some other way to dump incidental Connected: Bohemians. Since my only big cash-in options for Connected: Church or Bohemian are calling in favours, I can look at the net Connected shift to save time.

The EOR of 1 Connected CP is (2.4-(1*1.64))/120=0.0063 or 0.095 from a net shift of 15 CP.
The EOR of SiC, as previously calculated, is 0.718
And I get 0.6 echoes in the form of 60 Foxfire Candles.

When I add those values up, I get 1.413 for an EOR of -0.227. Not a profitable break from my usual grind, and so I might want to consider not playing the card at all and instead Calling in Favours with the Bohemians when my incidental Bohemian connections from the Repentant Forger eventually reach 15, assuming I can stand having an additional useless card in my deck. Am I calculating/understanding EOR correctly?

--
Just a simple doctor with a chess habit. Publisher of The Flit Dispatch. Available for lessons in the art of seeing things.

Average actions to complete Flash Lay: ~21
Number of Surprise Packages opened: 350 Average Value: 6.29 Echoes
http://fallenlondon.storynexus.com/Profile/Kaigen
0 link
xKiv
xKiv
Posts: 620

3 days ago
Kaigen wrote:
Am I calculating/understanding EOR correctly?


I think you got it mostly right.
One thing I would add is that 1 cp of SiC is actually worth a bit more, since you will often be able to use capering relicker (as a PoSI) for greater returns, when cashing in below SiC level 5. (at 14 cp - level 4 plus 4 points - IOW 1 cp below level 5 - that gives 0.74 EOR for 1 cp, and it even goes all the way up to 1.03 EOR/cp if you can cash in at exactly level 4. (But don't bother trying to align to that level, the extra actions are better spent on other things)).

Most connection-based conflict cards are poor for profit (siding with hell against church is one exception, assuming you don't mind the scandal), unless you want the connections, or the sideeffects.


As an example, applying the same to favours gives me the following interesting yields (at base 1.64 EPA) ...
Criminals: thieve's cache gives 12.4775 EOR/favor (but this is incorrect calculation, because it just assumes ~2.5 actions spent on getting the necessary dock favours ... so it's really closer to 3.42 EOR/favor for a 1:2.5 mix of criminals:docks; and it drops to only 2.6 when using only whispered hints grund at 150 hints/action)
... (alternative calculation: the other best cash-in for docks I have is vs. widow at 3.37, which means the criminal favour has about 12.4775-2.5*3.37=4.0525 EOR here)
Revolutionaries: flit cash-in is 3.88 EOR/f
Urchins: flit cash-in is 3.95 EOR/f (siding with hell against urchins is only 2.95 at 5 favours and 3.1 when done at 7 favours)
Tombs: 7.6 EOR/f via CoC gained in war of assassinss
Rubbery: 3.51 vs. revolutionaries (at 5 favours, but gives unaccounted-for scandal) or 3.25 vs. both constables and tombs (the last one also gives 10 carnival tickets, if you still have use for those)

--
http://fallenlondon.storynexus.com/Profile/xKiv - a witchful, percussive, dangermous and shadowry scholar of coexplodence, hopsidirean, and walker of fallen kitties.
+1 link
Kaigen
Kaigen
Posts: 423

3 days ago
This also places the value of Connections that have a partially updated conflict card (Society, Widow, Constables) at about 0.05/cp.

I used it to investigate the EOR of accepting A Polite Invitation. Assuming base 1.64 EPA, Talk of the Town can be valued at 0.6448/cp. An efficient route through the party looks like...

Accept Invitation: 1.29
Arrive Late: 2.58
(If rare success, gossip about the Brass Ambassador =1.93)
Argue with the Contrarian: 2.58
Help the Tentacled Entrepreneur: 4.01 (Rubbery Favour plus 50 deep amber)
Evening Draws to a Close: 1.29
Taking your Leave: 1.5 (from 30 cp of Connected: Society, assuming TotT isn't high enough to cash in)

That comes out to an average of 2.21 echoes, giving an EOR of 0.57.

Interestingly, the calculated value of Tomb-Colonist Favours suggests that fueling the CoC grind is in fact the best use of said favours, though I still hesitate to make that the base rate grind, given that I seem to grind Collections faster than Favours overall. When one has twenty Collections waiting to be traded in, it's hard to argue that the best use of one's actions is grinding still more Collections.
edited by Kaigen on 4/19/2017

--
Just a simple doctor with a chess habit. Publisher of The Flit Dispatch. Available for lessons in the art of seeing things.

Average actions to complete Flash Lay: ~21
Number of Surprise Packages opened: 350 Average Value: 6.29 Echoes
http://fallenlondon.storynexus.com/Profile/Kaigen
+1 link
Pumpkinhead
Pumpkinhead
Posts: 320

3 days ago
Kaigen wrote:

When one has twenty Collections waiting to be traded in, it's hard to argue that the best use of one's actions is grinding still more Collections.
edited by Kaigen on 4/19/2017

Yep, I ground about 20 a few months ago. I think I've managed to turn in 4 of them.
This is even with going to the colonies to get favors instead of using the card.

--
McGunn is open to social actions, although his plant isn’t fighting and he doesn’t really need to see the affluent photographer anymore.
Amanda Albright is a *spoiler* now. It’s what she always wanted. Follow your dreams, folks, and maybe you can be a *spoiler* too.
0 link
An Individual
An Individual
Posts: 515

2 days ago
This sounds similar to a calculation I made a few times during my cider grind. The question that comes up with you're only looking at EPA, as I understand the framing here, is "would you rather spend 1 action making 10 echoes of 10 actions making 5 echoes each?" If you decide to narrowly focus on single EPA the first seems like the better option. But the real question you should be asking is "what could I be doing with those extra 4 actions?"

My solution to this problem was to normalize the number of actions taken when comparing approaches, with the EPA of a repeatable grind like The Affairs of the Box or the Soul Trade being used to calculate the echoes gained with the extra actions. Here's basically how that math works.

P1 = profitability of option with more actions.
P2 = profitability of option with less actions.
A1 = actions taking in option 1.
A2 = actions taken in option 2.
G = EPA of repeatable grind

Option 2 is better than option 1 if.
P1 < P2 + G * (A1 - A2)

Unfortunately, this system is only useful for comparing 2 options and different values for "G" will give you different answers, especially if the numbers are close. You can't generate a number for a single option and then compare it universally against similar numbers. Unfortunately, I don't have time at the moment to run the math-foo necessary to know how this method compares to your EOR for comparing options with different numbers of actions. It handily answers the question "is this better than just doing my repeatable grind?" But I'm not sure how it works for the question "how many more echoes will I have if I take limited option X over limited option Y?" My instinct is that without normalizing the actions you won't get a good comparison but I'd put an extremely low confidence value on that statement.

--
An Individual's Profile
The RNG giveth and the RNG taketh away.
Goat Farming or Cider Brewing? This browser extension may help.
Want a Cider sip? Please refer to this guide before requesting.
Scholaring the Correspondence? A Brief Guide to Courier's Footprint.
Contemplating Oblivion? First Steps on the Seeking Road.
Gone NORTH? Opened the gate? Throw your character in a well.
0 link
xKiv
xKiv
Posts: 620

1 days ago
That's exactly the same thing:
EOR1=P1-G*A1, EOR2=P2-G*A2
EOR1 < EOR2 iff P1-G*A1 < P2-G*A2, which can be equivalently rearranged to P1 < P2 + G*(A1-A2).

--
http://fallenlondon.storynexus.com/Profile/xKiv - a witchful, percussive, dangermous and shadowry scholar of coexplodence, hopsidirean, and walker of fallen kitties.
0 link
12




Powered by Jitbit Forum 8.0.2.0 © 2006-2013 Jitbit Software