Math people needed to help in analysis for Omen custom mod....

I'm in germany and playing with the german edition - which is, AH+Dunwich. KiH and CotdP aren't available in germany yet - they are should be in a couple of months. Kingsport will probably take another year at least.
So, I for my part won't bother about the Kingsport AOs for a while and instead focus on completing the Omens for the other 12 AOs when I find the time. Feel free to make your own expanded omen deck variations, but I hope you understand that I won't take the new AOs into consideration when designing my cards.

MavHunter,

Making Odgru Jahad an AO is a freakin awesome idea. And while your at it make Rasputin his herald.

Hellboy has so much love craft in it there is no reason he cant easily be incorporated into this. This makes me think Hellboy, Kate, Liz, Abe, Roger, and Joana (sp?) should be made as investigators. I can already think of their special abilities. Of course some items will need to be made for their starting equipment. hellboys revolver. Liz can have a spell called pyroblast or some such. Joanas special ability could be something along the lines of psionic touch. Where instead of having an event in an unstable location he can roll a die and gain a clue token. on a 1-2 he also looses one sanity (basically he used his psycic powers to see what has happened or is going to happen and gains clues for it, but if it's overpoweringly bad (the bad roll) he takes a sanity hit for it).

Yhtill,

The thing i was saying was to consider the math for balanceing the deck to incorporate not the Kingsport AO's but any additional AO's. Think of it like this. Basic game plus DH = x amount of cards. in the ful deck each elder god gets x cards with their name on it.

If 2 more expansions come out each with 4 more AO. How many cards should be added and how can you balance the original deck so that adding these new cards does not throw off the balance of which gods get choosen at random.

My suggestion is less about kingsport specifically. And more about designing the basic set to be able to be expanded easily without big math problems.

Isn't it quite easy to add AO's in the deck? Now the deck consists 30 cards where each AO is 5 times (once in +3, twice in +2 and twice in +1). If you want to add four new AO's, just add those +3 cards and

+2 AO1, +2 AO2
+2 AO3, +2 AO4
+2 AO1, +2 AO4
+2 AO2, +2 AO3

cards and two

+1 AO1, +1AO2, +1AO3, +1AO4

cards. So it's 10 new cards for 4 new AO's. The probability is the same or am I missing something here? But of course adding more AO's and cards makes the AO's appear later. I think this way when the AO's are "grouped" it's easy to add or remove AO's in the game.

EDIT:

And I tried some time ago those Omen cards and really liked them. Although I was beated by Shudde M'ell after I drew the +3 Shudde card (add 2 cthonians) and +3 Tsathoggua (pay 1$ for use of special abilities at locations). Cthonians made me unconcious all the time after I had lost all my money in hospital

It was bad luck but that's what Arkham Horror is about

greenrowan wrote:MavHunter,

Making Odgru Jahad an AO is a freakin awesome idea. And while your at it make Rasputin his herald.

Hellboy has so much love craft in it there is no reason he cant easily be incorporated into this. This makes me think Hellboy, Kate, Liz, Abe, Roger, and Joana (sp?) should be made as investigators.

green- i've already started on Rasputin, and as for the BPRD, i've taken some ideas from the ones posted on BGG and revamped them to better fit how i see them:

http://www.fantasyflightgames.com/ffgforums/posts/list/105/109.page#71082

The problem with adding in new AO's outside the initial formula is that those new AO's will be statistically different than the existing build.

If you add 2 custom AO's, and make 3 or 4 cards that just reference them, you will not get the same statistical output as if you balanced all the AO's at one time. (and even doing that isn't giving exactly even results)

Also, the more customs you add, the longer it will take to reveal the AO, and I we need to keep the average down to 4-5 turns if possible.

I think the best solution is to have more AO's per Tier 1 and 2 cards, keeping the total number of turns down.

Yep, MillMaster described exactly why adding more Omen cards later on isn't that easy.

When looking for possible tier-1 / tier-2 combinations, I used a geometrical method which makes the balancing problem visible and reduces the amount of math needed:

- Just write all 12 AOs Names down in a circle
- make sure that those who share a tier-2 card are next to each other
- now take a color - say, green* - and connect thos next to each other. These green lines represent the tier-2 cards.
- now take another color - say, red* - for tier-1. Connect the 4 AOs which share one tier-1 card with a rectangle-thingie.

To get an unbiased distribution, try to draw all 6 tier-one-card-rectangles in a way that
- every AO has exactly 2 connections
- no line appears twice (which means, no pairing appears on two cards
- no AO is connected to a neighbouring AO (allready connected at tier-2)
- and the whole pattern is as symmetrical as possible

This isn't as easy as it sounds, but it is possible. And when you look at the resulting pattern, it is obvious that you can't just add another AO without totally breaking the balance and simmetry.
So, essentially, if you want an omen deck for another number of AOs, you have to start the process of figuring out the distribution again - else the whole deck will be unbalanced.

For this reason, I'll stick to the current 12 AOs - and may redesign the deck with more AOs when they become available.

@MillMaster:
Have you run your calculation with all tier-2 cards? As I understand it, your last test was done using only 6 of the 12 tier-2 cards. I'd like to know if the complete set turns out balanced or if there is still work to do.

__________________
* Please, do not use yellow. Starring at strange yellow patterns can damage your sanity.

I guess I am not sure if I have all the Tier 1 and 2's in their finalized order. Can you repost or email me something just to be sure, and then I will run it again. I should have the Standard Dev worked into by tomorrow so I will post those numbers too...

I was wondering if anyone finalized anything with this and had some file i could steal.

*Shrug*

I played with this idea long ago, and pretty much thought you'd just put two AO's on each card, and that was that. But I then ran into problems thinking up every single freaking theme effect. Much worse, I wanted a variant that allowed each player to *influence* which AO came up. If "their" AO appeared, then that investigator, and only that investigator, would have some immunity to the AO.

You could just have one card for each AO, but then draw two cards per turn. Each card would have three sections: One advances the AO, the second would be a "verb" and take effect if this were the first card drawn, and the third would be a "direct object" describing what the verb acts upon.

For example, the first drawn AO card's verb section says, "THIS causes an additional 1 sanity loss when encountered" and the second drawn AO card's direct object section says "Tomes". So, for the turn, whenever a player acquires a Tome, they lose 1 Sanity. Again, probably too much work to do right thematically.

Anyway, I ended up tossing the cards completely, and letting a player choose three AO's to advance. If "their" AO was eventually chosen, they gained immunity to it. This variant optionally also chooses which variant to play. It's pretty much done, but I still have to make a counter for each AO.

aka. Washu! ^O^