Excel Simulations and Ivory

by Slaven

With Ivory comes the single largest change our game has had in its 15+ year run.* (Never forget – 7/29/11)  With the changes to gold pooling, holdings are now more important than ever. Getting behind in gold production early cascades into future turns.  In order to provide insight on gold schemes in Ivory, I have rebuilt my old calculator, using the complicated new rules for gold schemes.

This article will focus on answering three of the most basic questions for the most common decks, which are military decks under 4g producing strongholds. Strongholds that produce 5 or 3 gold, decks that have ludicrously cheap personalities (Scorpion Dishonor), or unique abilities (going second for Unicorn/Mantis) are NOT covered in the scope of this article.

Myth 1:  15x holdings was good enough for Emperor, and it’s good enough in Ivory.

Running 15 holdings is -not- enough in Ivory. Given a 40 card deck and a 4 gold production stronghold, I ran the numbers on a 15-holding deck with three Famous Bazaars (note: in order to reduce complexity, the model actually understates the value of Famous Bazaar), three other 2-for-2s, three clan holdings, and six 4-for-4s. This deck, on average, purchases 4.21 gold in production by the end of turn 1, with an ~8.6% chance to have purchased only 3 or less gold. Assuming a 16 game tournament (swiss and elims through the finals), you only have a 24% chance at avoiding gold screw over all of your 16 games. Put differently, there’s a 76% chance that you’ll get gold screwed in at least one game that tourney (gold screw defined as being able to purchase 3 or less gold worth of production at the end of turn 1).

Compare that to a similar deck running 18 holdings. The holding setup is the same as before, but with three 2-for-2s thrown into the mix (three Famous Bazaars, six other 2-for-2s, three clan holdings, and six 4-for-4s). This 18 holding setup results in 4.35 gold in production on average, by the end of turn 1, with only a 4.1% chance to get gold screwed.  By adding those three 2-for-2s, you’ve halved your chance to get gold screwed.

Myth 2:  I have to have rares in order to have a competitive gold scheme.

False! The 18 holding setup I talked about previously makes an assumption that the player has two playsets of rare holdings (Famous Bazaar and Productive Mine, both of which appear in Ivory Edition starter decks). Given the price I believe these two cards will reach, this might be a hefty assumption. Is the budget-conscious player screwed, for lack of a better term?

Nope. Nexus of Lies is a common holding that produces 4 gold at a cost of 4 gold. Running the numbers on a 4 gold producing stronghold with 18 holdings (twelve 2-for-2s, three clan holdings, and three Nexus of Lies) we arrive at an average of 4.33 gold in production by the end of turn 1, and a ~5.6% chance to get gold screwed.
4.33 gold with ~5.6% chance of gold screw for a deck using no rares compares rather favorably with the deck that does run rares, which sees 4.35 gold and a 4.1% chance to get gold screwed.

Myth 3:  Then you mean rares are worthless?

False! Rares are still worthwhile, I promise. Remember that my calculator is only focusing on what happens at the end of turn 1. Both Famous Bazaar and Productive Mine really shine later on, when you’re sick of holdings. The utility in those abilities is impossible to model, and really comes down to player preference. I have not covered it here, but trust me when I say that it’s darn near impossible to run Secluded Shrine or Tea House efficiently without running Jade Pearl Inn.

In order to have the deck that is running all common/uncommon holdings equal the production of the deck running rares, you have to use 19 holdings (thirteen 2-for-2s, three clan holdings, and three Nexus of Lies). Doing so gives you the same 4.35 gold average as the 18-holding rare deck, and a gold screw chance of ~4.2%. Is that event or extra personality worth the cost of getting rares? That is a decision that each of you will have to make on your own, but if you do decide they’re worth it, I highly suggest Strange Assembly sponsor l5rshop.com, which puts cards in your hands quickly and economically (or to simply find out about unusual cards).
At the end of the day, I believe that design has done a fantastic job of making rare holdings strong, but not so strong that they are a requirement for tournament-level play.
Ivory1 Ivory2

11 thoughts on “Excel Simulations and Ivory

    1. I’m a little swamped in IRL at the moment, and would have to scrub some stuff from it before I did so. I’d be happy to run any simulations for you though, if you have a gold scheme.

  1. Hi, thanks for the great article !

    I have one more question : What about a deck with 9x444s ? Considering that one Jade Pearl Inn + one 141 holding = one 444 holding, that can already be done

    Something like :
    3x Nexus of Lies
    3x Productive Mine
    3x Jade Pearl Inn + 3x 141 holdings
    6x 242

    That’s 18 slots but actually it’s similar to 15 holdings, 9 of them being 444s

    Would this scheme require more/less holdings, or are the odds similar to a 18 holdings (with 6x 444s) gold scheme ?

  2. Thank you for the kind words! The 18 holding setup you list in your post has a 4.1 mean gold production purchased by the end of T1, with a 6.3% chance of only purchasing 3 or less gold.

    A setup I discuss in this article, the second one under ‘myth one’, has better numbers, and does not run JPI. It has around a 4.1% chance of seeing 3 or less gold, and buys 4.35 gold in production on average by the end of T1. That said, I absolutely love JPI, and love the ability to tutor out my meta-holdings (the 1 gold producers, as Yukihime’s Hot Springs is great meta) is going to be worth it to me. JPI also thins the deck while shuffling the personalities you put on the bottom of your deck back in to circulation, and so it’s an amazing card for me.

    I’d probably suggest running Famous Bazaar in the example you proposed in your post.

    I hope that answered your query. Let me know if you have other questions.

  3. Can you run a simulation on this scheme?

    # Holding (18)
    1 Bamboo Harvesters – exp
    1 Counting House
    2 Famous Bazaar
    3 Gold Mine
    3 Jade Pearl Inn
    3 Nexus of Lies
    3 Productive Mine
    2 Secluded Shrine

    It’s close to the col #3 (1 less bazaar, add 1 Bamboo) but I’m curious how close.

  4. 0 gold 0.69% of the time
    1 gold 0.91% of the time
    2 gold 2.49% of the time
    3 gold 1.41% of the time
    4 gold 74.18% of the time
    5 gold 14.29% of the time
    6 gold 6.03% of the time

    An average of 4.145g. 5.5% chance of getting 3 or less gold worth of production. It’s worth noting that in my model, Bamboo harvesters xp is essentially not a holding, as I’m only concerned about the amount of production in holdings you purchase by the end of T1 (and BHxp will always be 0). Apologies for the delay.

  5. Hi again,

    Thanks for your answer. I tried the 18-Holdings gold scheme (with 6 444) and indeed, the number of gold screws is really low. However now I face the opposite problem : Gold overflow !!! Which is quite as bad as gold screw, since you have loads of money and no personalities to spend it… I know Family library, House of Exotic Goods, Imperial Dojo or Kharmic holdings are meant to reduce this drawback, but I’m not sure that it would work very well.

    I know such a simulation is really hard to run as it addresses later turns in the games when lots of things could have happened, but I would appreciate your opinion on that matter !

    1. If you’re able to run holdings that thin your deck (Famous Bazaar, Jade Pearl Inn, or even holdings with Kharmic), it shouldn’t be too much of a problem. Good luck!

  6. With the new 444 holding previewed, what does this scheme look like:

    (15) holdings
    3 Bountiful Fields
    1 Counting House
    3 Jade Pearl Inn
    3 Nexus of Lies
    3 Productive Mine
    2 Yukihime’s Hot Springs

    What happens when you add 2 famous bazaar?

    Thanks for spending the time to humor me.

  7. Any chance you’d post your simulation? Would be a wonderful resource for a notoriously difficult environment to calculate gold schemes.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.