How is Ark Nova balanced?

[@thefacethatstares]

the strategy tip of "try to play animals with the highest possible requirements" is extremely valuable, thank you. time to win some games :)

[@3vilash]

Already my favourite channel!

[@IvanIurchenko]

Best strategy video I've seen for any boardgame

[@strategyadvisor5604]

Thank you so much! I put extra effort into this one.

[@timmaddams1093]

Amazing work. Someone in my play group assumed I had this already as I "knew" which were the better animals.

[@vincentskater]

This was so cool, watched the whole thing and was fascinated!

[@EmDubya1]

Fantastic insight! I would assume that the Fur Seal is undercosted in your model because the combination of a rock AND water requirement on a single animal makes it much more restrictive to play

[@strategyadvisor5604]

I think you're right about that!

[@riptorius-d2y]

I would also think that the association worker is less valuable as the game goes on, and it is difficult to get an upgraded animals action early.

[@林雨昇-n1r]

I looked through the entire internet and I am glad that I found this video

[@thraka2]

I appreciate the strategic implications you derive here. I have played maybe a dozen times and have only won maybe 3 games (usually a close 2nd), so it is helpful to understand the balance.

[@phoenicksgaming]

This is impressive stuff

[@kornelije.h]

Fantastic video, thank you. I would like to see an analysis like this for other popular euro games.

[@patrickschneider4015]

What an absolutely fantastic video and analysis. It feels like we are getting a full on peak behind the curtain. It also makes me appreciate the game more. Its complex execution has a deceptively simple underlying mechanic.

[@strategyadvisor5604]

Thank you for the kind words! My main focus is usually strategic conclusions, but this video I wanted to give a deeper understanding. Converted value is a useful tool for many games.

[@sammysammyson]

I have no idea what any of this means, but KZbin brought me here, and game design is fascinating. Neat.

[@mannycarrasco8377]

Oof, let me take you down the rabbit hole. It's for a board game called Ark Nova in which you design a zoo and place animals in enclosures to increase the appeal of your zoo, all while you fund conservation projects to increase your overall score. Highly recommend, super fun and very deep in strategy.

[@BeesAndSunshine]

A lot of valuable insights. Clever is hard to use properly, so it's good to hear it is valued lowly and I shouldn't be scared to take it. I'm really surprised that reputation is equivalent to appeal, most animals will give you the baseline appeal needed to get the most efficient early income, but early reputation is hard to find and can get you an early card unlock and worker, as well as giving you more selection on upgraded cards, animals, and sponsors, and endgame every two or three appeal gets you a conservation point on top of all the bonuses in between. Personally I value it somewhere between appeal and conservation points. I did not even know that animals with multiple continent icons got multiple partner ship zoo discounts! Assertion seems pretty undervalued as it can often net you 5 conservation points on the final turn (minus the value of your next best play) where your opponent either gets either 2 or 0 and if they get 2 it is at the cost of their next best play so probably close to 0. Posturing seems really good before you upgrade build since it's both action and cost save and fairly weak afterward.

[@kevinmcgaire8684]

For the ones that require partner zoos they do inherently become cheaper. I wonder if that should be factored into the bison for example

[@strategyadvisor5604]

It kinda is because the 2nd continent tag is worth 3.

[@wetgos8286]

Very interesting! Great work

[@林雨昇-n1r]

Hello Mr. Advisor, your analysis is incredible. Can I have a look at the original data? I would like to know more about the enclosure and jumping mechanic and the possibility that they can go beyond the times two calculation. However, the data is presented in the way that they are lumped together. I would want to take a look at the original data that has not been lumped together. Applying statistic to the board game is such an awesome approach! Keep up the good work. Cheers

[@strategyadvisor5604]

Thank you! That's the only data I have, I created it manually by viewing the cards.

[@verlausterhippie]

Thats cool. I think another interesting aspect would be to look at the ratio of "useful" cost on a card, e.g. cost for appeal, rep vs the dead cost like enclosure. Is there any trend, for example when comparing small and large animals?

[@strategyadvisor5604]

I'm not sure I understand correctly, but if you mean efficiency excluding requirements, then the best card is King Vulture with fairly low cost and high requirements. 30 value vs 16 average cost, 14 cost if in aviary.

[@verlausterhippie]

​@@strategyadvisor5604yea I guess somehow what I meant is already priced into it by the inclusion of enclosure size. Was in general wondering if maybe larger animals are more efficient but that would mean just comparing the requirement numbers then.

[@strategyadvisor5604]

@@verlausterhippie Larger/small animals are pretty balanced, except at the end of the game you want to play large animals for action efficiency.

[@arthasthelichking6213]

Very impressive! I was looking for the value balance articles for ark nova but I couldn't find any. Is this based on someones work or was it all your work? You might be the first person to solve the value balance to ark nova are share it!

[@strategyadvisor5604]

Thanks! I did it from scratch.

[@janikbuser2604]

Another takeaway from this is that bigger animals are stronger than smaller ones, cause there is no "base value" for the card that would value the cost of adding that card to your hand and using an action to play it (and if you haven't upgraded build multiple actions to build enclosures).

[@strategyadvisor5604]

Yes I thought about that, it's complicated. Each animal costs a card (-3) and gives 2 tags (+4). So the best size animals are whatever spends all your extra money. Late in the game, that's big animals.

[@markvanderwerf8592]

Big animals are more effective in terms of card draw actions, build actions and animals actions. The limiting factor tends to be money though. Smaller animals are more effective for icons / conservation projects.

[@Hoellenseher]

​​@@markvanderwerf8592in general smaller animals are mostly good for the icons, but have less requirements and profit more from build II Larger animals tend to brick your hand but are overall more valuable, while the small ones feel like fillers. Especially late in the game larger animals have requirements that can't be met if you had no time to plan around them. So there is a holding cost in hand size for larger animals

[@markvanderwerf8592]

@@Hoellenseher i think it often doesn't matter that much. Smaller animals you need more draws, builds and animals actions but you tend to have enough of those and early on money tends to limit you. Smaller animals help you get conservation projects faster and better kiosk income. But bigger ones save a build action usually. Later on money tends to be less limiting and actions become more making larger animals more attractive. But even then with animals, cards and build upgraded you still find yourself playing small animals often.

[@_DaemonCleaner]

Very cool. The negative values - I was under the impression that the game has uneven numbers of animals from various continents and types. I.e. Australia and Europe has different amounts of animals. Same for primates vs herbivores. That means some prerequisites are more difficult than others, even within the same category. Should that be added to your model?

[@strategyadvisor5604]

That's a good point! I just checked the Mandrill and Proboscis Monkey, which have multiple primate requirements, and those have 0 error. So while those requirements are more difficult, they don't give extra reward. Strategic implication: it's better to speculate on animals like Lion or King Vulture than those monkeys. This only applies when you don't have the requirements yet. Next I checked the iconic animals, expecting Australia and Europe to be the most undercosted, but it was actually Australia and Africa. The frequency of each tag is good to keep in mind, because it doesn't appear to affect how the game was balanced.

[@AdamHorton01]

This is pretty interesting stuff. When you calculated your total error did you take the absolute value of the error and add them all together? Or did you assign a penalty for an actual cost being really far away from the expected cost? Also, I assume you didn't try every possible combination of values for what things cost, but rather just tried stuff out until the error was as small as you could get it, right? There are a couple of things that aren't costed that I think are interesting as well (assuming you can't use them somehow to improve your error): enclosure size also requires build action strength but larger enclosures seem to "cost" the same as smaller enclosures. Extra actions, boosts, and multipliers seem to be costed the same even though they apply to different actions, and let's be real, boost animals is going to be good a lot less often than anything with association on it. There are probably more of these that are worth exploring

[@strategyadvisor5604]

For error I did sum of absolute values. I try to increase the value and decrease the value of each variable, when both of those result in higher error then I have the right value because it's linear. Enclosure size always as 2$ per hex seems to be most accurate. There are about 1 of each boost/multi/action for each type of action, so there's not enough data to give each one it's own value (overfitting). Also the association ones happen to have 0 error. Sun Bear (action: association) is probably the best late game animal, but it's a bad starting animal.

[@ZookeeperArthur]

@Strategy Advisor Can you please do Terraforming Mars next?

[@JDSecretPolice]

Very interesting. Confirmed my suspicion that snapping is over-valued. Also a bit shocked on how expensive constriction is, usually feels like good value

[@strategyadvisor5604]

All the constriction animals have requirements that at least cancel out the value of constriction. So on pure $ to appeal they are fair.

[@newmancl0]

This is great! I need more insights specifically for the solo game. I've played 4 dozen times on BGA and i'm just getting slaughtered. I do quite well in games with other people but the solo game.....

[@strategyadvisor5604]

In solo it's very important to have sponsors to play, because breaking for 5 doesn't do much. See my Ark Nova Tier Lists video to know which sponsors are worth building around. Snapping is bad and should only be used for key sponsors and projects if they are about to be discarded. Play animals directly from rep range instead of snapping them, this lets you get deeper in the deck. You are not competing for cards so there is no rush. Association is a worse upgrade (good donation spots get taken) and Sponsors is a better upgrade than usual so you can play more cards from the row.

[@newmancl0]

@@strategyadvisor5604 Thanks for the tips. I'll check that sponsors tier list out. So you would recommend upgrading Animals quickly then to start playing animals from rep range? I feel like I can get a good amount of appeal OR a good amount of conservation points. But Im just not efficient enough to get both. First time I play Ark Nova I really hated it :D. Now i've played about 10 games with other people and about 4 or 5 dozen solo games and i'm loving it.

[@strategyadvisor5604]

@@newmancl0 Yeah, Animals is good to upgrade first or second. You might need to work on your action efficiency. Play 2 animals at a time when you have enough money. Rarely take x tokens or 5 money as an action. Focus on animals and association actions, especially late game (best use of x-tokens).

[@Rhajalob]

will water worlds be added?

[@arivald6677]

Good job!

[@NoWondersTV]

Wow. This is great stuff! How long were you fiddling around with the formula to get accurate results? And there must be some discrepancy with Petting Zoo animals, right? Seeing as they are all the same price but some like the Horse are strictly better. CP makes sense as early CP are worth 2 appeal while late CP are 3. I think the biggest implication is how valuable animals that give reputation are. It being the equivalent of 1 appeal or 1 card draw is insanely undervalued.

[@strategyadvisor5604]

Finding the right numbers wasn't too hard once I set up all the formulas so it automatically calculates the error. I could just check whether the error went down after each change, which is basically gradient descent, the same algorithm used in machine learning. Petting zoo animals are excluded from the model because it depends so much on how many you play. Horse is 4 and those without a bonus are 7. If you play 2 they are a combined -3, and all 3 are -21. That includes the penalty of 3 for lacking a continent tag. Yes I expected reputation to be 4. The first 8 rep gives an upgrade and worker, which I thought would be worth closer to 32 than 24. After that most rep is worth 3 (x, card, appeal) except the 2 CP, but you could count that value as coming from the Cards upgrade rather than the rep.

[@embo1345]

@@strategyadvisor5604 with regards to petting zoo: -21 meaning they are neutral cost/gain because it cost 21 to play them? or -21 meaning they are very much cost effective?

[@strategyadvisor5604]

@@embo1345 Very cost effective, you save $21; or get 7 extra appeal for free.

[@GameTester-i4o]

Do one on terraforming mars! It would be great!

[@whothennow24]

I haven’t played in a while, but it seems as if you’re arbitrarily assigning symbols a value. Like how is an ability worth 4, and how is an appeal worth 3? What are you basing that on? Is that in the rulebook?

[@strategyadvisor5604]

No, I calculated it with gradient descent. It would be impossible to model so accurately if the designer was not using this kind of balance system.

[@lestar123]

Ark Nova Water World expansion!

[@grantofat6438]

So basically it doesn't matter what you get in the game, because they are all equal.

[@Naeddyr]

All animals are equal, but some animals are more equal than others.

[@strategyadvisor5604]

The animals are well balanced so synergy matters more than raw power. The sponsors are imbalanced, I recommend looking at my tier list video to know which sponsors to prioritize.

[@PaulSzkibik]

Cards being balanced is a good thing. Especially in a game with so much random draws. Like who wants to play a game where Lions (as an example) are really strong but in some games you never draw them but your opponent draws two? That would suck. It also doesn't mean that it doesn't matter what you play and when you play it. It's been clearly outlined in the video, that you hopefully watched completely, that some cards are better played early and some better played late.

[@Bittler87]

Hey the icons that are requirements on the left side (like animal and country tags) do not count toward total icons.

[@strategyadvisor5604]

Yes I know.

[@Bittler87]

@@strategyadvisor5604 I figured, maybe I misheard what you were saying in the vid.