Concerns regarding the Eureka mechanism

[kaspergm]

I think the Eureka mechanism is a nice idea, but I'm quite concerned about the way it is implemented in the build we've seen at this point. The way the idea was originally presented, this mechanism was supposed to shape your development depending on your specific starting conditions and also to reward you for pursuing a very specific game style. I think this is a great idea to make the game more varied.

However, for this to work, the Eurekas need to be designed so that either they have high opportunity cost in return for a high yield, or they have low opportunity cost in return for low yield. Unfortunately, that's not really what I see with the current Eureka conditions. Most of them are low opportunity cost in return for high yield, and that's very bad. Going through the list, most of them fall under one or several of these conditions:

1) Extremely generic, i.e. something you'll be able to do in almost any game, and/or
2) Extremely easy to do, i.e. comes at no opportunity cost, and/or
3) Is something you'll always aim to do and will do automatically.

The full list of Eurekas can be seen on Arioch's Well Of Souls, but let me just give some examples (denotes which of above criteria I think they fall under):

- Find a natural wonder (1, 2, 3)
- Meet another civ (1, 2, 3)
- Kill three barbarian units (1, 2, 3)
- Conquer a barbarian encampment (1, 2, 3)
- Mine a resource, build a quarry, farm a resource, build a lumbermill, build a pasture, etc. (2, 3)
- Make a trade route (3)

I really hope they'll do something to make many of these Eurekas less generic and something you'll have to work actively to achieve. If the Eureka happens automatically every single game at no cost, it's really pointless. They could work with a gradual mechanism, for instance you get a 10 % tech boost for each time you fulfill the criteria for a maximum of 40 %. This would emphasize the flavoring towards certain play stiles / starting conditions.

 

[gfeier]

This is the introduction of a new mechanic to the Civ series. I'm sure it will evolve over time, possibly before the game is released. I think it has some potential, although your concerns are valid.

 

[SupremacyKing2]

I do think that the eureka bonus might need to be toned down a bit. 50% is too high IMO.

But one thing that you seem to be missing is that eurekas do require timing. You can't get a eureka after you already research the tech. Marbozir in his videos explains that while players will get some eurekas by accident just from the regular play style, if you want to maximize your research by getting the most eurekas at the best time, you will need to play more carefully and plan about 10 turns in advance. For example, sure everybody will build a quarry at some point, but optimizing your play might require that you use your builder to get that quarry first before a farm in order to get the eureka at the right time.

 

[kamex]

I'm not too concerned. I imagine if you want to beeline a tech as soon as possible, you may not have enough to turns to mess about getting all the requirements for every boost?

Some will be 'luck' based such as finding a natural wonder early, but I guess that adds a bit of realism...

 

[stiiknafuulia]

I very much share this concern and hope most of these are placeholder Eurekas. Building 3 Tanks in the Modern Era is hardly an achievement worthy of a 50 % tech boost; the Eureka might as well not even exist at that point.

 

[stealth_nsk]

It doesn't look like that to me. The important point - getting boost before completing the tech/civic. And it's very tricky. If you watched the videos, even Marbozir who specifically targeted boosts on Prince level was able to get around 70% of them. With more polished builds and on higher difficulty levels this could require more efforts (although game speed could be a factor too).

I'd separate them differently:

1. Random/exploration like meeting civ or finding natural wonder. Since these are required for very early techs you could use the opportunity if one of the events occurs. I.e. if you found natural wonder early, you could aim for Stonehenge.

2. Requiring aiming for the boost. For example, normally you wouldn't want one of your first cities to be coastal, but if you're aiming Sailing, that's good idea. Building specific improvements is in this category.

3. Benefiting specific playstyle. Bonuses from killing with specific units are here to boost techs related to military, bonuses from building different kinds of districts for builder techs, etc.

 

[Hakan-i Cihan]

I have similar concerns as OP. But we still have time until game release. I think that a good indication for a proper balancing of the Eureka system should be that on average you should be able to acquire around 20-30% of eurekas and inspirations and if you plan it well the maximum should be around 40%.

 

[janboruta]

From what it seems so far, those eureka/inspiration boosts are getting progressively harder or more situational deeper into the trees. And some, like Computers, require the player to switch to a modern government form with 8-card slots to receive a boost, so if someone elects to stay with a 6-card mid-tier government like the Merchant Republic for the rest of the game, he will miss out on that. It's a nice opportunity cost kind of moment and I very much doubt getting most of the boosts in any given game will be viable. (although I suspect there will an achievement for getting 100% of boosts in one game )

I personally think that a 50% boost is too much, though, and some balancing will surely be needed, but as a concept it is really nice and engaging the player (maybe a 25% boost would be better, and China receives 50% due to their UA). As Marbozir said in one of his Brazil episodes, it is an improvement over the Civ 5 system, which only requires you to click what tech to research and forget about it until it's complete.

 

[demidyad]

But one thing that you seem to be missing is that eurekas do require timing. You can't get a eureka after you already research the tech. Marbozir in his videos explains that while players will get some eurekas by accident just from the regular play style, if you want to maximize your research by getting the most eurekas at the best time, you will need to play more carefully and plan about 10 turns in advance. For example, sure everybody will build a quarry at some point, but optimizing your play might require that you use your builder to get that quarry first before a farm in order to get the eureka at the right time.

I agree with this, and another thing to think about, is it's not just about yes/no whether you do or not, it's about the timing and how it affects the order you research your techs. You might think "Oh I really want Irrigation, but I've already completed the Eureka for Masonry, so should I quickly grab that now, or should I focus on getting Irrigation ASAP?" or "Hm, should I build another warrior to take out that encampment and boost Bronze Working? That seems risky when I could just pump out another galley and boost shipbuilding instead" and things like that, leading to a reactive and adaptive path through the tech tree.

They're not supposed to be rewards for achieving something remarkable, they're supposed to be an organic feedback loop between the game world and the tech tree.

 

[Ryika]

The really important thing will be how well the AI can deal with Eureka Bonuses. If they're good at it then it's mostly about tuning of the original cost based on how many Eurekas you're expected to get. If not, then there's again a large gap between AI and Players that must be filled with bonus modifiers.

That would then however indeed miss the original Design Idea of shaping your empire around your style of playing and mutate into a system that wants you to do a bit of everything to unlock as much as you can with an Eureka.

 

[SammyKhalifa]

I'm thinking of it in a different way I guess. We all know that the developers don't like adding penalties into the game if they can help it (they've said as much), but this might be a hidden set of them. Does it help to think of the Eurekas not as bonuses for fulfilling certain criteria, but penalties for NOT doing them? It's not easier for coastal people to make boats, but rather more difficult for landlocked ones to fully understand them.

 

[Acken]

To me that system look extremely gamey with its current design.
It will make people do random stuff not because they need it but because of optimizing boosts. Wich i think is unintuitive and contrary to the design of the system.

Like the op I think the rewards need to be lower or the tasks to be harder. It should either be more map based so that it varries a lot from map to map or should reward some strong detours out of an economic game.

 

[stealth_nsk]

The really important thing will be how well the AI can deal with Eureka Bonuses. If they're good at it then it's mostly about tuning of the original cost based on how many Eurekas you're expected to get. If not, then there's again a large gap between AI and Players that must be filled with bonus modifiers.

If you can't see it, it's not important. It's totally possible to fully disable eurekas for AI and replace them with some standard bonus like 30% for all techs/policies. It will not affect gameplay as you'll be mosty unable to tell the difference.

 

[Ryika]

If you can't see it, it's not important. It's totally possible to fully disable eurekas for AI and replace them with some standard bonus like 30% for all techs/policies. It will not affect gameplay.

That's nonsense. First of all, you can indirectly see it in the result, and inevitably disabling mechanics for %-modifiers leads to a simplification and homogenization of the AI, that will cause problems whenever an AI diverts from the "generic playstyle" that these modifiers were balanced for.

Take for example a military AI, with a high military flavor. It will produce tons of units, would fulfill many Eurekas that other AIs would not, but then NOT get the Eurekas for producing them, at the same time not produce much "natural" science compared to other Civs and automatically fall behind, while if Eurekas work it would open up a gap where it can keep up because of the Eureka.

Making the AI work with Eureka Bonuses is a pretty big priority in my opinion.

 

[stealth_nsk]

That's nonsense. First of all, you can indirectly see it in the result, and inevitably disabling mechanics for %-modifiers leads to a simplification and homogenization of the AI, that will cause problems whenever an AI diverts from the "generic playstyle" that these modifiers were balanced for.

The AI playstyle depends on completely different factors than one of human player. Imagine Aztecs discovering early natural wonder and they are like "screw military, let's play religion". And you encounter them later - of course you don't know what they found the natural wonder, you just see Montezuma going crazy - not attacking anyone, building Stonehenge and so on. Making AI actually play the boost will often make it less logical in most cases.

If part of AI behavior isn't directly exposed to human player, it should be coded as simple as possible. Spending a lot of work on such feature, spending even more work to adapt the feature to game changes and in best case getting nothing for the game (in worst case the AI will look more random) is surely a wrong thing to do. That's the first lesson AI developers learn. For many indie developers it ends up being a very painful one.

 

[Kouvb593kdnuewnd]

I think erukas are good as they are. Their purpose is to encourage people not go all out on science. Most erukas require a time investment and the more science you produce the less time you will have to fufill the eurkas. So if you focus on science you are likely to miss some erukas.

 

[masda_gib]

I'm not worried that there are generic, easy-to-do eurekas since I think that's part of why they are there: to help civs/players that are not focused on science to research some techs easier.

+ they have some thematicly fitting or funny ones there. So it's a good mix. Let's see how it turns out in play.

 

[Ryika]

The AI playstyle depends on completely different factors than one of human player. Imagine Aztecs discovering early natural wonder and they are like "screw military, let's play religion". And you encounter them later - of course you don't know what they found the natural wonder, you just see Montezuma going crazy - not attacking anyone, building Stonehenge and so on. Making AI actually play the boost will often make it less logical in most cases.

You can 100% see that AIs play by different rules when they do, even if you can't see their numbers. Two obvious examples:
- AIs can stack up a lot more cities than you could realistically get, because they play with happiness bonuses
- AIs will end up with more population than you in the midgame, because they have growth bonuses

Other examples are less obvious directly, but their results can still be recognized - like the fact that AIs were able to see what's happening in the fog in earlier games for example. You'd be that kind of guy* who would be like: "It's hard to make the AI work without being able to see in the fog and players don't -really- notice it!" - but thankfully that was changed. (Although it may make sense to allow them to see units that have left their visibility the previous turn to simulate human memory) Them playing by different rules when it comes to Eurekas would be the same thing.

(*edit: By this I mean if I "translate" your current position onto a person of the past, not that you specifically would stand for that decision)

Some of these bonuses are of course required to make the AI scale in difficulty after a certain point, completely disabling the Eureka system and just adding a flat science modifier should not be.

And again, I'm not on board with the "Get every Eureka!"-system, I would very much prefer the "you end up with bonuses for the type of Empire you are"-system - the AI would work automatically with it. It is only when most of the Eureka-Bonuses are stuff that realistically everybody can do if they spend a little bit of extra time on them that the AI starts having problems keeping up.

If part of AI behavior isn't directly exposed to human player, it should be coded as simple as possible.

From a budget position, sure. Not because it has "no effect", but simply because most players will not notice. Experienced players however will notice.

 

[Joketa]

I have the same concern, though it's hard to judge when we haven't played yet.

I'd also like to see random Eureka moments so that they aren't always the same tasks. You'd only need a pool of 3 to make it nice variety. In one game Masonry could be rushed by building a quarry, in another by extracting stone, etc.

Of course, I'd like random policy cards, too. I love the unknown. Kris Swordsmen were my favorite just because there was another unpredictable layer to them.

 

[SammyKhalifa]

Of course, I'd like random policy cards, too. I love the unknown. Kris Swordsmen were my favorite just because there was another unpredictable layer to them.

Unfortunately though the "average" civ player hates the unknown. I enjoy the challenge of rolling with the punches and taking what I get, but most don't seem to.