Thoughts on a tech tree

[Zygot]

It is indeed very elegant.

When the game begins and you first start researching, you'll be presented with a random selection of techs from Tier One in each field. You will always have at least one tech to research in each field, but you may have several or all, depending on the limited research list.

As soon as you complete research on a tech, you will be presented with all the previously available and unresearched techs for that field, plus a new random assortment from the next Tier. In other words, the instant you research one single tech in a Tier, you unlock the following Tier of that field for possible research.

If you choose to research old techs that reside in anything besides than your most advanced unlocked Tier, you do not get any unlocking benefit, you just get whatever benefit that individual tech provides; you must complete research on a single tech in your highest available Tier to unlock further options. I don't believe there is any lower bound at all.

The same rules apply if you acquire a tech through trade, espionage, or combat. The next time the scientist asks what to research, the highest Tier available will be the one following whichever Tier holds your most advanced owned tech.

---

As for the displayed Tech Level, you are correct that it has no direct relation to the Tiers. Here's how it works. What is the Tech Level of your single most advanced owned technology in a field? Multiply that value by .8 and round down, then add the total number of techs you own in that field.

Therefore if your highest tech in computers in Battle Computer III (TL 10), but your displayed level is 13, you must be a completionist:

int(10*.8) = 8
8 +5 = 13

In other words, you have four lesser techs than the Battle Computer III. Someone who tries to vault up the Tiers as quickly as possible (only owns one other tech, like ECM Jammer I), might have Battle Computer III but only be Tech Level 10:

int(10*.8) = 8
8+2 = 10

So your Tech Level in a field varies depending on how much of a completionist you are. Sometimes it is strategic to research as few techs as possible and vault up the Tiers, but sometimes it is advantageous to fill out old technologies.

[Arantor]

Thanks for the clarification - that does seem in line with my observations since I couldn't find any reference in the manual or OSG that explained exactly how the tiers worked, which is what I was really interested in. You're right, I was being a completionist, but only to see if I could make sense of how the setup worked.

The one thing that concerns me about having 7 tiers rather than 10 is that players might vault too fast up the tree, but having 7 branches rather than 6 should limit that potential, along with not having the strict tiers so much.

[RefSteel]

Thanks for the clarification - that does seem in line with my observations since I couldn't find any reference in the manual or OSG that explained exactly how the tiers worked, which is what I was really interested in. You're right, I was being a completionist, but only to see if I could make sense of how the setup worked.

The one thing that concerns me about having 7 tiers rather than 10 is that players might vault too fast up the tree, but having 7 branches rather than 6 should limit that potential, along with not having the strict tiers so much.

It may seem counter-intuitive, but having 7 tiers rather than 6 actually increases the potential to vault quickly up the tree for someone who wishes to do so: If they're expected to put spending into 7 fields at once instead of 6, it effectively means that - if they choose to put all their research eggs into one basket instead - they have almost 17% more research points available for tree-vaulting than they would with similar "expected" rates of progress in 6 fields.

In general though: Don't get too caught up on the number 7. The number of techs in each tier and the number of tiers in each field (and other factors) should be chosen on the basis of what works best, not on the basis of a "magic number."

Re: Magic numbers, also, supposing that the way I used "chunking" to memorize that number relates to 7 +/- 2 looks to me like confirmation bias; if I'd split it into 3 chunks of 5 or two of 7 and 8 apiece, you could have said, "Each chunk fits into 7 +/- 2." But of course I wouldn't have split it into two chunks with 7 and 8 because 8 (and even 7, really) is too large of a chunk. The +/- 2 thing has no cognitive basis that I know of, whereas 7 seems to be a real upper limit on how many discreet pieces of information we humans can normally deal with at once (without e.g. chunking). In any case, the fact that there were 5 chunks in my example is arbitrary. I've also have memorized a different "too-long" (14-digit) number by "chunking" for example, but use only three "chunks" of 4-5 numbers each. I'm sure the tendency for credit cards to split their 16-digit numbers into 4 sets of 4 is another example.

I'm probably not really telling you anything new, but I figured I'd put this stuff out there. Ultimately, the key is to decide what you want the tech tree to do, and let that determine how it's shaped, rather than the other way around. Do you want a chance of finding a tech of each type in each tier, for example? Missing a key tier of Eco Restoration or Space Scanner and knowing you'll have to research through two more tiers for a chance at the next upgrade can add a lot of tension in MoO. Should each tier's techs be major upgrades on the ones before, or small incremental changes as is often the case in MoO? How thoroughly should end-game techs outclass early-game ones? The answers to questions like those will tell you how many tiers you need in your tree and how many techs should be in each tier.

[Arantor]

Actually, the number everywhere was more a case of being coincidental rather than intentional and then I just sort of ran with it.

I wasn't actually planning to add a proportional RP amount into the economic model, as it happens, so vaulting up the ladder would be less of a risk in that respect. While there's 7 branches, there isn't quite the same complement of techs as before, some serious playtesting will be required to ensure the tech race occurs at roughly the 'right' place.

There are fewer of some techs: right now, I only have the equivalent of IIT9 through IIT3 in the tree, IIT2 doesn't exist, and they're further apart - IIT2 is level 38 for example, I have my equivalent of IIT3 currently as level 45, which may end up being too late in the game for it to be of any real use, needs more testing. There are also techs that didn't exist in MOO (the sociology branch includes spying related techs), for example right now none of the terraforming techs exist in my tech tree so something else is replacing them.

There's the colony techs, which covers all the knowledge of the planetary environments in question, there's some limited additional population mechanics (biodomes, allowing for +10 pop, currently level 6, as an example, testing will reveal if more are needed)... but the real key is getting the Genesis devices later on, currently levels 15, 18, 25 cover off deploying the terraforming which improves suitability of a planet for a given race - and it's race vs planetary type that really determines the pop max for that race on that planet. This definitely leads to careful UI design but I don't at this stage believe this mechanic is 'too complex' provided the UI provides a breakdown of what the planet is now and what that means for the player (including for potential planets)... but approaching mid-game might prove to be too late.

I'm all for keeping the incremental nature of technologies, because it *works*. But in the case of things, I suspect that I'm not exactly thinking in tiers in the way you are. In MOO, tiers are hard and defined groups of technologies; in Ad Astra they are more guidelines. They group technologies together but access to a technology is not based on your current-to-next tier state, the furthest you can see is always whatever's in the next 7 levels from your highest tech in that branch. While this sounds like falling back to 7-ness again, it's really not.

Consider MOO for a moment.

Any branch that's reasonably full of tech will do, but we'll run with planetology. So you do IER for tier 1, maybe you jump into tier 2 with Tundra env (level 6). This gives you the run of tiers 2-3 for the next research. Which means you can pick any tech up to 9 levels up from you. Suppose, though, you jump into tier 2 with the first bioweapons tech (death spores IIRC, anyway it's the first one you hit with -1 size, at level 10)... you still have the run of tiers 2-3, which means you have the run of the next 5 levels after that. 5-9... hey, that's 7±2 again! I'm just smoothing out the transition so it's always wherever-you-are-plus-7.

Am I proposing to drop the tension? Heck no! You're right, it's one of the *best* things about MOO is that if you miss a key tech, in almost every case you will be able to get a more powerful version later without having to trade, steal or acquire-by-combat, though these are all valid alternatives too. I don't see any reason to change this behaviour for the most part; for the times when there are few techs in a family that are to some degree essential (space scanners, xeno diplomacy), at least one of those will be guaranteed, but not because of presence in tiers. That's one thing that I need to be quite careful about, is ensuring that there will always be continuity of techs with the rolling 7 levels deal, since in MOO if you had only one tech in a field, I understood it would be guaranteed even if the LRL might otherwise exclude it as part of the 50/50 (or 75/25 for Psilons) chance of having a given tech.

Ultimately I'm not talking about 'researching through x tiers' but 'through x techs' to get to the next iteration of something; in practice they should amount to something comparable in terms of drip-feeding higher class versions of lower/existing techs.

The process of figuring out what's needed where is important as you've picked up on The headache for me is that because I'm so radically changing the way we deal with terraforming, I find myself with less of a guide to go on, but right now the mid-game is where all the fun is, where the end-game techs don't 'thoroughly' outclass mid-game but do for early-game techs; it would not feel right to have ships with early-game tech being able to meaningfully compete with ships equipped with end-game era tech, or indeed an early-era race being able to meaningfully compete with a mid-era race, save for racial bonuses making up some shortfalls.

That's also really where 7 tiers vs 10 tiers comes in; it's also a hint towards generations of tech and should also to a point correspond with the phases of the game in a way that 10 tiers doesn't (at least in my head). I did briefly consider a certain amount of romanticising with a parallel to the Seven Ages of Man but realised that might not work out so well in practice since ultimately the seventh age is one of dependence on others. But again it's more a case of the mechanics supporting the number 7 and then hanging something off it rather than going out of my way to 'make it fit'.

Pop and production increases are something I'm still working on, though... I don't have anything like the IRCs in my tree yet, partly because I'm not sure I want them but I don't know what to do in place of them yet. Lots of thinking still required.


In other news, I also dug out Space Empires IV. It is not for the faint of heart, btw. The interface is sadly one of the most cumbersome I have ever seen and has a little too much of playing-by-spreadsheet for my taste but there's some interesting ideas. In direct relevance here, research is not a tree but more a levelling system for 'skills' such as we should consider them, e.g. gaining Level 2 in Ship Types to unlock another ship size/class, while they don't seem to have the range of planet buffing (in either population max or production max) that MOO enjoys, though they do make up for it in other ways e.g. planetary morale being able to account for a 20% production bonus. Additionally, you can select multiple research projects at once, up to a maximum of 12. (Got other games to experiment with though)

It's almost a game in itself, this design lark.

[Zygot]

Something to consider: in MOO, the tier limitation creates a situation where you might fill out old research, even multiple ones, while remaining in the dark about future tiers. In fact I often research old techs simply because their value might be more important to me at that moment, especially given the exponentially growing cost of advanced research.

With a system like yours, always looking 7 rolling techs ahead, does filling out the lowliest old tech advance your knowledge about cutting-edge research? This is the opposite of how MOO handles a situation like this; your scenario seems to give a lot more information, so it's a quite different approach.