Science Pack progression in terms of materials needed.

[adoa]

Bob's mods have been around for a long time and touching something as fundamental as science pack recipes is possibly controversial, but when comparing to other science progression systems, Bob's Mods definitely lack behind. Let me explain.

In Vanilla, you have these costs for science packs: The progression of number of materials is: 2 → 2 → 5 → 5 → 6 → 6 → 7

In Bob's Mods you have these science costs (leaving out the first and last because they are not problematic): Numbers of materials: 2 → 3 → 5 → 13 → 12 → 18 → 20 → 32

Bob's Mods + Science Cost Tweaker: Numbers of materials: 2 → 5 → 10 → 19 → 20 → …

Several things stand out for Bob's mods:

0. Bob's mods adds an insane number of materials to the game.
1. You need only 3 of >30 materials to research the first two tiers of science, which cover a huge amount of technologies.
2. The jump in number of materials from military to chemical science is huge
3. Production Science requires fewer materials than chemical science, and all materials were already required for military & chemical science. In particular, that means you don't need to automate chemical science to even research all materials required for production science. That makes me think chemical & production science are basically the same tier.
4. SCT smooths out the above points quite nicely.

My suggestion: Modify the science pack recipes to increase number of materials needed for Transport Science, Military Science, and Production Science. However, this might have a big impact on general game balance when playing with biters and possibly Bob's enemies. So maybe don't increase the overall cost too much.

[adoa]

For reference, you could try to spread out science tiers vs. material tiers but it's not quite obvious how to even classify material tiers. (this started as an angel's discord discussion https://discord.com/channels/164794742893182976/692824284607021198/824793190154502205)

[mecejide]

This seems like a great excuse to finally do the science overhaul that Bob was talking about.

[atburghardt]

I don't see why science would need an overhaul. The increased difficulty between the sciences after blue isn't the Material cost, but the more complex variety of Materials and how to produce them.

[adoa]

@atburghardt that's exactly my point: purple science does not need any complex materials to be produced at all, and it's after blue

[mecejide]

Related to this, every material Bob’s Technology adds to a vanilla science pack is either from vanilla or from Bob’s MCI. The recipes could do with more variety.

[adoa]

@mecejide Please elaborate. The only materials Bob's Technology knows about are either from vanilla or MCI. What else do you want to throw in there?

[mecejide]

@adoa what do you mean by “knows about”?

[adoa]

essentially: dependencies, things you can expect will be there on most maps

[mecejide]

MCI isn’t a dependency of Technology.

[KiwiHawk]

Other of Bob's mods could be added as optional dependencies if required. Sensible fallbacks would need to be defined though.

These mods all add things that could be used as science pack ingredients:

• Assembling Machines
• Electronics
• Logistics
• MCI
• Mining
• Modules
• Warfare

[Qatavin]

I would suggest not doing too much with the transport science pack. There are not that many automation-science-only technologies, so adding significantly to the complexity could make things overly difficult for the early game. And they already require wood. Wood requires greenhouses, and greenhouses need glass, which is not reflected in the above diagrams. In the long term, wood could also end up being made with fertilizer, which increases the complexity further (particularly if fertilizer is rebalanced to make it more energy efficient to grow wood using it - this is an issue when using Revamp's Extra Chemistry, as the hydrogen peroxide needed to make nitric acid is energy expensive). Switching the transport science packs back to using yellow belts instead of gray belts would add tin to the mix, which would make things reasonably more complex.

On the topic of the production science pack, I'd like to present this chart I made a while ago of basic materials needed to make various machines, roughly organized according to which electronic board they require, or which other machines they upgrade from. I'm pretty sure it's still up-to-date.

So, the issue raised by adoa is that production science requires fewer materials than the chemical science that precedes it, and more importantly that it doesn't require any materials or production lines you wouldn't already have by the time you reach it. Basically, you need red circuit boards for both chemical and production science. Red circuit boards <- plastic <- oil <- trains <- steel. That accounts for basically everything production science needs. You could say that production science is still more complex because it takes both red and yellow circuit boards, but we want something completely new.

I think the chart I posted above illustrates the problem with the current recipes of both the science and the machines. In order to introduce new materials, you'd have to go down to assembling machine 4, chemical plant 3 and multi furnace 1 or electric furnace 2. This would get way into blue circuit board territory (and several technologies that currently require production science, screwing up the progression on the tech tree from chemical to production to utility science) and make the material and energy cost skyrocket.

The solution I would propose would be to increase the production science pack output to 4, and make the ingredients assembly 2, chemical plant 1 or 2, electrolyser 3, and electric furnace, while reducing the number of red circuit boards needed for electrolyser 3 from 10 to 5. This is not too out of control in terms of the number of steps required, and introduces invar, aluminium, bronze, and glass into the mix. This solution is also appealing to me because it does not include brass or cobalt, which creates a nice divide with the logistic science pack. Although, logistic science would likely need to be made more expensive to keep it on par, and utility as well to keep it costlier than those two.

There is, however, another solution. One that I came up with while working on my revamped Bob expansion mod. You could add a unit or two of nickel to the electric furnace recipe (and perhaps a couple others, where relevant). Nickel is used to make heating elements for furnaces like that because of its electrical resistance. Doing it this way would mean a smaller overall change to the recipes.

Either one of these routes would require nickel, and that would make a really substantial impact on how the game progresses. Nickel is supposedly an early-mid-game material, but you don't need it for anything at that point, and at least for me, something else always seems to be so much more urgent. Not getting it blocks you from some useful upgrades, but that's it. It isn't part of production, logistic, or utility science, nor is invar, so nothing really pushes you towards it until you get to space science and need nickel to make tungsten. I don't consider it great to force the player to go for two ores one right after the other. Requiring nickel/invar for production science, or just for something as important as electric furnaces, would be a game changer for this element that is currently so easy to overlook.

[mecejide]

One possible alternative to making transport science harder is to introduce a new science back between it and chemical science.

[KiwiHawk]

Suggestion from @LovelySanta :

With boblogistics, logistics science pack takes regular inserters (unlocked at game start), thus the green science technology does not need a dependency on electronics...

suggestion: require filter inserter instead of a regular inserter?

[Qatavin]

I still think replacing the gray belt with a yellow belt would be the way to go. There aren't that many SP1-only techs, so SP2 needs to be set up really early. Having the build order be something like

SP1->basic circuit boards->basic automation->SP2

seems like a better way to pace things than being forced to go

SP1->basic circuit boards->basic electronic boards->SP2

In the former case, you have some play time with just the basic circuit boards and the things you can make with them before having to go to the next step of basic electronic boards.

[adoa]

I think changing grey belt to yellow belt for SP2 is the minimum of increased complexity needed to smooth out things in the early game.

However, the other issue still remains: With SP1 & SP2 you can research more than 50% of all technologies. But SP1 & SP2 then require only 3 out of >30 materials in bobsmods: iron, copper, wood. Adding yellow belt would only pull in tin as a fourth material.

It might be worth testing out changing SP2 like this: yellow belt + filter inserter → 2x SP2

On the one hand, this would pull in an additional requirement of tin, carbon, and lead. On the other hand, increasing the output would make sure the cost is not too high. Maybe doubling the output should be paired with a doubling of recipe time, just to make sure you don't change the number of assemblers needed for a given SPM output.

I also like the idea of changing production science like @Qatavin suggested: at least pull in nickel for electric furnace.

[Qatavin]

If SP2 is changed to require a filter inserter, I'd suggest changing filter inserters to only take one yellow electronic board instead of four. That's a pretty substantial cost increase right there.

[Qatavin]

Now that news has come in that filter inserters are no longer going to be a separate thing from regular inserters, I'm even more inclined to say that SP2 should be left using regular inserters. No point in changing it just to change it back when 2.0 drops.