Friday Facts #275 - 0.17 Science changes

[Tricorius]

I think it's a really bad assumption to assume only experienced players use math.

Sure. They will eventually figure out ratios and such. Some may even desire to do it all up front. But I think the people inclined to do that will largely enjoy the challenge of figuring out a build to do what they want as they go. And if they find they can’t support science and have enough left over they will find a pretty simple solution.

The rest of the people (I personally think the majority) will simply not care so much and ramp up different parts of their factory as they go.

They might even do what I did and restart their base a few times before “winning” to try to get a better overall configuration now that they have some hindsight.

I simply think that most people posting (and to a lesser degree reading) are pretty advanced players and forget how long it takes the typical person to reach certain milestones and learn all the stuff you need to learn.

I’d be super curious to see the metrics around that stuff.

:: shrug ::

[Serenity]

I ratio my stuff as much as people. But that assumes again that things will always be running full time. Producing at the max rate and consuming at the max rate. But that's just not always the case. Personally I play rather slow and have lots of down time when I build stuff. Sometimes I don't research for a while.

[featherwinglove]

I think it's a really bad assumption to assume only experienced players use math.

Sure. They will eventually figure out ratios and such.

No, I mean right off the bat. I'm pretty sure I figured out the (then 0.12.33) 3:2 wire:circuit machine ratio while I was still in the demo and without even reaching for my calculator. A lot of players are going to be coming to the store off of a Nilaus episode or a tutorial specifically on machine ratios. Some will probably figure things out by poking around a primitive build and reinforcing the bottlenecks until everything's busy. Just 20SPM is enough to min/max a whole factory, and probably most new players will be paying attention to such things by the time they get to the production science pack in their first game. Even if that's not the case, it is still a bad basis for a science pack recipe design.

You have to realize that if a gamer doesn't ever reach for his calculator, Factorio is probably not interesting enough for him even to finish a single freeplay, and the demo, which won't (CMIIR) feature the production science pack, will probably warn him off.

[Dune]

I don't want to be negative, but I thought the goal of 0.17 was to polish for launch. These changes "should" take another year of testing. As you're essentially making a new tech tree.

I know developers have free reign to do as they please, but 0.16 is a fun game. As coders, don't you know the phrase, "If it ain't broke, don't fix it?"

[DaveMcW]

These changes "should" take another year of testing. As you're essentially making a new tech tree.

Don't worry, I'm sure we are still a year away from launch.

And the old tech tree definitely needed fixing.

[Rythe]

So what I've gotten from the thread at this point is that the reason rails are in the production pack is because:

1)You decided science would pull double duty and automate things you want the player to automate because reasons
2)Rails were on that list and fit into the production pack least badly
3)All other considerations were a mix of post-hoc rationalization, happy coincidence and fluff in order to pretend more went into it

Because if you wanted an interesting recipe, you would have gone with refined concrete.
If you wanted a tech appropriate recipe (how recent in the tech tree, how much factory development to produce), it would have been batteries, their derivative products (accumulator), or similar.
If you wanted an ingredient that made sense, it would have been the accumulator, combinator or, hell, even the laser turret would make more sense if it weren't so expensive.

So this bit:

One of the most important aspects of the science pack is its recipe as it defines how many resources it costs, how complicated is it to produce, how much do you need to research before being able to produce it (because of prerequisites), and how much time it takes to craft it including all the ingredients.

Was a bunch of meaningless, misleading fluff. And the reason we're arguing in circles to nowhere is because you didn't clearly give us the only criteria that actually mattered here.

Thanks.

Edit: Just realized the utility pack stole the processing unit and that the robot frame uses batteries, which complicates things in a way that suggests the ingredients chosen for the utility pack are making this harder than it should be too. Like maybe the utility should be running a Defender Capsule instead of processing unit to shift the processing unit to production because that would make a lot of sense too.

[Tricorius]

You have to realize that if a gamer doesn't ever reach for his calculator, Factorio is probably not interesting enough for him even to finish a single freeplay, and the demo, which won't (CMIIR) feature the production science pack, will probably warn him off.

Actually I’ll probably have to agree to disagree here. As previously mentioned. I still don’t care much about perfect ratios and I’m a 2000+ hour player.

I have never tossed anything related to Factorio into a calculator or spreadsheet and I doubt I ever will.

This is all speculation anyway. I have no idea why the “core factorio player” plays. I play because it is fun to have a giant digital “Erector Set” that also tickles my “fancy for trains”. (Also, robots. Robots are cool.)

Or perhaps I’m an anomaly and every other factorio player invests in detailed excel spreadsheets.

:: shrug ::

Even without perfect ratios my factories do just fine. I’m almost always “tapped-out” way ahead of the next tier of science (my “vanilla” bottleneck is not having bots soon enough to lay down stuff quickly enough to have blue science automated before I have fully researched red and green—but that is probably also due to a kinetic-defense based, Death World playstyle preference; which takes a while to setup a proper defensive perimeter and ammo feeder line).

Regardless, I agree the typical person drawn into Factorio probably has a predisposition toward “nerdy” things like math and science.

But I don’t know about assuming all of them are permanently attached to a TI-85 graphing calculator. (Yup, showing my age...used those in high school when they were just newly released.)

But don’t fret...almost all of us suffer from some form of false-consensus bias.

[Oktokolo]

Or perhaps I’m an anomaly and every other factorio player invests in detailed excel spreadsheets.

Nope - not using spreadsheets or calculators for Factorio and have >1k hours in the game too.

[5thHorseman]

Or perhaps I’m an anomaly and every other factorio player invests in detailed excel spreadsheets.

Not sure how many anecdotes you need to create data, but I'll throw my own in. I know the ratios for several recipes and when I do know the ratio I generally keep to it, but otherwise I'm in your camp. The way I balance ore to plates is, if I need more plates I make more furnaces. If they don't fill up I mine more ore. Then I go back to the fun part which is getting trains all jammed up and untangling spaghetti.

Not a 4-digit-hours player, but I did just break 600 the other day.

[morsk]

Let's imagine an alternative universe where they picked refined concrete. And we wanted rails. The iron ore would not be popular. We also might claim it's too complex.

I've never built refined concrete in any factory. This is something I need a push to make. Rails aren't. Is it good or bad to make me build something I'd never build?

Edit: Just realized the utility pack stole the processing unit and that the robot frame uses batteries, which complicates things in a way that suggests the ingredients chosen for the utility pack are making this harder than it should be too. Like maybe the utility should be running a Defender Capsule instead of processing unit to shift the processing unit to production because that would make a lot of sense too.

Putting fancy components in production science takes flexibility from new factories. It won't need sulfuric acid or lubricant, letting you get it early if you're willing to skip things. I wouldn't skip batteries, but a new player, or someone using the nanobots mod, might. And anyone might want assembler 3s before processing units, or coal liquification if they got a bad map.

Utility needs higher materials, but they're all things anyone rushing power armor needs to make anyway, so I still feel I can go for this "early" by skipping other things, and I'm pleased with it.

[featherwinglove]

These changes "should" take another year of testing. As you're essentially making a new tech tree.

Don't worry, I'm sure we are still a year away from launch.

And the old tech tree definitely needed fixing.

(signed - Tech Tree)

I wish I could quote this with the signature block, that was so epic

[quadrox]

While I don't condone the tone of the post quoted below, I have to agree with the message itself. I actually think it is a rather important point.

So what I've gotten from the thread at this point is that the reason rails are in the production pack is because:

1)You decided science would pull double duty and automate things you want the player to automate because reasons
2)Rails were on that list and fit into the production pack least badly
3)All other considerations were a mix of post-hoc rationalization, happy coincidence and fluff in order to pretend more went into it

Because if you wanted an interesting recipe, you would have gone with refined concrete.
If you wanted a tech appropriate recipe (how recent in the tech tree, how much factory development to produce), it would have been batteries, their derivative products (accumulator), or similar.
If you wanted an ingredient that made sense, it would have been the accumulator, combinator or, hell, even the laser turret would make more sense if it weren't so expensive.

So this bit:

One of the most important aspects of the science pack is its recipe as it defines how many resources it costs, how complicated is it to produce, how much do you need to research before being able to produce it (because of prerequisites), and how much time it takes to craft it including all the ingredients.

Was a bunch of meaningless, misleading fluff. And the reason we're arguing in circles to nowhere is because you didn't clearly give us the only criteria that actually mattered here.

Thanks.

Edit: Just realized the utility pack stole the processing unit and that the robot frame uses batteries, which complicates things in a way that suggests the ingredients chosen for the utility pack are making this harder than it should be too. Like maybe the utility should be running a Defender Capsule instead of processing unit to shift the processing unit to production because that would make a lot of sense too.

While I think that the vast majority of decisions made by Wube so far have been correct (even controversial ones, in hindsight), I fear that we are reaching a point of fatigue were developers just want to finish the thing, everything else be damned. I hope I am wrong, but even though I think there is much good in the changes from this FFF, there is also some bad, and it doesn't look like Wube is able to realize this.

[V453000]

While I don't condone the tone of the post quoted below, I have to agree with the message itself. I actually think it is a rather important point.

So what I've gotten from the thread at this point is that the reason rails are in the production pack is because:

1)You decided science would pull double duty and automate things you want the player to automate because reasons
2)Rails were on that list and fit into the production pack least badly
3)All other considerations were a mix of post-hoc rationalization, happy coincidence and fluff in order to pretend more went into it

Because if you wanted an interesting recipe, you would have gone with refined concrete.
If you wanted a tech appropriate recipe (how recent in the tech tree, how much factory development to produce), it would have been batteries, their derivative products (accumulator), or similar.
If you wanted an ingredient that made sense, it would have been the accumulator, combinator or, hell, even the laser turret would make more sense if it weren't so expensive.

So this bit:

One of the most important aspects of the science pack is its recipe as it defines how many resources it costs, how complicated is it to produce, how much do you need to research before being able to produce it (because of prerequisites), and how much time it takes to craft it including all the ingredients.

Was a bunch of meaningless, misleading fluff. And the reason we're arguing in circles to nowhere is because you didn't clearly give us the only criteria that actually mattered here.

Thanks.

Edit: Just realized the utility pack stole the processing unit and that the robot frame uses batteries, which complicates things in a way that suggests the ingredients chosen for the utility pack are making this harder than it should be too. Like maybe the utility should be running a Defender Capsule instead of processing unit to shift the processing unit to production because that would make a lot of sense too.

While I think that the vast majority of decisions made by Wube so far have been correct (even controversial ones, in hindsight), I fear that we are reaching a point of fatigue were developers just want to finish the thing, everything else be damned. I hope I am wrong, but even though I think there is much good in the changes from this FFF, there is also some bad, and it doesn't look like Wube is able to realize this.

If the list of important things a science pack recipe needs to do was not absolutely complete, I am sorry. However in the rails part, it is quite clearly explained why are they there. I do understand that it probably does not seem to match the list perfectly as the list does not include "item is actually useful" and "it's a nice motivator to automate it, and so on.

I do not think this is showing fatigue of the developers. If we were that fatigued we would not have done this whole rework in the first place, definitely not to this much detail. There would not be a 33k characters long FFF with tons of changes and information, and we would not have discussed this topic on 20 forum pages even through some utterly demotivating tone from some.

I just also see a large group of people who like the change (of specifically rails), alongside of reason. I still do not see a better alternative than rails.
Concrete not really an useful thing at the moment. Should that change? I'm not sure. Does it need to change AND change the science pack? No, we could still discuss that change separately.

[MicFac]

...
This way the player is required to automate all fundamental logistic "devices" very early game.

In a game about discovering, learning, inventing, solving puzzles, and playing freely all this new hand-holding is the opposite of a good thing.

I think it's not too much hand-holding and could encourage new players to automate other machines as well.

[Schallfalke]

I just also see a large group of people who like the change (of specifically rails), alongside of reason. I still do not see a better alternative than rails.
Concrete not really an useful thing at the moment. Should that change? I'm not sure. Does it need to change AND change the science pack? No, we could still discuss that change separately.

I have suggested a new intermediate product (like rocket control unit for rocket part) in previous reply, so I am not repeating here. Though I do not see anyone commenting it would be a good alternative or not.
(Someone said even the recipe of rocket control unit needs to be changed, but that's another story.)

Please do not use concrete in the recipe. I never like laying concrete all around. I want the ground to absorb pollution as much as possible. The only use of concrete to me is for the nuclear stuff, so not a good idea to "force" player automating that just for science pack. (Of course, if significant changes are made so concrete is used abundantly in various recipes, then it can be considered again.)
And in the "stone consumption" point of view, using concrete do not reduce the number of 30 there.

Side note: I am a math-minded person. On my first playthrough, I did my own spreadsheet calculating proper ratios. So even new players can do lots of calcualtions, while quite a good portion of "advanced players" do not care about ratios. So I guess it's more related to the player personality, rather than being new/advanced.

[V453000]

Side note:

There is of course also the new belt spacing mentioned in the latest FFF that I was counting with, that allows even more rails to be supplied to the production pack. Just in case you still worry about your throughput, worry a bit less.

[V453000]

I just also see a large group of people who like the change (of specifically rails), alongside of reason. I still do not see a better alternative than rails.
Concrete not really an useful thing at the moment. Should that change? I'm not sure. Does it need to change AND change the science pack? No, we could still discuss that change separately.

I have suggested a new intermediate product (like rocket control unit for rocket part) in previous reply, so I am not repeating here. Though I do not see anyone commenting it would be a good alternative or not.
(Someone said even the recipe of rocket control unit needs to be changed, but that's another story.)
...

I think artificially adding a new ingredient that wouldn't really do much, and we would have to inject it into as many as possible places to make it useful, is a bit of a crazy solution, and would it Really be better than rails?

[Schallfalke]

I just also see a large group of people who like the change (of specifically rails), alongside of reason. I still do not see a better alternative than rails.
Concrete not really an useful thing at the moment. Should that change? I'm not sure. Does it need to change AND change the science pack? No, we could still discuss that change separately.

I have suggested a new intermediate product (like rocket control unit for rocket part) in previous reply, so I am not repeating here. Though I do not see anyone commenting it would be a good alternative or not.
(Someone said even the recipe of rocket control unit needs to be changed, but that's another story.)
...

I think artificially adding a new ingredient that wouldn't really do much, and we would have to inject it into as many as possible places to make it useful, is a bit of a crazy solution, and would it Really be better than rails?

Currently in 0.16, the ONLY use of rocket control unit is for rocket part, and it is a simple 1+1 ->1 recipe. But I think it fits quite well there.
So why can't we have similar thing for production science pack? It could be something like 10 stone + 1 iron stick + 1 steel plate + anything -> 1 prefabricated concrete section.
If really some extra uses is needed, maybe replacing those concrete with this item, in recipes of centrifuge and nuclear reactor, rocket silo, or whatever advanced buildings in the future?

EDIT: To new players or anyone not playing this game before, "rails" is more on logistics side. "prefabricated section"/"prefabricated concrete section" is more on production side. Does it sound a good reason?

[V453000]

I just also see a large group of people who like the change (of specifically rails), alongside of reason. I still do not see a better alternative than rails.
Concrete not really an useful thing at the moment. Should that change? I'm not sure. Does it need to change AND change the science pack? No, we could still discuss that change separately.

I have suggested a new intermediate product (like rocket control unit for rocket part) in previous reply, so I am not repeating here. Though I do not see anyone commenting it would be a good alternative or not.
(Someone said even the recipe of rocket control unit needs to be changed, but that's another story.)
...

I think artificially adding a new ingredient that wouldn't really do much, and we would have to inject it into as many as possible places to make it useful, is a bit of a crazy solution, and would it Really be better than rails?

Currently in 0.16, the ONLY use of rocket control unit is for rocket part, and it is a simple 1+1 ->1 recipe. But I think it fits quite well there.
So why can't we have similar thing for production science pack? It could be something like 10 stone + 1 iron stick + 1 steel plate + anything -> 1 prefabricated concrete section.
If really some extra uses is needed, maybe replacing those concrete with this item, in recipes of centrifuge and nuclear reactor, rocket silo, or whatever advanced buildings in the future?

- rocket components are something really special, it's The Rocket after all
- in 0.17 they already are a component of atomic bomb
- maybe they could be somewhere else as well?
- having problem X1 being a reasoning for adding problem X2 is not great

[Oktokolo]

It could be something like 10 stone + 1 iron stick + 1 steel plate + anything -> 1 prefabricated concrete section.

"Prefabricated concrete section" sounds a lot like "wall segment" to me. Maybe they should change walls to use concrete instead of bricks.