Is mining productivity research worth it?

[AndrewIRL]

Edit: The answer is yes, it is worth it after redoing the math. Seems well balanced.

According to this tech tree:
http://davemcw.com/factorio/tech-tree/

Mining productivity 1 - 100 red, 100 green = 750 Iron, 250 Copper
Mining productivity 2 - 200 red, 200 green = 1500 Iron, 500 Copper
Mining productivity 3 - 300 red, 300 green = 2250 Iron, 750 Copper

Each level yields a 0.02 improvement. So to break even on my 750 investment back I need to mine another 37500 Iron (2% of 37500 = 750).

This part is wrong.
To get all three levels is 750+1500+2250 = 4500 and for breakeven on that you'd need to mine 225000.

So the first level calculation is correct. The second one is where the trouble starts. Say you have 4500 iron and you buy all three levels at once (for example). How much ore do you need to mine to get your 4500 back?

For the first 37500 each of the three productivity bonuses has to pay itself off, but they all do so simultaneously.

Code: Select all

After 37500 ore mined
-----------------------------------
Cost - Return = Remnant debt
1:  750 - 750 = 0
2: 1500 - 750 = 750
3: 2250 - 750 = 1500

Now the first productivity is paid off and making a profit. You can apply the 2% profit it makes to the remnant debt from the second productivity. You only need to mine another 18750 @ 4% = 750 return to pay off the second one.

Code: Select all

After 37500+18750 ore mined
-----------------------------------
Cost - Return = Remnant debt
1: 750 - 750 = 0  <-- This one brought in a profit of 375 applied below
2: 1500 - 750 - 375 - 375(1) = 0 
3: 2250 - 750 - 375 = 1125

At this point the remnant 1125 debt is paid off at 6% which means only 18750 more needed. But let's do level 4:

Code: Select all

After 37500+18750+18750 ore mined
-----------------------------------
Cost - Return = Remnant debt
1:  750 - 750 = 0  <-- Profit of 375 applied to (3)
2:  1500 - 750 - 375 - 375(1) = 0  <-- Profit of 375 applied to (3)
3:  2250 - 750 - 375 - 375 - 375(1) - 375(2)
4: 16600 - 750 - 375 - 375 = 15100

The 15100 debt is paid off @ 8% which requires 188750 ore to be mined and so on...

First 15 levels

Code: Select all

   Cost  Breakeven  Total
1  750     37500    37500
2  1500    18750    56250
3  2250    18750    75000
4  16600   188750   263750
5  20750   154750   418500
6  24900   137750   556250
7  29050   128036   684286
8  70400   354464   1038750
9  79200   324583   1363333
10 88000   303667   1667000
11 96800   288455   1955455
12 159000  499545   2455000
13 172250  473654   2928654
14 185500  453310   3381964
15 198750  437036   3819000

These are the numbers ONLY if you buy all these research items at once without mining at all in between. If you do it slower (necessity in-game) then it takes more ore total.

Mining drills are cheaper but dont compound

For comparison the raw material for an electric mining drill is 23 Iron and 4.5 Copper. If you took your investment here:
Mining productivity 3 - 300 red, 300 green = 2250 Iron, 750 Copper
and put it into mining drills instead you could buy 100 of them. For 100 mining drills to only be an equivalent 2% increase in your capacity you would have to already have 5000 mining drills in operation.

This part is wrong

This graph produced by Mavzor on Reddit does the math for higher levels:

HIZqud8Z4NgnP9LHc6DjpJzxJXjHUXrWGy2cBxcmqC8.png (185.85 KiB) Viewed 23281 times

[/url]

The breakeven for level 4 is over 1M Iron. That merely returns your initial investment. To get a return you need to keep mining beyond that.

For reference it takes ~100K Iron to launch a satellite. Maybe ~1M to build the factory and do the research to get that far.


Note: Assumes smelters, assembly machines and labs needed to make and consume the science packs are all free. So the iron and copper costs are a lower bound.


Edited after Aeternus' comment which lead to a math review. Unfortunately Mavzor's graph is wrong.

[Aeternus]

You fail to take into account that the total ore yield per field also increases, lessening the need for more and more and more expansion bases. That alone makes me consider this research valid. If I can get 16M ore out of a 10M field, I can worry about that other 6M field much later in the game. Mining productivity also applies to Uranium, Coal and Stone. Some players tend to have low amounts of Uranium, which makes this research useful to be able to utilize nuclear power longer.

[AndrewIRL]

If I can get 16M ore out of a 10M field, I can worry about that other 6M field much later in the game.

I see your point, research now can extend the life of your nearby ore field but your numbers are far too high. The achievable bonus early game is much lower than you've suggested.

To get 16M ore from a 10M field needs a 60% bonus.

Mining productivity level 16 requires 1500 space science packs. You get 1000 per launch, so 1.5 launches to get to level 16 which is only a 32% mining bonus.

It is an extra 100 per level after that totalling 33,000 space science packs needed = 33 satellite launches for the space science alone. There's not a lot of "later in the game" after your 33rd satellite has been launched.

[dragontamer5788]

It is an extra 100 per level after that totalling 33,000 space science packs needed = 33 satellite launches for the space science alone. There's not a lot of "later in the game" after your 33rd satellite has been launched.

Have you ever tried to build a Rocket-per-minute base?

There are lots of players around here who have achieved "Rocket-per-minute" (roughly a max-speed beacon rocket), and others still who have achieved 10-RPM. Mining productivity is most certainly worth it for these "Megabase" builders.

There's lots of things that aren't worth it in the short term. If you're going for a singular rocket, then Logistics, Solar, Nuclear, and even Trains are not very useful. You can basically launch the rocket with maybe just 3ish iron ore patches, and powering your base with just coal throughout the whole time is possible.

--------

In any case, the primary point of infinite research seems to be to give endgame players an arbitrary goal that they can set themselves. I've seen players post screenshots of ~200 mining productivity for example, which is an achievement in of itself.

[Aeternus]

If I can get 16M ore out of a 10M field, I can worry about that other 6M field much later in the game.

I see your point, research now can extend the life of your nearby ore field but your numbers are far too high. The achievable bonus early game is much lower than you've suggested.
To get 16M ore from a 10M field needs a 60% bonus.
Mining productivity level 16 requires 1500 space science packs. You get 1000 per launch, so 1.5 launches to get to level 16 which is only a 32% mining bonus.
It is an extra 100 per level after that totalling 33,000 space science packs needed = 33 satellite launches for the space science alone. There's not a lot of "later in the game" after your 33rd satellite has been launched.

30% from lvl 15 mining productivity. Requires no launches.
30% from 3 Productivity Module 3's.

Sure, it'll mine slower, but if you're sprawling across multiple mining fields, you shouldn't worry too much about mining speed. Just add more electric miners to keep the flow going.

[AndrewIRL]

30% from 3 Productivity Module 3's.

Sure, it'll mine slower, but if you're sprawling across multiple mining fields, you shouldn't worry too much about mining speed. Just add more electric miners to keep the flow going.

One productivity module on one miner gives you 10% extra resources, sure.

But it costs 1083 iron, 2000 copper and 4025 petroleum.

Because it only applies to one miner at a time it has to pay off the entire 3083 ore price (neglecting the petroleum). So you have to mine another 30,830 ore to get a return on your investment.

A miner runs at 0.525 ore per second. You'll be waiting 30,830 / 0.525 = 58723 seconds if there wasn't a slowdown.
Accounting for the 15% slowdown and you are at 58723/0.85 = 69085 seconds = 19 hrs.

It gets worse if you use two or three because although the payoff is simultaneous the 15% slowdown stacks.

2 modules = 58723/(1-0.15*2) = 23.3 hrs
3 modules = 58723/(1-0.15*3) = 29.7 hrs

That's a long time to wait for the payoff and I didn't assess the cost of 4025 petroleum it takes (don't know how to price it in ore terms).

The speed hit also affects your mining productivity research so your overall mining speed is
1.6*(1-0.15*3) = 0.88

I suppose over the long haul and where you are resource constrained this could be worth it but you are paying a big time penalty.

[Aeternus]

I wouldn't worry about petroleum much, oil doesn't run out after all. It's usually less of a problem then iron. And I admit, if you're looking for a quick "launch a rocket and done" game, don't bother with productivity. If you're looking for a "I'm gonna build me a whopping megafactory" kind of game, you'll easily use those modules for 29+ hours or longer. Then it starts to pay off big time. Besides, at endgame, in the current state of the game, you'll end up with an excess of copper and a shortage of iron usually. Productivity (and other) modules are a good sink for excess copper.

As for the time penalty - you can solve that by just making more mines, or plop a speed beacon in there.

[BlakeMW]

I was curious about the worthiness of these researches on marathon/deathworld:

Marathon science pack costs:
red: 5
green: 14.5
blue: 137

Mining productivity tech costs / break even:
lvl1: 400rg: 7,800 : 390,000 ore
lvl3: 1200rg: 23,400: 1,170,000 ore
lvl4: 1600rgb: 250,400: 12,520,000 ore

In Deathworld, in the transition from early to mid game I'm usually on the lookout for a convenient 1mil iron patch on the edge of the starting area to expand to. Clearly, lvl1-3 is worthwhile researching before doing major expansion out of the starting area - while still within the starting area, not so much, maybe the 1st would pay off.

lvl4 has a much steeper cost, possibly just barely within the scope of a reasonable Deathworld factory aiming at launching a rocket. But productivity modules can dramatically reduce the resource cost of techs and is a no-brainer in Marathon/Deathworld, there is literally no reason to not productivity the living hell out of everything higher than furnaces because the higher tech costs gives an insanely better ROI, even a Marathon speed run would probably benefit hugely from extensive use of productivity, if not prod3+speed beacons then at least prod1 spam. By the time one is considering Mining Productivity 4 it would be reasonable to expect there is at least +12% productivity over 3 tiers (1.40x), to as high as 40% over 3 tiers (3.3x). Assuming Prod3 modules in all the important places the break even for Mining4 could be ~4mil ore, which is very reasonable. Altough the higher tier ones are likely to be outside the bounds of reasonable except for Marathon/Deathworld megabases.

edit:

On the topic of productivity modules in miners.

I've done this on oldschool deathworlds, once I've got prod3 modules kicking around with nothing to do. But rather than eat the -45% speed penalty, I just used a beacon sandwich. A setup spans 10 tiles (assuming 2 gaps between rows of miners) which means 1/10 rows goes unmined. After dismantling the setup you can place un-beaconed miners to remove the last 10%. So you get 0.9 * 1.3 + 0.1 = 1.27 effective productivity multiplier without wasting prod3 modules running at -45% speed, and technically you could even mine the whole lot with prod3+speed beacon, if you don't mind coming back to move the setup after clearing the unmined strip.

This is clearly a lot more viable than the productivity mining research especially as the modules aren't even consumed, you can use them for something else later if you want.

[AndrewIRL]

Productivity (and other) modules are a good sink for excess copper.

Fingers crossed for a copper rebalance.

[pichutarius]

cross reference: viewtopic.php?f=5&t=48597&p=287076#p287076