Architect Archive

I could only get halfway through this thread, so forgive me if this has been said already. Frayo, the 'principle' you have discovered is the same line of thinking that keeps every casino in business. It's called the 'random positive reinforcement' principle in phycology. Essentially, the human mind can come to believe that, in cases of random events, the mind the result of the next roll of the dice is somehow dependant on the result of the last. Its such a strong allusion, it is the primary reason some people become addicted to gambling. Anyway, you're better off making as much as you can in factories (with 0% chance of failing) than risking failure on each componant as well as the final product. Make sense?

You will fail, but at no greater rate than before. Yes, 100% of your critical failures will be more costly than before, but you will still fail that one time out of ten regardless of how your components are produced. In the end, the math favors the factory-made components.

I'm not understanding you.

Let's say I do 100 combines. All 100 are for deeds.

If the failure rate is 10% then I will lose 10 deeds.

My next 100 combines is for 99 structural components and 1 deed.

I am much more likely to lose 10 structural components than the single deed. There's still a chance I'll lose a deed, but it's much smaller than in the first scenerio.

The deed is the only item with value.

So I don't see how using a factory to make the subcomponents is better.

I have a game called "100". You lose if roll a 100 on a 100-sided die. Roll it ten times. what is your chance to win the game? It's very low.

Now, I trade the old die for a new die. It's 100-sided, but it has a "100" on every side. What is your chance to lose the game? 100%.

im sorry but your way off on your assumption.

I make all of my structure mods and sub components with teh factory, and then do a final assembly by hand and i have YET to fail.

Mainly because i use only the best UT resources on the final combine.

Odds are if your failing every time, your UT isnt high on all resources used in the final assembly.

(well most of the schematics require high UT but not all, like factories.)

---

Fyro wrote:
Let's say I do 100 combines. All 100 are for deeds.

If the failure rate is 10% then I will lose 10 deeds.

---

The net product? 90 deeds.

---

Fyro wrote:
My next 100 combines is for 99 structural components and 1 deed.

I am much more likely to lose 10 structural components than the single deed. There's still a chance I'll lose a deed, but it's much smaller than in the first scenerio.

---

The net product? 1 deed.

Repeat that process over the long run, you'll still lose 10 out of the 100 deeds, plus resources from the components. You lose out.

I have made 15 or so Heavy Harvesters which requires Identical walls as well as identical components.

I have failed on exactly one Harvester.

I understand that there are more variables (like resource quality) involved in success than just the 10% failure rate.

But that still does not change the fact that if you create all subcomponents in a factory except the final deeds, you will always lose deeds when you fail on the final hand-combines.

Only the deeds have value.

Am I thinking about this completely wrong?

Ni'ger don't think that's true.

"Repeat that process over the long run, you'll still lose 10 out of the 100 deeds, plus resources from the components. You lose out."

So, if I do 10,000 combines and 100 of them are for deeds and 9,900 of them are for structural modules, aren't I less likely to fail on the deeds and more likely to fail on the structural modules?

---

Fyro wrote:


But that still does not change the fact that if you create all subcomponents in a factory except the final deeds, you will always lose deeds when you fail on the final hand-combines.

---

You are wrong.

All the components of a Heavy MineralHarvester are identical. Granted there is one spot for a single wall, and a spot of 6 identical walls. But that is irrelavent. When I make Heavy Mineral Harvesters every single component is identical right out of a factory.

I have failed one time out of about 15 times.

Out of 100 deeds I might fail on 5. Out of 100 structure modules, I might fail on 15. Resources do matter, as do the quality of the parts used. Experiment, get an excellent schematic, make a lot of that item. Have all the parts you need for structure X made by the factory with good stats and your crits will go way down. Also, nothing wrong with having a schematic say for a medium harvestors and have the factory pump out ten. If your failure rate is 10%, then you're going to lose a deed anyways, might as well be to a schematic.

Knocky, I'm not being clear.

When I say, "But that still does not change the fact that if you create all subcomponents in a factory except the final deeds, you will always lose deeds when you fail on the final hand-combines."

I don't mean you will lose the deed 100% of the time. I mean the only thing you can lose are deeds. Because the combines you are doing are for deeds alone. You will lose 10% (or whatever the crit fail rate is) of your deeds.

More attempts to make deeds alone result in more lost deeds than attempts to make deeds and subcomponents both by hand.

If you make the subcomponents by hand you spread the same 10% risk of crit failure over many more opportunites for failure, therefore you lose fewer deeds.

You can present all the emperical information you want, but all that will do is help to affirm or contestthe currently perceived critical failure rate.

(Assuming that all items have equal weight) the chances of critical failure remain constant, regardless of what has happened in the past. The chances of flippinga quarter and getting heads remains 50% no matter how many times you flip it. The benefits oftracking history is two fold, 1) reverse engineer the failure rate; and 2) determine just how lucky or unlucky you've been on stringing together a chain of critical events (1:4 chance of flippin two heads in a row, 1:8 three times in a row, etc.)

As Nirrill has pointed out, if the failure rate is 10%, this rate will only become evident if you track the histories of successes and failures over the course of an extended run. Just because you've only failed once so far out of so many times does not provide conclusive evidence of a 7% failure rate. Only after recording hundreds of events can one truly plot out and determine the actual mathematical values used by the developers.

Fyro, you do bring up a good point, but you're being a little too adamant in your thinking that we're just not seeing what you're saying. We are seeing it, but in practicality, you're just wrong... There are too many ways that a factory increases productivity and reduces overall critical failures for your grief to be validated. Spend some more time with your factory. Cuddle with it, get to know it better! You'll get to appreciate it as much as the rest of us do. I've had my factory since I got Buildings III and Installations I, and I couldn't think of business without it.