Dave's Galaxy

[Ramiel]

How can I move Steel from all my now colonized planets from arcs - back to my main planet?

[Nathan]

Either the % is capped at the old 20% or I need some other formula for getting past this barrier.

Best I've had is 25.9%, which slowly dropped as the population increased. This was on a society level 1 planet for an inactive person, so was only getting population increase. Currently at 22.2% for 35 Subspacers heading for a society level 1 planet with 63587 people. Still a few days before it arrives, so (if I remember) I'll let you know what it gets to.

How can I move Steel from all my now colonized planets from arcs - back to my main planet?

Merchantmen from your main planet and send them out to colonised planets. Or you can build scouts on the colonies and scrap them on the main. I think scouts are the most steel:people efficient way to move steel, but I could be wrong. Obviously that moves population too, so don't do that if there isn't a large population on the planet already. I'd recommend using merchantmen from your main planet as you probably don't need all that steel immediately and the merchantmen will make you some Quatloos too.

You could also just leave it there and build a few arcs on the colony when it's ready.

[fodder]

I haven't touched probability theory since high school, given my test numbers is 13.8% chance better then 5 * 7.4% for example?

can't remember much either..
but..

eh.. why is it 5x 7.4%?

you flip a coin 100 times and it's still 50% chances each time of 1 choice or another. surely it's still 7.4% chance?

anyway.. 1 bb is roughly 5 cruiser in cost (of steel) but roughly 7 bb = 10 cruiser in atk.. and your numbers seem to say roughly 3 bb = 4 cruiser in success % so perhaps it's linked to atk value?

so.. 1 bb is roughly 50 scout in steel but 1:10 in atk.. perhaps mass scout is indeed the way to go?

or even subspacer (~2.5 times the steel of scout, 4 times the atk).. but cost loads of money.

[egamma]

so.. 1 bb is roughly 50 scout in steel but 1:10 in atk.. perhaps mass scout is indeed the way to go?

There's code in place to limit the effectiveness of that.

[De-Legro]

can't remember much either..
but..

eh.. why is it 5x 7.4%?

you flip a coin 100 times and it's still 50% chances each time of 1 choice or another. surely it's still 7.4% chance?

anyway.. 1 bb is roughly 5 cruiser in cost (of steel) but roughly 7 bb = 10 cruiser in atk.. and your numbers seem to say roughly 3 bb = 4 cruiser in success % so perhaps it's linked to atk value?

so.. 1 bb is roughly 50 scout in steel but 1:10 in atk.. perhaps mass scout is indeed the way to go?

or even subspacer (~2.5 times the steel of scout, 4 times the atk).. but cost loads of money.

When I said 5 * 7.4% I was referring to the fact that you could have 5 fleets of 100 cruisers or 1 fleet of 500. Thus in a single turn you could have 5 attempts at the 7.4% chance. While each one taken on its own is 7.4%, the chance that 1 in the 5 is successful is not 7.4%, I just forget what it is. Again if you flip a coin 50 times, the chance that in any 1 flip you get tails is 50%, but what is the chance that at least 1 time in the 50 you get tails?

Anyway just an update, out of 65 assaults that I am currently running, 5 were defended, 1 was successful. The successful one was from one lone scout, go scout.

[Ramiel]

ta

[Chenier]

When I said 5 * 7.4% I was referring to the fact that you could have 5 fleets of 100 cruisers or 1 fleet of 500. Thus in a single turn you could have 5 attempts at the 7.4% chance. While each one taken on its own is 7.4%, the chance that 1 in the 5 is successful is not 7.4%, I just forget what it is. Again if you flip a coin 50 times, the chance that in any 1 flip you get tails is 50%, but what is the chance that at least 1 time in the 50 you get tails?

Anyway just an update, out of 65 assaults that I am currently running, 5 were defended, 1 was successful. The successful one was from one lone scout, go scout.

I don't remember those rules, which we learned pretty early on. Logic would say that if an event has 1% odds of resulting in something, doing it twice would double the odds of it happening at least once. However, it's obvious that if you do it 100 times, it's not 100% guaranteed to happen.

Indeed, if you flip two coins, you have 75% odds of having at least 1 heads (only 1 in the 4 possible combinations doesn't involve heads). Mathematically speaking, though, I'm not sure what the formula is.

[De-Legro]

I don't remember those rules, which we learned pretty early on. Logic would say that if an event has 1% odds of resulting in something, doing it twice would double the odds of it happening at least once. However, it's obvious that if you do it 100 times, it's not 100% guaranteed to happen.

Indeed, if you flip two coins, you have 75% odds of having at least 1 heads (only 1 in the 4 possible combinations doesn't involve heads). Mathematically speaking, though, I'm not sure what the formula is.

If I remember correctly the topic was called Permutations and Combinations.

Update by which I totally meant Binomial Theory

[Vellos]

The economist in me is laughing right now, but also very sad at the lack of stats knowledge in our player base.

Yes, if you do the same 1% chance thing over and over, your chance of eventually getting it will be higher.

And, if Dave's code is structured to relatively disadvantage large fleets (giving them increased chances of conquest but at a diminishing marginal rate with each additional ship), then, yes, the optimal strategy is to send lots of small fleets rather than a few big ones.

That is, of course, assuming a one-round game. Dave has clearly incorporated a code that damages planets and ups the odds of conquest. This feature might scale more linearly, or even convexly, making big fleets better and with an increasing marginal destructive efficiency. I don't know.

In sum, very simply, spamming small fleets is better than one big fleet, under the assumptions:
1. There are diminishing marginal returns (measured as % chance of captured added per ship) on each additional ship
2. There are diminishing marginal returns (measured as society level/population decreased per round) on each addition ship
3. There is no inhibiting code which creates diminishing marginal returns as the number of planetary capture attempts in one turn increases

Under those assumptions, spamming is better. A significant enough violation of any one of them, especially assumption 3, could put everything back in favor of big fleets. Big fleets also, of course, have the benefit of being able to kill defending fleets more effectively.

[fodder]

been something like 15 years since maths.

to be dead honest.. it would be better for all fleets in a spot to be counted as a giant fleet, unless the said fleet has an "evade" command or some such.

well, ok, you can make it a bit more effective by deciding that a fleet above certain size is not very effectively commanded by the commander,  whilst also deciding that too many fleets is counter productive, and that separate fleets can outflank/surround an enemy's fleet (though no reason why it can't do that automatically when you have only 1 nominal giant fleet that split itself for the battle to have such manoeuvres...)

[Tom]

Indeed, if you flip two coins, you have 75% odds of having at least 1 heads (only 1 in the 4 possible combinations doesn't involve heads). Mathematically speaking, though, I'm not sure what the formula is.

1-pow(1-odds, number-of-times)

[De-Legro]

The economist in me is laughing right now, but also very sad at the lack of stats knowledge in our player base.

That is such a stupid statement. Should I be said that the majority of the population doesn't understand how to design a robust state machine for reliable process control? Or the basic concepts behind PID controllers? How about what to me are very basic algebra systems, like partial differential equations? Peoples knowledge are bound to be tied to their professions and their interests.

Just in case it wasn't clear, we knew that the chance of a event with a fixed percentage happening at least once from multiple attempts was higher then that simple base chance, we couldn't off the top of our heads remember the exact formula to quantify the difference.

[Kai]

its funny because vellos showing pride in his ability to do a simple calculation is sadder than the inability to do that calculation

but i hope that you guys dont ever have to flip more than one coin. lol

[Chenier]

its funny because vellos showing pride in his ability to do a simple calculation is sadder than the inability to do that calculation

but i hope that you guys dont ever have to flip more than one coin. lol

Failure to remember a formula doesn't mean inability. It would be easy to look it up and actually calculate it, if so desired. I personally lack the motivation to do so, however.

[Vellos]

Should I be said that the majority of the population doesn't understand how to design a robust state machine for reliable process control? Or the basic concepts behind PID controllers? How about what to me are very basic algebra systems, like partial differential equations? Peoples knowledge are bound to be tied to their professions and their interests.

I generally think the ability to understand a coin-toss is slightly simpler.