colonization sucess percentages

[_Astolph_]

Hi,
I'm a newbie playing a game with Portugal using last patch and last beta, but colonization is annoying me. Are that "success percentages" correct? I am asking this because the colonization of Fernando Pó, Pernambuco and Ciskei were difficult.

For example, i had to send 6 colonists to estabilish a trading post at Ciskei. The success rate was 45%, so, if i remember statistcs from school well, the chance of this happens is:

0.55 * 0.55 * 0.55 * 0.55 * 0.55 * 100 = 5% !!

This makes me mad. Am i THAT unlucky? Note that similar things happened to other attempts of colonization. I don't believe this success percentage is correct...

Regards

 

[Freebot]

Improbable does not equal impossible. Also, the random number generator in EUII seems to be "trendy."

Edit: 5% is a 1 in 20 chance. Therefore, this should be a fairly common occurence.

 

[Slargos]

I failed 10 times in a row the other day trying to colonize an indian province as Mughals at 55-60%. Probability of 0.6%..

 

[JaxomCA]

Hi,
For example, i had to send 6 colonists to estabilish a trading post at Ciskei. The success rate was 45%, so, if i remember statistcs from school well, the chance of this happens is:

0.55 * 0.55 * 0.55 * 0.55 * 0.55 * 100 = 5% !!

This is an often misused concept of probability. Each colonisation event is an independent event, in this case slightly less than 1 chance on 2 to succeed. The fact that you failed in the previous, or 100 previous attempts has no impact on your next attempt.

Now that is true only for truly random events. A computer random number generator is not random. It follows a precise algorithm to generate a number. Unless a new seed is provided, each number is completely dependant on the previous results but not in the statistical sense of increasing probabilities.

We can't tell whether the RNG is re-seeded during the game, but it is certainly re-seeded when you load the game. So if I get too many consecutives annoying failures, I load from the autosave or even the previous autosave if nothing much has happened in the world during those 2 years. This guarentees a new seed, which may or may not produce a bad streak.

 

[unmerged(16181)]

We can't tell whether the RNG is re-seeded during the game, but it is certainly re-seeded when you load the game.

I believe Johan has said it is not, and recommends the reloading method you talked about....

 

[unmerged(25696)]

This is an often misused concept of probability. Each colonisation event is an independent event, in this case slightly less than 1 chance on 2 to succeed. The fact that you failed in the previous, or 100 previous attempts has no impact on your next attempt.

Now that is true only for truly random events. A computer random number generator is not random. It follows a precise algorithm to generate a number. Unless a new seed is provided, each number is completely dependant on the previous results but not in the statistical sense of increasing probabilities.

We can't tell whether the RNG is re-seeded during the game, but it is certainly re-seeded when you load the game. So if I get too many consecutives annoying failures, I load from the autosave or even the previous autosave if nothing much has happened in the world during those 2 years. This guarentees a new seed, which may or may not produce a bad streak.

It's true that dice don't have memory, but what Alstoph said is that there was only a 1 in 20 probability he would fail 5 attemps in a row. His formula is then correct, because in a group of 20 players, only one would normally end up with no TP (and about less than 1 would end up with a lvl 5 TP).

What you say is only true for a focused view. On a larger scale, probabilities are pretty well balanced. In other words, probabilities are only good for analizing after a large set of events, not at predicting the outcome of a single one.

 

[unmerged(19159)]

is number of colonies but not colonial cities decreasing chances of next colonization ?

 

[Castellon]

IIRC that shows in the displayed percentage.
Thus the members who complain about the 1% chance to colonize a province.

 

[_Astolph_]

Well, i think i was a bit histerical, sorry o
About the probability formula, i think it's correct, since it's for independent events (as colonizing is).

 

[unmerged(3825)]

IIRC, there are modifiers on the success percentage, as well. I could be wrong, but I think monarch admin skill is one modifier (could be pos or neg), and there are positive modifiers for army in the province, conquistador in the province, and explorer in the port of the province. It seems to me that religion and stability may have some impact as well, but I'm really not sure about that.

Bottom line, I don't think you can look at the percentage alone and assume that it tells the whole story.

 

[Castellon]

But the percentage does reflect all calculations.
See the FAQ for details.

 

[unmerged(19081)]

IIRC, armies does not increase your chances, only conquistadors.

 

[unmerged(7225)]

Armies don't affect the % chance of establishing, but they do protect the fledgling colony from native revolts; preventing the natives from destroying or damaging a <700 pop colony. Thus, in the second order, they DO help establish colonies. I _ALWAYS_ station an army anywhere I try to establish a colony if the aggressiveness is medium or better and I want to incorporate the natives. Armies DEFINATELY help these situation.

 

[unmerged(19081)]

Yes, it helped, it's an evidence, but it does not change the colonization chance percentage.