CIV IV makes me sad

[ChITty]

After playing again tonight... I suggest whoever ran those tests to confirm the percentages run them again with warlords.

[trev]

I find it hard to understand the problems people claim to have on noble and lower levels, I am not a person to continually check my cities, just do mouseovers every 10 - 20 turns and if hammer, food, commerce, or GP production not what I expected, then I will go into cityscreen and adjust things. With this attitude I can win most of the time on emperor and I am trying to win on immortal, almost there 3 times, just losing to spacerace by 12 turns 1 game, just failing domination when space race goes off by around 6% land area couple of times. Maybe if I did check things more thoroughly I would gain that extra bit, but I am not that motivated. But others with similar attitudes to microing should be able to win on noble, prince, monarch levels without too much trouble.

[padillah]

Originally posted by trev
I find it hard to understand the problems people claim to have on noble and lower levels...

Bobby Fisher couldn't understand why other people couldn't play more than one game at a time.

Did it ever occor to you that you're actually good?

"I'm not sure if I'm crazy because I can see the answer or the rest of the world is crazy because it can't."
Albert Einstein attr.

Tom P.

[sparkyal]

those are great quotes!

anyway, while i do not claim to be awesome at the game, I do tend to win more often then not. As an almost warmonger (my take overs come later in the game), I actually use a similar strategy with citites as descibed above. I check in on each one every so often making sure they are where I want/expect them to be.

Maybe this will help you and maybe it won't. but to me, the game is about planning.

so, for example, let's say it's 500 AD and you notice you have a pretty large border with one particular civ. At that point you should begin to think about how to crush them. If they are next to you, they should die. In one particular game I calculated that i needed 20 turns to build an army of SEAL's big enough to invade a particular country. this took into account the need for BB's, DD's, CV's and TRN's and such. After 20 turns it was time to launch the attack. did it matter that they were a good friend? no! Did it matter that they had done nothing wrong? no! they existed so they had to die.

So, plan out wars and attack. and, as you build your army, others will not want to attack you.

Other peices of advice, use your advisors a lot. Look at the power grach to determine where everyone is in ralation to you. Try to be on the top of every graph.

Also, go to war for resources. And don't trade horses for example with someone with whom you are going to fight.

I think those things should serve you well.

sparky

[MightyTiny]

Re: Re: CIV IV makes me sad

Originally posted by Bkeela


I hear you. Programmers have come here and stated that the AI doesn't cheat in regard to odds - others explain it away saying that you only notice negative losses but never remember when you achieve an against the odds victory. That is BULL****!

How is it bull****?

First, people haven't just arbitrarily asserted, as an item to be taken on faith, that the AI doesn't cheat in regard to odds - what people have been saying is that if you suspect otherwise, you can check it yourself using the worldbuilder to set up a great number of identical battle-pairings with you against the AI, tally the results, and you will find that they gravitate toward the reported percentages over large number of trials. In other words, the AI DEMONSTRABLY does not cheat in regards to battle odds - this is something that is universally testable by all players willing to go to the trouble.

So since this is a demonstrable fact, not conjecture or arbitrary assertion, the only question remaining is: "Why do many people FEEL cheated by perceiving poorer battle results than they FEEL they should be allowed to expect by the percentages?"

If you feel that way, and it can be demonstrated that no actual cheating occurs, then there must be an error in your perception - an illusion.

That illusion is explainable by two well known facts (not bull****), which I explained in more detail in my post in the first page of this thread: First, a lack of understanding of the streaky nature of randomness, and secondly, the phenomena of confirmation bias.

Either of these, or both in combination, can account for the feeling of being cheated by the odds in the face of facts demonstrating otherwise.

Confirmation bias is something ALL people are subject to to some degree - including myself, including scientists, including and anybody aware of this cognitive error of thought. That is why only controlled tests can provide reliable confirmation or disconfirmation of claims such as "The AI is cheating with battle odds."

It is essential to understand that nobody is immune to cognitive error, and only by being aware of the possibilities of error, and the common types of error, and acknowledging the possibility of one's self committing those errors, do we have a chance of arriving at truthful explanations. So it definitely is not bull****.

I'll put the link to the explanation of confirmation bias here again, if you're interested. It's really worth checking out: http://skepdic.com/confirmbias.html

Once people get into their heads an idea they tentatively believe to be true, they tend to unconsciously selectively pay more attention to evidence seeming to support this idea, than to evidence disconfirming it, thus strenthening the belief - everyone does this, the difference is only between people who acknowledge this weakness of human thought, and consciously try to control against it, and those who believe themselves above such errors, and so see no reason for controlled tests when it comes to their beliefs.

[Arrian]

Originally posted by DaveMcW


As stated previously, the AI get higher support costs on Noble.

I stand corrected. Then about the only thing they get is the discounted upgrades.

-Arrian

[Gibsie]

Originally posted by Arrian
3. The AI gets bonuses, even at Noble level. Among them are lower unit support costs and big discounts on upgrade costs. The AI is pretty dumb, though.

That's why when in a recent multiplayer game where a human player died out soon and took control of an AI not too far from me, I was none too pleased when the human sent the dozens of units he had floating around to wipe out half my cities...

[padillah]

To reenforce MightyTiny, I also think most people don’t fully understand the math behind odds and randomness.

Take my example of putting mountains on a standard map. A person looks at 1% and says to themselves "it's a lock! 1% is nothing, I'm not even sure you can beat 1%." But when fully realised, in 4368 tries 1% will crop up at least 43 times.

We bandi about the phrase "one in a million" but, given the current population of the Earth that can happen 6500 times!

70% means that if you had this battle a mear 100 times you'd get beat 30 times. You can also think of it as a 30% chance of loosing the battle. That's not insignificant.

Some are saying wait for 80% or even 90%. Well, at 90% that's still a 10% chance of loosing.

Maybe that'll help, when you get into a battle think about the chance of loosing. It's a much better indicator (psycologically).

Tom P.

[Shogun Gunner]

Quite frankly, I've witnessed being on the winning side of unfavorable odds more than a few times.

Especially when you send the cats or cannons into battle first in order to cause collateral damage to defenders. You don't expect to win the individual battles, it's just softening up the position and causing some damage to other stacked units. Every so often your artillery will win those unfavorable battles.

It's not unfair, it's probability.

[MightyTiny]

Good posts Padillah and Shogun Gunner.

Also, one thing to add popped into my mind, in the light of what was said above. Something that may contribute to the feeling of the AI winning more than it's fair share of low odd battles:

If, as a human player, you don't often engage in battles with low odds in your favour, then naturally, you won't see as many low-odd victories as the AI who frequently takes on battles with unfavourable odds: this is simply a matter of statistics - more tries at low odd battles means a lot of losses, but also more victories against the odds.

As Shogun Gunner said, this is why the human player mostly tends to get HIS surpricing victories with catapults or artillery units that are set against the enemy to soften them up, without an expectation that they'd win.

[Aro]

Originally posted by anhalibut
ChITty,

if you're interested, here's a turn 1 save game with a really good starting location.

Egypt, Noble, pangea, marathon

I don't think I've had a better capital, and there are lots of resources nearby.


http://apolyton.net/upload/view/5027....Civ4SavedGame

Thanks!

Great start, spiritual civ (no anarchy)...
Playing!

[Aro]

Originally posted by MightyTiny

As Shogun Gunner said, this is why the human player mostly tends to get HIS surpricing victories with catapults or artillery units that are set against the enemy to soften them up, without an expectation that they'd win.

Or defending, when you do not have any other option...

[Qwertqwert]

Originally posted by Will9
I once won a battle at 0.3%. The AI gets no combat bonus. The AI can win at 0.3% so can I.

Really? Wow, the lowest I won at was 9.something % (I didn't check what exactly it was, just saw about 9) but I will rarely attack below that, unless the goal is just to weaken that unit (when you know there's no chance of just 1 unit killing theirs)

[Dis]

at thread.

Which is funny, as I routinely get pissed of at losing my superior units.

But 70% is 70%. Not 100%. Of course you are going to lose them sometimes.

[Dis]

as for noble level. I haven't completely mastered it yet. But I can stay even with the AI with no problem. The problem is I love to dominate games. That's why I still play warlord.

The balance is building the right things, and not building too many wonders. I'm pretty sure these are mistakes I make. Building wonders gives you less time to build military units. This is why the AI's always seem to have larger armies. The AI only concentrates on getting 1 wonder at a time. Where as I often try to get 2 or 3 at a time.

Next up is military strategy. I admit I'm not the best at this. Sometimes I make really stupid military mistake. Like moving a unit outside a city to protect a resource, not realizing the AI had a 2 move unit able to take my city. The key is utilizing defensive terrain, and not leaving your offensive units (the units who do not recieve defensive bonuses) out in the open. You simply have to improve your strategy so you don't end up with only 70% odds. If your units are 2 techs higher than the AI's as you said, you should get better than 70% odds. You can't just throw your troops into the meat grinder. Use strategy.