Il Principe
Prince
I am aware that the AI needs further work, don´t worry. Just saying it already came a long way since RoM ( and even more since vanilla)
Railroads
- Thread starter Koshling
- Start date
@Thunderbrd, the abilities of CPU are not so great as you can think.Do you know the Go game? It is no such a complex game. But did you know, the greatest specialists can't build AI which would properly play it?
Actually some progress has been made recently. A 9 dan professional lost a 4 stone handicap game in a demo tournament against a machine earlier this year, and in June another 9 dan pro lost a 3 stone handicap game to another program. To put that in perspective, a 4 stone handicap from a 9 dan pro is about where a 6 dan amateur would need to be for a 50-50 chance of a win, and that's around European or US amateur champion strength.
References: http://en.wikipedia.org/wiki/Computer_Go#Recent_results
You might also be interested in: http://writingagoprogram.blogspot.com/
JosEPh_II
TBS WarLord
@n47,
I like oatmeal raisin cookies.
Burned out on chocolate chip even though I love chocolate. Mmm.. Dark chocolate covered raisins or peanuts...Mmmm!
JosEPh
I like oatmeal raisin cookies.
Burned out on chocolate chip even though I love chocolate. Mmm.. Dark chocolate covered raisins or peanuts...Mmmm!JosEPh
Dancing Hoskuld
Deity
Can't you talk a little bit about make up or cookies?
I once traced a recipe for ginger biscuits back to about 1600 West Country England (I am from Devon). Tradition has it they were made for the "Maid Hiring Fare". I then cooked a gross in total with the three variations over time. Honey, Treacle (Molasses in the US of A) and Golden Syrup. They were well received and I discovered that using that much fresh ginger can give you a slight high from absorbing the ginger through the skin.
Actually some progress has been made recently. A 9 dan professional lost a 4 stone handicap game in a demo tournament against a machine earlier this year, and in June another 9 dan pro lost a 3 stone handicap game to another program. To put that in perspective, a 4 stone handicap from a 9 dan pro is about where a 6 dan amateur would need to be for a 50-50 chance of a win, and that's around European or US amateur champion strength.
References: http://en.wikipedia.org/wiki/Computer_Go#Recent_results
You might also be interested in: http://writingagoprogram.blogspot.com/
I think you may have the wrong levels there. I made it up to about 5 ku (sp?) which means I would get 5 stones from a one dan player. I have seen and played against the computer recently and it has come a long way since I first saw a game. The free version I tried was about par with me. I suspect that if I forked out money I would loose so I didn't.
Thunderbrd
C2C War Dog
AI in Civ is really made up of the following:
1) Create generic means to identify the general values of particular choices, such as what promotion, what tech, what civic to adopt etc...
2) Enable a leader personality to tweak the way they value particular decisions by coloring those decisions with a 'flavor' paint (and often establishing a number for how influential that flavor is to the choice)
3) Generate large scale strategy concepts and let leaders choose them based on current status and personality.
4) Build in reactive strategies.
I'm sure Koshling can say more on this. Sure, AI could be an endless project in itself. The more its worked on, the better it is. There's very few limits as more and more strategies and perspectives may be taught to the system and ways to vary the ways those perspectives are adopted.
But if you're arguing that the AI can never match a player, you're probably right. But there are places where the AI's likely to confound the player with such an amazing ability to analyze through the complex patterns of this game that the player may end up having a lot to learn from the AI to be as good as he can be.
1) Create generic means to identify the general values of particular choices, such as what promotion, what tech, what civic to adopt etc...
2) Enable a leader personality to tweak the way they value particular decisions by coloring those decisions with a 'flavor' paint (and often establishing a number for how influential that flavor is to the choice)
3) Generate large scale strategy concepts and let leaders choose them based on current status and personality.
4) Build in reactive strategies.
I'm sure Koshling can say more on this. Sure, AI could be an endless project in itself. The more its worked on, the better it is. There's very few limits as more and more strategies and perspectives may be taught to the system and ways to vary the ways those perspectives are adopted.
But if you're arguing that the AI can never match a player, you're probably right. But there are places where the AI's likely to confound the player with such an amazing ability to analyze through the complex patterns of this game that the player may end up having a lot to learn from the AI to be as good as he can be.
No cheating! Dark chocolate covered raisins or peanuts are not cookies!@n47,
I like oatmeal raisin cookies.
JosEPh
But I checked oatmeal raisin cookies. They look so yummy, I fill like I will not be able to sleep this night. I must cook them!I thought, I felt in loveI once traced a recipe for ginger biscuits back to about 1600 West Country England (I am from Devon). Tradition has it they were made for the "Maid Hiring Fare". I then cooked a gross in total with the three variations over time. Honey, Treacle (Molasses in the US of A) and Golden Syrup. They were well received and I discovered that using that much fresh ginger can give you a slight high from absorbing the ginger through the skin.
... but then I realized that ginger is not an alcohol. But still, it looks pretty yummy too, so I guess I'll have two reasons to not sleep this night. (Should I say "to not" or "not to"? ... I'm such a dummy with English.
)Something like this. If there would be an AI scientist here, maybe there would a chance, but only very very maybe.
The thing is, it will always choose hard-coded pattern. It will not create a new one its own. Or it will even not learn some new ones from a human player. For such complex game as it is, there is so many possibilities, that you can not practically hard-code even a reasonable piece of good pattern, I'm afraid. Moreover, when the game is developed the old patterns may not be to nice any more. So there is a chance, that you will still and sill need to tweak the AI.
I would say, there are actually quite many limits. If you would have map 1000x1000 plots and would like to use some algorithm of cubic time on it (cubic time means that the algorithm needs something like x * size_of_data * size_of_data * size_of_data CPU operations to complete the computation) you would need to wait something like x * 8 years with 4GHz processor to finish. And such algorithms are nothing outstanding, there are much worse. (Luckily we rather do not need to use a whole map as our data.)
No way! I will not do the dirty job. I am already specialized in too many things and still need to learn some. No no no, I am not from AI.You might also be interested in: http://writingagoprogram.blogspot.com/
Professional levels are about 1/3 stone per level, whereas amateur are one stone per. Professional shodan (1 dan) is about amateur 6 dan. 9 dan is the top recognized rank, so it has a little more span than lower ranks. It is believed that a top professional would need less than a 2 stone handicap to beat God (i.e. - a perfect player). Of course in an even game ANY mistake is a loss, so they'd never have any chance at all evens
Last I was graded I was 2 dan, and once made it into the British top 8 playoffs for the British championship (I was probably close to 3 dan at the time, but stopped playing shortly afterwards).
It's spelled 'kyu' BTW
Dancing Hoskuld
Deity
Professional levels are about 1/3 stone per level, whereas amateur are one stone per. Professional shodan (1 dan) is about amateur 6 dan. 9 dan is the top recognized rank, so it has a little more span than lower ranks. It is believed that a top professional would need less than a 2 stone handicap to beat God (i.e. - a perfect player). Of course in an even game ANY mistake is a loss, so they'd never have any chance at all evens
Last I was graded I was 2 dan, and once made it into the British top 8 playoffs for the British championship (I was probably close to 3 dan at the time, but stopped playing shortly afterwards).
We had a top player out from Japan and I was sure he was only 5 dan, but I could be wrong. He played 6 of us at once. With me as the lowest. He beat each of us by 5 points exactly except the best player who was being tested for 2 dan. He beat him by 10 points.
Being Australia's capital we have the embassies here so every now and then people from the embassies sponsor a tournament. The play stiles of the different countries are very different.I thought so but what with this head cold and hay fever it is hard enough keeping my eyes clear so I can see the screen
Dancing Hoskuld
Deity
I thought, I felt in love
(Should I say "to not" or "not to"? ... I'm such a dummy with English.
In that sentence both are are valid and have the same meaning. Sometimes English is like that.
Getting back to the original topic of the thread...
http://www.wired.com/autopia/2013/09/maglev-310-mph/
http://www.wired.com/autopia/2013/09/maglev-310-mph/
But does it need steel!?
Pretty sure it does 
.
Early railroads however, were made from simple iron (and even wood, but that didn't work very well). IMO, Railroad should require Iron, and Steel should just allow you to build it faster/cheaper.
http://explorepahistory.com/hmarker.php?markerId=1-A-1A8
.Early railroads however, were made from simple iron (and even wood, but that didn't work very well). IMO, Railroad should require Iron, and Steel should just allow you to build it faster/cheaper.
http://explorepahistory.com/hmarker.php?markerId=1-A-1A8
Thunderbrd
C2C War Dog
Footnote on the original topic: this Steel prereq on railroads is driving my wife nuts right now!
Il Principe
Prince
Could alternative fuel requirements be considered? Right now it needs coal. Which is historicaly fine. But charocal would work too. Ok that would make access too easy gameplay wise, but how about adding oil/oil products as an alternative to coal?
They come later in the tech tree, but somehow I often find myself having them long befor I can build railroads, due too the lack of coal.
They come later in the tech tree, but somehow I often find myself having them long befor I can build railroads, due too the lack of coal.
Yup. I'm imagining a locomotive pulling 6 cars of charcoal and one of people.
Have you considered how light is charcoal? It is something like 5.5 times lighter then coal.Il Principe
Prince
Yup. I'm imagining a locomotive pulling 6 cars of charcoal and one of people.
Not saying it would be super effective.
But in theory you could burn even wood to power a steam engine. Lots of wood.
But like I said: gameplay wise charcoal would make it too easy anyway.
It is effective. But per kg.Not saying it would be super effective.
It sometimes can provide even more energy then ordinary coal. (per kg)No! One is -- you must store the fuel in a train. So even if we need the same mass of charcoal as coal, it brings more mass to the train by a need of more cars. And the second -- you would need to build 5.5x bigger engine. -- It would be damn heavy train. And if the mass goes over some point the train is not even going to move. So it very probably would be a damn heavy not moving train.
Or maybe maybe -- a damn heavy train with one usable car. 
Yeh, me too, which was the reason I posted this in the first place. I really think we should just change it to iron and be done.
Similar threads
