The Mechanics of Overflow Inflation

[Hammer Rabbi]

I have to imagine this is one of those things people knew about but kept under wraps like the old Research Agreement overflows giving way too many beakers.

Thanks to the OP for bringing this to light. I'll watch Moriate's vid when I have a computer again

maybe. but i'd think it would have reared its head in the competitions, particularly HoF where you can construct your settings for the desired goal. i havent seen any LPs from here doing it. I wonder if this can be done in mp.

 

[Browd]

Just tested this the same way Moriarte did (used IGE to give myself 3 GSs), but without Scholars in Residence. As is typical on Deity, I was 7th in literacy (33%, 7% behind the leader), with just over 300 bpt. Bulbed the 3 GSs, generating about 6000 beakers. After sequencing the lower cost techs, I peaked at 16,400 of beaker overflow, which was enough to take me to Radio. This is broken even without Scholars in Residence. Can only imagine what Babylon could do with this.

 

[ravenmagus]

Just tested this the same way Moriarte did (used IGE to give myself 3 GSs), but without Scholars in Residence. As is typical on Deity, I was 7th in literacy (33%, 7% behind the leader), with just over 300 bpt. Bulbed the 3 GSs, generating about 6000 beakers. After sequencing the lower cost techs, I peaked at 16,400 of beaker overflow, which was enough to take me to Radio. This is broken even without Scholars in Residence. Can only imagine what Babylon could do with this.

I played a game with Babylon and did as much exploiting as possible. Turns out you don't need to be Babylon at all and that extra great scientist is a waste.

I saved my first Great Scientist, went for Education to get a few more, and then slow-teched my way through Architecture while I sent out diplomats to get Scholars in Residence passed.

When Scholars was ready, I was last place in tech by a fair margin: I had Architecture just completed, but no bronze working or sailing. I used one Great Scientist with 150 bpt, and that gave me enough science to research the entire tech tree and then get Future Tech 10 times before running out of beakers. Off of ONE scientist. I had two others prepared that sat around unused because they would have caused my beakers to overflow into negative.

 

[slowcar]

I used one Great Scientist with 150 bpt, and that gave me enough science to research the entire tech tree and then get Future Tech 10 times before running out of beakers.

crass

 

[Paszczak]

So, instead of depleting, current overflow is increased to 1+0.3*7/8 = 126.25% of its value every time a tech known by 7 (out of 8) opponents is researched and it's further inscreased to (1+0.3*7/8)*1.2 = 151.50% if Scholars in Residence was also passed? Am I getting this right? If yes, that's one huge-ass bug .

The question is, can you actually play the game using it, like I dunno - expand or conquer someone , or is this restricted to 4-city turtle and science victory? Because that's already easy enough without exploiting the game.

 

[grades]

So, instead of depleting, current overflow is increased to 1+0.3*7/8 = 126.25% of its value every time a tech known by 7 (out of 8) opponents is researched and it's further inscreased to (1+0.3*7/8)*1.2 = 151.50% if Scholars in Residence was also passed? Am I getting this right? If yes, that's one huge-ass bug .

The question is, can you actually play the game using it, like I dunno - expand or conquer someone , or is this restricted to 4-city turtle and science victory? Because that's already easy enough without exploiting the game.

It's viable for any strategy. In fact, since having researched fewer technologies increases how well this works playing a tradition 4-city science game actually makes this less effective.

Really though this is rather exploitative of the game's mechanics and feels cheap when you jump from the renaissance era to the information. So if you're wanting to play a legit game you probably just want to avoid doing this altogether.

 

[Ronza149]

Where does this "0.3*percentage of players discovered this tech" come from?I never see this mentioned anywhere in the game.

 

[Fallen_Kings]

Where does this "0.3*percentage of players discovered this tech" come from?I never see this mentioned anywhere in the game.

It was first discovered in "Research Discount"-the cost of a technology goes lower each time another player discovers it.

Actual Cost = Original cost/(1+0.3*players discovered it/players alive)

sometimes use it to predict AI's technology alignment and see if it can build a wonder we want or new military units we may encounter at war.

 

[joshua43214]

Does this really surprise anyone?

We are talking about a game that alphabetizes the Inca by the t in "the," instead of the i in "Inca."
If Firaxis sinks any lower in competence, they will be competing with EA for the bottom rung of game developers.

 

[YourHarry]

Haha. This is pretty awesome!

I just thought of a crazy strat using this. Holy crap I'm so excited.

 

[YourHarry]

So I just tried this with Babylon. Built 3 cities each with library. I had calendar tech to maintain the happiness and gave away all my money (luckily Venice was my friend). This created a deficit in gold flow, slowing down my science. To do this, I kept making more units to increase maintenance cost. This also allowed faster population growth and increased happiness, as I was able to complete city states' quests. I grew 3 cities to decent size, "wasted" whatever little science output on philosophy. Researched every other immediate techs down to "1 turn" remaining. So at this time, I basically had: Granary, Writing, and Calendar.

I declared war on Venice, suicided units, to return the gold flow to positive. Waited 8 turns, researched drama and poetry. And then bulbed GS.

I have to go to school and don't know how well this will work. Science overflow is definitely increasing. This was around turn 75 deity

 

[Fallen_Kings]

Does this really surprise anyone?

We are talking about a game that alphabetizes the Inca by the t in "the," instead of the i in "Inca."
If Firaxis sinks any lower in competence, they will be competing with EA for the bottom rung of game developers.

Like C&C F2P?

 

[Cromagnus]

Moriarte: Watching your video (thanks for posting it) I noticed you missed an opportunity to get more beaker growth... not sure if it was on purpose or if you didn't realize this, but... at any given level of the tech tree you can tell which tech has been researched by more people, just by looking at the cost. Navigation and Metallurgy both have the same default cost (1150) but in your game, Navigation was noticeably cheaper than Metallurgy. I noticed you eventually theorized that if you had taken Navigation earlier you would have gotten more overflow, but I don't know if you realized that you can determine this in general for every tech. If it has a lower cost than other techs at the same level, then you know for certain that at least one player has researched it.

So, theoretically, you could use this to "ride the limit" of 210,000: Check the costs and stop taking techs that have already been researched when you're at risk of going over 210k. Start again when you drop safely down far enough, say 150k beakers.

Of course, you were planning on stopping at some point to Dominate(tm), but still. The overflow multiplier for taking Navigation would have been higher than Metallurgy.

Anyway, I'm guessing the optimal approach, if used correctly, would allow you to complete the entire tech tree, the limiting factor just being the number of techs. Furthermore, I'm guessing it's better to bulb your GSes on separate turns. I think the overflow applies each time you research a tech, and by bulbing your GS's on three separate turns, you research two extra techs. Since the overflow from ONE GS is enough, you'll finish whatever path you're trying to finish 2 turns faster that way.

 

[Cromagnus]

Does this really surprise anyone?

We are talking about a game that alphabetizes the Inca by the t in "the," instead of the i in "Inca."
If Firaxis sinks any lower in competence, they will be competing with EA for the bottom rung of game developers.

I think that was logical and on purpose, actually. I think, with the exception of "The Netherlands", which I believe is the correct name for that country in English, they chose "The" for civilizations that don't have a well-known country name. England is a country. China is a country. When the Aztecs and Inca were dominant civilizations, what were the names of those countries? I don't think, by today's standards, they had "country names", or at least I've never heard of them. I have however heard or seen it written as "The Aztecs" and "The Inca", so that naming convention makes sense to me. It took me off guard at first too, but I think it's logical.

/shrug

 

[joshua43214]

Just played Aztecs (random), Deity, on Pangaea, decided to try a "worst case scenario" with this.
2 city coastal (sucks to not research sailing), typical tradition set up, planted 1 GS, saved 2, lost WC vote for scholars by one vote (reloaded and bought a city state instead of a second school). Bulbed both GS's together with 238 bpt, finished Stealth, Nanotech, and Nuclear Fusion by T201.

So, there I am, 2 cities, all the elite military techs done on T201, missing 4 trade routes, less than 100 gpt, less than 1500 gold saved, finished Big Ben, half way through Brandenburg Gate. Dom win on T242. A very nice win time, but about the same as doing a 3-4 city arty rush.

This feels a lot less like and exploit than I expected it to. Will try it again but with a less isolated inland start and see if I can get a better economy going. Being coastal, and not researching sailing early on Deity killed my economy in a major way, and then dealing with the industrial era dogpile made things even worse. Coastal starts just suck for Pangaea domination, but it was fun to sink Boudicca's navy with a single missile cruiser

 

[Cromagnus]

I think it's *very* likely that with an inland Pangaea start you could do better than that. Bulb XCOM, send one to each capital, game over on t205. Of course, you can pull off a sub-t200 Deity domination win without cheating, but the reason this exploit is *so bad* IMHO is that it makes a t240 Deity Victory *trivial* for anyone to achieve. It takes skill completely out of the equation.

 

[Cromagnus]

joshua: I actually think it's worse that you get similar results. It makes it very hard to tell whether someone legitimately posted a win or cheated... for things like GoTM, HOF this is especially bad. Especially since you can be subtle about it. You don't need Scholars in Residence. You can pretty much do this at any point once you get universities up. It won't complete the tree, but you could probably get super-early artillery even with a "bad start". Like t135. I think... I haven't personally had a chance to try this... yet. You can bet I WILL try this as soon as I get a chance. Sadly, I won't be able to play much Civ for the next few weeks. Real life sucks.

My point is, someone who played optimally might get t145-150 artillery. Someone who played very much not optimally will probably get the same result with this technique. That's why it irks me.

 

[joshua43214]

I think that was logical and on purpose, actually. I think, with the exception of "The Netherlands", which I believe is the correct name for that country in English, they chose "The" for civilizations that don't have a well-known country name. England is a country. China is a country. When the Aztecs and Inca were dominant civilizations, what were the names of those countries? I don't think, by today's standards, they had "country names", or at least I've never heard of them. I have however heard or seen it written as "The Aztecs" and "The Inca", so that naming convention makes sense to me. It took me off guard at first too, but I think it's logical.

/shrug

No, it is simply wrong.
"The Aztecs" goes under "A," not "T," same as The Netherlands goes under "N." It is simple incompetent and lazy programing with no oversight. I don't program games, but I write a lot of MATLAB code for numerical analysis of data, sorting is something I do a lot of. Writing simple if/then logic loops is one of the first things you have to learn, I can't imagine other programming is any different.

Writing code that lets math go unbounded will get you fired in many jobs.

 

[Karmah]

What an horrible bug , it needs an hotfix right now !

 

[Cromagnus]

No, it is simply wrong.
"The Aztecs" goes under "A," not "T," same as The Netherlands goes under "N." It is simple incompetent and lazy programing with no oversight. I don't program games, but I write a lot of MATLAB code for numerical analysis of data, sorting is something I do a lot of. Writing simple if/then logic loops is one of the first things you have to learn, I can't imagine other programming is any different.

Writing code that lets math go unbounded will get you fired in many jobs.

Oh, my bad, I thought you meant the inconsistency in naming them "The Aztecs" instead of "Aztecs". Yes, the alphabetization is wrong, I agree.

However I must disagree that writing bad code like that will get you fired in many jobs. Sadly, in my experience that is simply not the case. Especially in the game industry.

Also, bear in mind that at many game companies (and I'm guessing Firaxis is no different) it's designers, not programmers, who often end up writing a lot of the logic bugs by "coding" in various scripting languages... so the error in logic may not be on the part of of someone with a computer science degree.

EDIT: In fact, I'd go so far as to say there's no such thing as "good code" and "bad code"... there's "well-tested" code, and "poorly-tested" code. All code has bugs. Companies that have good a good *QA* process, and have the *luxury* of enough time (read: money) in their development schedule to find and fix those bugs release software with less bugs. So IMHO a buggy game is in no way an indication of bad programming. It's an indication that they were in a hurry to get the game out the door, so they didn't have time to find/fix all the bugs.

Furthermore, some bugs are extremely HARD to find.... How many thousands of hours of QA would be needed by expert civ players before all such exploits could be tracked down? If it takes tommynt or moriarte to play at a high enough level to discover an exploit, and they don't have such a player on their QA team... they won't find it.

Just my two cents.