1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

Meta Analysis: How popular IS civ 5?

Discussion in 'Civ5 - General Discussions' started by Zechnophobe, Oct 19, 2010.

  1. PestWulf

    PestWulf Chieftain

    Joined:
    Sep 28, 2010
    Messages:
    22
    Holy Crap!

    I just found out I can click on the posters name and add them to an ignore list!

    Suddenly this thread became interesting again! Thought I'd share in case any one else out there is kind of new to CivFanatics and do not care to read wall of text attacks made by forum trolls.

    To the OP, thanks for starting the thread, an interesting read to be sure.
     
  2. AlpsStranger

    AlpsStranger Jump jump on the tiger!

    Joined:
    Feb 8, 2009
    Messages:
    5,820
    This may not contribute much to the discussion but it is worth pointing out. If you take *all* of the facets of a single person they are, indeed, unique. If on the other hand I want to know what percentage of adult males are between 5'5" and 5'8" this fact is irrelevant. You may be quite unique taken as a whole, but your height is not. You also share your annual birthday with, roughly(if they were evenly distributed, they probably aren't), 1/365th of the population.
     
  3. Crezth

    Crezth What's Up, Danger?

    Joined:
    May 26, 2006
    Messages:
    10,352
    Location:
    Brooklyn
    Thanks, this is the point I am trying to make.
     
  4. JLoZeppeli

    JLoZeppeli Chieftain

    Joined:
    May 11, 2009
    Messages:
    598
    I starting to not understand what are we speaking about:confused:

    We started with some issues about popularity of Civilization V and we ended in complainig each other and some strange statements about the value of statistics and polls:eek:

    Ok, gathering samples is something we do in archeology too, we split a map in squares and pick up the samples from them so we can establish with numbers the probably living sites or, better, activity sites...

    It's necessary often to speed up work and to schematize the life progression of the site (it's a bit difficult to explain, i'm much better in italian, sorry).

    So, it may not be 100% careful, but it is largely usable to have a rough idea of the situation...
     
  5. DaveShack

    DaveShack Inventor Retired Moderator

    Joined:
    Feb 2, 2003
    Messages:
    13,108
    Location:
    Arizona, USA (it's a dry heat)
    Minor addition, the population and the sample must be "large enough" relative to what is being measured, and there are guidelines on how large the sample needs to be in relation to the population.
     
  6. Crezth

    Crezth What's Up, Danger?

    Joined:
    May 26, 2006
    Messages:
    10,352
    Location:
    Brooklyn
    Thank you, this is also relevant to the point I'm trying to make.

    The main idea here is that these guidelines tell us what a good sample is as opposed to a "bad" sample, and really my entire point is that the polling data from the first post doesn't give very much to go off of - at least reliably, and certainly not in any way of understanding the demographics on a larger scale.

    The population of Houston is not, for example, representative of the population of the entire United States. It is, however, representative of the population of Houston. Bringing this discrepancy into view is important.
     
  7. Zechnophobe

    Zechnophobe Strategy Lich

    Joined:
    Sep 1, 2006
    Messages:
    1,866
    Location:
    Goleta, California
    Uh, wow, what happened here?

    So, I'm going to only respond to one actual thing. For those that don't know:

    A Strawman is an argument where you attack a weak argument that wasn't actually made. For instance, if someone where to say "You claimed that this poll represented all civ players, here's proof you did not" and you had not made such a claim, that is a straw man. On the other hand, when you attack a person, instead of the argument they present, that's called 'ad hominem'. Confusing the two can make for very weird misunderstandings, some of which have occurred in this thread.


    As for the data. I think we are sorta looking at this the wrong way. Clearly these polls do not give us any certainty. They have a specific population they work on, their language isn't neutral.. there are lots of possible flaws.

    But there is a huge chasm of difference between scientific certainty, and just good indicators.

    Consider it from a logician's standpoint. It is well known that certain types of data aren't really the best. For instance, just because someone is a complete idiot, doesn't mean he couldn't provide good relationship advice. But it does sorta indicate you should maybe check further.

    We HAVE data here. We argue about the extent of its validity, but throwing it out because it isn't bullet proof is folly. Just because a poll only reflects the opinion of a subset of the population doesn't make it meaningless. Just don't go trying to convince someone it means something it doesn't.
     
  8. Axxon

    Axxon Chieftain

    Joined:
    Oct 2, 2008
    Messages:
    46
    The OP's question can simply be answered through Google trends, which graphs popularity of web searches. That should be a better proxy to "popularity" than polls and the likes. Not that I think popularity actually correlates with "good game" necessarily.

    http://www.google.com/trends?q="civ...vilization+iv"&ctab=0&geo=all&date=all&sort=1

    Civilization V is the small blue line, graphed around a few other well known franchises, and compared with Civilization IV.

    It's hard to see in the graph, but if you measure, the Civ V interest spike on release is clearly smaller than the Civ IV spike. About half as much. For whatever reason, Civ V generated less interest on the web than Civ IV at release.

    The Starcraft 2 and Call of Duty spikes are roughly 20x the Civ V spike.

    In terms of sales, VgChartz is notoriously inaccurate, but let us hope they can still get the orders of magnitude roughly accurate.

    Civ V is in the 100s of 1000s.
    Starcraft 2 is in the millions.
    Call of Duty Modern Warfare 2 is in the 10s of millions of units sold.


    So there's roughly 100 CoD players for every 1 Civilization player. There you have it. Keep in mind Call of Duty is up there as of the top 10 franchises of all time, so that's a harsh comparison.

    I still don't think that proves anything about the quality of the game though. There's 10,000 GM cars sold for every Ferrari. There's a lot of factors in play.
     
  9. ShuShu62

    ShuShu62 Chieftain

    Joined:
    Oct 14, 2010
    Messages:
    183
    Lightning does, in fact strike twice. :)

    Beer bet time. Guy flips a coin and it comes up heads, which way should you bet the next flip will be...

    Law of averages: Tails
    Law of Probability: Heads
    Law of Observation: Don't bet, the guy flipping the coin cheats, and that's the only scientificaly provable fact... welll, if you believe in that sort of thing at least.
     
  10. Zechnophobe

    Zechnophobe Strategy Lich

    Joined:
    Sep 1, 2006
    Messages:
    1,866
    Location:
    Goleta, California
    Yeah, you'd want to vote heads. All data points suggest it is a more likely outcome!
     
  11. r_rolo1

    r_rolo1 King of myself

    Joined:
    May 19, 2006
    Messages:
    13,818
    Location:
    Lisbon, Portugal
    Law of the wise @ss: catch the coin in the air and run away. It is far safer to ensure a win that way than betting in heads or tails :p
     
  12. Crezth

    Crezth What's Up, Danger?

    Joined:
    May 26, 2006
    Messages:
    10,352
    Location:
    Brooklyn
    Thanks for responding. I want to clear this up.

    I never said the information was worthless - in fact, multiple times I stated ways the information could be used. I also mentioned ways it could not be used, this time with the intention of correcting the many people who posted before me confusing the two population pools.

    *facepalm*

    Read up, boyo. You've got a lot to do.

    (You can skip ahead to the part about the law of averages.)

    Specifically, understand that each action's probability is independent of each other. You can do a lot of reading on this subject (or take a prob and stat course) to understand how you can never be said to be "due" a result with any certainty - this is a statistical fallacy that results from a fundamental misunderstanding of the law of large numbers.

    That is, the law of averages dictates "tails" but it is no more accurate than any other prediction and the real probability sits at approximately ~50%. The returns you get will always be near this "average."
     
  13. Zechnophobe

    Zechnophobe Strategy Lich

    Joined:
    Sep 1, 2006
    Messages:
    1,866
    Location:
    Goleta, California
    This is now completely off topic, but I *love* math, so I have to interject here. What if we took this to a different context. Let's say you are given a bag that has either red or green marbles in it. You reach in and blindly pull out a red one. What is more likely to come out next if you return the marble first?

    While we assume the 'flipping of a coin' scenario has equal odds for each side (even though not actually stated) we do NOT assume there are equal numbers of each marble in the bag, but we do know that we got a red one.

    In the real world, a coin is actually not going to exactly even, due to the weight distribution and shape of it. This is the same as not knowing how many marbles are in the bag, and their color distribution. So therefore it always makes sense to assume that all existing data is representative of the actual distribution of possible results. If we flipped the coin 1 million times, and 900k of them it came up heads, empirically we are very likely making the right decision to choose heads again.


    Oh, and ShuShu62 was correct, the "Law of Averages" does in fact state that the correct guess is tails. Um, but it is important to know that the Law of Averages is actually a term to describe the WRONG way that people often think about probability, so as usual, it is incorrect here as it normally is.
     
  14. Crezth

    Crezth What's Up, Danger?

    Joined:
    May 26, 2006
    Messages:
    10,352
    Location:
    Brooklyn
    Hey, I love math too! I love it so much I made a career out of it. Let's talk math.

    Marbles are of a finite quantity and if we remove a red marble from the bag it can be said that there is one less red marble in the bag. We thus adjust the probability to account for the different distribution of marbles. I assume this is what you were going for.

    We do not base probability purely after results obtained from empirical observation, however: probability is a synthesis of a number of different factors, and if we had a perfect, unweighted coin being flipped in a perfect world, the probability is always going to be 50%.

    This remains true even in an example of extreme probabilistic unlikelihood (the scenario you're describing is so massively unlikely it is almost quantum - I haven't crunched the numbers of 9:10 results in a 1:2 probability environment, but let's just say that's in the neighborhood of falling through the floor due to quantum displacement). It may appeal to common sense to make probability judgments based on empirical reasoning, but it is not mathematic. When you can choose between the mathematical prediction and the prediction derived from observation (ie the law of averages) the mathematical prediction is always, always the better choice.

    It has been tested and through sufficient repetitions the mathematical prediction is more reliable (a 50% success rate) versus the law of averages (less than that - I do not remember the exact study).

    If you like, I can try to fish up my source, but really the math speaks for itself.

    I am also aware of the trappings of living in a non-ideal world - it is my business to be aware of such things. But when dealing with mathematical theory we assume purity, because that gives us a much better starting point for when we apply it to real-world situations.

    I was unaware if he was touting it as fact or not. It remains true that the law of averages is a statistical fallacy, however.
     
  15. Charvel1

    Charvel1 Chieftain

    Joined:
    Jun 23, 2009
    Messages:
    65
    Location:
    Parker, CO
    Not sure if your English skills are getting in the way but what you're saying doesn't make any sense.

    Also, you failed the class. I'll try to teach it again. There is no 'average' that you can calculate here. The numbers are for THAT moment in time. It's entirely possible that all 300,000 owners are playing every single day. However, the Steam stats show that say at 2:15 PM 25k people were playing, then at 2:45 PM there could be an entirely different group of people playing the game, as many as 40k let's say. That adds up to 65k doesn't it? Then at 4:30 PM another group of 35k could be playing the game, making our new total 100k. So, again, all 300k owners of Civ V could be playing every single day (it's not likely, but possible). So you can't make any kind of claim about the 'average' number of people that are still playing Civ V.

    Finally, your little comment to Piece of Mind is baseless. He made a very good point and all you're doing is trolling. If that's what you're all about so be it.
     
  16. Zechnophobe

    Zechnophobe Strategy Lich

    Joined:
    Sep 1, 2006
    Messages:
    1,866
    Location:
    Goleta, California
    So, look at that post you made, and notice the words of mine you quoted? I say something about putting the marble back, right?[/quote]

    Yes, otherwise known as a 'fair coin'. But we don't know that is the case, and in fact, in the real world never is.

    I'm really not disputing any of this.

    The point is you have two sides of a coin. They each have a chance to come up , let those chances be A and B. A comes up. If A > B, this first event happens, on average, more often. That will remain true on the second flip. And since A > B, it is more likely that the first flip returns the result that is more likely.

    See what I'm getting at?
     
  17. AlpsStranger

    AlpsStranger Jump jump on the tiger!

    Joined:
    Feb 8, 2009
    Messages:
    5,820
    At the risk of missing the point and looking stupid in front of everyone, I think I get what you're saying. If there was something skewed about the coin (to a noticeable level) then you'd be safe betting the side it 'favored' would be the result of the first flip. If the coin was, to use a nice round number, 60/40 in favor of one side or the other, odds would be on your side to bet that the first result was going to match the second.

    Forgive me for making an ass out of myself if I misunderstood.
     
  18. Zechnophobe

    Zechnophobe Strategy Lich

    Joined:
    Sep 1, 2006
    Messages:
    1,866
    Location:
    Goleta, California
    That's pretty much it. If we are working from the standpoint of already knowing the exact chances of things to turn out one way or another, then we should use those.

    But if we don't, and we can only determine them by experimentation, we should always assume our previous results are indicative of this hitherto unknown process.
     
  19. Crezth

    Crezth What's Up, Danger?

    Joined:
    May 26, 2006
    Messages:
    10,352
    Location:
    Brooklyn
    Ah, well, there you have it.

    No, indeed, but the principles of probability (specifically that of a 50-50 chance, which is fairly well-modeled by a coin) are important in other sciences, especially quantum mechanics.

    I can vaguely see where you're going with this but I'm not sure what your point is. If A > B, then A is more likely for every single coin flip, regardless of what the first, second, etc. coin flip is.
     
  20. Zechnophobe

    Zechnophobe Strategy Lich

    Joined:
    Sep 1, 2006
    Messages:
    1,866
    Location:
    Goleta, California
    The point is that if someone in a bar flips a coin, and then asks you to pick a side for it to come up next, you should go with the side flipped the first time, since you don't know if there are other factors that might be making it more likely. This whole sorta discussion started with 'beer bet' on flipping coins :p.
     

Share This Page