"Formal" definition of a popular game

Edit: Alright dumb idea. I still think it has value, but that's irrelevant if no one cares.

IMGO that's a bit too partial to modern "play and ditch" type PC and console games, which have a very quick peak period after launch and are quickly superseded within a matter of months. It also doesn't account for these games that remain popular for years even without having a modding scene, or that remain popular in their "standard" form.

Examples include older arcade games that could remain favorites for years (also due to a lower renewal/turnaround rate, I remember Bubble Bobble and Puzzle Bobble being tremendously popular in every arcade room that had them, even when newer games and even releases of the same were available side-by-side) despite them being unmoddable by definition).

Also, games like Warcraft II, Counterstrike etc. while both are to some degree moddable, they enjoyed long popularity in their original forms and versions.

And what to say about Myst? No multiplayer, no modding scene, not even gameplay by some accounts ;-) yet it remained top-selling for years.

Maes said:

IMGO that's a bit too partial to modern "play and ditch" type PC and console games, which have a very quick peak period after launch and are quickly superseded within a matter of months. It also doesn't account for these games that remain popular for years even without having a modding scene, or that remain popular in their "standard" form.

Examples include older arcade games that could remain favorites for years (also due to a lower renewal/turnaround rate, I remember Bubble Bobble and Puzzle Bobble being tremendously popular in every arcade room that had them, even when newer games and even releases of the same were available side-by-side) despite them being unmoddable by definition).

Also, games like Warcraft II, Counterstrike etc. while both are to some degree moddable, they enjoyed long popularity in their original forms and versions.

I edited it just before you posted to factor in Replay value separately from modability for that very reason. Thanks for bringing it up.

Maes said:

And what to say about Myst? No multiplayer, no modding scene, not even gameplay by some accounts ;-) yet it remained top-selling for years.

Hmm. I'll need to think on that one.

Edit: Well it had the first impression value, and the gameplay I guess translates into the abstract variety that involves a really large reliance upon some stellar level design (which I guess includes game's the puzzle elements).


I guess it holds up.

Hmm...so with your formula, replay value can at most halt popularity decline (even if it's r->infinite, at most it will bring that (1 + (t / r)) to 1.0, and if it's r < 1 it will make it greater than 1.0 thus reducing overall popularity. IMHO replay value should be a direct multiplying factor, not a cushioning one.

Let alone that popularity can actually increase over time, especially if the game entered the market quietly, something that the t/r relationship just can't model.

And don't forget totally unmoddable games that still have great replay value and popularity (adventure games are a very good example of that), there should be an exclusion term or distinct formulae.

Maes said:

Hmm...so with your formula, replay value can at most halt popularity decline (even if it's r->infinite, at most it will bring that (1 + (t / r)) to 1.0, and if it's r < 1 it will make it greater than 1.0 thus reducing overall popularity. IMHO replay value should be a direct multiplying factor, not a cushioning one.

Well I suppose it CAN be an additive element (Left 4 Dead I think makes a good recent example), but I disagree about making it a multiplying factor. The other elements (gameplay, etc.) have to be in place first before replay value can have an effect, usually.

Maes said:

Let alone that popularity can actually increase over time, especially if the game entered the market quietly, something that the t/r relationship just can't model.

Yes, that's happened before. I guess I could call it the 'cult' factor or the 'meme' factor or something. Not an element so much as a result of the game as it is a result of a slow trickle of its first impression value, I guess.

Lemme think on this one for a bit.

Maes said:

And don't forget totally unmoddable games that still have great replay value and popularity (adventure games are a very good example of that), there should be an exclusion term or distinct formulae.

Yeah. That's why I separated modability from replay.

Maes said:

And what to say about Myst? No multiplayer, no modding scene, not even gameplay by some accounts ;-) yet it remained top-selling for years.

Anoraks who don't like to play games finally got a product released for them, so they all jumped on board :P

Seriously though, it would be interesting to apply these formulas to actual games.

I think you need a value for fanaticism as well, the human equation does matter in the popularity of some games. For instance, I don't think Starcraft would be anywhere near as popular today if not for the legions of Korean "cyberathletes" who raised the bar of multiplayer RTS strategy to levels previously unseen in the Western world.

These sums hold up well for a game like Doom, which was a large technological jump and has both great first impression and modability, and especially when comparing it to similarly-built games of the era like ROTT or Duke3D.

I think there is a bit more to it, or at least a different formula, for other more games like Tetris, Mario, or Bejewelled.

There's another factor too: For PC games only, a game can become more popular as more people are able to afford systems that can actually play it and not just run it. E.g. the first Quake and even Doom didn't enjoy maximum popularity immediately upon release, while when more and more people had MPC2 or MPC3 level machines, there was much more room for enjoyment. This is a non-issue for console or arcade games, of course.

The popularity of multi-platform titles is another matter too: a game could be a hit in the arcades years before it found its way to PCs and consoles, and then live a second "hit" period or just flunk it completely.

And there's also the case of the "dusty diamonds" kind of games: they may have all of the replay, gameplay, moddability etc. factors, but remain largely unknown to the public.

IMHO you can't really quantify game popularity, any more than you can quantify teen pop idol, fashion top-model, or hollywood star popularity. In those areas, even if you have what it takes, you may still end up being a nobody. While if *somehow* you make it, then it's about marketing and presence skills to keep popular for a while.

In other words, popularity is an "a posteriori" quality you attribute to a game. It can't be predicted if a title will ever be popular, or for how long/under which circumstances. Even "a posteriori" popularity is hard to quantify. You consider sales, or total number of installations? Over which period of time? E.g. footy games enjoy incredible sales peaks when they are new while Myst sold millions of copies of the same exact game over a number of years etc.

Super Jamie said:

I think you need a value for fanaticism as well, the human equation does matter in the popularity of some games. For instance, I don't think Starcraft would be anywhere near as popular today if not for the legions of Korean "cyberathletes" who raised the bar of multiplayer RTS strategy to levels previously unseen in the Western world.

Yeah, that was a bit of a freak occurence, wasn't it? :P Surely Starcraft would have held fast in popularity regardless, but those bloody Koreans really gave it boost. I'll account for this with a separate function I'll call 'cult' value that factors in some degree of probability (just because trying to track all possible human elements is waaaay beyond the scope of what I could do in any reasonable amount of time :P).

Super Jamie said:

These sums hold up well for a game like Doom, which was a large technological jump and has both great first impression and modability, and especially when comparing it to similarly-built games of the era like ROTT or Duke3D.

I think there is a bit more to it, or at least a different formula, for other more games like Tetris, Mario, or Bejewelled.

Well for something like tetris, let's see. Tetris was mainly popular because of the timing (and game platform, too) of its inception, right? That could also be a factor. I think my equations hold for Mario games, though...

Maes said:

There's another factor too: For PC games only, a game can become more popular as more people are able to afford systems that can actually play it and not just run it. E.g. the first Quake and even Doom didn't enjoy maximum popularity immediately upon release, while when more and more people had MPC2 or MPC3 level machines, there was much more room for enjoyment. This is a non-issue for console or arcade games, of course.

Good point. Without tracking each individual platform, I guess I could also factor in an equation for the current platform generation and available technology in relation to how hard the game tries to push its technological bounds. Lemme think on this, too.

Maes said:

The popularity of multi-platform titles is another matter too: a game could be a hit in the arcades years before it found its way to PCs and consoles, and then live a second "hit" period or just flunk it completely.

Yeah this introduces elements beyond my scope. I'll have to chalk a lot of that up to chance for my purposes, for now.

Maes said:

And there's also the case of the "dusty diamonds" kind of games: they may have all of the replay, gameplay, moddability etc. factors, but remain largely unknown to the public.

I'll call that poor marketing. :P

Maes said:

IMHO you can't really quantify game popularity, any more than you can quantify teen pop idol, fashion top-model, or hollywood star popularity. In those areas, even if you have what it takes, you may still end up being a nobody. While if *somehow* you make it, then it's about marketing and presence skills to keep popular for a while.

In other words, popularity is an "a posteriori" quality you attribute to a game. It can't be predicted if a title will ever be popular, or for how long/under which circumstances. Even "a posteriori" popularity is hard to quantify. You consider sales, or total number of installations? Over which period of time? E.g. footy games enjoy incredible sales peaks when they are new while Myst sold millions of copies of the same exact game over a number of years etc.

Also piracy, game patches, obsolescence, etc.

Yeah, I know not every value can be accounted for in every case. Well, maybe they could, but for the practical purposes of an equation you could use to, say, tell if a game you're making is going to suck or not in advance, this could potentially be useful.

(Sorry triple post)

Added a 'cult' factor and Marketing/public exposure factor. Marketing I think is a huge one that stands to affect virtually everything else. I'm not sure if I should define 'cult' as a part of it or an additional, separate element of its own, though.

Edit 1: Improved 'cult' factor to account for time wear and tear and also the cases of games that occasionally get blown way out of proportion (f^2 wahaha!).

Stop abusing mathematics.

exp(x) said:

Stop abusing mathematics.

In what way are they being abused?

Granted it's been a while since I've formally used math, so I'm writing it out as if I were to eventually stick it into a computer application or something.

I was told there would be no math.

TheeXile said:

In what way are they being abused?

Because all you're doing is adding and multiplying meaningless subjective values to give other meaningless subjective values.

It's impossible to objectively quantify "first impression value", "cult value" or "atmosphere". Your equations are meaningless.

This sort of nonsense is sometimes called "Tesco Value Science". It's pseudoscientific fluff.

fraggle said:

Because all you're doing is adding and multiplying meaningless subjective values to give other meaningless subjective values.

It's impossible to objectively quantify "first impression value", "cult value" or "atmosphere". Your equations are meaningless.

This sort of nonsense is sometimes called "Tesco Value Science". It's pseudoscientific fluff.

It's not about objective definitions. It's assuming the dominant subjective stance of any given value is at least somewhat knowable (Doom has better gameplay than Daikatana, for instance) and then finding a way to plug it into the formula. The goal being then that you can paint a picture of what constitutes a good (or at least a 'popular') game or not based on its elements rather than by its outcome. If that makes sense.

It's a brand of thinking I'm trying to iron out to see if I could make it practical for use at some point. Because I feel like it.

The only way this could be made "scientific" is through gathering actual statistical data relative to sales, community sizes, geographic reach etc. etc. and trying to find a correlation between these factors and games that are "successful" and "less successful".

Once again, it would be an a posteriori evaluation of data, at most you would be able to find an empirical formula that says: "if a game uses what's considered X in latest technology, spends more than Y $$$ in marketing, is targeted at an age group Z to Z' etc. etc..... then it's going to be a success/fail".

But there are professionals who already do that stuff better than you or anyone here will ever be able to. Not surprising, since certain types of risky investments amount to making decisions with uncertain outcomes, of which at most you can estimate their probabilities and costs associated with them happening or not.

If you are serious about studying and quantifying trends, you should strive to become an economics or statistics analyst.

Maes said:

The only way this could be made "scientific" is through gathering actual statistical data relative to sales, community sizes, geographic reach etc. etc. and trying to find a correlation between these factors and games that are "successful" and "less successful".

Once again, it would be an a posteriori evaluation of data, at most you would be able to find an empirical formula that says: "if a game uses what's considered X in latest technology, spends more than Y $$$ in marketing, is targeted at an age group Z to Z' etc. etc..... then it's going to be a success/fail".

But there are professionals who already do that stuff better than you or anyone here will ever be able to. Not surprising, since certain types of risky investments amount to making decisions with uncertain outcomes, of which at most you can estimate their probabilities and costs associated with them happening or not.

If you are serious about studying and quantifying trends, you should strive to become an economics or statistics analyst.

I'm learning about that right now, actually.

TheeXile said:

It's not about objective definitions. It's assuming the dominant subjective stance of any given value is at least somewhat knowable (Doom has better gameplay than Daikatana, for instance) and then finding a way to plug it into the formula.

Without some kind of objective measure I don't see any point. Your equations also seem to be completely arbitrary. For example, this is your "equation" for "first impression value":

f = g + (i * 2a)

I'm going to assume here that you just made this up and there's no experimental basis for it at all. Assuming that you can measure "technological innovation" and "atmosphere", what does it even mean to multiply those values together? Why are they then multiplied by two? why not three? 1.5?

Perhaps most absurd of all is the "cult value" equation, where the entire thing is multiplied by a random value between 0 and 1, essentially rendering the entire thing completely random and the measure irrelevant. You might as well just say:

c = (random)

You could conceivably do some kind of statistical analysis of peoples' subjective opinions of games, and then identify relationships between those values. For example, you could ask people to quantify the "replay value" and "cult value" for different games and see if there is any statistical relationship between the two. That would be proper mathematics. You can't work out relationships by just devising arbitrary equations without any real analysis, whether of objective measurements or subjective opinions.

This reminds me of Drake's equation. A long string of factors, most of which are impossible to estimate precisely, yielding a result that is in the range from "one" to "infinity".

fraggle said:

Without some kind of objective measure I don't see any point. Your equations also seem to be completely arbitrary. For example, this is your "equation" for "first impression value":

f = g + (i * 2a)

I'm going to assume here that you just made this up and there's no experimental basis for it at all. Assuming that you can measure "technological innovation" and "atmosphere", what does it even mean to multiply those values together? Why are they then multiplied by two? why not three? 1.5?

Well sure, that's where I'd go if I were intent on making this into a formula that could actually measure something like this. That is, homing in on what factors to use and how to define them. I'm just doing it here because it's a brain exercise that I like to think could be useful at some point.

For instance, I added 2a because I reasoned that atmosphere is worth more towards a game's first impression than its technology (though I can't know yet if it's worth twice as much, specifically). I did that by asking the question: Would you say it was Doom's atmosphere (music, sound, graphics, etc.) alone or the knowledge you were controlling a character in a 3d-ish environment that was the bigger part of its first impression when you first played it? I just made a subjective guess upon what sounded right. The assumption being that if I were to zoom in on that guess to discern specifics (which, as Maes said, would enter into reasoning about statistical information and whatnot; the actual science you're referring to), the correlation might emerge roughly in line to what I'm guessing.


It's not really science at all in this form, but it is the kind of critical thinking that's behind solving a problem like this. At least, I would imagine so.

TheeXile said:

I'm just doing it here because it's a brain exercise that I like to think could be useful at some point.

Duh...then you should post this in Blogs or study privately and write a journal/conference paper on the argument.

Maes said:

Duh...then you should post this in Blogs or study privately and write a journal/conference paper on the argument.

Sure, but since it's just guessing I figured it could be something others could join in wasting time having fun with in general. :P

TheeXile said:

For instance, I added 2a because I reasoned that atmosphere is worth more towards a game's first impression than its technology (though I can't know yet if it's worth twice as much, specifically).

TheeXile said:

First impression value[/b]
f: First impression value
g: Gameplay value
i: Technological innovation
a: Atmosphere (graphics, animations, sounds, music, story, etc.)

f = g + (i * 2a)

Um. No offence, but I think you need to go back and learn basic algebra :-)

fraggle said:

Um. No offence, but I think you need to go back and learn basic algebra :-)

Yeah I noticed. Rub it in plz.

I haven't used this stuff in years. I'll clean it up later.

To respond to what you actually said in your post beyond the mathematical mistake :-)

TheeXile said:

I just made a subjective guess upon what sounded right. The assumption being that if I were to zoom in on that guess to discern specifics (which, as Maes said, would enter into reasoning about statistical information and whatnot; the actual science you're referring to), the correlation might emerge roughly in line to what I'm guessing.

It's not really science at all in this form, but it is the kind of critical thinking that's behind solving a problem like this. At least, I would imagine so.

I can see that you are trying to apply what you believe to be critical thinking to the issue, but what you say really is correct: this isn't science at all. I don't believe that if you did a survey that you would see relations anything like what your equations predict, no matter how much you tweaked them. The reason is that they were just made up: they aren't based on any real analysis of evidence.

What you're doing looks "scientific" in that you're identifying variables and drawing up equations, but it misses the substance of real scientific investigation. I recommend reading Feynman's comments on "Cargo Cult" science. You might find it interesting.

It's not science. If it were, I wouldn't waste time posting it on a Doom forum with serious intent for debate.

But it's not arbitrary guesswork either. It's guesswork derived from general, common knowledge about subjective topics translated into likely equations that could predicatively describe the data on the assumption that the subjective information its based on could be provided in specific, concrete form through further effort (and thus rendered into real science).

TheeXile said:

But it's not arbitrary guesswork either. It's guesswork derived from general, common knowledge about subjective topics translated into likely equations that could predicatively describe the data on the assumption that the subjective information its based on could be provided in specific, concrete form through further effort (and thus rendered into real science).

Unless you examine real world evidence, it makes no difference. While you may think that your reasoning "makes sense", things are often far more complicated than you would realise. It's naive to think that you can just guess something like this and that it will be accurate.

This is especially the case in this example, where you're measuring subjective human opinions. People are unpredictable and irrational, so the variables will be linked in unexpected and complicated ways that you would never anticipate.

No matter how much thought you put into it, or how carefully you construct your equations, any model derived in this way (ie. guesswork) is doomed to failure. They are not based on reality, only your preconceptions and assumptions.

Danarchy said:

I was told there would be no math.

I'm hoping this isn't followed by a test.

TheeXile said:

It's not science. If it were, I wouldn't waste time posting it on a Doom forum with serious intent for debate.

But it's not arbitrary guesswork either. It's guesswork derived from general, common knowledge about subjective topics translated into likely equations that could predicatively describe the data on the assumption that the subjective information its based on could be provided in specific, concrete form through further effort (and thus rendered into real science).

Lol. Stream of consciousness mathematics?