Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[catpaw]

Ok. In other news, I'm gonna start up a colloquial usage of "proton" that actually means "electron" and vice versa. Physicists can just shove it.

Electrical engineering does exactly this since ever. Due to convention colloquially assuming electric current to consist of positive charges flowing from + to -. 

[Fen]

In my experience, the term 'learning curve' in relation to games isn't drawn out in terms of how much you learned over how much time, but instead how much you need to learn, and when. In the Angry Birds/ Dwarf Fortress example, this is very simple:

In angry birds, you are introduced to concepts one at a time, and thus only need to (in fact, only can, in normal play) learn one of these concepts at a time. Thus, you only need to know as much as you've already been told for each stage, and only need to learn anything new when it's been introduced in the game. The curve for this game starts low and stays low, going up very slightly as each individual element is brought in.

In dwarf fortress, however, you need to know a very large number of things, all at the very beginning. For starters, you need to learn to navigate the menus, instruct a dwarf to begin digging out your fortress, make a room, and designate a meeting area- all so your dwarves won't be out in the open. Then you need to make sure they'll have access to food and drink so they don't starve or dehydrate, and the pile of things you need to do to keep your dwarves alive only grows higher with each step. The curve for Dwarf Fortress raises drastically, immediately, and irrevocably.

Ultimately, it's not an actual 'learning curve' but the phrase has been used for this method of looking at games for so long now that it doesn't really matter that it's not actually correct; more people will recognize this as being what you mean when you say 'learning curve' in regards to games than if you described it another way. And because people think of this as the 'learning curve', they think of games like Dwarf Fortress having a really steep one- because it does, even if that's not the terminology we should have used for it.

[Orange Wizard]

I agreee with what Fen just said. "Learning curve" doesn't refer to a graph, it's just a silly idiom. It is technically wrong, like the electrical engineers thing, but people don't change it because it would take too much effort to convince everyone that what they thought was wrong and the opposite is true.

[catpaw]

It is not just games. For example I remember hearing the term "learning curve" used that way also for english as foreign language. "It is shallow in the beginning and steep later on" As in easy to get into since english does not have gendered articles like most other languages but difficult to master as for example the illogical use of prepositions. That would btw. a life filling task for GavJ, spending your time at the airport and everytime somebody says, he or she is getting *on* the plane, to correct them that they most likely actually want to get *in* the plane, since it is going to be kind of windy on it. When there are no guests around, tell the airport, that it is not "boarding" the plane, as since centuries there is no board involved, but people are using a ramp, so its "ramping" the plane.

In fact, I don't even know what the proper term would be. "Shallow learning curve" wouldn't be correct, since you are not learning more when playing Angry Birds, you just need to learn less. "teaching curve" also wouldn't be correct, as DF is quite bad at teaching you what you need to learn - you'll mostly need an external tutorial to come by. "stuff you need to learn to keep enjoying the game" would getting closer, but maybe some dude enjoys watching the dwarves die outside without doing anything?

Language is fuzzy not just because people are idiots, it is also fuzzy because our world is fuzzy.

[Matt_S]

As a physicist, comparing learning curves to potential energies is pretty natural for me.  So steep learning curves are harder.  Until I switch fields and publish psychology papers, I don't get why anyone would care that I'm using the intuitive, colloqual definition.

Ok. In other news, I'm gonna start up a colloquial usage of "proton" that actually means "electron" and vice versa. Physicists can just shove it.

Ignoring the fact (as already mentioned by catpaw) that the sign of the electrical charge is defined in a totally bonkers way originating from before people actually knew anything about what electrical charge was, I'm not sure why you think that would bother me.

[GavJ]

Until I switch fields and publish psychology papers, I don't get why anyone would care that I'm using the intuitive, colloqual definition.

Honestly, I don't really care if you use it in colloquial parlance in discussions. And I guess I agree that in that sense, there is no clear right answer, as it is established enough to be a valid thing to say outside the context of the original psychology research sort of situation.

What I DO quite care about (as I obliquely but not entirely suggested in the OP) is when you draw out an actual graph, and then plot things in the way that seems intuitive based on the idiom, but is glaringly, objectively wrong given the actual axis labels you used. Which has been the case for EVERY single attempt I've seen in comics or similar that have tried to do this. Not a one of them even approaches internal consistency.

That stops being a fuzzy matter of opinion or language quirks, and instead just becomes a clear failure to properly know how to graph things, and counterproductive for general public science skills and data interpretation, and just makes you look like a moron.  When you'd raw axis labels, you need to actually properly stick to them.  And even if people here might arguably be able to come up with some hypothetical axis labels that would yield a steep curve for DF, none of the ones that might work that have been suggested are ones that authors of those comics actually used.

[catpaw]

x-axis: progress within the game
y-axis: amount of stuff the game expects you to learn at once to reward you with a feeling of accomplishment.

DF is in that graph very steep, however you call it, in the beginning as you need to know a lot of stuff to have your fort surving somewhat, while there are a few things you can ignore at first and master later, like aquivers, minecarts or lava pistons.

The "learning curve" would be much more shallow if there would be a campaign with mostly already running forts which ask you do accomplish the one or other thing.

[Miuramir]

Part of the difficulty is that I think that when most people say "steep learning curve", they aren't actually talking about a learning curve, but an amount needed to be learned to get anywhere curve.  To phrase differently, they are not talking about X = time, Y = amount learned, but something closer to X = progress in game, Y = amount you need to learn to increment to the next X. 

In particular, most "modern" games try to avoid having you need to know anything before you start; the expectation is that you will be drip-fed things in artificially-simplified scenarios.  Graph "Game Progress" on X, and "New Things You Need To Know To Progress" on Y; the usual goal is that the line starts pretty close to 0, and climbs gradually (a shallow slope).  An example might be a flight combat game where the first tutorial starts you out already in the air, with engine, flaps, radar, etc. already configured in sensible ways for where you are, and pointed in the general direction of some sort of simple drone target that doesn't shoot back; and the scenario ends with a win before you have to land. 

DF would be more like one of the old Microprose simulators, where you're expected to read several dozen manual pages and install the keyboard overlay before being able to even get your plane off the deck into the air, and there are several difficult-to-digest walls of "you need to know new things to get further or Fun Happens" in your immediate future.  E.g. you've managed to survive a season or two sort of OK, then winter comes, you have no unfrozen water and haven't figured out booze yet, and bad things can cross the ice to your fortress.  Compare to managing to just about survive your first air combat encounter, and now your first carrier landing is with a damaged plane...

Graphed as above, "New Things..." per unit game progress increases much more rapidly in this sort of old-school gaming, a "steep curve".  Modern games try very hard not to do that to their players, because they're likely to rage-quit and go away, and that's a problem because modern games tend to need you to keep playing to make money (subscription, DLC, expansions, in-game purchases, ad revenue, whatever). 

This is a potentially important point I'd not really thought through before.  When you paid your $50+ or so in 1994 ($80+ in today's dollars) for F-14 Fleet Defender, you had the install disks, the 123 page manual, the keyboard overlay, etc. and Microprose had your money.  End of transaction; if you throw up your hands with the manual in your lap somewhere in the process of trying to figure out how to best switch to the RIO "back seat" cockpit, adjust the AWG-9 radar from PDSTT (Pulse Doppler Single-Target-Track) mode to a TWS (Track While Scan) mode and figure out how many bars and degrees you needed to reliably lock up a new target while still keeping a tactically useful scan rate and, you know, not crash the plane or get blown out of the sky in the meantime... at some level Microprose didn't care.  They had your money, you couldn't get it back, it was up to you to learn the game or not. 

The vast majority of modern games, in contrast, depend on a continued relationship between the player and the developer; whether it's a trial version to induce you to buy the game, or one of the many ways games are cheaper up front but extract money from you over the long term.  If you walk away, the developers, studio, distributors, etc. loose some or all of their profit; it's so vitally important to all of them that you don't give up, that there is tremendous pressure to make the game easy to get into, potentially with considerable opportunity costs elsewhere that could have made a better game for experienced players. 

This is why fans of more complex games have needed to turn to alternate funding models like Kickstarter, "Early Access", and direct patronage of developers.  The old "up front" distribution model admittedly allowed for some real disasters to get out, but it also permitted some rare gems that the current "pay as you go" model makes difficult or impossible. 

P.S.: I don't want to sound like I'm dissing F-14; on the contrary, it's one of the pinnacles of PC simulation gaming, in many ways not surpassed despite the passing of two decades.  But in Authentic mode in full Campaign, it's as close to flying a real plane as you're likely to experience outside of a multi-million dollar simulator (or a couple of years in the military), for a plane that had a reputation as one of the most complex electronic suites flown in something ever expected to get into a dogfight.  It's arguably the commercially released game that goes the furthest in the direction of simulation realism at the possible cost of being too far over the head of most potential players.  In their own words, p.21: "If you're a pilot with a lot of flight sim time under your belt, this chapter is designed to allow you to begin play almost immediately. On the other hand, if you're new to flight simulations, bear with it. ... FLEET DEFENDER, like the F-14 it portrays, isn't always user-friendly (especially the portions dealing with radar).  ... Just remember, the Navy spends years teaching its naval aviators the same information that FLEET DEFENDER tries to teach in hours (and days)."

[Kappas]

Yeah, I also think the "steep learning curve" is quite precise expression: New players are initially faced with a tall cliff to climb, which in turn is followed by more gentle ascend.

As said, this would of course depend on the Y-axis. Original post commented on how much road is traveled, while more useful variable would be how much RPM you need at any point of the hill. Some would-be-players have weaker engines than others.

[GavJ]

Part of the difficulty is that I think that when most people say "steep learning curve", they aren't actually talking about a learning curve, but an amount needed to be learned to get anywhere curve.  To phrase differently, they are not talking about X = time, Y = amount learned, but something closer to X = progress in game, Y = amount you need to learn to increment to the next X. 

I'm pretty sure most people aren't thinking of either of the above things, but are just visualizing a steep hill to metaphorically climb.

However, the axis you describe here does sound like it would work. I know it's similar to one or two others mentioned earlier, but you went into good un-ambiguating detail, and that sounds reasonable. Probably not what anybody is actually visualizing... but reasonable.

At least if I/somebody wants to make a new comic in the future, this would be a good route to go. Thanks.




Although still there's a BIT of a hiccup in that for it to be steep, you'd have to be able to survive awhile without knowing anything, in order for the graph to start low rather than just being at the top initially. I guess you can argue that your dwarves won't literally starve to death for a season, but... yeah. Depends how you think about it -- you need to do a ton of stuff IMMEDIATELY to get a healthy long term fort (thus not much of a "curve," just starts high) but on the other hand, you don't need to do a ton of stuff immediately to literally survive a few months?  Seems like massaging the data a bit to go with the latter interpretation (merely not losing officially with a "you lose" screen), but I suppose as long as you specify clearly what you mean, it works.

[Putnam]

This topic runs entirely on the assumption that the Y value on the "learning curve" graph is what you think it means and not what it actually means.

The learning curve refers to how much you have to learn, not how much you do learn. That simple.

[GavJ]

This topic runs entirely on the assumption that the Y value on the "learning curve" graph is what you think it means and not what it actually means.

The learning curve refers to how much you have to learn, not how much you do learn. That simple.

1) It's not an "assumption." It's the one and only technical, formal definition of the term. It's a measurement tool invented by, coined by, still used by psychological researchers, and they measure it as that: amount learned.

I understand if you want to argue colloquial metaphorical usage, fine.
I understand if you want to argue "here's an axis that COULD theoretically work," fine.
But saying that somehow that's "the" right labeling, as if it has any sort of actual technical pedigree, is simply incorrect. The formal axis is "how much you heave actually learned." So your two choices are:
* Use the formal definition, which establishes DF as a shallow learning curve, OR
* Argue for informal usages, metaphors, hypothetical, etc.

2) The way you just stated it, it wouldn't lead to a steep curve anyway! "Amount you have to learn" would simply be this, the graph I posted earlier, which is ALSO shallow and doesn't change anything:
(At time zero, the amount you have to learn is... everything, thus it's at the top. And there's a lot mroe to learn than in angry brids so the intercept is higher. And then it's harder to learn every new bit, because of worse UI and worse tutorials, so the amount left remains higher longer, i.e. shallow curve)


The only way you could possibly make it work would be if you start adding in half a dozen caveats and stipulations like "amount you have to learn, but only as applies to the next timestep, and only if you're talking about physical basic survival of dwarves not a well oiled fortress, and only if you are including moving average smoothing of the curve, and.... blah blah"

Which is basically what the guy above was getting at. But although you CAN make it work that way, that's clearly not what any normal Joe using the term is/was actually thinking of when they use it. NOR is it the formal definition. So that's all still a pretty academic thought experiment to make it actually work out steep.

[Putnam]

Ah, so I just looked it up, and yes, you're right! Your pedantry surpasses mine immensely, and clearly my reliance on colloquial language is errant. O, please correct our problematic ways of communication, for clearly to be understood is not nearly as important as to be correct.

[GavJ]

clearly my reliance on colloquial language is errant.

I understand if you want to argue colloquial metaphorical usage, fine.

[Putnam]

Then why post this topic at all if all you're going to do is argue semantics over a colloquial term?