Jump to content

How Pgi Saved Cw


54 replies to this topic

#4 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 7,500 posts
  • LocationParanoiaville

Posted 16 May 2015 - 12:41 PM

This again? All you're suggesting is that people be punished for playing 'mechs other people like or find effective - and heaven help them if they don't have enough of the "right" 'mechs to avoid the penalty.

This doesn't even prevent there being a meta - all it does is spread the meta out over multiple chassis, because weapons drive the meta more than 'mechs do. Sure, hitboxes and hardpoint locations matter - and sometimes are the deciding factor that put a chassis on top (consider the Stalker.) But if I recall correctly, when the Highlander was initially nerfed, players simply shifted over to the Victor; the "poptart" meta didn't go away until the core system that enabled it was nerfed - jump jets.

The only thing this is likely to accomplish - aside from annoying everyone and confusing new players even more - is to shift the meta to collections of chassis which share similar characteristics and hardpoints. Thus, the meta will still exist; certain chassis will still be played less than others - and the entry fee for getting into the meta will skyrocket. See, the end result of what you're proposing is an inflation of the cost to "buy in" to whatever the meta turns out to be. That won't be much of a problem for very long-term and/or high-volume players, but it will quickly become burdensome to new players and those wanting a less time-intensive commitment.

In the end, you're "saving" Community Warfare by making the meta harder for the average player to reach - while doing little to "solve" the meta issue.

Edited by Void Angel, 16 May 2015 - 12:41 PM.


#5 Adamski

    Member

  • PipPipPipPipPipPipPipPip
  • Mercenary
  • Mercenary
  • 1,071 posts
  • LocationCanada

Posted 16 May 2015 - 12:44 PM

Everyone wants to drive viable mechs, especially when you can only take 1 at a time.

While people occasionally take a perverse joy in seeing just how much they can do with nonviable / suboptimal decks, its not fair to force everyone to drive them because the developer doesnt want to balance them.

We are already in a fairly good state, with the STK-4N == WHK, and SCR <= TDR <= TBR

If PGI could stop listening to the whiners they could continue along that path so that more mechs become viable.

#6 Koniving

    Welcoming Committee

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • The Guide
  • The Guide
  • 23,384 posts

Posted 16 May 2015 - 12:56 PM

It's a start... but the thing is PGI needs to actually work on balancing things without over buffing things. Extreme percentages are ******. "Quirked" stuff is also screwed. Between a dynamic skill tree per mech variant without super-enhancing or even enhancing weapon types and carefully monitoring real issues and addressing them directly rather than bandaging, we'd get something good situated in here.

Balancing by market value is only a first step as you implied by Pretty Baby's tweet. It isn't quite how Russ can save CW, but definitely would make me want to play more. The issue is having the value of the mech change dynamically would screw with how people arrange their dropdecks. What if while waiting, your 'market value' of the mechs you have chosen has changed and now you're over value, 3 seconds before launch and therefore can't launch?

Dynamic R&R by availability of parts (with them being unique to each chassis/variant) would be a similar way of doing things; the more 'meta' it is the harder it would be to upkeep and at some point it'll be too expensive to do. Less issue with suddenly being over market before dropping. Still it has its own issues.

You're on to something... but while it's a good idea for diversity, it is merely a bandaid for the underlying problems with the game.

Edited by Koniving, 16 May 2015 - 01:00 PM.


#7 Dolph Hoskins

    Member

  • PipPipPipPipPipPip
  • The Territorial
  • The Territorial
  • 499 posts
  • LocationThe Machine

Posted 16 May 2015 - 01:05 PM

I liked the story for it's dramatic points and memorable characters. Very well written :)

#8 Freebrewer Bmore

    Member

  • PipPipPip
  • The Patron Saint
  • The Patron Saint
  • 64 posts
  • LocationBaltimore, MD, USA

Posted 16 May 2015 - 01:41 PM

View PostVoid Angel, on 16 May 2015 - 12:41 PM, said:

All you're suggesting is that people be punished for playing 'mechs other people like or find effective - and heaven help them if they don't have enough of the "right" 'mechs to avoid the penalty.

Once prices adjust, there won't even be any 'wrong' mech. Everything will cost what it's worth. Nobody's STK4N (or TDR9S, rust in peace) will become unplayable, because if it does, its price goes down until it's playable again. Nobody gets punished.


View PostVoid Angel, on 16 May 2015 - 12:41 PM, said:

the "poptart" meta didn't go away until the core system that enabled it was nerfed - jump jets.

Right, and so the effort to balance those core systems is a better place to apply developer resources than trying (and always failing) to balance every chassis against every other. Let the dynamic system manage the chassis balance so you can use developer intervention where it's really necessary.


View PostVoid Angel, on 16 May 2015 - 12:41 PM, said:

the end result of what you're proposing is an inflation of the cost to "buy in" to whatever the meta turns out to be.

This system doesn't change C-Bill mech ownership costs at all. So the issue you're raising here is only a concern if dropdeck costs were wildly fluctuating, requiring people to keep buying new mechs every time the meta wind blows a different way. However, that volatility is eliminated by a sufficiently large rolling window for the usage stats. With stable prices, 4 mechs under this system costs no more than 4 mechs now.

Again, I'd contrast this with the way things happen now, which is far worse: PGI decides to make a balance change and suddenly your superquirked mech becomes unplayable, or your whole 250-ton dropdeck now needs reworking to fit 240, etc. Now that gets expensive, and the dynamic system wouldn't cause that suffering.

Have I addressed your concerns? You're making good points, but I don't think they're reasons to fear.

#9 Freebrewer Bmore

    Member

  • PipPipPip
  • The Patron Saint
  • The Patron Saint
  • 64 posts
  • LocationBaltimore, MD, USA

Posted 16 May 2015 - 02:15 PM

View PostKoniving, on 16 May 2015 - 12:56 PM, said:

The issue is having the value of the mech change dynamically would screw with how people arrange their dropdecks. What if while waiting, your 'market value' of the mechs you have chosen has changed and now you're over value, 3 seconds before launch and therefore can't launch?

Yeah, as I was saying above, you'll need a sufficiently long rolling window for the usage stats so they don't fluctuate wildly. I'd personally be inclined to initally try a 1-month usage window with prices updating at a fixed time once per day (e.g. at the same time your daily double XP is updated on your mechs). Since only 1/30th of the total body of usage data is being updated, the costs should move slowly day by day (if it doesn't, OK, use a longer window; PGI has access to the data needed to set it at a good length).

If your dropdeck had been set up at the very edge of what was allowable, maybe that one day's new data would push it over the edge, yeah, but it would happen at a set time of day. You also could (and should) insulate yourself from that by building your deck with a little cost wiggle room, and/or using mechs that are clearly on a downward market cost trajectory, and/or simply having one cheaper mech to slot in as a price substitute when you get pushed over the edge.


View PostAdamski, on 16 May 2015 - 12:44 PM, said:

Everyone wants to drive viable mechs, especially when you can only take 1 at a time.

In the CW context we're talking about taking 4 at a time, not 1 (tho Roland's thread explores possibilities in group queues, even he never proposes using this idea for the solo queue). As mentioned in the story, you'll always be able to take any 1 mech, it's just that you might need to make some sacrifices in the rest of your dropdeck if you're trying to bring OP stuff.

And here's the thing: the weaker mechs will become less weak (and more fun) simply because they won't just be surrounded by only the strongest few variants anymore. Instead of just getting their butts handed to them by OP stuff, they'll be able to hold their own against a more diverse field of mechs.

#10 Void Angel

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 7,500 posts
  • LocationParanoiaville

Posted 16 May 2015 - 02:49 PM

Magical thinking about market economics, or hand-waving away problems like buy-in costs to get to the meta with an unsupported (and unsupportable) assumption that the meta will never shift, isn't going to cut the mustard. For example, re-setting my drop deck to 240 was trivial, and while it's not a top-performing chassis any longer, the Thunderbolt 9S (for example) is still viable. But even ceding that claim for the sake of argument, you're still looking at the same thing, just on a grander scale, and without human oversight. Instead of PGI looking at 'mechs and possibly overshooting a balance point, you're going to have whole ranges of 'mechs nerfed automatically - based not on how good they actually are, but on how much people use them. It's game balance by democracy, and if you think that phrase is a positive descriptor, you need to take more political science.

Your proposed system contains the same kind of overgeneralizations and possibly (in this case certainly) faulty assumptions that are often used in publicly-discussed economics. You assume that the great mass of people operate according to enlightened self-interest, for example. This is emphatically not the case; look at all the people who use missile boats, or refuse to move from their favorite camping spot in order to counter the enemy's movements. I'm not sure I'd like your system anyway, even if it could work. But since it has insurmountable obstacles between paper and the real world, that question is truly moot.

PS: Use of slurs like "tryhard" reduces your credibility out of the gate. See also: Ethos.

#11 Brizna

    Member

  • PipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 1,367 posts
  • LocationCatalonia

Posted 16 May 2015 - 02:56 PM

While I give the OP a big BROFIST for a great opening post. A system like that would filter mechs by good and bad but also by fun and boring, effectively pushing people to play a boring game. People play mechs because they are good but also because they are fun and conversely don't play them because they are unexciting.

#12 Skarlock

    Member

  • PipPipPipPipPipPip
  • The Defiant
  • The Defiant
  • 328 posts

Posted 16 May 2015 - 03:05 PM

View PostVoid Angel, on 16 May 2015 - 02:49 PM, said:

It's game balance by democracy, and if you think that phrase is a positive descriptor, you need to take more political science.


This, exactly. Peoples subjective opinions are the worst way to achieve balance. Balance is not a popularity contest, and most of the players in this game don't even base their decision making on empirical data. Because the players balance the game by choosing their mechs, they are the ones really in control of it.

Also, this system would only work in an environment where the playerbase as a whole is pretty educated about what is actually good. Otherwise, the majority of players will simply take bad mechs and enable the meta players allowing them to take the top tier mechs. I'd love for PGI to actually try something like this though, simply for the schadenfreude of watching it fail spectacularly on the IS side by enabling good players and units to insane levels, and more or less destroying the viability and competitive ability of the clan decks in one fell swoop. Why would the clans be destroyed? Simple. The bad clan mechs are *really* bad, whereas IS has a ton more variety and viable mechs to choose from. If the clans can only win an event with the majority of the high level comp teams going to the clan side, and IS having a 10 ton advantage, and clanners spamming timber wolves and storm crows all day every day, and removing the IS ability to just rush generators with light mechs by forcing them on defense, by SEVEN PERCENT, they'll get absolutely, hilariously demolished after implementing this suggestion.

Not to mention that in order to compete, as people shift their mechs around, you will need an extraordinary amount of mechs to switch to because the system is now dynamic. So your perfectly legal drop deck one week becomes overpowered or underpowered and you need to optimize it. Thus you can only do well at CW if you have a veritable ton of mechs to choose from to take the optimal deck, which of course would favor the "tryhard" teams who actually earn the c-bills to do so...

Like I said, I really want PGI to try something like this, I really do. As a learning experience...

Edited by Skarlock, 16 May 2015 - 03:08 PM.


#13 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,687 posts

Posted 16 May 2015 - 06:41 PM

View PostThe Ripper13, on 16 May 2015 - 01:05 PM, said:

I liked the story for it's dramatic points and memorable characters. Very well written :)


too bad it will be forever relegated to the fantasy section of the library.

#14 sycocys

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Moderate Giver
  • Moderate Giver
  • 7,698 posts

Posted 16 May 2015 - 08:01 PM

Like the way the post was written, don't care for the idea he is suggesting.

Also "easy to implement".... if you've been tracking PGI you'd know that nothing is easy for them to implement and complex solutions are little more than dreams.

#15 Repasy Cooper

    Member

  • PipPipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 1,131 posts
  • LocationAlpheratz

Posted 16 May 2015 - 08:32 PM

Meta is dull. So many people are limiting themselves by obsessing over the "OP" builds, and running the same mech over and over and over and over and over and over and over and over......

Before they know it, the game ceases to be fun. It becomes routine grinding to earn more C-bills to buy more mechs to turn into the same mech to grind with some more.

I beseech you all to add more flavour into the game. Don't wait for some super-amazing update that fixes everything. JUST DO IT. Don't limit yourselves to some cookie-cutter loadout that some deadbeat basement squatter dreamed up with all their unemployed-free time on their hands. Dream up your own build! If somebody criticizes your build for not taking full advantage of a mech's capability well SCREW THEM! because in the end it is not the mech's capability that matters at all. It is your own.

I guarantee 100% that any player who does this will increase their piloting skill. Because when it comes down to the hard truth, meta is a crutch that you really don't need if you're a good player.

And you'll have more fun too.

Edited by Repasy, 16 May 2015 - 08:32 PM.


#16 Skarlock

    Member

  • PipPipPipPipPipPip
  • The Defiant
  • The Defiant
  • 328 posts

Posted 16 May 2015 - 09:33 PM

View PostRepasy, on 16 May 2015 - 08:32 PM, said:

Meta is dull. So many people are limiting themselves by obsessing over the "OP" builds, and running the same mech over and over and over and over and over and over and over and over......

Before they know it, the game ceases to be fun. It becomes routine grinding to earn more C-bills to buy more mechs to turn into the same mech to grind with some more.

I beseech you all to add more flavour into the game. Don't wait for some super-amazing update that fixes everything. JUST DO IT. Don't limit yourselves to some cookie-cutter loadout that some deadbeat basement squatter dreamed up with all their unemployed-free time on their hands. Dream up your own build! If somebody criticizes your build for not taking full advantage of a mech's capability well SCREW THEM! because in the end it is not the mech's capability that matters at all. It is your own.

I guarantee 100% that any player who does this will increase their piloting skill. Because when it comes down to the hard truth, meta is a crutch that you really don't need if you're a good player.

And you'll have more fun too.


Meta builds are used when you want to win. For example, the MLMW tournament. Lots of people have fun with under powered, silly, or quirky builds, but they ALSO have fun competing at a high level against other skilled opponents, using what they consider the best possible drop decks in these competitions. Finding ways to break or refine the meta is now the challenge, and your skills at this are pitted against other players. If the meta is dull, it's because it has become static. Nothing changes, everyone uses the same strategies, same mechs, and same weapon loadouts because they have been *proven* through exhaustive testing and competition to be the best. As the game grows more stagnant, so does the meta, because there is less to discover as the players discover over time everything there is to discover. Hopefully the game makers add new things to keep the game fresh and interesting, and then a new meta will be born.

Your dismissiveness of the meta as being the concoction of a single "basement dwelling unemployed" person though is rather ridiculous and insulting. The meta was established not by a single person with a web site, but by the entire collective of the top players of this game. It's also wrong to assume some people just want to play to "have fun". For some people, competition *is* the drive to play, and trying to win *is* the fun part. Telling someone what they find fun and not fun is absurd, as is ignoring other peoples input on your builds.

So in closing, play whatever you want to play. But don't tell other people how they are supposed to enjoy the game.

Edited by Skarlock, 16 May 2015 - 09:36 PM.


#17 Crockdaddy

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,684 posts
  • Twitch: Link
  • LocationSaint Louis

Posted 17 May 2015 - 12:25 AM

View PostAdamski, on 16 May 2015 - 11:24 AM, said:

And lo I drop in nothing but the most rare hero mechs that few players have or would spend the money on, and so are critically undervalued.

No, seriously, your idea is dumb.


Actually in fact it is rather brilliant. Why not use statistics to generate your relative BV within the game. Whom but the community knows what works vs what doesn't? One of the community driven leagues Proxis attempted to take this into account. it was fascinating as obviously everyone taking the very best of something generated scarcity of a thing. Creativity would indeed have to flourish as us tryhards figured out every devious way we could to exploit the system while maximising value.

#18 zortesh

    Member

  • PipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 624 posts

Posted 17 May 2015 - 01:58 AM

View PostCrockdaddy, on 17 May 2015 - 12:25 AM, said:


Actually in fact it is rather brilliant. Why not use statistics to generate your relative BV within the game. Whom but the community knows what works vs what doesn't? One of the community driven leagues Proxis attempted to take this into account. it was fascinating as obviously everyone taking the very best of something generated scarcity of a thing. Creativity would indeed have to flourish as us tryhards figured out every devious way we could to exploit the system while maximising value.



Scarcity of overused things would be hilarious.

For some reason i imagine some poptart trying to equip there mech, telling the tech what they want..... and the tech being all.
"hahahah... autocannons.. jumpjets... ppcs... pfft good luck buddy anything like that has been sold out for months!"
and then a single nerf later and theres piles of used autoocannons and ppc's sitting in budget bins, and a huge black friday style mob fighting a melee over the few remaining lasers in stock.

#19 Smith Gibson

    Member

  • PipPipPipPipPipPip
  • 214 posts

Posted 17 May 2015 - 05:03 AM

The only change I would make to the idea is that the formula become:

[Initial Battle Value] x [Usage Percentage] = [Modified Battle Value]

You will need to start with a Battle Value based on Weight, Weapons, Armor, Engine, etc. And then modify it with usage percentage. That way someone that plays a "less than optimal" build of the most optimal 'Mech will have a smaller modified battle value. Since Hitboxes and Hardpoint placement is not something we can actually calculate to any usable degree like we can Weapons, Armor, and Engine, usage percentage will have to be used to fill that role.

The end result would be:

Initial Battle Value Modifying Variables:
Chassis Weight
Armor
Weapons
Ammo
Equipment
Engine
Endo Steel
Ferro Fibrous
Artemis

Usage Percentage Variables:
Quirks
Hitboxes
Hardpoint Placement
Available Missile Tubes
Hill Climbing

We start with a Modified Battle Value equal to the Initial Battle Value, then after the usage percentage information is collected the Modified battle value either goes up because it has a superior combination of Quirks, Hitboxes, and Hardpoints, or the Modified Battle Value goes down because it has arm weapons placed lower than it's knees or something equally ineffective.

There are quite a few 'Mechs that need to have the Quirk of "-10 tons for CW purposes" and this will effectively give us that but with the added advantage of the numbers being generated by actual popularity instead of being semi-arbitrary.

Edited by Smith Gibson, 17 May 2015 - 05:17 AM.


#20 Tywren

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • 296 posts

Posted 17 May 2015 - 07:44 AM

I still like the idea of repair time between CW drops myself.

Every point of armor damage taken = 30 seconds repair.
Every point of internal damamge taken = 1 min of repair.
Every weapon would have it's own time base (possably with modifiers if it's non-stock equipment for the mech in question), and things like ECM, and TCs would have a 30 min repair time.

Sure there are those who will just buy more of the meta mechs, and rotate them out between drops, but those people have to consider that they'll be doubling down on the current meta, and may be left behind once that meta shifts. For everyone else, it becomes a question of "do i wait for my good mech to come out of repairs, or do i go digging through the mechbays and find something else to drop with in the mean time".

#21 HC Harlequin

    Member

  • PipPipPipPipPipPipPip
  • Mercenary
  • Mercenary
  • 655 posts

Posted 17 May 2015 - 12:19 PM

ya... doesn't work.

you see, battletech is about putting stuff onto a generic chassis.

Back in the day you had a generic battletech battlemech sheet of paper that you filled the holes in and added stuff to the hit locations. There were very general rules. Each hex moved was a factor of how much tonnage over how big engine. Jump jets was a factor of how many jump jets based on walking distance.

The actual figurine you put onto the board had no relation to the battletech battlemech sheet of paper unless you were playing in a tournament that was wysiwyg. Or you were playing a stock mech only tournament.

Sure, MWO has streamlined that a bit with engine limits and weapon hardpoints. But even the quirks aren't a requirement to build the mech.

So for a 35 ton mech, according to your system, I can take a raven, or a jenner, or a firestarter. Spoiled for choice, completely.

65 ton mechs, Thunderbolt, catapult, Jaeger. Spoiled for choice.

The only people that your system would brutally punish would be new players who don't have a large mech collection..

The very people we don't want to ostracize.

And the people that could take an overwhelming advantage out of this would be the guys with the all the mech belong to us stable of mechs.

So a very very bad idea all around

Edited by HC Harlequin, 17 May 2015 - 12:28 PM.


#22 Freebrewer Bmore

    Member

  • PipPipPip
  • The Patron Saint
  • The Patron Saint
  • 64 posts
  • LocationBaltimore, MD, USA

Posted 17 May 2015 - 02:27 PM

View PostVoid Angel, on 16 May 2015 - 02:49 PM, said:

It's game balance by democracy, and if you think that phrase is a positive descriptor, you need to take more political science.

Well, if you don't think that additional perspectives have anything to add, I'm really not sure why you bother saying anything. By that logic we should want Russ to ignore everyone's opinions and just get on with it and finish making his perfect game already, right? He must be omniscient, whereas we all know how dumb we are...

There are all sorts of problems with global economics (distribution of pricing power, ugh) and democratic politics (OMG majority voting mechanisms SMH), but the basic idea of using distributed decisionmaking devices like market pricing to enable a system to gather and act upon more and better information is not one of them. Yes, we are all dumb monkeys, which is precisely why we need to find ways to combine our meager wisdom rather than handing the job to just one dumb monkey. Again, Roland already did a nice job of explaining these conceptual underpinnings in his thread.


View PostVoid Angel, on 16 May 2015 - 02:49 PM, said:

Magical thinking about market economics, or hand-waving away problems like buy-in costs to get to the meta with an unsupported (and unsupportable) assumption that the meta will never shift, isn't going to cut the mustard. For example, re-setting my drop deck to 240 was trivial, and while it's not a top-performing chassis any longer, the Thunderbolt 9S (for example) is still viable.

The thought that you can go into a store and buy something for a relatively stable price that in large part reflects its cost and value is hardly magical. It happens every day.

I wouldn't assert that the meta will never shift. In fact I expect the tryhardiest folks will want to continuously tweak their decks seeking the tiniest advantage, and yes that'll cause price fluctuations. However, once the dynamic system has been fully phased in and the prices have had a chance to stabilize at the approximate value of each variant, the rolling usage data window will prevent those fluctuations from being disruptive to average players. If the competitive players were to devise and propagate some completely revolutionary approach to team composition that nobody had thought of before (and thus was not already priced into the market), yes that would impact average players both by kicking their butt in the short term and altering prices in the long term, but that kind of meta shift would already cause disruption under our current system anyway (inducing everyone to run out and buy that trendy new stuff or lose).

Speaking of current disruptions, I actually thought the TDR9S de-buff was rather well done myself, and neither that nor the 10-ton drop caused me any personal hardship... but I'm not the only one playing this game. I'm not one of those unfortunate souls who bought heroes during the Tuk sale specifically for a 250-ton deck and then suddenly didn't have a place for them. I still use my 9S sometimes, but the fact that there are so many fewer on the field nowadays tells you that they're spending time rusting in their mechbays or have been resold. So there are a couple concrete examples of how the present regime regularly causes the inflated buy-in that you're worried about.

All you can really do is speculate that the dynamic system might also incur inflated costs. Neither of us can prove exactly what will or won't happen. However, I can point to self-correcting features of the dynamic system that are absent from the current system and which you have yet to substantively address (merely being contemptuously dismissive of them as magical thinking and hand-waving isn't going to cut the mustard).


View PostVoid Angel, on 16 May 2015 - 02:49 PM, said:

But even ceding that claim for the sake of argument, you're still looking at the same thing, just on a grander scale, and without human oversight. Instead of PGI looking at 'mechs and possibly overshooting a balance point, you're going to have whole ranges of 'mechs nerfed automatically - based not on how good they actually are, but on how much people use them.

PGI isn't going anywhere, so there would still be human oversight should it be needed. And we're not talking about "nerfing" insofar as this doesn't touch combat effectiveness, it only affects meta effectiveness. And overshooting balance points is not merely "possible" with static balance mechanisms, it's guaranteed. And how much people use them already takes into account how good they are. And again this stuff was already well-addressed by Roland.


View PostVoid Angel, on 16 May 2015 - 02:49 PM, said:

Your proposed system contains the same kind of overgeneralizations and possibly (in this case certainly) faulty assumptions that are often used in publicly-discussed economics. You assume that the great mass of people operate according to enlightened self-interest, for example. This is emphatically not the case; look at all the people who use missile boats, or refuse to move from their favorite camping spot in order to counter the enemy's movements.

No, actually I assume that none of us knows everything, so we need to be able to try things and screw up and learn from each other's boneheaded mistakes (and from each other's brilliant insights that we initially boneheadedly dismissed as boneheaded mistakes). It's not that I like market pricing because I have any delusions of optimal allocations via perfect information or even classically "rational" behavior; I like it because it's a way of collectively compensating for our individual idiocy (maybe you could say I'm from the Black Swan school of anti-economist economics.)


View PostVoid Angel, on 16 May 2015 - 02:49 PM, said:

PS: Use of slurs like "tryhard" reduces your credibility out of the gate.

Insofar as the story is clearly parody, it was my hope that those who read it in good faith would recognize a flagrantly cartoonish phrase like "the evil Tryhard Cartels" as being part of the fun, rather than a straightfaced indictment of anyone's lifestyle choices. In fact, my unit exists for the explicit purpose of making a home to both casual and competitive players. So you might say that some of my best friends are tryhards. :)

Edited by Freebrewer Bmore, 18 May 2015 - 10:48 PM.


#23 Moldur

    Member

  • PipPipPipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 2,241 posts

Posted 17 May 2015 - 03:05 PM

They shut it down and open it again when it works/ is fun/ is worthplaying





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users