Crit­i­cal Mass

With con­sumers able to ex­ert more pres­sure on de­vel­op­ers than ever, how is the game in­dus­try re­spond­ing? And is the cus­tomer al­ways right?

EDGE - - HYPE - BY SI­MON PARKIN

86

On March 12, 2012, a dis­grun­tled fan of the

Mass Ef­fect se­ries posted a mes­sage on the of­fi­cial BioWare fo­rum out­lin­ing the rea­sons why he was frus­trated with the Shep­ard tril­ogy’s end­ing. The post is ad­dressed to

Mass Ef­fect 3’ s de­vel­op­ers and writ­ers and out­lines, un­der a se­ries of bolded head­ings, a litany of their sup­posed fail­ings. Some of the au­thor’s frus­tra­tions sprung from the short­fall be­tween the prom­ise of mak­ing mean­ing­ful choices across three games and the re­al­ity of how play­ers’ ac­tions af­fected the three end­ings pro­vided. Other griev­ances were based on per­ceived in­con­sis­ten­cies in the science fic­tion. In con­clu­sion, he wrote, “BioWare… this game de­serves a bet­ter end­ing. We know you can do bet­ter than this. Please, do not let us down in this way.”

Later that week, another mem­ber of the same fo­rum posted a mes­sage say­ing that he had taken his com­plaint to a higher power than even the devel­oper: the US Fed­eral Trade Com­mis­sion. He called on any other un­happy fans who be­lieved that BioWare had some­how failed to de­liver a sat­is­fac­tory ex­pe­ri­ence to do the same. Another user posted a fo­rum poll in which you could vote on whether or not the devel­oper should patch the end­ing in or­der to re­place it with a new, “brighter” one. Im­plau­si­bly, al­most 100,000 peo­ple voted in the poll, an over­whelm­ing 91 per cent in favour of the no­tion that BioWare should re­write and re-re­lease the end­ing. Yet another player launched a char­ity drive, ask­ing play­ers to do­nate money to Child’s Play as a way to demon­strate the depth of their dis­con­tent and put a cu­ri­ous kind of emo­tional pres­sure on the com­pany (‘Fix the end­ing or else the chil­dren get… sup­port?’). The fundraiser col­lected more than $80,000 be­fore be­ing shut down by Child’s Play.

All this rab­ble-rous­ing worked. Seven days af­ter that first fo­rum post, Dr Ray Muzyka, co-founder of BioWare and its then-CEO, re­leased a strange state­ment in which he si­mul­ta­ne­ously de­fended the work of Mass

Ef­fect 3’ s writ­ing and de­sign team, while pledg­ing to “ac­cept the crit­i­cism and feed­back with hu­mil­ity”. Muzyka also re­vealed that, as a re­sult of the pres­sure, his staff were now work­ing on sup­ple­men­tary ma­te­rial that would “pro­vide more clar­ity” to play­ers “seek­ing clo­sure” to their Mass Ef­fect jour­ney. In his post, Muzyka ar­tic­u­lated a co­nun­drum that cre­ators of pop­u­lar fic­tion have wres­tled with through the cen­turies: how to square the artis­tic vi­sion of a cre­ator with the de­mands of the au­di­ence who pay for the work. “We’re work­ing hard to main­tain the right bal­ance be­tween the artis­tic in­tegrity of the orig­i­nal story while ad­dress­ing the fan feed­back we’ve re­ceived,” he wrote.

This tus­sle be­tween an artist and their au­di­ence has his­tor­i­cal prece­dent. For in­stance, Charles Dick­ens’ The Old Cu­rios­ity Shop was pub­lished in in­stal­ments in the au­thor’s own weekly pe­ri­od­i­cal, Master Humphrey’s Clock. Within three days of pub­lish­ing Chap­ter 53, in which a lov­able

char­ac­ter, Lit­tle Nell, vis­its an old church and has a con­ver­sa­tion in a grave­yard, Dick­ens had re­ceived sev­eral letters warn­ing the writer to re­frain from what many be­lieved he was plan­ning: mur­der. These read­ers took Nell’s visit to the grave­yard in con­junc­tion with the line that she looked “pale but very happy” as a fore­shad­ow­ing of her death.

There’s an anec­do­tal, pos­si­bly apoc­ryphal, story that in sub­se­quent weeks Amer­i­can read­ers stormed New York City’s piers, de­mand­ing to know from visi­tors from Eng­land whether or not Nell dies. And Dick­ens was later swarmed with letters ex­press­ing anger and heart­break over her even­tual fate. Sim­i­larly, af­ter find­ing out that the char­ac­ter Aeris dies mid­way through 1997’s Fi­nal Fan­tasy VII, fans sent letters and emails to the game’s devel­oper, Square, ei­ther be­rat­ing the team’s de­ci­sion or de­mand­ing to know how it might be un­done.

While nei­ther Dick­ens nor Square acted upon their au­di­ence’s com­plaints, the gap be­tween videogame cre­ator and videogame con­sumer has closed con­sid­er­ably since

FFVII’s re­lease, mak­ing it far harder for stu­dios to ig­nore such im­pas­sioned pleas. No more does the game-maker who kills off a beloved set of char­ac­ters, or who adds a con­tro­ver­sial de­sign, re­ceive a mere sack full of letters. They are now in­un­dated with pe­ti­tions, fo­rum posts and even guiltin­duc­ing char­ity drives. In the most ex­treme cases, some de­vel­op­ers have re­ceived per­sonal threats and online ha­rass­ment. In 2013, for in­stance, David Von­der­haar, de­sign di­rec­tor of Call Of Duty: Black Ops II was told he should “die in a fire” or “kill him­self” by up­set play­ers af­ter he an­nounced via Twit­ter that the fir­ing rate of one of the game’s most pop­u­lar weapons had been tweaked from 0.2 sec­onds to 0.4 sec­onds. The vol­ume of ha­rass­ment be­came so great that Ac­tivi­sion com­mu­nity man­ager Dan Am­rich wrote a blog post in which he called the at­tack­ers “im­ma­ture, whiny ass­holes”.

“The way in which game de­vel­op­ers in­ter­act with play­ers has been the most sig­nif­i­cant change in the in­dus­try in re­cent years,” says Zac An­tonaci, who is head of com­mu­nity man­age­ment at Fron­tier De­vel­op­ments and acts as a go-be­tween, man­ag­ing the de­mands of Elite: Dan­ger­ous’s player­base and the de­vel­op­ment team. “To­day, ev­ery el­e­ment of game pro­duc­tion has to con­sider and be mind­ful of the play­ers. With all games mov­ing more online and the de­vel­op­ment of the in­dus­try in so many ways be­com­ing a more so­cial ex­pe­ri­ence, the value of a strong re­la­tion­ship be­tween game maker and play­ers is more im­por­tant than ever.”

Things have grown yet more fraught in the crowd­fund­ing era and with the rise of the am­a­teur con­sumer-pub­lisher. Many back­ers of nascent videogames such as Fron­tier’s Elite: Dan­ger­ous (which was funded via Kick­starter) hold the belief that their pa­tron­age should be­stow spe­cial priv­i­leges – as stake­hold­ers, they think that they should be given a say in a game’s di­rec­tion, even at the gran­u­lar level of game and char­ac­ter de­sign. “[Crowd­fund­ing] does change the ex­pec­ta­tions of both the game stu­dio and the play­ers,” says An­tonaci. “De­vel­op­ers un­der­stand that by run­ning a crowd­fund­ing cam­paign, they are invit­ing play­ers even closer to the process, which, as a com­mu­nity man­ager, is fan­tas­ti­cally ex­cit­ing.”

Per­haps, but as another com­mu­nity man­ager (who asked to re­main anony­mous) puts it, the re­la­tion­ship is also com­pli­cated and chal­leng­ing. “Some com­mu­nity mem­bers who fund a game be­come en­ti­tled and think that they know bet­ter than the de­sign­ers,” he says. “If they feel like their de­mands are be­ing ig­nored, things can be­come ugly.”

Garth DeAn­ge­lis, lead pro­ducer on XCOM 2, agrees: “Lis­ten­ing to the com­mu­nity has be­come ab­so­lutely crit­i­cal for game de­sign­ers in re­cent years. But there is a fine line. The team mak­ing the game un­der­stand the ram­i­fi­ca­tions of cer­tain fea­tures and the knock-on ef­fects in a way that play­ers look­ing in from the out­side sim­ply can­not. We can’t act upon ev­ery sin­gle thing that the com­mu­nity asks for. De­sign by com­mit­tee never works.”

Com­mu­ni­cat­ing this sen­si­tively to irate play­ers is a chal­lenge. “The most pas­sion­ate fans are of­ten the most vo­cal,” says DeAn­ge­lis. “These char­ac­ters are usu­ally scream­ing for fea­tures that we have al­ready ex­am­ined very

“Some com­mu­nity mem­bers who fund a game be­come en­ti­tled and think that they know bet­ter than the de­sign­ers”

care­fully in­ter­nally. There are log­i­cal rea­sons why we have cho­sen to not go down cer­tain paths. That can be tricky to com­mu­ni­cate to the com­mu­nity.” How­ever, when the feed­back com­ing from play­ers is unan­i­mous, it can of­fer one of the most use­ful gauges for a devel­oper to use in pri­ori­tis­ing a se­quel’s de­sign or fea­tures. “While work­ing on XCOM: En­emy

Un­known, we had in the back of our minds that it would be cool to have pro­ce­dural lev­els,” DeAn­ge­lis ex­plains. “Af­ter the game launched, we ag­gre­gated all of the re­views, the fo­rum posts and the YouTube com­men­taries. Then we took them apart and cre­ated piles of feed­back so we could see which crit­i­cisms were most widely main­tained. We found that play­ers wanted more va­ri­ety in lev­els. That al­lowed us to pri­ori­tise the pro­ce­dural fea­ture for the se­quel based on hard data. For us, that’s how it tends to work. Player feed­back al­lows us to pri­ori­tise fea­tures on which we were al­ready work­ing.”

Other stu­dios have taken this process of gath­er­ing and im­ple­ment­ing feed­back to an in­dus­trial-scale level. John Hopson heads up the user re­search team at Bungie in Belle­vue, Washington. He leads a team of re­searchers who pro­vide Des­tiny’s de­sign­ers with player feed­back, de­liv­er­ing data that can be used to in­form de­sign de­ci­sions. The stu­dio in­vited play­ers into the stu­dio to be­gin this process long be­fore the game was fin­ished, us­ing a range of tech­niques, from sim­ple fun­da­men­tals to the highly ad­vanced.

“We ob­served par­tic­i­pants di­rectly and made notes,” Hopson ex­plains. “Then we had the par­tic­i­pants fill out sur­veys to let us know what they thought. We also recorded ev­ery play ses­sion and con­nected it to our log­ging data so we could go back and watch key mo­ments. Those video clips be­came one of our most ef­fec­tive com­mu­ni­ca­tions tools with the rest of the stu­dio, be­cause when you can watch a dozen or so videos of play­ers ex­pe­ri­enc­ing the same prob­lem, that’s much more per­sua­sive than any re­port.”

Hopson’s team de­signed a tool that would al­low play­ers to of­fer in­stant feed­back while they were play­ing. “Par­tic­i­pants could press com­bi­na­tions of but­tons on their con­troller to tell us that the game at that mo­ment was ‘awe­some’ or ‘frus­trat­ing’ and so on.” These in­stant feed­back re­ports were then con­nected to the video record­ings, so a de­signer could sit down af­ter a test and watch the ex­act mo­ments where play­ers said they were hav­ing a prob­lem, or even the mo­ments where play­ers said they were hav­ing the most fun.

As well as ac­tive feed­back of­fered by play­ers, Hopson and his team col­lected and ex­am­ined pas­sive feed­back. Des­tiny was the first Bungie game for which the team made ex­ten­sive use of eye-track­ing data, which showed how ef­fec­tively play­ers were able to nav­i­gate the game’s com­pli­cated user in­ter­face. “Know­ing where play­ers were look­ing and what they were read­ing was re­ally use­ful in telling when play­ers were hav­ing trou­ble un­der­stand­ing some­thing, and when they were sim­ply not notic­ing it,” he says.

Hopson and his team’s work had only just be­gun when Des­tiny launched, how­ever, and he es­ti­mates that to­day they spend around a third of their time mon­i­tor­ing feed­back from the public, and two-thirds test­ing fu­ture ad­dons. “We start ev­ery morn­ing by look­ing at data re­ports of how peo­ple are play­ing

the game right now and mon­i­tor­ing fo­rum dis­cus­sions to keep abreast of player at­ti­tudes. We use this data to tell which changes will have the most im­pact and which fixes are the most ur­gent. For ex­am­ple, back when the Vex Mytho­clast was first used in PVP, In­ter­net fo­rums were filled with play­ers claim­ing that it was a su­per-weapon that made the Cru­cible un­playable. But from look­ing at the data, we could see that the ac­tual im­pact of the weapon was a lot smaller than the emo­tional re­ac­tion.”

In the in­ci­dent of the Vex Mytho­clast, the cus­tomer, it turned out, was ob­jec­tively wrong. But, ac­cord­ing to Hopson, that doesn’t nec­es­sar­ily mat­ter. “The emo­tional re­ac­tion was still true,” he says. “But know­ing that [the out­cry] was based on emo­tion rather than fact meant we could take the time to make the right fix in­stead of rush­ing some­thing out the door and pos­si­bly break­ing the game in some other fash­ion.”

The struc­tured in­flu­ence of player data and feed­back on game de­sign is a rel­a­tively re­cent de­vel­op­ment. “While analysing player data has been around for quite a while in games, it re­ally took off with the suc­cess of the free-to-play genre,” says

Seth Kil­lian, for­mer spe­cial com­bat ad­vi­sor at Cap­com, who worked on games such as

Street Fighter IV and Ul­ti­mate Marvel Vs Cap­com 3. “In the world of pack­aged con­sole games on a shelf, hav­ing a ton of data never meant that much, be­cause you could only make lim­ited changes to your game af­ter it had shipped. Up­dates re­quired a lot of time for ap­provals, test­ing by the plat­form holder, and of­ten in­volved a sig­nif­i­cant fee. The devel­oper also had less in­cen­tive to of­fer up­dates as the play­ers you were study­ing had al­ready spent the money to pur­chase your game, so there was lim­ited or no op­por­tu­nity to make ad­di­tional money. With com­pet­i­tive free-to-play games, you ex­pect to be mak­ing a lot of up­dates, so you have a chance to ac­tu­ally use the data you col­lect to ad­dress play­ers’ prob­lems and im­prove in-game sales.”

This may be true, but there was one far ear­lier era when player feed­back ex­erted a sig­nif­i­cant in­flu­ence on game de­sign: the ar­cade hey­day. Back in the in­dus­try’s ear­li­est days, Atari would mon­i­tor how much money new games made at test lo­ca­tions. If these cab­i­nets failed to pass a cer­tain fi­nan­cial thresh­old, the game would be qui­etly canned. This rather sim­plis­tic im­ple­men­ta­tion of player feed­back be­come more so­phis­ti­cated with the ear­lier

Street Fight­ers. Cap­com’s de­sign­ers would launch a new en­try in the se­ries, and then care­fully mon­i­tor top-level play­ers in or­der to see if cer­tain char­ac­ters were over­pow­ered. They could then fine tune the ros­ter bal­ance in the sub­se­quent it­er­a­tion.

“For us, us­ing player feed­back to tweak and im­prove com­pet­i­tive videogames isn’t a new phe­nom­e­non,” says Peter ‘Com­bofiend’ Rosas, as­so­ciate pro­ducer on the forth­com­ing

Street Fighter V. In more re­cent years, how­ever, the process has been im­proved thanks to data gath­ered from the game’s servers as well as the var­i­ous com­mu­ni­ca­tion chan­nels that play­ers around the world use to com­mu­ni­cate with the com­pany. “Our method­ol­ogy and data tar­gets have shifted re­cently,” Rosas says. “Nowa­days, with YouTube, Twit­ter, Cap­com-Unity, and Face­book, there are many ways that play­ers can reach out to us to share feed­back. A lot of those play­ers aren’t pro­fes­sion­als, so it

“In the world of pack­aged con­sole games, hav­ing a ton of data never meant that much, be­cause you could only make lim­ited changes”

pro­vides us with a va­ri­ety of data, from those just get­ting started all the way to our hard­core.” Bal­anc­ing a game like Street

Fighter, which is played in high­stakes cash-prize tour­na­ments and which fea­tures a large char­ac­ter ros­ter, is a tremen­dous chal­lenge, but it’s un­doubt­edly one of the most im­por­tant el­e­ments of the se­ries’ de­sign. If the com­mu­nity comes to col­lec­tively con­sider a com­pet­i­tive fight­ing game as be­ing im­bal­anced, it can ruin its rep­u­ta­tion ir­re­deemably. “We’ve found [from look­ing at the data] that there are cer­tain types of losses to play­ers that are more ac­cept­able than oth­ers,” says Rosas. “By this, I mean that if a player loses to some­thing their char­ac­ter does not have the tools to stop, or in an im­bal­anced match-up, the dis­par­ity will dis­suade them from play­ing the game fur­ther or from tak­ing the loss; the blame is shifted from them to what they might per­ceive as a poorly de­signed game. If they feel as though the loss was jus­ti­fied, the odds of them stick­ing around to play again greatly in­crease: it be­comes a per­sonal chal­lenge that they want to over­come.”

CCP, devel­oper of EVE Online, has gone fur­ther than most in cre­at­ing sys­tems whereby its au­di­ence can feel heard in steer­ing the game’s di­rec­tion. In 2008, the devel­oper com­mis­sioned a study into the po­lit­i­cal state of the MMOG’s game world. The study’s au­thors ar­gued the need for a player-run po­lit­i­cal body, con­clud­ing: “EVE’s so­ci­ety must be granted a larger role in ex­ert­ing in­flu­ence on the leg­isla­tive pow­ers of CCP.” In re­sponse, CCP es­tab­lished the Coun­cil Of Stel­lar Man­age­ment, cur­rently the only ex­am­ple of a game-based demo­cratic or­gan­i­sa­tion de­signed to rep­re­sent a player­base. The coun­cil has 14 places, and each year scores of can­di­dates stand for elec­tion, cam­paign­ing in the game on par­tic­u­lar is­sues, mak­ing prom­ises to other play­ers to ef­fect cer­tain changes.

Ev­ery six months, CCP flies the suc­cess­ful can­di­dates to its head­quar­ters in Reyk­javik for three days. Dur­ing that time, the coun­cil meets with the stu­dio’s staff and hears about the new fea­tures planned for the game’s fu­ture. Much like on In­ter­net fo­rums or in Kick­starter cam­paign com­ment sec­tions, the de­bates can be­come heated; of­ten coun­cil mem­bers dis­agree with one another. But the coun­cil per­forms a cru­cial role: it bridges the gap be­tween the game’s play­ers and cre­ators.

Some play­ers be­lieve the coun­cil is lit­tle more than a to­ken ges­ture, but in 2011 they were of­fered some strong ev­i­dence to the con­trary. At the time, CCP in­tro­duced a new fea­ture, one that had been 18 months in the mak­ing: a dig­i­tal store for in-game items that could be pur­chased for real money, in­clud­ing – brazenly – a mon­o­cle with a $70 price tag. The fea­ture was an­nounced in a grand re­veal by CCP, but it was the last thing that the ma­jor­ity of play­ers were in­ter­ested in, es­pe­cially when the game was, at the time, suf­fer­ing from nu­mer­ous tech­ni­cal is­sues. Play­ers com­plained and these com­plaints turned to anger when an in­ter­nal memo from CCP’s CEO, Hil­mar Pé­turs­son, leaked. In it, he dis­missed the play­ers’ re­ac­tion to the new store as mere noise.

That week, thou­sands of play­ers gath­ered in the game and be­gan stag­ing sym­bolic ri­ots. Many fired their ships’ weapons on a gi­ant mon­u­ment sta­tioned out­side a ma­jor trad­ing hub. The protest marked the be­gin­ning of what later be­came known as the ‘sum­mer of rage’. CCP im­me­di­ately called an emer­gency sum­mit in Ice­land for the coun­cil. It proved ef­fec­tive. Act­ing on the coun­cil’s ad­vice, Pé­turs­son wrote an open let­ter to the game’s play­ers ad­mit­ting that he had made a mis­take. Fur­ther­more, the devel­oper ac­knowl­edged the events in game: it erected a bat­tered me­mo­rial at the place where the ma­jor­ity of the protests had taken place.

For Fron­tier’s An­tonaci, re­gard­less of whether or not a devel­oper chooses to act upon the feed­back given by a game’s play­ers, it’s cru­cial to demon­strate that griev­ances have been heard. “I think it’s about in­tel­li­gent lis­ten­ing,” he says. “Any idea that is prof­fered in a con­sid­er­ate way is a good idea, since it means some­one ac­tu­ally loves your prod­uct and wants to be in­volved. Ideas need to be fil­tered, but re­gard­less of how you do this, or whether or not you im­ple­ment the idea, the key is to al­ways be sure to let the com­mu­nity know how the process is work­ing. Al­ways, al­ways close that loop.”

BioWare co-founder Dr Ray Muzyka

Fron­tier De­vel­op­ments com­mu­nity man­ager Zac An­tonaci

This Des­tiny heat map shows player ac­tiv­ity (and, in red, com­mon snip­ing spots) on The Burn­ing Shrine map dur­ing Tri­als Of Osiris, a weekly, high-level, 3v3 mul­ti­player con­test

Garth DeAn­ge­lis, lead pro­ducer on XCOM2

Seth Kil­lian, for­mer spe­cial com­bat ad­vi­sor at Cap­com

Peter Rosas, as­so­ciate pro­ducer on SFV

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.