Sure, if you're EA and have a very fat wallet to work with.
Judging by the state of their recent releases I'd say that's exactly the thing they don't do.
Estimating for more and aiming for less means estimating for more money than they have.
Eh... We have different meaning of word "estimate"? You estimate project cost, time, etc, based on data provided. Or you estimate what could be done for budget presented, and how fast, because there is threshold after which increasing the number of workers has no positive impact on schedule, and impact could be even negative, simply because working ground will be overcrowded.
So I really do not understand why devs cannot estimate their budget or development time. For example, "basic" development time" for project like this is 2 years. Add 1 year on top for all possible hiccups, caused by the fact that team is small, new, lacks experience, and has no cohesion between members. Then add 6 month more for unforeseen problems and/or "final" polishing. Here we have 42 months, this is best (shortest) result.
Money wise it's a bit similar. First you calculate one-time expenses, like software and hardware. Since there are far more of them than I could possibly imagine, say 50K per employee. Then calculate regular, unavoidable expenses, like rent and utility bills, food, transport costs (if they are not living in the place where they work), salary, if they pay them to external workers, or for freelancers. Don't forget to add possible inflation, and add some emergency fund, either flat value, or in percentage from regular payments.
Let's say team has 4 constant members, they rent a flat, they are single, so they do not need to feed their families (yeah, "ideal conditions" for calculations, but like I said, I have little knowledge on development). Rent could be 1K/month, food 1.5K, utility 1K, tranport fees 0.5K. They bought 4 new PCs with software for 50K each. So we have 200K budget spent there. Then monthly costs, 4K. Say, inflation is 10%, meaning these costs will result in sum under 190K for 42 months. Add 6% of monthly expenses as emergency fund, for 42 months it will be equal to 11.5K. Round sum of those up to 200K.
Additional expenses include hiring freelancers or external employees, buying assets from stores, if this is more effective (faster/cheaper) than letting one of team members to perform this task. Since I'm not from devs, I have no idea how much it will cost, but if we assume 1K/months totally, it will result in 46K for 42 months. So, for 42 months, this small team of four would need 450K. Of course, it is very simple layout, I didn't included legal advisor, PR, some other stuff I never heard of, but that's about principle. Reverse process - determation of for how long they could sustain themselves with sum present is quite similar, as well as productivity rate (probably for everyone but programmers, go figure their mumbo-jumbo of "ultimate code"). Is it that difficult, even if development is not so much "concrete" process? IMHO it's better to have unfavorable estimate and finish ahead of schedule (unless you're drivers, working with tachograph:)), than to have positive estimate and be delayed.
They are literally one poor selling game from shutting down. All the time.
Yeah, sure, and construction industry workers are swimming in banknotes like Scrooge McDuck. They don't have problems with "muse absense", they have problems with broken equipment (like, "wanna have a real sex?"). Like trucks. With "broken part being so rarely braking", so no dealership having these spare parts, and logistic expenses, in terms of forced downtime, are about week (or more). And you can't just buy another truck "just in case": it costs more than 100000€. Good luck finding money to buy one, good luck finding truck at dealership, especially with manufacturer's backlog being about several weeks. You bought it? Oh, crap, bad paperwork, police require trucks relocation for re-registration (two more weeks). Trucks finally arrived? Lucky you - board computer shows "Stop, Service". Just for the record - newly purchased truck, driven by official drivers only. Several days more (and several thousands of Euros less) problem was found: disconnected wire. Not enough? "Incorrect antifreeze" replaced during service visitation. Official service, by the way. Wow, so quality, much assurance, very service.
Or how would you like lack of materials caused by various reasons? "We don't have wires, because there is deficit of firearms ammunition and our metal suppliers cannot provide necessary materials because of deficit, and Americans invading another country, so oil price increase caused decline of chemical components supply, and plants can't provide enough source materials for insulation". Not exact quote, but leifmotif...
Or a more basic "this game has four years of work to do it properly, but marketing said it's coming out in October 2015."
Guess who wins that battle?
Nobody.
Though Ubisoft finally understood something and withheld South Park release, saving us from more "why obsidian games are so buggy" queries in Google.
There are a lot of external constraints on construction that limit what a single all-star worker can do.
Yes, and you can calculate of all them, starting from climate, environment (clothing/equipment, hindering movement), cramped surroundings, different materials, different tools, logistic, etc.
Depending on where you put someone in development, those constraints don't exist. A top tier engine developer will deliver you better performing code with fewer bugs in less time than multiple average developers will. That then impacts everything else along development.
Houston, we have a problem, two in fact: first, go find that top tier developer (buying third-party engine isn't a solution?), second, there is no official table grading developers by their skills. Another reason why I like military - short glance on someone and you know exactly what that person in front of you is capable of. Developers look murky waters in this aspect: "I participated in development of "Go kick that nasty rabbit with a shovel while riding on pink elephant's radar dish: absolute droolage", and I was senior assistant of junior lead on "funky fishes chunks". Well, maybe someone could understand that, I don't know. I can't.
Aiming for lowest possible performance means you're planning on the assumption that you don't actually have the talent required to build the thing, most likely.
If we talk about new team, without experience, cohesion, organized working order, and teamwork habits, wouldn't it be presumptuously dangerous to assume they could beat maitres in that area? Of course, maybe that particular team is super talented, but... have you seen many of those? It's not gods who make pots, but the "average Joe", honestly doing his drudgery. No need to go all Yoda on me.
Most developers aren't making Call of Duty 82. They need to deliver experiences that are new and interesting to stay in business.
Seeing massive migration of developers first into MMO, then into social, then into mobile, I tend to believe they actually do Call of Battlefield 84. Or, currently, happy candy airportville 7, or whatever. Definitely new experience.
Failures in this industry are extremely common.
Yet it doesn't save us from tonnes of shovelware "because it's safe to release another sequel". Alas, that shovelware is far from being stellar.
Brad has talked about his experience selling games at Stardock, where Walmart basically says "we want it by this date or you forefit your shelf space and have to pay us a penalty
I never worked with Walmart, but tight schedules are fairly common in construction. That's why everything is calculated, and calculated a lot. Including soil load bearing capacity and behavior during different seasons, and meteo history for this area to approximate what will happen during building time, and which risks there are. Nobody wants to see 50 tonnes road train with expensive equipment stuck in the mud that resembles mixture of clay and glue so even high pressure washer can't remove it. It's getting even more fun when it comes to oversized cargo haulage.
Not exactly.
But one day/month on site isn't much of anything when you need help with something like multiplayer NAT issues, or server scaling, or things like that.
Planning issues?
No thanks. That bloated piece of substance that incapable of sinking... Don't let me started. They may have a MASSIVE support, but when it comes to actual support they... I don't know English that well to find proper civil replacement. "They useless" is simply not reflecting the whole uselessness of them. Three (yes, just three) guys from Warszawa (I mean GoG support) work better and faster. And they help. I don't know how many employees Stardock support has, but they work fast, and helping too (should I be wearing a hat, I'd tipped it off for you, guys). Steam is slow, indeed, but they help in the end. And Valve has less than three hundred employers everywhere. I wonder if EA has more in just support stuff.
They're huge, have an entire network operations team, and still managed to totally screw up the server side of the Simcity launch.
Can less be more? Yes. See above. Many members in team is not equals increased performance. First Eador may be ugly (enough for me), yet it worked, despite being made by one guy. Second Eador, basically remake, looked prettier, but was it bugridden!
Blizzard had infamous problem with Diablo 3 before EA. EA itself had problems with BF3 at start. And they still never tested peak load with simstreet accordingly, or attempted to mitigate it with additional servers, when problem arised. Saved few grands on server rent? Remind me one SUV owner I know. Meaning, EA either greedy, lazy, just stupid outright, or all of above. They barge in like they own the place and know everything, yet it doesn't match results of their actions. I don't even mention their lies about mandatory online "because potato calculations".
There's lots of reasons for that, but none of them are "they didn't bother to hire a consultant."
Sure, they didn't bothered to listen to that consultant. Apparently he told them they can eat shi should increase number of active servers, but they considered that to be unnecessary expenses. /pokerface
For someone making GalCiv 3, there's no "gameplay" bricks they can just use.
Cells, turns, moves, races (with their abilities), techs, mechanics - all that doesn't count?
They have to build that stuff, along with the art, music, writing, and the UI.
They make UI for a living (by the way, one more happy guy with start button on Win8 (why people buy it is beyond me, but world is full of perversions) asked to say "thanks"), so I guess it's a bit easier for them. Though other components needs to be made, it's not "naked start", they already have foundation of overall design, rather large and well thought background history. So it's not like it's totally new project, made by totally new people.
You can add more stuff later.
My point exactly. From my, non-dev perspective, creating such "prototype" should be less expensive than creating of "more complete" project, thus allowing releasing it to public earlier, hence allowing testing it, and proving concept principles to be working (or not). I'd rather see you dead, little girl small part of map that works, rather than huge, yet empty map with occassional "T-pose" people walking sideways or backwards and random trees or tombstones flying around.
To an extent Telltales release their "episodes" similarly, in the meantime adjusting problematic areas. At least I've heard they did with with Sam and Max, I think - haven't played their recent "games".
Witcher sold a complete game, then used the sales from that to build a new engine for a better game. That's not the same thing as polishing up a prototype for one game, only to then throw it away and redo everything for the same game.
Isn't that just a different scale of exactly the same thing? Replace W1 with Early Access prototype (I don't know, current state of DayZ? Early builds of Space Engineers?), replace revenue from sales with, heh, revenues from sales of Early Access version, replace W2 with "same game on different engine". IMHO the same, no? Of course, there was time, 3.5 years or what? Don't forget, they used around 9 months (hm!) each time to roll out "enhanced editions", which were like 100 times better than "extended cut" (that took 4 months) for one particular 75 times perfectly scored game.
They spent time and money polishing up a version o the game to demo at PAX, and none of that work is actually in the game anymore.
I think I've heard/read that before. If in construction that usually been decided during design/planning/approval phase ("before anyone grabs their shovel"), gamedev, apparently, does that during "live prototype" stage, not early prototype, basically with paper and dices.
In addition, thank for the links, I'll read them later, if you don't mind.
So it was wasted time in terms of developing the game.
They don't, they got experience. I do not remember where I've heard or read about that, but I think there was interview with one developer, who told they've started to add stuff to their prototype and that new stuff ruined gameplay flow, therefore ruining game itself - first demo was accepted very warm, yet second one was met with less enthusiasm. They still decided to retain new features in final version of game, but they never had time to polish them properly, basically turning them into features for features sake. Things I never understood IRL too, like those "ruche thingies" interior designers so like to add, saying "this demonstrates youth!" Really, old broken bottle poured with eye-gouging pink paint? Gah, if I got 1 dollar for understanding every single interior design idea I saw I could finance non-Steam version of GC3.
In the realm of AAA game development, $30 million isn't as much as it sounds like.
Is Game dev tycoon supports this statement?
On serious side, if they perfectly knew their expenses rate, productivity rate, they know what they already created, and they are sure in their model of gameplay, they probably sure on new content additions. If it worked with dices, it probably will work with real (virtual of course:)) models. I never paid much attention to their milestones, but if they use a lot of templates in certain areas of production, that could save a lot of time and money. Don't forget outsource somewhere to poorer areas of the world where 100K$ is huge sum of money.
And I say that as a backer that really, REALLY wants it to succeed.
I'm not backer, but I don't mind them succeeding. Getting more good games is always win-win, even if games are from genre I don't particularly interested in. That's not true with space sims, I found a lot of amusement and entertainment in X-series, even in X-rebirth (still better than supreme commander 2 and ParagonRenegade, do not, I say again, DO NOT read nex text! mass effect 3:D). To an extent, of course, I'm not fun of what they did with X-R, and, apparently, they would need a truckload of time to fix that.
And just to make it clear: I do not hate developers (unless they decide to move the foundations and remove start button, ahem:)). I like them, I appreciate their work, because they do something I like. I do not like developers who, instead of standing up and admitting their failure, prefer to chicken away, hide behind "art" and you not appreciating art. I do appreciate art, but games are not art, even best of them (no Todd Howard, please). Games are games. Art is art. van Gogh is art. van Ruisdael is art. Someone's paint sneeze on the wall is not an art. So I'm pretty friendly guy, who have problems understanding some aspects of gamedev.