Todd Howard says it took 7 years to make Starfield fun to play: "I thought we would find the answers faster"

I'm not surprised about this. The game was developed entirely around what it would have rather than around what the player would do and you can tell.

I can imagine the initial pitch meetings, with everyone going "whoaaa it will have hundreds of solar systems and biomes whoaaah" and no one going "ok, but what does the player do in them". A few other guys enthusiastically saying "There will be spaceship building and you will get a crew and explore with it" and not a soul in the room thinking of "ok, but how will we make space travel work within our current systems and technology? Can we make it substantial?". And this way of thinking probably permeated every second of development for the first few years.

The game is chockful of vestigial systems that they had obviously intended to be more significant and in depth, but ultimately decided not to develop further, yet still maintained in the game in a manner that only harms the game. The fuel "system", the contraband "system"... So many examples of stuff that doesn't add anything to the game, yet was still maintained because man-hours and money went into it I guess, and because the "and it will have that and that" mentality tool a priority over player experience, player agency, and actual game design.

If I can circlejerk for a bit, this is one of the reasons why Baldurs Gate 3's release and success is so timely. How many areas, how many biomes, how many systems, how many quests and how many square kilometers does that game have versus Starfield? 30 times less? 50 times less? Yet it had an overwhelmingly positive reception where Starfield didn't because its elements put player experience first. Yes it has less quests, but most are super modular and super reactive and not afraid to let you solve them in janky or silly ways that go out of the suggested solutions; yes it has fewer areas smaller in size, but you are constantly coming across stuff to do. Etc etc etc.

I'm really hoping that that contrast changes design philosophies just a tad in the future. Start with how a normal hour for your player looks like. Confirm that your technology can deliver your vision before committing to it, experience be damned. Don't reach for the stars, because contrary to what they say, it won't at least get you the moon, it will just leave you stranded in the middle of bumfucknowhere in space.

And, as we saw in Starfield, that means you get yet another annoying load cutscene.

You got the analogy backwards, it's "Aim for the moon. If you miss, you'll end up among the stars." The thing is, they didn't aim for the moon, they aimed for the stars and somehow missed.

As you rightly pointed out, the game lacks focus. It's not a procedurally generated exploration game like No Man's Sky, it's partially procedurally generated, and they didn't commit enough to make it compelling. It's not a space shooter, but it has occasional space battles, but they didn't commit enough to make that compelling. And so on. It's a game with a lot of ideas, but no direction. It's like they threw in the kitchen sink thinking that it would be fun, but all the dishes are chipped and mismatched.
- I don't even think they aimed for the stars, they built a building and then "aimed for the planetarium" that they shoved in that building that wasn't built to hold a planetarium.
  
  There are some people that think cloud imperium games (star citizen) is aiming to sell their game engine (see: starengine video) and personally I really hope they do. Then all these so-called "AAA" publishers can be as money hungry and lazy as they want and we'll still have an amazing platform for devs that actually give a shit to work off of.
- You got the analogy backwards, it’s “Aim for the moon. If you miss, you’ll end up among the stars.”
  
  waitwaitwait I swear to you on the grave of my budgies that I have always seen it the other way around
  
  I'm worried now. What other things do I have a warped understanding of? Has my life been a lie until now?! Is Starfield actually secretly a great game?!

I think it could've used a few more years, because its still not that fun.

Exploation is meaningless, which completely takes the fun out of it. There's nothing interesting to discover.

The amount of loading screens and forced cutscenes was just painful. I was so thankful for the gamepass trial because I would have been so upset if I bought this game.
- I would just love an option to stay in first person. Make the loading screens show the entire takeoffs and landings. We know the engine can already do most of it because of the glitch when you get into a random landed pirate ship and it takes off.
  
  The cutscenes look good the first 20 times you see them but I don't need the ground perspective of watching my ship take off and turn left every time.
Give them 7 more years.
- For modders to fix it.
- To ruin Elder Scrolls 6

No they didn't.

They took 7 years to release a game that isn't that fun to play.

I wonder how the game was at first if this is what they consider "fun"

I feel like this is emblematic of why many AAA titles are so dull.

I mean, you gotta give Bethesda some props here for developing their own engine. Indies don't do that.
But still, 8 years ago, they had this idea of a Bethesda game in space. Maybe they should have seen it coming that this concept won't work out terribly well, but ultimately someone decided to go ahead with it and then they spent 7 years building a space physics simulation, procedural planet generation and so on.

There was no way, they could have not released this game after realizing the concept doesn't work out terribly well. Or taken a step back and shifted the focus of the game towards space flight. Or taken a step back and deviate from the Bethesda-typical formula for this space theme.
These are options you have, when you've spent a few months prototyping, not after multiple years. They had to roll with the concept and basically try to bruteforce the fun into it.

Blizzard, back in the day, was willing to simply can games, even highly anticipated ones, when they didn't meet their standards, even after a couple years of work. StarCraft: Nova, Lord of the Clans...

And Square-Enix managed to take an MMORPG that was already released, tear it down to bare bones and completely rebuild it to make it good, with FFXIV: A Realm Reborn.

So it is possible to completely redo something if it doesn't work out...
Anyone who wants to make a space exploration game needs to play The Outer Wilds. It has one star system all the planets are about 10 miles across and the game resets every 20 minutes and there is no combat, and yet it's still an infinitely better game than StarField. Why?

Because it actually has things to do and explore and interact with. Everything you do has reason to it, It has intention. There's nothing in the sidelines that was added just a pad the experience out because it's embarrassing having 11 quadrillion planets and nothing to do on any of them, so better add some random quests.
They do not get props for trying to recycle their engine for the 100th time, because Creation sucked when they used it for Oblivion and it sucked when they upgraded it for Skyrim and it continued to suck through Fallout 4 into Fallout 76 and is very clearly not an engine designed to support a large game in space. Same bugs all the way through like five consecutive games.

Starfield was the least rocky release probably in Bethesda history in terms of bugs, but that's only because MS took literally the entire QA team from Xbox and assigned them to Starfield and brute forced a lot of the initial bugs out of the launch. A good engine doesn't need an entire megacorp's fucking quality assurance department to get ironed out.

Ultimately it feels like the same engine, despite having been improved to 64 bit for SKSE, upgraded even further for FO4, and then slapped with netcode for FO76 - it's still not good. It's unbelievable that we can have games with life-scale cities and zero loading screens, while Bethesda still needed to cut Neon in half and instance basically everything behind a billion loading screens. Even Jemison is like, 4 separate zones and not just one whole city.
- Believe me, I really don't care to defend Bethesda. I'm not saying their engine is incredibly good.
  I'm mostly saying, I feel like their games would be different and even more AAA-generic, if they built it on top of Unreal or Unity. And I'm giving them mild props for not just buying into the duopoly.
  
  But I'm also just saying that, as a result of building their own engine, Bethesda can't just quickly prototype something. To see what the final game looks/feels like, they have to invest years into engine development.

7 years wasted I'm afraid

I'm not quite sure what took them seven years. It's not exactly what you'd call fleshed out, is it?

It would have been a better game if it had been a smaller game, as it is all of the content is kind of sprayed out over a vast area.

Yup, remove the space part and just fast travel between planets. Then spend whatever resources they spent on doing the space stuff on making the dungeons more interesting and character animations better.

Wait starfield is fun to play? News to me

So it will be good and fun in 7 years then?

Yeah if we trust the modders, I suppose.
- This is the first Bethesda game that I'm not sure mods can truly save... At least not quickly. They'd almost have to make a totally different game :/ At least with other Bethesda games, the core gameplay loop is fun. With starfield, ehhhhh...

It still feels like fallout/elder scrolls in space. It feels like the same games with just a new outfit on.

About as fun as NMS on release.

Space games really aren’t that much fun. They seem fun when I read about them, but they’re not.

NMS is the least worst analogue to Starfield, and I can’t play that anymore because it’s the same thing over and over.

I feel like it's not impossible for them to be fun, for example the Everspace series does focus on action gameplay, but yeah, tons of titles try to go for realism and showing off the scale correctly, which is neat for space nerds, but quite contrary to actual fun.
- tons of titles try to go for realism and showing off the scale correctly, which is neat for space nerd
  
  As one of those space nerds, I'm glad we have games like Elite: Dangerous, Starfield, X series, Independence War etc. Choice is good and I, along with many others, love 1:1 scale sandboxes to fly a virtual spaceship in, fight , trade and explore. There are plenty of fast action games including space shooters like Star Wars Squadrons for those who don't appreciate the emptiness and loneliness of space and don't want the travel-and-life-in-space part in a space game.
  
  Starfield is the only new game from past 5 years I'm excited about and going to buy once I upgrade my GPU. It's a life-in-space sandbox that complements E:D well by doing things the latter does not.
There will be no Freelancer slander!
- Freelancer rocked so hard, and no other game has come close to how much I enjoyed space combat in it.
  
  Star Citizen is the 'spiritual successor' but I wish we had a quick and dirty standalone game that wasn't in development hell.
- To be honest, Freelancer was kind of "meh". Graphics and character animation were very nice, but ship and station design was weird and the combat felt shallow and one-dimensional. In short, too arcade-y. No joystick support was the real downer, space ship combat never feels good with mouse.
  
  For the perspective, though, before Freelancer I played and modded the absolute crap out of Independence War 2 and that is still the pinnacle of space combat that doesn't feel like WWII dogfight arcade in space while still being rather accessible and intuitive.
Permanently Deleted
- Elite is one of those ungodly good games, but it appeals to a very specific type of person. It's like a bottle of a very good alcohol with a very acquired taste. And something that requires thought, not something you get stuck into at 3pm for a handful of minutes.
  
  It's how I met the love of my life, haha. I've put well over a thousand hours into it, but remembering the controls and wondering what happened to my carrier gives me anxiety xD
  
  Still...it's one of those games I still dream about sometimes. It gives me a deep longing to see Sirus, or Canis Majoris, or the core once again, with all the stars overhead. Sometimes, we still talk about it between talks about getting engaged soon. :)
  
  God, maybe I should give it a go again. I fucking love that game.
Space games are one of my favorite genres and they can absolutely be fun. But space is really really vast and mostly empty and most space games are too shallow and/or constrain the player too much. Starfield does have some good polish and stories but is so frustratingly restrictive and non-immersive with traveling.
I think you'll find that what's actually not fun is boring games. There are plenty of fun space games but the requirement is that they are fun, and have things to do rather than just being big.

As a general rule if it tells you how many star systems there are, it's going to be a boring space game because it means that they all are procedurally generated which means they're all boring.