
This is a response to the excellent article over at one of my favorite sites, Explorminate. In the article, author Oliver Kiley laments on what he sees as the current unpolished state of the modern 4X market.
He defines what he defines what he means by polish:
Before going further, I should clarify what, in my mind, a “polished” game has:
- No major imbalances or exploits in the game that undermine its intended gameplay
- No major bugs – particularly the obvious and game breaking sort
- No major performance issues (late game lag, memory leaks, poor optimization, etc.)
- No underdeveloped mechanics that leave you thinking that something was only half-implemented
- An iteratively refined gameplay loop and an engaging overall game pace
I agree that the above is a good start to what would constitute polish. But the devil is in the details. For example, I would argue that Galactic Civilizations III: Crusade was unpolished at release yet it it would have passed this definition by most standards. Thus I would also add:
- Gameplay systems are intuitive and highly usable
- Documentation is complete and easily accessible
- The tutorial is both informative and inviting
- The game has no obvious "how could they miss that?" [typos, missing assets, etc.] bugs
So why are so many recent games having a hard time delivering a polished experience out of the gate?
The short answer is: cost. People pay for gameplay, not polish.

Galactic Civilizations III: Crusade shipped hundreds of amazing features including free-form ship design..and then misspelled "infidel"
Offworld Trading Company is an excellent game that was extremely polished upon release. And yet, many players balked at its initial price point of $39.99 even though it's an excellent game and one of the few economic RTS games on the market. Polish is very expensive and generally undervalued. Whether we like it or not, people buy games based on feature checklists and not polish.
Budgets vs. Sales
Now, let's have a cold, hard look at the game industry. Galactic Civilizations II cost $600,000 to make. Sins of a Solar Empire cost $800,000. By contrast, Galactic Civilizations III cost 5X as much as GalCiv II and not only has the market not gotten bigger but price pressure is greatly increased.
For example, Galactic Civilizations II sold over 3 million copies during its lifespan (over 700k on Steam where it wasn't added until it was over 6 years old). That's more than all the current crop of space 4X games combined. We're a long way from the days of games being on the shelves of Walmart and Best Buy and Steam has not filled that void completely yet (especially given its discoverability issues).

Stardock's popular, MULE-inspired RTS, Offworld Trading Company is loved by many...but frequently down-voted over its price.
So what is the answer?
Each game has its own unique story.
Galactic Civilizations III, when it first came out, had nearly half its budget consumed by the development of a brand new, multi-core, 64-bit, engine. That meant throwing out all of the GalCiv II source code base (in multicore, you're not even supposed to use pointers to give you an idea of what's involved). So the design was a lot more conservative than it otherwise would have been.
Galactic Civilizations III: Crusade is outstanding and only cost $400k or so to make because its focus was purely on innovative gameplay additions. It wasn't nearly as polished as I would like it to be but it does mean there's hope in the future (I felt GalCiv III need a lot of gameplay additions and I chose to sacrifice polish for more features) to being find a better balance between polish and innovation.

Stardock is hoping that modding, rather than DLC, is the future.
Which brings us back to the question: What is the answer? Fundamentally it involves an understanding between developers and players about the strategy game market:
- No, gaming is not "big business". Let me put this right out there: Endless Space 2 + Stellaris + GalCiv III combined will almost certainly never make as much money as Start8. Enterprise software is big business. 4X strategy games (outside Civ) not so much. There is a huge disconnect between players and developers on this issue. I see the term "money grab" regularly used in response to a $4 DLC. Do these same people give their barista this kind of grief? In an age where Steam Spy exists, there's no excuse for people not knowing that niche game development, especially today, is not a get rich scheme. Developers make the games because they love making them.
- Understand the trade offs. I am not privy to the budget of ES2 or Stellaris or MOO or what have you, but I would bet that the budgets for each of those were well over $5 million. If you want to know how much a game makes, take the list price, divide it in half for the average price, then multiply it by 0.7. So imagine a game with a $5 million budget. Let's say its list price is $40 and it has sold 200,000 units. That means it is only about half way to breaking even on the development cost (let alone marketing, etc.).
The new Master of Orion game has sold about 200,000 units. Wargaming.net paid $2 million just for the trademark. Or put another way, the game hasn't yet sold enough to pay back the cost of the trademark acquisition let alone the development budget. But as anyone who played it can tell you, it was very polished at release and relatively in expensive. The criticism directed towards it is that it wasn't ambitious enough. Would the new MOO have sold better if it had been less polished but more ambitious? I think so.
The point being: Developers have to be very careful where they invest their resources.
- Understand why timing matters. Have you noticed that May is the new release date for many games? That's because in the Steam universe, if you don't release your game by mid May, you have to wait until September. June is the Steam sale month and July and August are dead months effectively. If your studio has a $500k per month burn rate, you are asking them to lay off employees to delay. I delayed Sorcerer King until August specifically for polish (and it's one of Explorminate's favorite titles). But polish hasn't made Sorcerer King popular. It's sold only around 60,000 units on a $2 million budget. Instead, Sorcerer King would have been better served having a lot more depth and features rather than going smaller and more polished. Thus, if you have a game that is basically done but could use more polish and your choice is to release it in May or wait until August and polish it, then you should release it in May (we waited until August and had to lay people off when it didn't sell as well as we hoped -- if we had released it in May and sold the same quantity, the studio wouldn't have had to lay anyone off).
- Patience. As an industry, we are migrating from a 32-bit, single core code base to a multi-core, 64-bit code base. It's worth noting that the most "innovative" game the author calls out is the one that hasn't begun that transition yet and thus could take advantage of a mature code base. But ES2 and GalCiv III both had to make that transition and it's non-trivial. It just means that the next set of games will be much more polished. Going from from single core to multi-core is very hard. All that code where you're passing around pointers? Yea, that's gotta go for the most part. That's a really bitter pill to swallow.
So the good news is that I think players will see a substantial improvement in polish going forward as the transition from 32-bit, single-core to 64-bit multicore is completed. But in the meantime, we had to pick between gameplay, cost and polish and we can only pick two.

GalCiv III: Crusade. Some argue this is what GalCiv III should have been. But many of those players never played GalCiv II (the base game) but rather started at the Ultimate Edition.
I really enjoyed Mez's article. It also highlights the core disconnect between gamers and developers. The best way to think of those of us making these games is that we're gamers who happen to know how to code. We aren't in this for the money. We're in this because we love the games and the gaming community.
Cheers!
-brad