Click here to view with black text on white background
Here we go again. Civilization 6 is about to release in about a week, and I've had a number of different people asking me for my initial thoughts and impressions. I have an interesting perspective on this series, as someone who played Civ3 to death, worked on the development team for Civ4, and then found Civ5 to be a disappointing and unsatisfying title. Don't get me wrong, I know very well that Civ5 was a huge commercial success and most people seem to have liked the game just fine, but it never managed to capture me as a fan, for a variety of reasons that I've written about here on my website in exhaustive detail. Anyway, with Civ6 about to crest over the horizon, I'll pause here and offer my pre-release impressions from reading some of the previews and watching a bit of the gameplay videos. As with all thoughts from someone who hasn't actually played the game in question, readers should understand that I could be completely wrong about basically everything involving Civ6. That won't stop me from taking my best guesses ahead of time though. We're all hoping that Civ6 turns out to be a good game - let's think some positive thoughts here. 
I'll start with some more philosophical questions. What exactly is Civilization as a series? What makes a Civ game "good" to play? My answer to the first question is relatively straightforward. The Civilization games are obviously turn-based strategy games, but strategy games of a very specific sort. They use history as a theming mechanism to create a unique kind of gameplay, one where you manage a collection of people who start from humble beginnings and expand over time. These titles are often referred to as "4X games" or "God games" or something of that sort. I've always preferred the term "empire-building games", however, taking the name straight from the box artwork of the original Civilization, which read: "Build an empire to stand the test of time." To me, that's what the Civilization series is all about, creating and developing a nation over time while making all of the strategic decisions that factor into its growth.
The question of what makes an empire-building game "good" is a bit harder to define. I tend to go with the definition that Sid Meier himself came up with: a good strategy game consists of a series of interesting and meaningful decisions. In other words, the player is presented with a series of decisions that don't necessarily have a right or wrong answer, but which lead the player down different and equally valid paths. The richness in the gameplay comes from exploring those different paths and finding new ways to be successful nonetheless. The Civilization games have generally been very good at creating deep and memorable gameplay in exactly this fashion. Is the goal of the game to pursue advanced technology and win by spaceship, forge diplomatic alliances and win through the United Nations, or kill everyone else in a brutal military conquest? What techs should the player research, and in what order? What to build in each city, more units or more wonders? There are always competing priorities, and due to the random maps, no two games ever play out in the same fashion. The Civilization games have always been exceedingly good at creating that One More Turn feeling and keeping it rolling along for hours on end.
I didn't enjoy Civ5 because it failed to satisfy this premise of creating interesting and meaningful strategic choices. That's not to say that there are no choices to make in Civ5 - far from it. My problem with the gameplay in Civ5 was that so many of the choices didn't ultimately end up mattering very much, and that the gameplay kept forcing the player back into the same mostly linear path over and over again. It's not that Civ5 was "dumbed down" or anything like that; I've never made that criticism of Civ5. My issues with the game have always been that the mechanics as designed simply don't work in practice, and the gameplay ultimately felt boring to me as a result. Civ5 is a game where things happen to the player, as opposed to the older Civilization games where the player makes things happen, and has much greater freedom to do so. Civ5 tries to do a lot of neat new things, and John Shafer's design deserves credit for trying to innovate. Unfortunately, those innovations ultimately failed to work, and it dragged the rest of the game down with it.
Civ6's mechanics have been designed as an iterative improvement from Civ5, just as Civ4 was an iterative improvement from Civ3 and Civ2 was an iterative improvement from Civ1. The games in the series seem to run in pairs, it seems. As someone who spent untold hours playing Civ3 and then worked on developing Civ4, I can attest to how heavily Civ4's design was based around fixing the issues present in Civ3. Soren Johnson, the lead designer for Civ4, wrote extensively about this in the Civ4 instruction manual, discussing how Civ4's design tried to address lots of the tedious and unfun aspects of Civ3. He succeeded wildly in this endeavor, to the point that I've never gone back and played a single game of Civ3 since the release of Civ4. (Going back to corruption and no research/production overflow? No way. Not happening.) Civ4 was a near-perfect refinement of the gameplay that existed in Civ3. Rather than trying to refine this gameplay further, Civ5 wisely tried to develop entirely new gameplay by shaking up the formula of the series. The game added the One Unit Per Tile restriction, switched to the hexagon grid system, allowed cities to work 37 tiles instead of 21 tiles, introduced city states and social policies, did away with the research/wealth slider, and so on. In practice, the intent was good but the execution was lacking. In Civ5's defense, Civ3 was in a similar position, and it was also a *VERY* rough game when it was released. The difference is that Civ3, for all its flaws, was still a noticeable improvement on Civ2. Civ5 had the bad luck of being a direct sequel to Civ4, one of the best strategy games ever, and the inevitable comparison made it look much worse. As I wrote in my final review of the Brave New World expansion, Civ5 ended up being an average strategy game in terms of quality. Yeah, the mechanics didn't really work in the end, but it did enough things right to be a fun diversion for a while. If Civ5 had come out before Civ4, I probably would have loved it. However, we don't live in that alternate world, and Civ5 suffered in my eyes for failing to live up to Civ4. It simply does not have the same depth in the gameplay, and no one will ever be able to convince me otherwise.
Anyway, Civ6 has the chance to benefit from the rather messy implementation of new mechanics in Civ5. Ed Beach is heading the design this time around, a man known for creating some very intricate board games, and he's had the luxury of getting to pick and choose from what worked in Civ5 and what didn't work. I'm cautiously optimistic that this will make for a pretty good game. I'm going to look at Civ6 from the perspective of what didn't work as intended in Civ5, and offer some very preliminary thoughts on how Civ6 might be able to do better. Let's get started.
My single least favorite thing about Civ5 is the way in which expansion was handled. This has always been a problem for the games in the Civilization series: if one city is good, then ten cities is better, and a hundred cities is better yet. What stops the player from blanketing the map with an endless carpet of cities and winning the game that way? The early Civ games tried to use corruption as a mechanic to reign in expansion, in which cities lost production and commerce the further they were away from the capital. This was a profoundly anti-fun mechanic for players, and it also didn't work in practice, since even corrupt cities were still a net positive overall. It was always better for you to control the territory with a corrupt city than let another empire control that territory and use it against you. Thus mass expansion dominated the gameplay in Civ1-Civ3. In Civ4, expansion was halted through use of a city maintenance mechanic: each additional city cost more and more gold to control. The player still wanted to build as large of an empire as possible, but expanding too fast without the proper economic infrastructure in place would cause bankruptcy. Although not a perfect system, I believe that this has been the best limiting brake on expansion, by far, of any game in the series.
Civ5 used global happiness as its anti-expansion mechanic. All cities share a common happiness meter, and growing additional population points or planting additional cities eats up that happiness. When it goes negative, growth eventually stops until more happiness can be found. Unfortunately, this mechanic was broken on Civ5's release and it was never fixed despite a great deal of tweaking in patches and expansions. In the release version of Civ5, the brakes on expansion were set too weak, and it was trivially easy to just expand forever, throwing down an endless sea of cities to carpet the land. This infinite city sprawl was formulaic and boring to play, going back to Civ3 days where more cities was always the right answer in all situations.
The Civ5 design team tried a number of fixes to the global happiness mechanic over the course of patches and expansions: changing the minimum city distance from 3 tiles to 4 tiles, weakening a number of the per-city happiness tools, and so on. Nothing really worked until the Brave New World expansion, which implemented a crippling new penalty: every additional city increased the cost of all techs by 5 percent. This poor decision turned expansion into a net negative, as any additional cities beyond the first 50-100 turns of the game would *NEVER* become net positive on science. T-Hawk has explored the economics of this in great detail, concluding that strictly limiting the number of cities settled is the optimal way to play the game. On maps that aren't re-rolled endlessly to get ideal conditions, this typically translates into 3-5 cities. Anything beyond that is unnecessary, and probably a net negative in terms of winning the game. This solution is worse than the disease, as it completely destroys the whole reason for playing a Civilization game. There is no need to expand beyond a handful of early cities. That means there's no competition for land, no reason for conflict with the other civilizations. In fact, going to war at all is pointless, as any cities gained by conquest will be economic net negatives after the science (and culture) per-city penalties are applied. As a result, in Civ5 Brave New World, vast expanses of fertile land go unsettled by anyone, unneeded and unwanted. A Civilization game where expansion and warfare have been rendered meaningless is a Civilization game neutered of everything that makes it fun to play.
So for me, the single best thing that I've heard about Civ6 is that the global happiness mechanic has been removed. Thank goodness! 
It was a poor idea that was never balanced correctly, and probably never could be balanced correctly. Civ6 is going back to a per-city version of happiness, this time using "amenities" to refer to things that provide happiness for individual cities. It looks as though each city will need its own amenities and housing (a health-like mechanic) to be able to grow upwards in size, in the same fashion as Civ1-Civ4. While I haven't read a ton of details about how this will work in practice, the basic idea seems sound.
I haven't heard anything about per-city penalties to science or culture in Civ6. That doesn't mean they aren't there, but I'm hoping the design team has avoided that particular pitfall. The one brake on expansion that I've seen thus far is that each settler is supposed to escalate the cost of additional settlers, making them more and more expensive to produce over time. That's an interesting idea, making expansion more costly the more the player invests into it. This mechanic also suggests that warfare might become more attractive, since it might become cheaper to go capture someone else's cities rather than try to build additional (really expensive) settlers of your own. It also appears as though districts will scale up in cost over time, to the point where not every city will build districts for itself. It's hard to say how this will play out in practice with so many details still missing. The developer Livestreams released thus far suggest that the anti-expansion mechanics are a lot weaker in Civ6 than in Civ5, and that larger empires should be common again. I'm hoping that Civ6 will follow the dictum that I laid out for strategy games in my Brave New World review: A BIGGER EMPIRE IS BETTER, BUT IT IS HARD TO GET BIG. The Civilization games are empire-building games, and they fall apart when the motivation for expansion is removed. I'm hopeful that Civ6 has learned a lesson from Civ5 here, and won't fall into the same trap as its predecessor.
One of the issues that I wrote about in my initial Civ5 review was the prevalence of many punitive aspects in the gameplay, things that needlessly punished the player at every turn, often to make up for shortcomings in the mechanics. The per-city penalties to culture and science are a glaring example of this, creating anti-fun aspects to research and culture generation to make up for the poorly implemented global happiness mechanic. I cited plenty of other examples: roads costing gold/turn despite being necessary to connect cities, city buildings costing maintenance each turn (often making the weaker buildings an economic net negative), the per-city requirements on national wonders like the National College, the unfun choice on capturing enemy cities between razing/puppeting/annexing, and the fact that new social policies had to be chosen immediately and could never be changed once taken. In all of these cases, the player was penalized for taking positive actions. In Civ5, I want to connect my cities together, but I have to pay gold for the roads to make the connection. I want to build the national wonders, but the fact that I need a library in EVERY city to build the National College deters me from expanding. I want to grab additional social policies, but if I tick over the culture counter one turn before my research enters the Renaissance era, I can't take a policy in Rationalism and have to wait a dozen turns for the culture bucket to fill up again. It's certainly possible to quibble over these details, and a reasonable case can be made that some of these are strategic tradeoffs for the player to make. However, I still believe that a lot of the aspects of Civ5's design are needlessly punishing to the player, and could have been avoided with better design decisions.
Civ6 appears to have avoided many of these issues. (I'm sure it will introduce more annoying stuff that we can't anticipate yet, of course!) Roads are now created through trade routes and presumably don't cost gold/turn to maintain. I've seen no mention of per-city requirements for national wonders, or even confirmation that the game has national wonders at this point. There's no word about puppeting captured enemy cities, which will be a good thing if true. (I have no idea who thought it would be fun to have "turn captured cities over to AI governors where you can't control anything they build" as one of the default options. 
) The buildings in Civ6 will be split off into separate "districts" outside of the main city tile, which is a really interesting idea that I can't wait to see in practice. How that will play out is hard to predict. I'm optimistic that district placement will make for some fun city-building management decisions.
I also wanted to mention the revamp to the social policy system in Civ6. The social policies were the single best thing about Civ5, finally establishing a greater reason for wanting to have culture beyond the "expand borders" implementation in Civ3 and Civ4. Civ5's social policies were effectively a second tech tree, and one where the designers placed most of the cool stuff in Civ5's design. T-Hawk remarked that one of the reasons Civ5's actual tech tree is so boring is because all of the good stuff was moved over to the social policies. The problem is that the social policy trees have all sorts of weird restrictions in place to disguise the fact that they are a second tech tree. The policies themselves have no fixed cost, instead only escalating in cost with each new policy taken. (This is the same way the tech tree worked in the original Civilization, for the curious. It wasn't a good idea in Civ1 either.) The policies were capped by technological era, which was an awkward kludge to keep players from taking the good stuff in the later policy trees at the start of the game. Since science and culture were wholly divorced from one another in Civ5, there was no real reason to tie them together in this way other than not granting access to the lategame policy trees at the start of the game. And as already mentioned, the whole system of "you must take a policy immediately when the culture container is full" was another wonky fit. (Yes, I realize this could be turned off as a pre-game option.) Perhaps most annoying was the fact that social policies could never be changed once taken, which was a major contrast from the civics system in Civ4. While the social policies were a great idea, their inflexibility helped contribute to the often rigid gameplay in Civ5. There were no levers to pull or switches to flip. Once you took a policy, it was permanent.
However, the social policies were a great idea despite these flaws in implementation. What was needed was a refinement of the system to keep the best features in place (culture being much more important, setting up culture as a second tech tree) while smoothing out the rough and awkward spots in the gameplay. On paper, Civ6 looks like it has done exactly that. Rather than disguise the fact that culture is a second tech tree, Civ6 now embraces that fact openly. There is a second tree for culture that looks almost identical to the science tech tree, and culture is used to research those new social policies, now called "civics" in Civ6. The really cool thing about this is you can now have empires that are strong in science but weak in culture, or strong in culture but weak in science, with according benefits to either side. I imagine the Soviet Union of the mid-20th century as being a good example of the former (everyone respected Soviet science but no one really wanted to be Soviet themselves), while China in the late 19th century as a good example of the latter (massive culture influence despite technological backwardness). The designers have also stated that the civics tree will also unlock later "eras", so it will be possible to enter the Industrial era via culture even if science is lagging behind. Neat.
The civics tree then unlocks the government system in Civ6, which looks to be a combination of the social policies from Civ5 with the civics system from Civ4. Each government has a set number of policy slots, which are designated as Military, Diplomatic, Economic, or Wildcard (which can be any of the first three). For example, the Classical Republic government (which is researched via the civics tree with culture, remember) has 2 economic slots, 1 diplomatic slot, and 1 wildcard slot. These slots are then filled with individual policies unlocked by the civics tree, which have all sorts of different benefits. The player has the option to change their policies for free every time that they unlock a new civic, or can pay gold to change at other times. This allows for the Civ4 gameplay of changing governments/policies throughout the game as conditions change, while keeping Civ5's culture tree system. Even better, sticking with a single government for a long period of time accumulates additional bonuses; for example, Classical Republic provides a bonus to Great Person generation for staying in it over time. In other words, the player can hop around between military and economic and diplomatic stuff if desired, or the player can stick with one system and accrue permanent benefits for doing so. I love everything about this system, and it sounds like a total winner on paper. 
I hope it plays out this way in practice!
Civ5 had terrible diplomacy on release. The design team had the stated goal of making diplomacy "mysterious" by removing all of the information provided in Civ4 that helped to explain why the AI leaders liked or disliked other civs. They did succeed in making diplomacy "surprising", but not in a good way. In the release version of Civ5, it was impossible to understand why anything happened diplomatically. The AI leaders all appeared to be insane as best I could tell, warring endlessly for unknowable reasons. I still have no idea what the "Pacts of Secrecy" meant, and no one else appears to know either. As it turns out, it's not a good thing to have AI leaders who are complete black boxes in a strategy game. If the player can't understand why they do what they do, then they might as well be acting randomly based on hidden dice rolls. The Civ5 design team admitted that they made a mistake and backtracked in patches, re-introducing some of the diplomatic information from Civ4 and making it somewhat more clear why the AI felt the way that it did. However, since the patches were working off of a flawed starting assumption, the diplomatic interactions in Civ5 always remained a bit of a mess. It didn't help either that all of the lessons that were learned in the transition from Civ3 to Civ4, about why players should not be allowed to bilk the AI out of cash/units/cities by paying gold per turn credit only to be immediately canceled by declaring war, had to be relearned all over again in Civ5. Jon Shafer played a ton of Civ3, so how he allowed those same diplomatic exploits to reappear in Civ5 after they had been corrected in Civ4 will always remain a mystery to me.
This general principle applied to the city states as well in Civ5. I'm on record as being opposed to the concept of city states; I don't like the idea of AI entities that sit around on the map not expanding and not trying to win the game. However, I do think that city states could potentially be implemented in an interesting way, if they could be used to drive conflict between the major civilizations. Civ5 certainly did not achieve this. Civ5 may have introduced the city states but didn't seem to have any idea what to do with them, beyond providing magical free food or culture or military units out of nowhere to the player's civ. The mechanic for gaining those free bonuses was particularly stupid, either consisting of throwing gold at the city state for an alliance or fulfilling completely random quests for the city state. I mean, when you can get a full alliance with a city state for clearing out a barbarian camp or connecting a silver resource in your territory, for doing stuff that the player would be doing anyway... ugh. 
Basically, in Civ5 the player would or would not get very significant free bonuses essentially based on the random whims of the city states. This was a disastrous implementation, and suggested that the designers had no idea what to do with the city states once they were added to the game.
I'm hopeful that Civ6 will address a lot of these issues. The AI leaders have diplomatic agendas that they try to achieve, each AI leader with one default agenda that they will always pursue and one hidden agenda that can be determined through espionage. I like this idea a lot, as it helps to give each AI leader some personality and flavor, while also keeping some things held back from the player, at least initially. If I know that Ragnar will always try to build a large navy, and that he hates civs that fail to build a navy, then I can plan around that accordingly. Having some predictability on the part of the AIs makes the gameplay stronger here, not weaker. If every AI leader just does random stuff, then what's the point of diplomacy at all?
Civ6 is also introducing a "casus belli" system for warfare, straight out of the Paradox strategy games. In other words, you incur a much larger diplomatic penalty for attacking a nation without a just cause for war, and a much smaller penalty for attacking a hated enemy. Interestingly, the diplomatic repercussions for warfare are also supposed to increase over time as the era advances, with the Ancient era being a rough free for all and the Modern era a much more structued system of alliances and negotiations. I have no idea if this will work in practice, and it's certainly new for the series. I'm always hesitant about mechanics that limit what the player can do, so I hope the anti-warfare screws don't get set too tight in the later eras. We'll have to see. I'm cautiously hopeful that this will add some interesting elements to the formation of alliances between civs.
The city state system is also fortunately getting a complete rework in Civ6. Influence with city states is no longer gained by throwing gold at them (thank goodness) but through sending diplomatic envoys. How these envoys are generated is not exactly clear to me, although some governments and civics generate more of them. City states grant benefits based on how many envoys are sent to them, with more envoys granting more benefits. The civ with the most envoys sent becomes the "suzerain" of the city state, and gains additional benefits; multiple civs can gain the basic bonuses from a city state, but there can only be one suzerain at a time. City states are also more unique this time around, and every city state has its own unique benefit that falls into one of six categories: religious, commercial, scientific, militaristic, industrial, or cultural. (Fun fact: those are almost the exact six civ traits from Civ3, just with cultural in place of expansionist. The Civ series tends to re-use the same words a lot.) There are supposed to be more city states quests in Civ6, which I find to be somewhat unfortunate. I'm hoping they aren't as random and silly as the ones in Civ5. Still, this is a much more structured and detailed approach to the city states than the one employed in Civ5. I'm imagining that some of the city states will be very desirable to own, and players will have to weigh the benefits of sending lots of envoys to one city state against heavy competition, as opposed to allying with a series of less desirable city states against less competition. And the unique bonuses of certain city states will naturally favor some strategies more than others; in a military game, I might really want an alliance with Preslav (+5 Strength to cavalry units in hills) while in an economic game I likely wouldn't care about that. So there's a lot of potentially good stuff here, and at a bare minimum the system should be far superior to what Civ5 was offering. That's my hope anyway.
This is perhaps getting a bit deep into the specifics of the gameplay, but another thing that I disliked above Civ5 was the appearance of so many "free" things. The Liberty social policy tree is one of the best examples of this, with the Collective Rule policy granting a free settler. Culture fills up in the background, the player selects this policy when it's available, and poof! A settler magically appears next to the capital. This is hardly unique either; magical freebies are everywhere throughout Civ5's gameplay. The ancient ruins at the start of the game throw out tons of goodies, the city states have magical food/culture/units/faith, the social policies produce free units and Great People, and so on. It never ends from the beginning to end. Civ5 is a game all about exploiting the free stuff that appears on the map as much as possible.
So why exactly is this a problem? Well, it undercuts the actual gameplay, that's why. If I can pop 20 faith out of a goody hut and found a pantheon that way, then it belittles the need to build a shrine and produce 1 faith/turn. If a settler magically appears upon taking a social policy, then it devalues the need to build the settler in a city. If allying with a series of cultural city states can produce more culture than all of the cities in my own territory combined, or I can feed my entire empire off of magical maritime food, then what's the point of even having my own cities? As I've written before, Civ5 is supposed to create meaningful and interesting decisions in the gameplay... and then it undercuts all of those interesting decisions by throwing out freebies that render those tradeoffs pointless. The gameplay is simply more compelling when the player is building their own units, reseaching their own techs, and producing their own Great People than when all of those things are appearing out of the ether. This is a good general principle for budding strategy game designers: don't put a lot of magical free stuff in your game! It's low-hanging fruit to draw in new audiences, but it doesn't make for good gameplay.
I'm excited about Civ6 in part because it simply doesn't do this. According to the pre-release information that we can read right now, there are no civics that cause units to magically appear upon adoption or anything like that. This is partly because the civics can be swapped in and out as needed, as opposed to the one-time adoption in Civ5, which rules out that sort of thing. Still, there's a qualitative difference here. Civ5 has a social policy that grants a free settler. Civ6 has a social policy that provides +50% to settler production. Big difference. The former is fundamentally PASSIVE in nature: it does something to the player. The latter is fundamentally ACTIVE: it allows the player to do something better. This gets into the next major point that I want to make.
One of the striking differences between Civ5 and previous games in the series was how reactive the gameplay felt. In Civ1-Civ4, there's always something going on to occupy the player's attention. Cities and workers are built at a much faster pace, there are far more units to move around on the map, there are government swaps to consider, the research/wealth/luxuries slider to manage, etc. Civ5 was a stark contrast, with a significantly slower pace in the early game. Civ5 cities grow slower, there are fewer of them to manage, fewer units to move, smaller tile yield outputs, slower worker improvement speed, no spending sliders or government/civic choices to manage, and so on. To me, the game mostly felt boring. There just isn't much going on, and a lack of interesting and meaningful decisions to make. The player spends much of the time sitting back and waiting for various buckets to fill up in the background, and then choosing a prize when they do. The research bucket fills up, a new technology is discovered, on to the next one. The culture bucket fills up, a new policy is chosen, on to the next one. The happiness bucket fills up, a Golden Age is launched regardless of whether the player wants one or not. The faith bucket fills up, the tourism bucket fills up, etc. The player is sitting and waiting for things to happen. With no sliders to manage, no Slavery civic for whipping, and fewer workers than ever before, the turns are often quite dull. 
This is the passive gameplay that I keep mentioning when I discuss Civ5.
Because so little of consequence is taking place on a typical turn in Civ5, the game instead turns to trivial filler to occupy the player's time. I've called this the Civilization version of "flair" in the past, and I think it's an apt analogy. The religious conflict in Civ5 typically turns into a Red Queen's Race, requiring more and more faith to keep running in place. Once another religion is established, it's nearly impossible to make any headway in converting their cities, and the player ends up spending lots of faith in pointless battles that go nowhere. Brave New World added more of this fluff material, with World Congresses meeting periodically to do... something? Aside from getting the resolution passed that increases Great Scientist or Great Artist generation for science and culture wins respectively, the organization seems to be pretty pointless. The addition of the archaeologist unit is probably the epitome of this time-wasting nonsense. When I'm playing a Civilization game, I should be building an empire, not walking a researcher around to dig in old ruins! That sort of thing belongs in an entirely different genre of game. Anyway, the main idea is that all this stuff is largely pointless fluff designed to take up the player's time and disguise the fact that very little is actually happening while waiting for all those research/culture/faith/tourism/etc. containers to fill up in the background. Unlike past Civ games, there's no reason to go to war and no incentive for conflict, since we've already established that any cities gained after the early game are economic negatives for your empire. The result is a lot of little decisions taking place, almost none of which are meaningful in any sense, serving only to waste time and distract the player.
Furthermore, once the player drills down beyond the surface, Civ5 proves to be a very linear game. The gameplay simply won't let you get too far away from a predefined course of events. Take the social policies, for example. Due to the scaling per-city cost of the social policies, they tend to come in at roughly the same pace in every game, regardless of the size of your empire. The gameplay has been hardwired to produce an outcome where new policies arrive at a rate of roughly every 15 turns, and it's very difficult to shift this pattern outside of a deliberate cultural victory attempt. The religious system works the same way. Pantheons also scale by cost, and are founded roughly every 5 turns in the early game. The gameplay forces an outcome where most, but not all, civs will eventually get their own religion, and they all tend to spread at roughly even rates. Note how this is a dramatic difference from Civ4, where religions would be founded at wide intervals, some religions would be far more popular than others, and it was possible for a single civilization to found multiple religions (or even all of the religions!) if desired, something that isn't even allowed in Civ5. The per-city sliding tech costs do much the same thing with regards to research, trying to force research to proceed at the same pace regardless of empire size. Whereas Civ4 lets the player go off into all sorts of crazy directions, Civ5 forces the gameplay back into the same predefined channels over and over again. I don't think the average player is even aware that this is happening, but when you've played a ton of these games, the pattern starts to become obvious.
To bring this back to Civ6, I'm encouraged by a lot of the developer comments regarding asymmetric gameplay. Ed Beach and his team have suggested that they want players to be able to take the gameplay in a lot of different directions, and the information released so far seems to back that up. If the player wants to emphasize pure research, there are a lot of ways to do that: choosing to build the research campus as the first district in cities, picking science-generating faith options, picking science-generating civics, and so on. Remember, cities are limited in how many districts they can build in the early game, so if you're building a campus then you aren't building an encampment (military) or a holy site (religion) or a commercial hub (wealth/trade). This design decision will force city specialization and should make for some fun tradeoffs. If you don't build an encampment, you can't build a barracks at all, for example! 
The "eureka" and "inspiration" quests are another example of promoting active gameplay over passive gameplay. The way this mechanic works is that every tech (science) and civic (culture) on their respective trees has an individual quest associated with it. If the player achieves this quest, then the cost in beakers/culture is reduced by half. These are generally logical things that you'd want to be doing anyway, like contacting other civilizations to provide a discount on a diplomatic tech, that sort of thing. I think the key difference from the city state quests in Civ5 is that these aren't random and they are known ahead of time, which means they can therefore be planned around and made a part of the player's strategy. The mechanic even winds up adapting itself to the way the player is progressing through the game. If you're pursuing a military strategy, then you're going to be getting the eureka quests for the military techs, making them cheaper to research. This should help make different playthroughs of the game feel unique from one another. (I do have some reservations about this mechanic, as it seems as though players are often getting many of the eureka/inspiration bonuses without even trying, which is not a good thing. We'll see how it plays out.)
Similarly, it appears as though the gameplay will allow the player to push for some very different paths to victory. There's a religious victory now in Civ6, and there are a whole bunch of buildings, civics, and religious enchancements designed to improve religious combat. There are similarly a lot of different things to favor military conquest, and economic development, and diplomacy with city states. Someone who is going for a religious victory is going to be playing a very different game from someone who is pursuing pure science. This focus on asymmetrical gameplay is highly encouraging to me, at least on paper. I'm hopeful that it's going to create a lot of replay value, and create some tense confrontations between empires pursuing different strategies. What happens when the religious powerhouse confronts the economic juggernaut? Who knows! Should be fun to find out. So while I don't know if Civ6 is going to create interesting gameplay on a turn-by-turn basis, I'm hopeful that it will lean more towards the Civ1-Civ4 gameplay where the player is free to go in lots of different directions. The variability of the districts and the civics has a ton of potential if they can be implemented well.
Civ5 is a poor Multiplayer (MP) game. Yes, I know there's a group of players in the No Quitters group who enjoy Civ5 MP, and the game has some popular streamers on Twitch. That does not change the fact that Civ5 is inferior to Civ4 in virtually every way when it comes to MP, and that desite the fact that Civ5 is five years more recent and has innate Steam functionality. Civ5 was almost unplayable for MP when it released, and took many patches and expansions to reach the still weak state in which it exists today. The number of missing MP features in Civ5 compared to its predecessor is long indeed: no 8 second unit delay to stop double-moves across the turn window, no ability to ping the map, no ability to see teammate research or production, no shared wonder benefits for teams, no varying turn timer options, no city elimination options, no Pitboss functionality, no online lobby or staging area, no capacity to handle players who drop from the game, no ability to save MP games or use mods for months after release, no ability to run games with more than 5-6 players without causing massive game instability... As I said, it's a long list. Kudos to the small community who has managed to make Civ5 MP work despite all of these problems (I understand that they use a custom mod to make Civ5 more playable for MP). But Civ5 is still a bad game for MP, and considering its huge sales, the fact that the Civ5 MP community is significantly smaller than the Civ4 MP community was back in the day demonstrates the weakness of the gameplay. Our attempts at Civ5 MP games at Realms Beyond have been largely unsuccessful, and our community has stuck with Civ4 MP instead.
So this is where I talk about how Civ6 is going to fix all those problems and be an awesome MP game, right? Uh, not so much. 
The designers have said almost nothing about MP thus far, which matches how the Civ5 designers also said essentially nothing about MP before release. That tells me that there has been very little attention to MP in designing Civ6, and it will probably be an afterthought on release. That's a major disappointment. One of the reasons why Civ4 turned out so well was due to the heavy MP testing that took place before release. The stress that MP testing put the game through fixed a lot of potential issues, particularly in the realm of combat. Civ6 appears to have ignored this area again, and that's a real missed opportunity.
Long story short: I think Civ6 will be a very good game for Single Player variants. We may see a renaissance of that sort of endeavor, which never seemed to get too popular under Civ5, probably because Civ5 doesn't really allow for much deviation in the gameplay. As far as Civ6 MP is concerned though, I wouldn't hold my breath. I think it's probably going to be awful at release, and slowly improve via patching. Expectations are not high here.
OK, let's briefly discuss the most famous addition in Civ5, the adoption of the One Unit Per Tile restriction. I've repeatedly stated that I think this is a mistake: Civilization is a series with a macro focus, not a micro focus. These are not Total War games where the player does some crazy tactical moves and out fights an army five times its size. In fact, if the player can do that, then it undercuts the strategic side of the game and makes it weaker. Civilization should be more about who can build a dozen knights and bring them to the battlefield first, not who can do the awesome tactical stuff to allow two knights to beat a dozen knights. That's perfectly valid in its own right, but it's a diferent kind of game, a tactical combat game rather than a strategic empire-building game.
I'm not going to run through the full discussion of why One Unit Per Tile had all manner of negative effects on Civ5's gameplay once again. That's been up on my website for years and it still very much remains valid. In a nutshell, the map in Civ5 isn't large enough to allow the kind of maneuvering needed to make the tactical side of the combat work, and the strategic side of the gameplay had to suffer major consequences to implement the new rules (tile yields much smaller, units much more expensive, many fewer total units, etc.) The One Unit Per Tile restictions also tend to break down when playing on high difficulty, when the AI's production discounts cause it to overwhelm the map with units in the infamous Carpet of Doom phenomenon. And the AI is also dumb as a brick when it comes to playing the game under this system, completely failing to understand the movement rules and largely sitting around to get cleaned up by the player over and over again. It trivializes the gameplay when the player can be outnumbered 5:1 or even 10:1 and still keep winning because the AI is just that dumb about moving its units. Overall, the One Unit Per Tile mechanic is just a bad fit for the series, and I believe the games would be better served by ditching it.
Civ6... also uses the One Unit Per Tile system. 
This is probably my least favorite thing about the game thus far, although I can't say that I'm surprised to see that it's still around. Civ6 is an incremental improvement from Civ5, and removing the One Unit Per Tile restriction would have been a drastic change. There are a couple things that do give me a little room for optimism though. For one thing, hopefully the AI will be a little bit smarter in Civ6, now that the design team has had years to work with this system and do the programming accordingly. Maybe? I mean, I'm not expecting much here, and the AI is still going to be really stupid so long as it has to work within the system. But it can't be worse than the Civ5 combat AI, right? Right?
There are also some small changes to the overall mechanic that make Civ6's combat "Not Quite" One Unit Per Tile. Although only one combat unit can be placed on each tile, Civ6 will allow them to be stacked with a support unit. There's not a lot of information available about these support units right now; examples include a battering ram support unit to help take down city walls, and a medic support unit to help heal units in the field. How these support units will play out in practice remains to be determined. Civ6 will also allow units of the same type to be combined together in the later portions of the game, with two units combining together to form a corps and three units combining together to create an army. Their combined strength is supposed to be lower than a straight addition of their individual strength values, which means that combining units together will represent a bit of a strategic choice. This is still a major improvement though, as it should allow Civ6 to get around the worst aspects of Civ5's combat, when terrain created bottlenecks and traffic jams. Civ5's combat model badly needed a way to combine unit strengths together to power through those obstacles, and this is one solution. The corps/army system should also help out the AI's combat performance, by allowing them to brute force their way through opposition by combining units together. In Civ5 at high difficulty, far too often the AI would have massive unit advantages and be completely unable to take advantage of them due to unit traffic jams. Armies will hopefully allow the AI to use its unit production advantage to bulldozer through defenses, rather than having half of their military uselessly milling around in the background doing nothing.
I'm also encouraged by the presence of the district system, and the placement of so much of a city's infrastructure outside the center tile. The AI has always been good at producing a lot of units, and it always seems to love pillaging in past Civ games. Pillaging never seemed to matter that much in most games though, since tile improvements could be quickly replaced. Losing mature Civ4 cottages was about the only thing that really hurt. In Civ5, as long as the AI units couldn't capture the city itself, there was very little danger to your overall empire. In Civ6, however, almost all of the buildings are going to be out there on the map in districts, and everything other than encampments are supposed to be defenseless. If an attacker walks in and pillages all the buildings off the map, well, that's a pretty devastating blow, almost as bad as losing the city itself. I'm hoping that even if an AI attacker remains clueless about capturing cities, a swarm of their units might be able to inflict real damage by pillaging everything in sight. We'll see how this works in practice.
Civ6 will not be a perfect game. There will be a lot of things that are poorly balanced, or mechanics that don't work as intended, or flaws in the gameplay that can't be seen before release. I expect to spend a lot of time frustrated that different elements don't work the way that I want them to work, and many of the things that I hoped to see in this pre-release article probably won't happen in the actual game. With those significant caveats in place though... I'm encouraged by what I've seen thus far. It's a major departure from the period leading up to Civ5's release, where I kept reading the promotional materials and thinking, "That doesn't make sense! Why are they doing that?" Instead, I'm getting a lot of the same feeling with Civ6 that I had when I was reading the Civ4 design documents in early 2005, unconsciously nodding my head and thinking, "I can see why they're doing that, making that change will fix a lot of the problems that were in Civ5."
I think a lot of this has to do with the design team in charge of the respective titles. Soren Johnson had been brought in to work on Civ3 mid-project, and was heavily involved in fixing its flaws during the patching process. When he was picked to lead the Civ4 project, he had a long list of issues to solve and could design the game from the ground up to remedy those concerns. Civ5 was Jon Shafer's first title as lead designer, and it showed in a lot of ways. The game was a total mess on release, and even Civ5's biggest fans will concede that it took a lot of patches and fixes to bring the game into its current state. To Jon's credit, he's been very open about how some of his initial ideas didn't work out, and I think his current gaming project is stronger as a result of the experience he gained with Civ5. And we also need to remember that Civ5 has been a huge commercial success, selling something around 8 million copies in total, compared to Civ4's roughly 3 million copies. Most developers would kill for that kind of "failure".
Fortunately for Civ6, lead designer Ed Beach already has long experience in developing complex strategy game simulations. He's produced a series of historically-themed board games, the kind with the really niche audiences with extremely complicated rules, including the highly touted Here I Stand. Bacchus, one of the posters at our Realms Beyond forums, did an investigation into these games before it was even announced that Ed Beach was heading the Civ6 design, as it seemed likely that he would be the one in charge. According to his findings, these board games are heavy on historically appropriate theming elements and asymmetrical gameplay that still remain (relatively) balanced between the different potential paths. Among the tiny community of board gaming enthusiasts who buy these things, Ed's games are rated highly. Bacchus' conclusion from looking at these games was that Ed Beach's Civilization would likely be "messy and overwrought, yet engrossing and enjoyable." That seems to match what I've been seeing from the Civ6 previews thus far. If you were picking someone to design a new Civilization game, a history buff who loves creating complicated (and highly rated!) historical simulation board games would seem to make a pretty good choice.
With that in mind, let's cross our fingers and hope for the best. The next few months will be a fun time for the Civilization community. Let's hope Civ6 can live up to its promise. 
