sithys

In 2015 I made a post titled "Some Thoughts on UI and Level Design" in which I said this: The user electricwolf said in a reply: I was tickled a little in 2019 when FE16 featured the ability to see where the enemies were going to attack next turn. This feature is orders of magnitude more powerful than the heatmap I was suggesting. I think this is the primary reason why FE16 is just so trivial, even on high difficulty settings. Here I see a missed opportunity. From a game design standpoint, they had solved a significant problem with the Fire Emblem franchise (in that the chores the player had to do at the start of each map were easily accomplished using algorithms). However, they added nothing new, they did not use the principles of game design to explore the possibility space afforded by eliminating the tedium of manually checking enemy ranges (this was acute in FE10 which was the context of that original thread). After 2015 at some point I discovered Nassim Taleb, an American essayist that wrote a book called The Black Swan, which concerns itself with the normal distribution and the failings of statisticians and economists. This book dramatically changed my outlook on many things, though in retrospect reading that old 2015 thread I find that I stand behind what I was saying and that I would extend my opinion further. Imagine that you are trying to design a military simulator. The player of this simulator must react to changing circumstances and formulate a strategy. The game designer creates many surprising situations for the player, for which they cannot be prepared. What is the game trying to teach the player? By surprising the player, they may challenge the player to become better in some way. To improve their own assumptions about the world. Thus Fire Emblem as a game is about formulating a strategy, or more specifically, developing your skills at formulating strategies given some arbitrary situation. My question that I have now after a few years of contemplating Taleb's book, is this: has any human in history ever actually improved their own assumptions about the world and improved at formulating new strategies in novel situations? I think the answer to this is no. In practice, in war the side with the better a priori strategy in a given situation wins, relative to the other side, and any improvement in military strategy as a whole is the result of accident combined with historical documentation and studying. The better strategist is therefore, the more informed one. When peering into the history of various armies, one ignores all the armies that were exterminated just as one ignores all the times the game needs to be reset because of a mistake. The thing is, those a priori strategies are often generated at random by the belligerents, either by pure emotion or by the experiences of a single commander. In a single lifetime therefore, no commander can ever be guaranteed to ever improve their arbitrary strategy generation skills, because strategies are either good or bad at random and history only pays attention to the victors and not the graveyard. If you wanted to simulate this phenomenon in Fire Emblem, you would have to release a game that had lots of surprises, only had a single difficulty and only supported ironman mode, and if you lose you can never play the game again for the rest of your life. The game would be released to a population of players at the same time and they could not communicate with each other in anyway. Only a handful of players would beat the game out of hundreds of thousands of players, because their a priori strategy was optimal for the requirements of the game. The players who lost form the graveyard. You then deny that you ever let anyone in the graveyard play your game, thus the winners learned correct strategy from your game. This is how history actually played out, and the graveyard is massive. My argument is that, you cannot say that Fire Emblem is about strategy. Even proper military strategy is not about strategy, it is more about survivorship bias. No human can ever become more skilled at military strategy in general except in relation to another army in a given circumstance with relevant historical perspective available only to the friendly side. Thus Fire Emblem as a concept for a game is total nonsense. Fire Emblem cannot be about anything more than it's own game design, and the principles of game design as such apply here. Fire Emblem is a game and the traditions of game designers are the only traditions that matter. Bad game design is always bad, even if it is "historically accurate," good game design is sometimes good, and if you know something is bad you should avoid it. Removing enemy range indicators in FE10 hard mode was bad, it did not make the game more difficult, just more tedious. The next-turn-move indicators in FE16 are nice but the design of the game fails to capitalize on them.

That is correct. It's a combination of narrative, abnegation, and challenge that attract people. This is the unique signature of Fire Emblem, though even these aspects are also decorative. Fire Emblem isn't far removed from a game of chance played with cards, one in which probabilities can be evaluated and maximized. I chose to highlight one aspect to point out the specific requirements of a software product and how development on a console does not facilitate the requirements. One could conceivably reuse the avatar editor in Awakening, but with many more options, to design all the non-generic characters in the game. Art assets are actually one of the easier problems to solve, though the product the player gets might not meed their standards of artistic expression. There are many editors that would be required to make the game, including a narrative editor, a unit growth rate/base stat editor, an overworld/world map editor (think campaign cartographer, use Perlin Noise on the edges to make it look like coast) and an editor that lets players string chapters together and insert gaiden chapters/alternate paths.

The set of people who debate statistics on this website is not a representative sample of the people who buy Fire Emblem. Yes, challenge is another important one, though not all players actually enjoy the challenge. It's important to remember that the people on this website are living in a fish tank and they actually can't see the water. To my point though, challenge arises from a configuration of the game systems, and that configuration would require the use of the aforementioned controls (combo boxes, etc) which are difficult to use with a controller. Super Mario Maker has one combo box (the theme selector) but a game as complex as fire emblem could have dozens, not to mention the need to input text. Jotari, there are other keyboard styles (such as Dvorak) which take only a few hours to learn and are much faster to use when typing, and despite these advantages the number of people who use this layout is statistically negligible when designing products (left handedness is much more significant). In order to make a profit you are going to need to appeal to lots of people, not just the handful who are willing to learn how to type using a controller.

The core gameplay aesthetics of Fire Emblem are narrative and abnegation (or submission, vicariousness, pastime). Both of these things manifest themselves over the course of long single-player epics. In order to create an epic narrative, the player would need to type in a lot of characterization, narrative text, flavor text, etc. This would not be easy to do using a console controller or even the Wii U game pad. A keyboard would be better sorted to the task. For creating abnegation in a game, you must design an XP and difficulty curve, classes, stats, and the like. You are talking about sliders, combo boxes, check boxes, basic web form or windows forms controls. Once again, these controls are much better suited to a mouse on a PC. If we did get such an editor, you would be able to create maps with no story or balancing, and at that point it's not the same game. So I can't see Nintendo devoting resources to such a project. If such a product existed, it would exist on PC (and would likely be developed by the community, as was mentioned).

Hey guys, I wanted to mention that a lot of my ideas are motivated by the philosophy of Jonathan Blow, he talks about some of the stuff we are talking about in this video (it's almost 2 hours long but it is fantastic!)

Thank you for your thoughtful post. I can see where you are going with this and I can also see that in specific instances what you say might be true. But if you give a copy of Monopoly to millions of players and somehow collect metrics on the outcomes of all the games played with those copies, you are absolutely guaranteed to find that the feedback loop has a sort of mechanistic, soul-crushing, totalitarian authority over the outcomes of the games on average. Game designers, who need to design products for millions of people, are paid to take such factors into account. What you are saying is absolutely true that we need to think about player emotion and social systems at work, but those factors are decorative. Feedback loops are the cupcake, emotion is the frosting. You can have a cupcake without frosting, but you can't have a cupcake without the cupcake.

You make some good points and people who play Fire Emblem might treat the game differently than say, Resident Evil 4 (which is the classic example of a game which uses Dynamic Difficulty). An Ulduar-style positive feedback loop would be better for increasing difficulty, especially if you can find efficient ways to include the narrative system in the feedback loop somehow. An example of where Fire Emblem does exactly that is the secret second playthrough objectives in Radiant Dawn that give extra cut-scenes in the endgame. Some of the tasks you have to complete to get those cut-scenes are somewhat difficult to accomplish. You could take this idea further by adding transparent secondary objectives, as was mentioned. You could look at it as a string of pearls with a few branches at the gaiden chapters. The gaiden chapters would be "hard mode" and if you meet the objectives of that chapter you move on to the "hard mode" version of the next few chapters on the main string of pearls. One system that I did not like at all was that found in Elder Scrolls IV: Oblivion. The enemies, items, and NPCs in the entire world scaled up as you leveled, which meant that there was no actual sense of progression (and the goblins had like a 10,000% HP growth, you could break your sword trying to kill a single goblin). The system should be far more subtle than that. In general, I also think that there needs to be many more options when starting a new game. For example, XCOM: Enemy Unknown offers Ironman mode and also many Second Wave options that change the way the game feels. I think it would be interesting if you could configure a playthrough that is hard-mode only, ironman, inverted weapon triangle, randomized classes for all units, with swapped growth rates and bases for defense and resistance, or 0% growths, etc. In this way the player can make the game as difficult as they want while also adding surprise and replayability to the game. You still want some systems in place for new players who want to play the game for the narrative but who are not skilled enough to play it normally. Fire Emblem does this with "Casual" mode but I think that is somewhat overkill. During playtesting you can find the areas that players have trouble with and tune them better. This approach makes the assumption that there are rational reasons why players have a hard time with Fire Emblem, and it attempts to address those reasons directly. I want to repeat that for emphasis. This approach makes the assumption that there are rational reasons why players have a hard time with Fire Emblem, and it attempts to address those reasons directly.

That is another solution, the dungeon Ulduar used this method in World of Warcraft.

The player growths is a reward system that keeps the player motivated to continue playing the game (beyond the intrinsic value of the probabilities and spacial tactics). Removing the positive feedback loop would make the game feel 1-dimensional.

I have thought about that exact idea before. I do think that there needs to be a neutral AI that can make subtle changes to difficulty in various ways. Choosing Easy/Normal/Hard at the start of the game locks you into that experience for the rest of the game. For new players it would definitely be better to have some sort of dynamic scaling, the game would in the first few stages calculate various metrics like move efficiency and from those metrics it would determine how difficult the game should be moving forward. If the player is known to not be very skilled the game could eliminate fog-of-war cheating for the AI, enemy unit compositions would include more units that your weaker units (which you are known to use) are good at countering to make it a bit easier to get the XP loop started, and it could force the enemy AI to be less optimal in it's quest to kill one of your units. As you pointed out this is a negative feedback loop that acts on the punishments the player receives. I hadn't thought of if that way so thanks for bringing it up. The more that the player is punished for their mistakes, the fewer potential mistakes the game presents to the player. And as you said more advanced players should be able to turn the system off entirely. This type of dynamic difficulty can be a very good thing if done well. If done poorly, it will seem patronizing and cheap to the player. The easiest way to make changes without the player noticing is to only make changes to systems which are already opaque (such as enemy movement/attack behavior and "True Hit" rates).

Because I think she is cool in general: Soldier/Halberdier/Sentinel Nephanee Because of the story/characterization: Mage/Sage/Archmage Soren Because of the uniqueness, one of the most iconic classes in Fire Emblem: Pegasus Knight/Falcoknight/Seraph Knight Florina

So I would like to talk a little bit about feedback loops. My experience with Fire Emblem is limited to playing the game normally on average difficulty settings, so I will leave it up to the community to point out any areas which I do not have direct experience with where feedback loops are used. Feedback is a term that comes from biology and it refers to the relationships between predators and prey. If ecological conditions change and a species which is preyed upon becomes very successful and reproduces rapidly, the predators which eat that species will also be very successful and act as a dampener on the success of the prey. This is a negative feedback loop. Nearly all board games use positive feedback loops as the core experience. For example, in Settlers of Catan players can build holdings which allow them to extract resources. The more holdings they have, the more resources they can extract, which in turns allows the player to build more holdings. The board game Monopoly uses a positive feedback loop and nothing else as a core game mechanic (at least in the early days). Players can purchase property which increases income and that income can be used to purchase more property. The vast majority of Monopoly games turn out the same way: a few turns into the game one player acquires enough property and enough income to dominate the rest of the game, and at that point there is no real point to continue the game. The original author of Monopoly created this simple, runaway positive feedback loop as a metaphor for the dangers of capitalism, and the game itself is not very rewarding. Games become interesting when you have multiple feedback loops with different rates of return and different risks associated with each strategy. In Starcraft, you can invest in worker units, which increase your income and allow you to reach more advanced stages of the tech tree, but it puts you at risk from being killed by an aggressive opponent early. The board game 7 Wonders features a symphony of positive feedback loops. Players can invest in resources, trade buildings, or buildings which grant free buildings as the game progresses, and they can invest in science, war, or luxury buildings which each have their own rate of granting points. But no game uses feedback loops quite as much as the Civilization series, which has so many different feedback loops that players can play through many 6+ hour games and never explore the full space of what the game has to offer. Fire Emblem has one primary feedback loop that the player can interact with, and that is the experience and progression system. Players can choose to invest a limited resource, experience, into certain units and those units will become stronger and be capable of acquiring more experience. If a character gets a bad level up, or several, they can be so crippled that you cannot reasonably use them for the rest of the game. People who run through the game with 0% growths don't necessarily worry about this scenario, but for the average player it is true. Being able to grind enemies makes the game significantly less compelling because once experience becomes unlimited, all depth is lost from the core feedback loop in the game. There are no other feedback loops to take place of the experience loop, and so the game loses an entire dimension. With this in mind, it is quite easy to conclude that a game like FE7 is objectively better than a game like FE8 if you accept the design goal of providing interesting feedback loops for the player. This also means that arenas, boss healing exploits, and infinite spawns must be ruthlessly cut from the game in the interest of preserving that core mechanic. There are other feedback loops in Fire Emblem as well. The stronger your units are, the faster you can complete levels and the easier it is to achieve secondary objectives. Many side-quests provide additional experience and powerful items that make your units even stronger. In Fire Emblem: Radiant Dawn, the knowledgeable player is presented with a source of conflict by the progression system. Players can spend a lot of time in a chapter killing all the units in an efficient way (a way that produces the most XP for the relevant units), or the player can beat the level as quickly as possible to get more bonus XP (assuming they know about it). This source of conflict is affords the player the ability to formulate a strategy and provides another level of depth to the game. Even if bonus XP is less efficient than simply killing units during a chapter, it does open up the unique possibility of inflating the growth rates of characters near max level with several capped stats. Weapons, terrain, positioning, and class bonuses all provide a source of positive feedback. If you have ever started a level, gotten halfway through and thought the level was impossible, then restarted only to find that a few better moves on the first turn make the level much easier, that is because you are taking advantage of positive feedback. So what other sources of feedback exist in Fire Emblem? Where would you create more sources of feedback in the game if you had the opportunity?

The coat of destiny. Nice avatar.

I would start from Radiant Dawn and branch from there. So leaving everything else the same I might: Make the game more like FE7 in terms of pacing, character growth, and relative unit strength. The only unit forced into the endgame chapters would be the lords. Bring back mercenaries/heroes. Enemy Swordmasters would be rare and placed in positions which challenge the player to come up with a strategy. All units would have one useful, balanced, and interesting skill that cannot be removed and which serves to give that character more character. To get other skills the player must find scrolls. Make dragons weak to arrows again. If the player is going to get stuck at some future point in the game, the game should instead force the player to get stuck immediately so that they must change their behavior until they are guaranteed to not get stuck at that future point. On the easier difficulties it should be impossible to get stuck. Make sure that speed growths are fair and balanced so that the player can reasonably form a team that they like and which fits their strategy using units that they like. Improve the stat caps for minor characters/classes relative to the most powerful units (Laguz Royals in Radiant Dawn) Don't put Tier 1 and Tier 3/Royal units in the same chapter. Normalize unit availability. Maybe get rid of Biorhythm (it seems somewhat unconsequential) More dark magic users earlier in the game. Nosferatu would be dark magic again. Lightning magic would have much higher crit. There would be much stricter limits on forging and you would not be able to bypass weapon disadvantages through forging. Maybe remove the Vine and 4 Ravens cards, remove the player's choice of improving stats, and make all forging require a coin, but make coins more common on enemy units to steal. Remove harmful skills (bane, mercy) and frustrating skills (canto) I would also add the following brand new features: Community modes, such as a randomized mode that is available after the end of the game that randomizes allied unit classes, or 0% growth LTC mode built-in. A Timesplitters 2-style mapmaker would be cool. I would change the mechanics further to add new gameplay mechanics: I feel the story must be linear (string of pearls) in order to create a proper dramatic experience, with no grinding/backtracking/world map for the same reason. But the gameplay can get a bit samey after a while, so I would change the map objective and unit count in different maps to make the game feel more varied. For example, there could be a 5v5 map against powerful enemies with abilities and unique personalities, or lone WoW-style raid bosses that require time-bound strategies to defeat. This would help break up the experience while leaving the existing game unaffected, provided these types of objective would only be present rarely. (2-3 times each out of 30 chapters) I might add new weapon types, maces, morning stars, longswords, halberds, battle staff. There would be a minor weapon triangle internal to each weapon category (swords beat daggers, daggers beat longswords, longswords beat swords) with a 5% bonus. I would make javelins/hand axes 1 use only, remove the hit penalty and immobilize units struck by such weapons for 1 turn. With the exception of javelins/hand axes, weapon durability would have 2 levels. Existing durability numbers would remain the same, but after a weapon reaches 0 durability in the current system, it would have an additional 50% uses before breaking with dramatically reduced effectiveness. These changes especially would need to be playtested extensively to see if they create a good experience.

Camel archers.