I wrote this in response to the vision document, but ended up getting much more specific than the vision document was ever meant to be. Sorry for the very long post, and for x-posting under "What to Include".Play Balancing and Sense of AccomplishmentWasteland was a sandbox game, meaning that that you could go anywhere at any time. However, the game’s difficulty didn’t scale, so if you wandered into an area for which you weren’t prepared, you would die fast and know that you were supposed to come back later. The result was, effectively, a linear sequence of events.
Since Wasteland 2 will be even larger, and presumably even more open-ended, there’s a good chance that the designers are spending a lot of time thinking about how to introduce some sort of play balancing. They want to make sure that players who zip through the main plot aren’t overwhelmed by late-game difficulty, while at the same time ensuring that completionists who grind it out aren’t bored with end-game encounters.
I think some sort of play-balancing is probably necessary, but it should be done with extreme care. One of the things that nearly ruined Oblivion for me was that due to constant balancing, you literally couldn’t find an encounter that was too hard for your character – regardless of your level. You also couldn’t sneak into areas that were too hard for you and snatch slightly too-powerful items. There was no sense of accomplishment when you developed your character, because all enemies developed right along with you.
In open games like Wasteland, I think it’s vital to always have areas that are way too hard for your party. Upon finding such an area, you leave, grind it out for a while in easier environs, and come back to smite the formerly un-smitable. Remember how satisfying it was to return to the Guardian Citadel in Wasteland, armed to the teeth and ready to take out the unholy saints? That’s a feeling I miss.
Getting that sense of accomplishment doesn’t mean throwing play-balancing out the window. One way to have your cake and eat it, too is to have dynamic leveling of enemies. For example, an easy, early game area might have enemies that are second level, or two levels lower than your party, whichever is higher. This way, if you find that area later than the designers planned, you will still have a pretty easy time, without being totally bored. On the other end of the spectrum, you could have an area with tenth level enemies, or two levels above your party, whichever is higher. If wander into this area way too early, you’ll be smashed. If you wait until later, you’ll still have a nice challenge.
This dynamic leveling should be combined with some kind of level-locking, meaning that once you visit an area, the enemies there stay at the level you first found them. This lends consistency to the game, and also allows players to experience the satisfaction of demolishing baddies that they formerly couldn’t even scratch. It also keeps the dynamic level balancing described above comfortably behind the scenes – it would look pretty silly if every time you returned to an area, it was populated by totally different enemies.
Placing Enemies in a Believable ContextWhen enemies in an area are scaled up or down to match a desired level of difficulty, it’s very important that their appearance and equipment make sense given their adjusted capabilities. In other words, I shouldn’t encounter a rat that can bite through power armor or a common thug with a rocket launcher.
Players use visual and other context clues to evaluate the strength of enemies. If a level two wolf looks and behaves like a level 20 wolf, it can be very frustrating. It’s important that designers don’t just boost the stats of existing enemies when they require higher difficulty – they need to replace these enemies with more difficult varieties, and help develop a rich visual context for players to use in anticipating the challenges they face. All the better if that context fits into the local story and environment.
Other Game Difficulty IssuesIf players are allowed to choose a difficulty level, there should be a reward for choosing the more difficult ones. For example, the game could provide an XP bonus per kill to players choosing a higher difficulty level.
Combat: Line of SightDuring combat, it’s essential to know what characters can see (i.e. what they can shoot). An easy way to facilitate this is to have a “line of sight” hotkey that, when pressed, shows what the active character can see by highlighting anything in his line of sight. This feature might be enhanced by color-coding the highlighted area to reflect the chance a character would have to hit a target with his current weapon, at different ranges. So the area in a character’s line of sight adjacent to him might be shaded green to reflect that he has a high chance to hit, orange as you move farther away, and red where he has little chance of hitting a target.
Just as importantly, players need to know what areas of the map will be visible to their characters after they’ve moved. The line of sight feature described above could have an alternate function by which the player would hold another hotkey, and simply move his mouse around the map to discover what his characters would be able to see from various positions. This feature would also accommodate checking the hypothetical line of site from a given position while standing, crouched, or prone.
Combat: Action PointsOne problem with the early Fallout games is that one character can be twice as lethal as another, by virtue of having just one more action point. Since weapons take a fixed number of points to fire, and action points do not carry over into subsequent rounds, one action point can become incredibly valuable. This is a major game balancing problem, and resulted in players investing an undue amount of ability points into AP-boosting traits. That ain’t right.
One way to fix the problem is to move away from an action point system. Another solution would be to allow action points to carry over, not just into the enemy’s round (see Reaction Moves below) but also into the player’s next round. Obviously you’d want to set limits on this concept.
Combat: Reaction MovesIn turn-based combat, allowing the player to take actions during an opponent's round is critically important. One reason for this is that reaction moves allow for a leap-frog style of advance. That is, one character spends all of his action points moving towards the enemy, while another character hangs back and waits for an opponent to reveal himself during the enemy’s turn. The next round, the roles are reversed. This is how troops move in real life, and there’s a reason for it. It’s hard to pull off without allowing for actions during your opponent’s turn.
X-COM handled this particularly well. Any character that had reserved enough action points to fire their weapon could do so during the enemy’s turn, with the reaction chance increasing with the number of action points reserved. Reaction was a separate character skill, as well – a very important one.
Combat: Behind the CurtainI think players appreciate being able to see how their decisions are being interpreted by the game’s various ghosts in the machine. When I see that I have a 55% chance to hit that bandit crouching behind a crate on the other side of the warehouse, I want to know what goes into that calculation. How much of it is about my weapon? My ability? The range? His cover? His armor and agility? Make it an option to display all of this in detail, with an ability to drill down to even more detail, for players that are interested in doing so.
Combat: Mop UpOne great feature would be a “mop up” button that fully automates all characters, basically having them attack the nearest enemy. It may make sense to disable turn-based mode while this is happening, to further speed things up. If a player is severely damaged, or a new enemy appears, the game would pause and mop-up mode would be suspended. Also, players could check a “melee only” box during mop up operations that would force all characters to conserve ammo by putting away ranged weapons and drawing their best melee weapon.
Combat: Damage and HealingWhen a character is hit during combat, he should suffer penalties other than just a hit point reduction. For example, he could have fewer action points during his next round, and depending on the type and severity of the damage taken, perhaps have temporary penalties applied to certain skills. If he is hit during his own turn by an enemy interrupt move, these penalties would apply immediately.
Along similar lines, minor penalties could be applied to characters who are being shot at, even if they aren’t hit. The idea would be to model the way suppressive fire works in the real world. This concept has been discussed on the forum at
viewtopic.php?f=7&t=2234 and
viewtopic.php?f=7&t=1958.
As far as healing, there should be limits on how many points can be recovered during a given time period. Spamming stimpacks is not only unrealistic, but somewhat game-breaking. Allowing characters to overdose on healing items is one way to limit this. This is discussed in some detail at forum post
viewtopic.php?f=7&t=697 and
viewtopic.php?f=8&t=91.
I also believe that a really robust location-specific damage model (e.g. get shot in leg, walk slower) is worthwhile. There is a relevant forum post at
viewtopic.php?f=7&t=1337.
Combat: General Forum PostsThe posts below have some good general discussions regarding how the combat system should work:
viewtopic.php?f=7&t=1066viewtopic.php?f=7&t=315Tons of GunsThe traditional RPG approach to weapon strength is to have weapons of increasing power discovered by the player over the course of his adventure. This steady escalation in lethality is fairly disconnected from any real-life features of the weapons. The damage model is based more on smooth and steady progression than on any sort of real-world logic.
Another approach is to go for a more realistic damage model, similar to the Stalker series of survival games. With this model, two guns using the same ammunition generally have the same range and do about the same amount of damage. Features like longer barrels might slightly increase power, effective range, and accuracy, while penalizing weight and maneuverability. Weapons using more powerful ammunition do more damage, but are heavier, more likely to overheat, and harder to fire rapidly (due to recoil). In short, damage in the game world is believable and jibes with what we know about weapons and their ballistics in the real world.
I find that the second approach is a lot more fun, but I realize that this is a very personal preference. The Wasteland world already has a robust inventory of weapons that fall closer to the traditional RPG approach, and that might be a design decision that’s already been made.
If not, I’d love to see tons and tons of real-world guns, realistically modeled for appearance, functionality, and power. I don’t think I’m alone in saying that when it comes to tactical combat games, you can never have too many guns. And while Wasteland absolutely needs science fiction weapons (like laser rifles) and hobbled-together makeshift weapons (like leaf blower based launchers) in an actual post-apocalyptic America, the great majority of fighting would be with traditional firearms stockpiled before the war. I hope Wasteland reflects that fact.
There are several forum posts about this topic:
viewtopic.php?f=7&t=1687viewtopic.php?f=7&t=1137viewtopic.php?f=7&t=1877viewtopic.php?f=7&t=962Weapon Statistics As I explained above, I’m a big proponent of pulling back the curtain on calculations used by the game, for those players who wish to know them. For guns, this would mean showing chance to jam per shot, damage at different ranges, noise created during firing… pretty much everything relevant to the game world. One reason why I feel that this is so important is that it allows players to see the impact of things they do. It’s a lot more rewarding to install a rifle component listing “increased reliability” as a benefit, if you can actually see the impact the modification has on the weapon’s more obscure statistics.
Assuming that Wasteland 2 continues the Brian Fargo tradition of allowing burst or single fire, it would be helpful to know the chance of each round hitting an average target at different ranges, in different fire modes. For example, a weapon firing an eight-round burst would likely have a first-round hit probability equal to single fire, with each subsequent round having a lower and lower chance to hit, up to some point. Players could view the “expected damage” of a given attack based upon bullet damage, and each round’s chance to hit (e.g. a weapon doing 10 damage with a 50% chance to-hit has an expected damage of 5).
Character DevelopmentThe ability to name and write backstories for your characters is a nice touch. That freedom could be enhanced by some pre-determined character backstory “tags” that the game understands, which the player could select during character creation. Examples could include things like:
• Orphaned at a young age
• Arachnophobe
• Shy with women
• Spent time in prison
Players would select whichever predetermined traits match up with their custom back story, and the game world would tailor unique events around those traits. For example, when talking with an orphan, a character who was an orphan herself might have new dialog options. An arachnophobe might freak out and run away when a giant spider moves within melee range.
Several people on the Wasteland 2 forums have mentioned cybernetic implants or other ways of having machine / human hybrids. I am all for this. And I’d love to see the designers explore how these might affect combat with robots, who would presumably get more sympathy from the increasingly robotic humans…
Map ConstructionSeveral recent games, most notably the Torchlight and Diablo series, have risen to great success based on randomly generated maps. I don’t believe that random maps have any place in Wasteland 2. Moreover, I don’t think static maps that feel random (e.g. dungeons in Oblivion) have any place in this game.
An enormous part of the texture and “feel” of a game is based on the layout of environments, and the placement of random non-interactive items (e.g. trash, furniture, etc). So far, I have yet to see a random map generator that comes anywhere close to the level of fidelity offered by hand-built, hand-placed game areas. Given the enormous community support behind Wasteland 2, which provides no shortage of free level building services by competent and committed fans… there just isn’t any excuse for not building custom, hand-crafting maps. Tons of them. Aim for Skyrim, but more varied.
Keeping with the always-(un)popular “expand on what Skyrim got right” theme, players love finding little sub-plots and back-stories that are totally off the beaten path. It is indescribably rewarding to discover an elaborate side-story that you know 90% of other players will never find. Again, utilize your rabid fan base to help build out huge volumes of this sort of content, and subordinate yourselves to revision and review for these less-prominent game elements.
Destructible EnvironmentsDestructible environments are very difficult to build well, but add a tremendous amount of realism. It’s very frustrating to shoot a LAW rocket at drywall and not even leave a scratch. Similarly, guys hiding behind a couch shouldn’t be safe from my high-powered rifle. Bullet penetration through world objects should be modeled, and ideally, just about anything should be able to be blown up. This could be balanced by a) making some walls very strong, and b) making explosives really rare. A good discussion is on the forum at
viewtopic.php?f=7&t=47.
A realistic fire modeling system could also be amazing. Far Cry 2 (a totally different game type from Wasteland, admittedly) did this very well, and it was always a blast to use fire strategically.
Leaving it up to ChanceRandomness is a large part of what makes RPGs fun, particularly in combat. However, randomness can also be a drag, particularly in two types of situations.
The first is character creation. The idea of allowing players to roll digital dice for their initial stats is a good one, but it should be tempered with some balancing. The total number of points assigned to all stats should be the same for all characters, for instance. It might also be nice to have players choose a distribution of skills (e.g. “expert in one skill”, “jack of all trades”, etc) that would influence the range of their rolls. You could even allow players select a trait or two that they want to be near maximum for each character before they begin rolling, which would allow them to build a diverse team more quickly without rolling for hours and hours.
The second situation where randomness isn’t fun is when any difficult action can be successfully completed with patient repetition. Lockpicking is a great example. In the first Wasteland, you could set up a macro for lockpicking and get through just about any door with minimal skill, so long as you were willing to sit there holding down an F key long enough. Having lockpicks break is a great solution to this problem, and every potentially repetitious skill test should have a similar limiter placed on it. Transferring items between characters was another Wasteland example. It initially seemed like a neat touch that NPCs sometimes wouldn’t give up their ammo when asked, but the fact that they would always give in after a few tries totally negated any real impact, and their behavior eventually just became an annoyance that added nothing to the game.
Survival SkillsThe decision whether or not to require players to manage food, sleep, and other minutiae seems like a personal one. My preference would be to have this be an option for players at the start of their game. Some ultra-realistic game elements that might be neat to offer:
• Weapon clips for each caliber, which must be pre-loaded with rounds before each fight.
• Realistic fire-starting, requiring something to make a spark, dry brush or other fuel, etc.
• More realistic wound management, paired to a host of specific medical supplies.
The forum post at
viewtopic.php?f=7&t=41 has other well-constructed ideas for simulating a survival situation.
Loot, Glorious Loot!Finding valuable loot, and figuring out how best to use it, is one of the most rewarding aspects of any RPG. Diablo and WoW have built huge franchises offering little else. The way that Wasteland 2 deals with loot will be a major factor in its success, and it seems like there are a few important considerations.
The first, which is becoming a common theme in this document, is how aggressively to pursue realism. In a real post-apocalyptic situation, many bad guys would have lots of valuable loot. If the bad guy in question was a survivalist in the wilderness, like you, he’d probably have an inventory of items very similar to your own – multiple weapons, food and water for several days, medicine, a big stack of ammo, and so on. In real life, you’d take whatever was most useful, and if you had some carrying capacity left, whatever was most valuable for trade. Much would have to be left behind.
In a game world, however, there are a few problems. First, at the rate you’ll be killing bad guys, a realistic amount of loot could make you a very rich person, very quickly, and diminish the element of scarcity from the game. Second, having loads of loot forces inventory management to the fore – a feature which some people enjoy, but others find tedious. Finally, the more loot you provide, the more you encourage players to spend inordinate amounts of time revisiting old locations to ferry items back to town for sale. That’s not the way most designers want their players to spend time.
One solution to this problem, and the one taken by most game designers in this genre, is to filter loot so that there are fewer items in the inventory of fallen foes, and those that are there tend to be hand-picked. A major boss might provide a big wad of cash and a powerful new weapon, while his minions might just have a few rounds of ammo, or nothing at all.
One bummer with this system is that players have to suspend disbelief after fights. If a bad guy shoots me for major damage, and then his corpse has only a paperclip and $3, what did he shoot me with?
Another possibility would be to go ahead and provide all of the loot, and force people to make decisions about what they carry. One inventory management feature that would facilitate this, and which I can’t believe hasn’t made its way into an RPG before, would be a “to sell” tag that can be applied to items in your inventory. The idea would be to select a group of inventory items, at which point a “to sell” button would illuminate. Clicking that button would tag all of the selected items.
From that point on, if a character is carrying too much, he automatically distributes “to sell” items to other party members. When the entire party’s inventory is full, “to sell” items are automatically dropped to make room for new items, in ascending order of value per pound. So if you pick up a one-pound knife worth $10, you might automatically drop, say, a two-pound pair of boxing gloves worth $10 (assuming they were tagged as “to sell”). A notification could warn players that this was happening.
Mina86 on the forums posted a similar concept to my “to sell” button, above:
viewtopic.php?f=7&t=1334To address the problem of getting rich too fast, there are several options:
• Common items, like standard-issue military pistols and food, could be worth very little at shops. Not the most realistic option, and frustrating, as it requires shopkeepers to have unbelievably high mark-ups.
• Carrying capacity could be limited. This takes away some of the fun that comes with carrying tons of loots around. It also rewards ferrying items back to town over multiple trips, though that could be counteracted if items disappear after leaving an area, a feature which would provide both its own advantages, and frustrations.
• The availability of high-end items in shops could be limited, so being rich isn’t a huge advantage. This takes away the fun of lusting after something in a shopkeeper’s window that you can’t yet afford. Unless it’s done just right, this can make money meaningless, which usually isn’t fun.
• Make nearly everyone in the game world really, really poor. The lethality of bad guys could be based more on their physical traits than on items they carry. The weapons most commonly used could be crude ad-hoc contraptions that have little or no value to shop owners. Every dead baddie could have a realistic inventory load out, but most items would be discarded by players. This is my favorite option.
It’s worth mentioning that if an “ultra-realistic” game mode is offered, having at least the option for tons of loot seems like a necessity.
ThieveryIn many RPGs, thievery is far too easy. You simply wait for an NPC to leave the room or turn their back, and rob them blind. What are the odds that a shopkeeper would actually leave a group of strangers unattended around his valuable merchandise?
Stealing from a shop should require breaking into that shop after hours, and even then, should be a challenge appropriate for the value of the items inside the shop. And the community should react to the theft, even if you aren’t caught in the act. After a few robberies, people should be able to connect the dots and grow suspicious of the new folks in town.
Gothic III did a nice job of this; if you continuously stole items, the town guard would confront you, even if no one saw you actually taking anything.
There’s a good forum post about thievery at
viewtopic.php?f=7&t=1322.
Inventory ManagementGiven how easy it is to build inventory organization tools for players, there’s no excuse for failing to do so. I don’t understand why so many games get this so wrong.
Players should be able to view the party’s inventory by character, or grouped together into one giant list. They should be able filter what they see by item type (weapon, armor, etc), item status (equipped, broken, etc), and any other relevant feature. They should be able to sort items by weight, value, value-per-pound, name, and more.
When trading with a vendor, players should be able to designate a representative for their party (probably the one with the highest charisma and / or barter skills) and have that character trade any item owned by the party, regardless of which character actually holds the item in question.
Players should also be able to quickly see the basic statistics for an item, even if they haven’t bought it yet. This would include whether or not the item is usable by each party character, along with damage, accuracy, and other stats specific to each character when using the item in question. Comparisons could be made to the items currently equipped by each character. For ranged weapons, accuracy stats might be given for both short and long range, as well as burst and single fire.
See my recommended feature for designating items “to sell” under Loot, Glorious Loot! above.
NPC BehaviorGuns are loud. If you fire a weapon inside a gangster’s hideout, everyone in the building will know that there’s a gunfight going on. RPGs in the past have handled this situation in two ways. Either a) all the bad guys come running towards the noise and are easily picked off one-by-one, or b) everyone stays put and keeps doing whatever it was they were doing, allowing you to similarly isolate and destroy them.
I’d love to see a tactical turn-based combat game where opponents react realistically to things like gunfire down the hall. Some bad guys would indeed run towards the noise. Some would run towards their responsibilities (“Oh crap, the vault!”). Bodyguards would move themselves and their boss to a safe area, and prepare for an assault (flipping over tables, blocking entrances, etc). Cowards and hired guns might run away, hide, or even offer to help the player’s cause.
Difficult DecisionsGames love to present players with important decisions that end up being arbitrary. For example, you might have to decide whether or not a thug should be hanged. Only after you make this decision do you learn the entire village is populated by vampires and the thug was actually your brother.
Plenty of games advertise that “your decisions matter”. However, if you don’t have adequate information to make a decision, it isn’t really meaningful. Sometimes, it makes sense that you don’t have all the information, but have to make a quick call based on your gut. But most of the time, you should always be able to delay and gather more intelligence before making an important decision.
SoundGiven how inexpensive it is to hire voice actors these days, I’d love to see the entire game fully recorded. That being said, it is way more annoying to hear the same five voices, or to have less content, than it is to read a little. If you’re ever in a position where you have to choose between making the game bigger and giving it a polished voiceover track, please go for “bigger”.
Also, please give us a way to disable specific sounds. For example, if my characters say something like, “Ready to go,” whenever I click on them, I’d like to be able to disable that without disabling all voice. Ditto background sounds like the constant whistling of the wind (think Fallout Tactics).