Compilation of Feedback & Ideas (stress test)

[Woody]

Below is a collection of feedback that I've gathered while playing the stress test so far. If I update the OP, I will highlight it in green.
*Note some of these points are not all mine but I still support so I have incorporated them into my feedback.

Combat

• Replace weak-spot chance with a low amount of scaling flat armor penetration %, reducing the defense value of the piece hit before damage calculation
  • RNG does not feel good given the high amount of skill based combat in the game. Some fights you won't get any weakspot hits and others you will get very few. Flat armor pen would make Dex builds more viable and engaging especially against standard high Strength foot fighter builds.
  • Armor pen % scales well by only reducing the defense value of the armor piece hit, therefore doesn't add direct damage and is less impactful against light armored players.
  • Armor penetration only applies to melee damage
    • Ranged armor penetration see arrow properties below
• Add aim sway when drawing/aiming a bow
  • This change aims to balance out high damage/strength archer builds by putting both a movement penalty and a general accuracy penalty when using bows without Dex. Dex lowers the amount of visible aim sway. In addition, this doesn't completely remove the skill aspect of aiming but instead aims to balance out the player aim and player build aspects of the archer game-play.
    • Greatly increase the degree of aim sway further when drawing/firing with a bow while moving/running.
      • Dexterity should further reduce the amount of aim sway of bows while moving/running.
      • Higher Dexterity should completely remove aim sway at a certain level of Dex when standing still e.g. 110-120?
      • At the highest levels of dexterity, there should still be a minor sway while moving and running.
    • Mounted archery should benefit from dexterity and the mounted archery skill should lower sway further while mounted.
    • Longbows should not be usable while mounted or have a large penalty via aim sway.
• Add properties to arrow types:
  • Broadhead
    • Add bleed effect to broadhead arrows to give them more usefulness against fleeing targets/prey e.g. for hunting fleeing animals.
    • Lower base damage
  • Bodkin
    • Standard base damage
    • Adds flat armor penetration % useful against armored infantry
  • Long bow arrow
    • Highest base damage however:
      • Greatly penalize the damage, range and arc when fired from a non-longbow.

Murder/Flagging/Rep

• Murder Status should only apply when
  • a player finishes off another player in mercy mode.
  • a player has put the other player into mercy mode, and they then die from any source of dmg (e.g animals, AI).
  • a player has inflicted damage of between 5%-100% of your max HP (so handle hits are negated) - the more damage you deal to them the longer the duration for the MC to clear e.g exponentially:
    • 5% = 2 mins (prevents handle hits)
    • 10% = 4 mins
    • 20% = 8 mins
    • 40% = 16 mins
    • 80%+ = 32 mins
• Remove murder count timers from refreshing on a new MC - this just feels plain bad mechanically and only encourages the player to not roam/PvP (even if they want to) when the timer gets close to dropping a stack, less they run the risk of refreshing the FULL 8 hours.
• Horses/pets should by default share all eligibility of crime/murder as if you were hitting the player that owns it. Likewise pets should flag the owner for the same when attacking another player/pet.
• Criminal Flag/Murder Eligibility flag should also share with:
  • a) players that heals you
  • b) players that complete (finalize and finish) a trade with you
• Reds (murderers) should be able to fight back in defense if attacked in town without guard repercussion - but only if hit first by the other player(s).
  • Also add local grey for that player to add risk that the (blue) attacking player still loses their gear if they should die to the player who defends.
    • This is important to encourage thoughtful decision making about when it's appropriate to attack and drive a red out of town, which in turn will help counteract the potential for mindless griefing via this mechanic.
• When you die to another player as a blue (non-murderer), the server needs to evaluate the closest blue priest to your position and should you then death run across the map to another region, it should then remember which priest you died at - removing reputation from the murderers in that region.
• Murderers should be able to res in a blue GY so long as their reputation is positive with the region.
• All reputation should passively increase back to 5 over time or allow players to pay a gold tribute (money sink). Paying gold as tribute to a nation should cap at 5.
• Add a criminal flag to a player that pushes/boops another player 5 or more times in the course of 120s.
• Add additional ways to gain reputation with each nation and further, add a tooltip on the reputation tab that allows the player to learn how to increase/decrease their rep with each nation.

HUD

• Stack counts shown on the inventory icon need to display weight value and have a decimal e.g. 10000 -> 100.00kg or 3456 -> 34.56kg. New players are constantly confused about the "stack" number.
• Right clicking weapons/equipment while a bank or sell/buy window is open should not default to equip but default to move to bank/bag/sell.
• When aiming/targeting a player, show a small sound icon beside the player name when they are using VOIP (this is so we can see who is talking in large crowds - useful to also be able to report bigoted/abusive language over voip)
• Add drag and drop detection so that deleting items when dragging is only possible when you drop the icon anywhere outside of a UI window that is currently active.

Graphical

• Trees meshes/textures that have volumetric clouds behind them produce unpleasant artifacting/shimmering.
• Draw/load distance of character models/equipped gear needs to be prioritized or re enabled post stress test. 10ft is too short.
  • Prioritize loading at longer distances at the very least such as Armour/Weapons > Animations > Race > Name > Features (face details etc)
• All tree textures/meshes need to extend down further so that roots clip through the terrain and look natural on slopes.

Crafting

• Extraction times feel slightly too long in comparison to other process times e.g. crafting times.
• Crafting times feel slightly too short in comparison to every other action in the game.

Animal AI/Non Humanoid AI

• Attacks by AI without a melee weapon should be telegraphed (not traditional telegraphing on the ground but via the AI animation).
  • E.g. wolves do damage to anyone hit in a line by telegraphing/charging a leap animation, then leaping forward hitting and damaging players struck in it's path.
  • Dodging or moving out of the telegraphed window (with lag compensation detection) negates the hit/damage.
• Each animal or AI should have their own AI package e.g:
  • Wolves rely on hit and run tactics, circling and strength in their pack.
    • Wolves flee when their pack dies.
    • Telegraphed lunge/pounce
  • Bears overwhelm the target and swipe, slam.
    • Allow partial blocking (blunt damage) of swipes.
    • Telegraphed swipes and Top down slams with paws that can be dodged/parried.

 

[Woody]

Tutorial

• Use tutorial island from OSRS as an example of how to explain some of the games core mechanics guiding the player on a relatively linear path while still allowing them to play naturally e.g:
  • Player can stop following the tutorial and deviate to explore and experiment
  • Come back to the steps of the tutorial anytime if they should deviate from the steps
• Add 3d waypoints to the HUD that guide the player to the objectives of the tutorial.
• Have a list of steps to complete and tick off that is visible on the hud (minimalistic - use the 3d waypoint and list of steps in conjunction with one another)
• At each objective, provide a voice over and as much interaction with the game systems between each step. Meaningful progression through the steps in the tutorial will help reinforce what the player did as being correct.
• Only allow the player to leave the tutorial once they've completed all the steps.
  • Only allow the player to skip the tutorial if this is not their first character or if they've got a tutorial achievement against their account.
• Use an order of steps that is varied while still making sense in terms of a progression of steps e.g:
  • Start by explaining controls
  • Introduce the player to the character sheet and equipping and interacting with items.
  • Have the player swing their sword at a target dummy and fire a few arrows of their bow.
  • Have them cast a spell on themselves and a friendly NPC (e.g a heal)
  • Have them head out of town and gather materials, kill animals etc. Explain the different tools.
  • Now introduce them to breaking down materials, butcher table and the crushers/extractors.
  • Introduce them to the skill system and crafting, etc.
  • Then move toward Graveyard with newly crafted gear and weapons.
  • If the player dies, have them easily recognize the tools to return to priest and ensure they can always return and find their corpse with a 3d waypoint.
  • etc.
• Reinforce previously taught mechanics over the course of the tutorial that make sense to do so e.g. how to quick swing after parrying a zombie.

Guards

• Static guards aren't a good idea as they don't allow for emergent gameplay to occur in certain places where they are positioned.
  • These buildings should instead be included in the path of patroling guards giving "windows" of opportunity and crime in settlements.
  • All guards should be on a patrol unless absolute necessary to have fixed guards (e.g banks).
  • Static guards should chase criminals but always return to their static position.
  • Static guards do however, make good additions to settlement entrances.
• Lictors shouldn't be permanently positioned at graveyards
  • Lictors should patrol out from the town to the graveyard in the morning and leave in the evening returning to the city. Graveyards will therefore not be guarded at night.
    • This gives a nice middle ground to the attempt that has been made to create relative safety for the noob farm while allowing for PKing and sandbox gameplay to still occur at night.
• Leashes need to be established for all guards. If a guard always has a set patrol/pathing, any deviation from their path such as chasing a player or npc should then resume back to that patrol path they left off.
• Leashes need to be shorter and be based on the distance exceeded from their patrol point. Exceeding a certain distance (250m?) should trigger this.
• Guards attack hostile mobs on sight e.g. if a zombie or a bandit wanders into town, they will attack.
• Guards ignore non-aggressive mobs unless the mob has dealt damage to a positive rep player, e.g. if the player initiated combat and the mob can attack back.
• Guards need a much smaller aggro radius so that crime can still occur in town
  • When an individual guard is aggro'd, they should aggro other guards to join in the chase should they overlap each other.
• Calling for guards should bring in all guards in a short distance (30m?) for a short period of time (for the sole purpose of having them enter the criminal's aggro radius). If no aggro occurs in say 30s, they should return to their patrol.
  • Will need a cooldown as to how often a player can call for guards (cry wolf).
• Infrequent patrols need to exist that extend out of the cities gates and down connecting roads within 1-2km of the town. This is mostly flavour and helps give a feeling of growing security as you approach a town.
• Reds should be able to fight back in self defense and guards will not intervene, but only against players that have damaged them first.
• Patrols should differ in size from 1-3 guards at a time.

 

[Sauron]

The part about guards sounds great

 

[Farligbonde]

Really good feedback.

 

[Truthshower]

actually a lot of good suggestions here tbh.

 

[Sauron]

Guards only attacking mobs that damage players is better than what we have now, but i think the better alternative is:
- Guards attack hostile mobs on sight (if a zombie or a bandit wanders into town guards shouldn't just watch them chasing after people)
- Guards ignore harmless animals (pigs, deer) unless they somehow damage positive rep players

 

[Woody]

Guards only attacking mobs that damage players is better than what we have now, but i think the better alternative is:
- Guards attack hostile mobs on sight (if a zombie or a bandit wanders into town guards shouldn't just watch them chasing after people)
- Guards ignore harmless animals (pigs, deer) unless they somehow damage positive rep players

Great point. Will update

 

[Woody]

Added the following to the OP:

Experience/Skill Leveling(UPDATED 10/4)

• (Bug?) The experience bar needs to represent exp required to reach the next level not reach max cap. We already know max cap is 100 levels, make use of the exp bar to show progress toward next level.
  • The tooltip of the experience bar suggests this should be the case, but clearly with the x100 skill gain in testing (however it might have been implemented) instead of it giving experience at 100x the rate, it's not even utilising or functioning correctly. Makes more sense this would've been giving us more exp?

• To mitigate against and make macro leveling less desirable, apply a ramping diminishing return on the amount of exp gained capping at 99% when a player gains experience from any source. This should be represented by a new stat under Hunger and Thirst called "Mental fatigue" which reduces the amount of experience gained by the % amount of the stat.
  • Exponential scale, incrementing the rate of fatigue gained at a much slower rate when at lower levels of fatigue, but increasing faster at higher levels of fatigue. E.g (10% fatigue increases to 11% slower than 80% to 81%.)
  • Linear recovery rate over time. Resting doubles the recovery rate.
    • 30s to recover from 1% of fatigue.
      • 15s when resting.
  • Increase the amount of fatigue not based on the amount of exp gained but the frequency of experience gained (so any amount e.g 1exp or 200exp increases fatigue the same. This is good because larger exp chunks are already balanced around time/cost investment so won't incur as much of a penalty from frequent leveling of these skills).
  • Does not apply to reading and is not increased by experience gained by reading.
  • At Max fatigue, would take 50mins (49.5) to go from 99% to 0%.
    • Resting doubles the rate of recovery resulting in 25mins (24.75) to recover from 99% to 0%.
  • This approach is performance efficient with regards to the server to track, as it doesn't rely on complex checks and timers for every player on the server - e.g. a different approach to this could be where it keeps track of who you've leveled off or what node you've hit which would be costly to performance.
    • One timer per character to reduce a percentage amount e.g every 30s to lose 1%. Performanent.
    • Incrementing the fatigue % pivots naturally off of the experience gain event on the server.