Sindusk

Members
  • Content Count

    875
  • Joined

  • Last visited

  • Days Won

    32

Sindusk last won the day on August 12 2021

Sindusk had the most liked content!

Community Reputation

2215 Rare

About Sindusk

  • Rank
    Villager

Recent Profile Visitors

10146 profile views
  1. The base kingdoms will not be fine. They will not be fine even with the map restricted. Players will gravitate towards PMK's for increased membership control and there's no way to avoid this. They'll have the higher player base and, as you mentioned, stronger organization. They'll run the map and the base kingdoms will be drained of all players. If your goal is to keep base kingdoms relevant, you cannot add PMK's at all. The real problem is that I don't understand the goals here. The PMK map restriction is artificial and doesn't solve any problems from a design perspective.
  2. Quote from Defiance article on the wiki: There was already an implementation for "base kingdoms people first join have plenty of space and are not going to get bullied off the map." They have an entire island to themselves when they first get to the server. This artificial PMK map is entirely unnecessary. The problem you're presenting has already been solved by the starter zones at the inception of Defiance.
  3. Based on latest WU sources should be 1,075 unique crafting recipes. This is strictly for the "create ->" menus, and ignores anything that doesn't use it, such as cooking or natural substances (mix isn't a creation). This also ignores the fact that many different materials can be used to make the same item.
  4. WurmNode

    When I wrote the post, I was thinking solution 1 would be sufficient. After sleeping on it and reading your post, I believe you are right. Solution 2 and removing the multipliers is the best choice at this time. It's definitely a weak argument on my part to say "I did the math and research, I just can't prove it!" I wouldn't want a system to be dictated by someone saying "trust me" instead of providing hard evidence, and that's what I've been doing recently. I may have been able to stand my ground and defend it if I still had the resource to produce the results like I did years ago... but without it, there's really no point in attempting to keep it the way it is. There would always be an asterisk on it, questioning the methodology, the assumptions, and the results. The next update to the STP will remove all the combat multipliers. This will affect all weapon and shield skills dramatically, as well as Mind Speed. I'm also very much in favor of a "raw STP" option for displaying the leaderboard. While I think that the current system is a general improvement over the raw values, it's better to provide the users some choice in how they want information displayed. There's merits to looking at raw STP as well, and would provide some interesting perspectives and comparisons. Finally, your previous post did not come off as aggressive. It came off as passionate, and I welcome opposing opinions and challenges to the system. You were right to do so - you had valid points and I'm not always correct. I don't want the STP calculations to be my opinion or biased in any way, and the only way to prevent that is to have others challenge the system in areas that look incorrect. Without your foundation none of this would be possible, and we'd still be looking at total skill points instead. I'd take the Niarja STP over that any day.
  5. WurmNode

    A really detailed post and one with a lot of useful information. Quite a lot to get to, mostly what I've highlighted in bold. Everything that is not in bold generally revolves around tick sizes or optimal skilling methods. This is skewed by the existence of skill gain multipliers and the usage of tick size/rate instead of skill per time. To quote Chakron's original explanation: STP is based around the amount of time spent on the action timer, but not in the most efficient way. It does not consider positive influences like skill gain multipliers, sweet spot, and tick rates. It also doesn't consider hinderances such as cooldowns or infrequency of use (like exorcism). STP as a whole will never be perfect, and should be used not as a definitive measure of investment, but instead a way to contextualize the numbers displayed on an account to a reasonable degree. Weapon skills can be skilled faster in isolation than a normal skill, that's undeniable. When in combat and dealing significant portions of creature health, you're going to see massive tick sizes that raise the average skill/time for weapons. It's even possible to gain 60 seconds of skill gain every ~6 seconds. This is obviously a flawed way to assume that players skill weapons, and giving a 0.1x multiplier on weapon skills just because you can get that amount of skill in short periods of time would be frowned upon heavily. The key here is to figure out a proper assumption for how much weapon skill is gained over a 10 second interval. This amount changes based on how much damage you're outputting, and that's where the simulation assumption has failed the calculations here. Everything I see in your post is correct as far as I'm aware, and I'm not in denial of what is absolutely true and exists in the game. But how much damage should be expected to be dealt to a creature over 10 seconds of combat? 10%? 50%? 100%? This type of assumption must be made at some level in order to keep the system sane. Weapons are not 1:1 with normal skills, I think everyone can agree on that. But to what degree? As you stated, the combat skill gained changes dramatically based on the CR differential compared to the opponent. Should we assume the player has higher CR than the opponent? It's becoming increasingly clear that players are seeing the value of combat skills is too high right now, and I'm inclined to agree based on the evidence provided. I can see several approaches: Solution 1: A simple mathematical improvement for existing simulation data would be to assume the player has double the CR of the opponent. This would lead to a 2/3 hit chance instead of a 1/2 hit chance. We can reflect this in the multipliers by reducing combat skills to 3/4 of their existing values to simulate the increase in hit chance. This would bring all the combat skills that are 3x to 2.25x, and the 2x skills to 1.5x. Upsides: No additional work required to be done, simply a change of assumption that should be mathematically accurate, while still reflecting the increased difficulty of leveling weapon skills. Downsides: This is still fairly generic and doesn't actually solve the root of the problem that combat is quite dynamic in nature. Solution 2: Remove all multipliers on the combat skills and reset them to 1x. Upsides: Removes any bias, simulation, or external thoughts from the process and allows the game system to determine the value of the skill. Downsides: Combat skills are definitely slower to skill than normal skills, and this would result in devaluing combat skills in the STP calculations compared to actual time spent to acquire them. Solution 3: Zero out combat skills and remove them from STP calculations altogether. Upsides: Keeps the STP system technically accurate, as it's meant to simulate actions on the action timer, not combat skills in combat. Downsides: Removes a huge portion of the game from the STP calculations. For me, solution 3 isn't really a solution. Either of the first two options seems relatively reasonable at this stage. To The Bolded Points I bolded several sections of your post and want to respond to each one individually here. Stance skills being extremely undervalued: This is definitely true. Stance skills are currently under their proper value. I can't recall why they don't have the combat multiplier and don't remember how stance skills advance in combat. I can look at that. Shield bashing over 50 should be valued more: I believe this uses the action timer, so it was simply left alone. Skilling methods, cooldowns, etc. are ignored in STP calculations. If it doesn't use the action timer for its times parameter, then it can be adjusted accordingly. Misc Items showing up twice: That would need to be answered by Drogos. Baking being panfilled: I recall this being mentioned to me, and saying I would fix that. Then I didn't. It'll be in the next iteration. Skill graph characteristics over time: Another one for Drogos. Archery: I believe this also uses the action timer, which is why it was also left alone. Skilling methods, cooldowns, etc are ignored in STP calculations. If it doesn't use the action timer for it's times parameter, then it can also be adjusted accordingly.
  6. WurmNode

    I only today just realized that I noticed Rocklobster has more soul strength than the other two. Why are you and Rocklobstar basically having the same name? Lol.
  7. WurmNode

    Then it would take 101 days of grinding to get to their skills. If that's commonplace, then I'm shocked that more players don't have similar skills by now. The STP calculations aren't based on 1 every 5 minutes anyway. They're based on actual in-game combat time. The goal was to get 10 seconds of combat = 10 seconds of smithing/mining/etc. Admittedly, it could be off and my simulations could have had false assumptions in them. The way I simulated it was through my program that sandboxed the WU combat/action/skill code into a standalone program, then ran tests with player combat to see the results. It's the same program I used to get many of the results found in the WU Datamining thread I made years ago. It allowed me to run thousands of "loops" of combat in seconds and analyze skill gains during that time period. It gave direct access to the multiplier passed through the times parameter and into the skill system. These were then averaged to a 10 second interval, then compared to what was gained from 10 seconds of action bar skilling normally through the skill system. The simulations came out to be around 3x lower than what would be expected in that timeframe if you were skilling something like smithing or mining. That's how the multipliers ended up where they are today. I really wish I still had that program and could run it again, so I could give accurate simulations and raw data to support the conclusions that I came to years ago. The sad truth is that this was all designed years ago when I was both more invested as a player and more interested as a developer than I am today. However, proving it accurate or inaccurate doesn't require such drastic measures. Players can present results of how much skill was gained over periods of time in combat through in-game data collection. If players track their amount of time in combat and their start/end skills after a hunting session, it would go a long way towards improving the system. This requires a lot of attention to detail since sleep bonus, weapon CoC, and other skill gain multipliers would need to be accounted for and reversed to ensure accuracy. Regardless, there was an immense amount of time put into these top accounts on WurmNode and it's not surprising they have the highest STP. That's the only point I was going for.
  8. WurmNode

    Mind Logic and Soul Depth are not affected by external multipliers. At one point early on they were, but not anymore. All characteristics are identical outside of Mind Speed. This is an exception because it can only be gained in combat, which has bad times multipliers. The STP calculations for combat are based on existing mechanics for combat skilling. The top accounts did not skill through current mechanics. To achieve them using the method above, I estimate that they'd have to kill about 880,000 creatures. Assuming you kill a creature using the method described above at a rate of once per 5 minutes (24 hours a day), you would need to do this for about 8.3 years straight in order to obtain their current skill levels. Personal Opinion: the top two accounts on WurmNode right now - Silakka and Rocklobstar - are way ahead of the accounts below them. If you select any characteristic that is not Mind Logic or Soul Depth, they're #1 and #2 in all of them (outside of Body Control, honorable mention to Spacelamp). They also are the only accounts with over 90 in characteristics outside Mind Logic and Soul Depth. They are 2 of the 3 accounts with 100 mining, and 2 of the 5 accounts with 100 digging. They also have the highest shovel skill among those accounts with 100 digging. While there are better accounts for general crafting - Finn for example - they do not have multiple skills at 100 like the current top accounts. Niarja had several accounts well known for high crafting, but they're not uploaded to WurmNode as of this time. If the old Niarja crafters such as Figs or Dadd were uploaded to WurmNode, you'd be able to see that it's not heavily skewed towards STP for combat. It's just that the current combat accounts dwarf the existing uploads of crafting accounts.
  9. WurmNode

    I haven't looked for the exact values recently, but that does sound right to me. The calculations for the combat multipliers were done with an assumption of 1 vs 1 combat with an opponent of equal skill. It's totally possible that it's a faulty assumption to make and I'd be willing to revise it if a better assumption scenario is presented. This is logical and I'm not opposed to splitting the value between the two skills equally instead of having it weighted to one side. The original thinking was that digging was the primary skill, so the STP should be valued fully. The shovel was simply a tool and could have it's value reduced to compensate without hurting the primary skill. If it's better to split their value equally, then that's logical as well and should have no negative effects since the two skills are so intertwined. If there's no objections to it, then I can adjust it alongside all similar skills (like mining and farming).
  10. WurmNode

    This portion of the response I separated because it will be much more technical and go into some of the mathematics and philosophy behind the calculations that determined the current state of the system. From the outside looking in, it does seem weird that the parent stats - the most difficult ones - are valued the lowest. However, the logic behind it was actually the basis behind the system. Problem: Skills that gave multiple characteristics would be multiple times more efficient in STP grinding than skills with less or no characteristics. The solution to this was far more trial and error than calculation to overcome when designing the values and basis for the STP system here. In the end, my approach was to essentially combine skill + characteristic STP into one total value. Instead of 1 skill + 1 characteristic = 2 STP, I broke it into a split. 0.5 skill + 0.5 characteristic = 1 STP. This was the initial idea, so for example prospecting would be valued at 100%, then something such as a nature skill would be valued at 50% and the other half would be covered by soul depth (which comes from nature) giving the remaining 50%. What I noticed was that this didn't really solve the problem. Mining was still immensely better because it would be granting 0.5 * 3 = 1.5, without even accounting for mining itself. The multiplier for mining was irrelevant. I then kept progressing this split with less of the multiplier on the characteristics and more towards the skills. 60/40, 70/30, so on and so forth. This continued until skills with less characteristics started to equalize with multi-characteristic gains of other skills. Off the top of my head, remembering years ago, I believe the test case was a nature skill versus mining. In the current system, leveling a nature skill is roughly equivalent in STP value to mining. Drawing the slider any further towards characteristics (80/20 for example) made mining more efficient. Drawing the slider any further towards skills (95/5 for example) made nature more efficient. The current is what was found to be an estimation of the best equalization for these purposes. Each skill multiplier is a calculation of the remaining split after the base 10% value of each parent. This is equivalent to 2 - (1.1^(numParents)). For mining: 2 - (1.1 ^ 3) = 0.669. This leaves room for the 10% valuation of the primary characteristics to fill in the missing amount from skilling each skill. This, I concede, is perhaps the greatest flaw of the system as it stands. I made a minor improvement here over the original Niarja version by reducing the mutiplier of skills based on how many parent skills they had. However, The unfortunate truth is that there's problems at play that I had no solution to, so I left it alone. Problem: Skills that have a parent give significantly more STP than those that do not. I could not find an elegant solution to this problem. There are skills such as Weapon smithing which prevent an elegant solution. If you mathematically reduce the parent skill STP multipliers based on all skills below them, it leaves weapon smithing sorely lacking. Players do not typically skill their subskill. If you reduce the value of STP, you're doing a disservice to anyone skilling it. This is additionally problematic due to the fact that some accounts are soft locked, unable to skill their parent skills for weapons. I did a quick google search trying to find the where that topic originated, and found this thread if you're interested in what's happening there. On one hand, you have a parent skill which must be valued fully otherwise you're doing a disservice. On another, you have parent skills which don't raise properly because the way the combat system works. So if you value them appropriately, they don't actually function correctly for the system. It's a mess that I could not find an elegant solution to, and am happy to hear opinions on how to resolve. This was the most difficult portion of the system to resolve. The fundamental truth is that weapon skills do not skill at the same rate as other skills. When you parry, how big of a skill tick do you get for your sword? Is that 5 seconds worth? 2 seconds worth? If you parry once every 10 seconds, are you training your weapon skill efficiently? Is there a difference between parry skilling and dealing damage? If you equip the same weapon in both hands, do they skill at twice the rate? I took a long, deep dive into the combat system a while ago, and really dug deep into getting these types of answers. In order to answer one of the questions I've asked: it's 1 second. You get 1 second of action time per parry. Is it even possible to parry once per second? Kind of. It's a soft cooldown, where your chance to parry improves as time goes on. If you're lucky you could be fighting 20 trolls and get 20 parries in under a second. How do you account for that? I took a simulation of assumed combat: 1 versus 1 combat of equally skilled combatants - and used that as the foundation to build the skill multipliers for weapon skills and combat skills in general. In my simulations I found that weapon subskills and shield subskills raised at roughly 1/3 the rate of normal skills when in constant 1 versus 1 combat with an equal skilled opponent. That is why they all have 3x the value of similar difficulty skills. It could be argued that this is subjective or based on a faulty assumption. That's a reasonable argument to make and I'd be happy to re-evaluate the calculations done if a case is made that there's a more accurate scenario to assume, or an analysis of the combat and skill system is made which proves my results inaccurate. On the topic of fighting skilling multiple skills, this is a problem throughout all of Wurm which, as stated in the second block, doesn't really have an elegant solution. When you skill platesmithing, you're also getting armoursmithing, smithing, hammer, and miscellaneous items. The philosophy is that combat is not much different to this type of scenario. If all of the fighting skills increase independently at 1/3 the rate, shouldn't they all get a 3x multiplier to compensate? If that logic is faulty, then it should be reconsidered and I will do so.
  11. WurmNode

    Why is weaponsmithing worth 31 million but small axe is worth 33 million? Assuming you are in combat at all times, weapon skills raise at roughly 1/3 the rate of normal skills. Combine that with the fact that small axe has a 7k difficulty instead of the standard 4k like most other skills, it's basically slightly harder than weapon smithing. Small Axe: 7,000 * 3 = 21,000. Weapon Smithing: 20,000. This probably isn't even enough, as small axe cannot be skilled at all times unlike weapon smithing. However, this dives into the philosophy of action time versus infrequency which I've gone over before. Shovel being 4.4 million with digging at 4.1 million, even though shovel takes roughly 2.3x as much effort, why's that? Because they level hand-in-hand. You cannot skill one without the other, and it essentially counts as two skills instead of one. Valuing shovel at it's normal rate would make digging a more valuable STP action over other skills. It's probably still more valuable in spite of the reduced STP of shovel. Also, Preaching is worth 3 million. Exorcismis worth 3 million.... Prayer? 6 million. Why? This is covered by the philosophy of following actual game time used for the action as opposed to accounting for cooldowns/infrequency of executing the skill. Refer to previous post for a more detailed explanation of that philosophy. Huge axe? 19 million. Carpentry? 5.5 million. Again covered by weapon skills being roughly 3 times slower than normal skills. This is covered by the philosophy of action time instead of infrequency of the skill. Unlike most other criticisms, this is fully identical to how Niarja valued these skills. In my previous post, I made this exact statement: "As of now, there are 3 skills (Cooking, HFC, and Beverages) that have multipliers that are arbitrary and not calculated." There are no other values "plucked from thin air" - at worst they are rounded. These are calculated modifiers to the difficulty system with the intent of balancing skilling, regardless of what skill it is, in regards to how much STP it gives per time. Imagine grinding prospecting to 100 versus grinding mining to 100. Prospecting gives no characteristics and a single difficulty skill. Meanwhile mining gives 3 characteristics (body strength, body stamina, soul strength) on top of a tertiary skill (pickaxe) which provides yet another characteristic (mind logic). Every action of mining would give 10 (11 if you count body double, which it does in the skill system) skills worth of STP while prospecting gives 1. The reason these multipliers are so dramatic is to balance out the actions. Niarja System Values (Prospecting vs Mining): Prospecting: 1.0 Total: 1.0 Mining: 1.0 Mining -> Body Strength: 1.0 Mining -> Body Strength -> Body: 1.0 Mining -> Body Stamina: 1.0 Mining -> Body Stamina -> Body: 1.0 Mining -> Soul Depth: 1.0 Mining -> Soul Depth -> Soul: 1.0 Pickaxe: 1.0 Pickaxe -> Misc Items: 1.0 Pickaxe -> Misc Items -> Mind Logic: 1.0 Pickaxe -> Misc Items -> Mind Logic -> Mind: 1.0 Total: 11.0 Sindusk System Values (Prospecting vs Mining): Prospecting: 1.0 Total: 1.0 Mining: 0.669 Mining -> Body Strength: 0.18 Mining -> Body Strength -> Body: 0.1 Mining -> Body Stamina: 0.18 Mining -> Body Stamina -> Body: 0.1 Mining -> Soul Depth: 0.18 Mining -> Soul Depth -> Soul: 0.1 Pickaxe: 0.36 Pickaxe -> Misc Items: 0.45 Pickaxe -> Misc Items -> Mind Logic: 0.18 Pickaxe -> Misc Items -> Mind Logic -> Mind: 0.1 Total: 2.599 Previously, every action of prospecting was equivalent to 1/11 of an action of mining. In current values, as a direct comparison, it's worth about 1/2.6. This is offset by the fact that it's unlikely to be within proper skilling difficulty for all 11 skills that mining provides. Mining is still better, but it's no longer overwhelmingly better. I want to further clarify the point that these values are not arbitrary but instead calculated in this post, since it's important to me that it's not seen as my opinion. Instead, they are a calculated response to how Wurm skilling works on a fundamental level. You had much more to say and I intend to respond to the rest of your post in a future reply. I just did not want to lengthen this post further. Some Closing Thoughts I'm happy to answer questions regarding the STP system used when I can. It might not be fast but I'll check back every now and then and answer when I can. That said, the reason values look strange is because they were not so easily compared until now. WurmNode does a great job displaying all the information without spreading it in different layers of the UI. When everything is on display side by side, it's easier to pick out what looks like a discrepancy or inaccuracy. With that said, I find it hard to argue that the rudimentary Niarja STP system was more accurate. It's well known that heavy HFC & mining grinders were always at the top. I was in the top 10 on NFI for a year after I stopped playing because I had over 90 mining and HFC. The way it valued player time was silly, and the goals behind the changes were to equalize it so it wouldn't matter if you grinded plate smithing, fine carpentry, mining, leatherworking, or coal-making. The amount of STP you gain by doing the action is roughly equivalent between all of them. The system you see was not done overnight. The multipliers you see here was after an immense amount of research and reverse engineering of multiple systems of the game. Combat, skills, action timers, etc. It lead to the discovery (or re-discovery) of the action timer overlap and was patched out of Wurm Online several weeks later. Again, this system is not perfect. It's an attempt at equalizing STP across all the different skills instead of allowing a few simplistic skills that provide far too much. If you'd like to see changes, I need hard evidence and real research into the reasons why my calculations are off. I can't really act on statements that simply assume "this looks off to me, it should be like this" without supporting material. If anyone is willing to invest their time to show calculations of skills that are being incorrectly weighted, I'm happy to admit I am wrong and will change values accordingly. I know this is a hard ask since I don't have all of my research written down to be used as a reference. The closest I can provide is my spreadsheet with some sporadic comments and tons of unorganized values that was utilized to create most of the system. STP Worksheet (Google Sheets)
  12. This poll seems... kind of over the top. It's not providing much useful information because it's too bloated. There's too many options for fairly simple questions. Additionally, the proposed solutions are unable to be described properly and lack context for the purposes of this poll. I would much rather see the proposed solutions replaced with perceived issues. Players are generally quite good at identifying problems, but lack foresight about how changes can affect tertiary systems. This is shown by some of these solutions causing multiple issues that cannot be properly expressed in the poll. An example of a more easily digestible format for this poll that I think would perform better and give meaningful results and statistics to analyze: NFI Character Have Set Don't Have Set SFI Character Have Set Don't Have Set Frequency of Private Hunts Very Active (once a month or more) Active (about once per 3 months) Infrequent (a few times per year) Inactive (never) Opinions of System Issues I want to earn a set, but don't feel I have the opportunity to do so Hunts are too time consuming Hunts are not enjoyable Sets are too rare etc. etc.
  13. Updated the original post with the new information of optional mods. Also linked directly to the /releases page of Tyoda fork instead of the main github page to reduce confusion.
  14. Give the enemies significantly reduced damage versus mountable combatants so the mounts don't get slaughtered immediately. If the mounts are able to be armored up (good for pve to finally have a use for horse armor) and tank damage from the encounter, they're actually useful. If they get killed quickly by comparison to their rider, then nobody will bring them into combat, main horse or otherwise. This should solve the real problem here: Reducing the mobility bonus of having a mount in combat (by injuring the mount), while simultaneously preventing players from walking back to their deed (by killing the mount outright). This will not upset the balance of other components of the game (namely pvp) since it's not a broad-stroke increase to the durability of mounts. Instead, it's directly targeted at the creatures for this specific content.
  15. This is an unfortunate combination of mechanics which is more likely to cause irritation rather than intrigue. Starting with archery, you will always need to deal with the melee monsters before you can even bother with archery. This leaves only a few viable strategies: If the melee creatures can be kited to the point where they remain aggressive, while being outside the range of archery, this is the most efficient way to approach the camp. If the melee creatures cannot be kited away from the archers, and the archers do significant damage, then you'll need to use a shield to block the arrows. If the melee creatures cannot be kited, but the archers are dealing insignificant damage, then you'll just ignore them and they aren't relevant to the encounter anyway. Once the melee are dealt with, then you'll be attacking the ranged. I can't see a design where the opposite occurs. On the topic of the creatures targeting mounts, this should be removed outright. Players will just leave their horses at the border. The differential between mount damage and player damage is immense. Mounts take multiple times more damage than players even with the best armor you can get. When they get injured, your movement speed is penalized harder than if you were on foot. So then, why be on a mount in the first place? Everyone will just leave their mounts at the door and enter the camp on foot. And if you're entering on foot, you're going to remain as close to the border as possible when fighting so you can leash the mobs if combat is going poorly. Unless the tick is equal for all players, it's likely players will just solo the camps anyways to get the most out of it. That's why players don't hunt in groups to begin with. Fight skill gains when 2 or more players are splitting it is significantly lower than if both were to solo hunt individually. Strong players will be able to cheese the camps through efficient kiting and a healing alt. If you want the camps to be approached anywhere near the way that's intended, these should give increased skill ticks that are independent of the amount of players that participated so that it's encouraged to group, since it definitely wont be required.