|Paizo Pathfinder® Paizo Games|
|About Paizo Messageboards News Paizo Blog Help/FAQ|
A lot of folks talking about these issues so I thought I'd provide some clarification on what is happening that will make it a bit easier to understand what you're seeing, and how to report things to us as "bugs".
There are neither "teleport" nor "rollback" bugs in the game (that we know of :) ). What you are seeing is the manifestation of the server becoming overloaded and becoming desynchronized from the client. So you're seeing a symptom not the problem.
Here is what is happening.
The logical "server" is a large number of physical server units, each of which is responsible for running the game in a number of hexes. The servers keep track of every moving thing - all the characters and monsters and NPCs. They are also responsible for the inventory and local vaults for the characters in the hexes they are responsible for managing.
The servers cache a lot of the game state data in RAM, and they periodically write that data to persistent storage (hard drives, basically although things are actually a bit more complex than that). The more often they write out the game state, the more overhead they consume. So the programmers make choices about frequency of these writes vs. performance. During the time they're doing the writing you may see more things like framerate drops, characters skipping from point to point across the landscape, delays in logging in or accessing vaults or inventory, etc.
The client cannot be trusted - it is in the hands of people who would like to cheat by manipulating the client's data. So the client and the server have to be kept "in synch"; periodically the server tells the client what its actual data should be and the client resets itself to reflect that update from the server. The server never trusts that something has happened just because the client says it has. Your client only has to process what you are doing. The server has to do exactly that same work, but for every client. So the load on the server is the multiple of the number of clients connected to it.
As the number of entities in a hex increases, the load also increases on the physical server that is running that hex. As more monsters are spawned, and more characters enter, the load increases. As those entities interact, like in combat, the load increases. At a certain point, the server becomes overloaded and it cannot process data fast enough to keep up with the game world.
That is the point where Bad Things Start To Happen. This is not a binary state where all entities are working or no entities are working. Instead, some, but not all, of the interactions are being processed by the server. So different people might see different effects of the server load impact them differently.
The clients and the servers begin to get more and more "out of synch". The server has to send resynch data to the clients more and more often, which consumes more server capacity. Eventually, the server determines that a client has become unresponsive and it disconnects.
Right now we don't have a smooth disconnect process, and instead of just being dumped back to the login screen with some kind of error message, the clients enter some very strange conditions. Sometimes the character just falls through the ground, on an endless loop. Sometimes everything else in the world vanishes and the client is left alone in the geography with no other entities. Sometimes the system hangs and sometimes the system crashes with a crash report.
Once disconnected, people try to reconnect. But the server they're trying to log back into has overloaded (which is why they disconnected in the first place). The overloaded server might start to long in the client, and fail, putting the client back in one of the failure states described above. Sometimes the server thinks the client was not disconnected in the first place and that there was just some network interruption and it tries to restore the client to a previous known good state. And sometimes it just crashes out again. These are all symptoms of the server being overloaded.
Sometimes the server tries to log you in, determines that it is overloaded, and tries to hand you off to another server. That server might have the capacity to accept and process the login and it will do so. However, the hexes that new server is responsible for managing are not the hexes that your former server was managing. The new server picks one of its hexes and spawns you into the game - you appear to have "teleported" across the map.
Sometimes after you are reconnected, the server determines that it can't trust the data you sent it while it was overloaded. In those cases, it dumps the transactions of your actions it was holding in RAM back to the last point where it knows it was safe to write to disk. When that happens your character is essentially "rolled back" to that previous state. This can result in Very Bad Things like item duplications and coin duplications and this is a classic attack vector against MMOs which we're well aware of.
We are attacking these problems in two directions.
First, the team is working to increase the server capacity. Basically we're finding ways to make the system work more efficiently so that the physical servers can handle more entities before they get impacted. We have a theoretical idea of how much load they can handle, and we're trying to find and fix problems to get closer and closer to that ideal.
Second, we're talking about how to have the client "fail more gracefully". The best, worst case scenario is that when the system can't handle the load it simply returns you to the login screen with an error message and until that physical server has had its load reduced (by other characters leaving the hexes it is managing and monsters despawning) you'll be unable to log in again. Yeah, lots of obvious problems we know all about, but perhaps better than randomly being teleported across the map.
These two issues are linked by programmer time. There are a lot of either/or choices here and we're trying to map out the best way to move forward.
This is the whole point of the Alpha Stress Test - to see how the server behaves under real world loads, to see how it fails, and then to come up with plans on how to address those failure modes and reduce those loads. So while it sucks that folks are falling through the sky, having to re-train Feats and re-earn Achievements, and be teleported all over the map, we are gathering really important data that will help us fix these issues.
New and Unusual Bugs: After the foregoing, you can see why it doesn't help us when you submit a bug about these basic failure modes. However, there are potentially things you can help us with that can look like these modes, but aren't.
If you find an actual hole in the terrain - a place that you can consistently walk to and fall through the ground - we want to know about that. There should not be any such places but if there are, we need to find them. (Falling "into a hole", a dip in the ground that you don't fall through but just can't walk or jump out of is important to know about too.)
If you are able to cause the system to reproducibly duplicate coin or an item, we want to know about that. This is more than "I got a duplicate". It means "if you follow these precise steps, x, y, z, etc. a duplicate is created." The former is probably a desynch problem. The latter is a bug.
If you are seeing monsters stand around and not attack you, no monsters, very laggy response to any input, repeatedly falling through the world, or teleports, the servers are stressed and no bug you report under those conditions will be particularly useful.
Our best advice if you are experiencing these desynch issues, is to log out and play later. Running out of the hex you are in may help, but you may not be able to tell for some time if you have actually gotten to a less-loaded server or if you're going to experience what appears to be a rollback.
Just FYI, in the industry, "AAA" means "a budget at the top end of what similar games spend on development. It doesn't have anything to do with quality ratings. There are many "failed" AAA game that were just not good.
So the "top end" for MMOs today is between $100 million and $200 million, compared with Pathfinder Online at about $5 million.
Continuing my series on how to rapidly increase a Role to 8 ranks, here's the Rogue info.
As before I used a developer (dev) command to give myself a lot of XP. That was the only dev command I used. I did not buy anything from any auction house.
I did the Tutorial and slotted the Tutorial Feats.
I got a Short Bow from the Short Bow supplier at the Thieves Guild.
The Rogue Role focuses on Rogue Kit Implement, Power, Stealth, Hit Points and Reflex Bonus, plus a little Perception. I took levels in those as necessary to advance the Rogue Roles. Occasionally I also took Base Attack Bonus Feats as required.
I killed 250 units of anything that moved with the Short Bow to get 15 Subterfuge points. Along the way I spent XP on Rogue Role Feats. That got me to Rogue 5.
I also got Heavy Armor Proficiency and Swashbuckler, slotted Swashbuckler, and equipped Hide & Steel Banded (from loot) as soon as I got some. This made the killin' much easier.
I got a Short Sword as loot, and switched to the Short Sword. I killed 100 units of anything that moved with the Short Sword to get 10 Subterfuge Points (total of 25). That got me to Rogue 7.
Killin' things took less than 3 hours.
At that point I needed 11 Con to complete the build. I trained Bow Weapon Proficiency, and 9 Short Bow Feats. Then I trained Ranged Attack Bonus 1, and trained the 9 Feats to Rank 2. Trained Ranged Attack Bonus 2, and trained the 9 Feats to Rank 3.
I trained Light Blade Weapon Proficiency, and 8 Short Sword Feats. Then I trained Light Melee Attack Bonus 1, trained the 8 Feats to Rank 2, trained Light Melee Attack Bonus 2, trained the 8 Feats to Rank 3.
Trained Base Attack Bonus 3, trained Stealth 8, and had Rogue 8.
Overall I killed 350 monsters or NPCs.
Hit Points: 560
324 copper pieces recovered as loot.
6 Recipes including two +1 and +3 Steel Plate (UNGH!)
I also got the very first Spell I have ever recovered as loot. Bless.
2 weeks of heavy play (whatever that means) limits your ability to buy a character with a lot of unspent XP, and deploy it quickly into a fluid situation. It substantially limits your ability to "buy win".
If you don't see why limiting that potential is "good business sense", I don't really have a foundation to even have a conversation about the system.
The amount of time required to clear an achievement gate increases like a hockey stick. So does the amount of XP required to learn higher ranks of Feats.
These tests show that you can make a reasonably trained character with a month or so of game time, and a reasonable amount of game play during that time. My opinion is that average gamers would not see the time they spent playing the game in that manner as "a grind".
But those are not the characters people will sell. Anyone can make those characters so they'll have a low value. The valuable characters will be the 1-year plus characters. Those are the characters that will be getting into the upper reaches of the game mechanic. To reach those upper reaches will require a reasonable amount of game play, but "reasonable" for a year old plus character is not going to be a Saturday afternoon. To clear a gating system reasonable, but meaningful, to the depth of XP those characters will have without effort would be very poor design.
The average people who actually play one year old plus characters will not feel those gates are grinds in the same way they don't feel the earlier gates are grinds. But someone who wants to purchase an untrained one year old plus character and get the advantages of a year of banked XP will not be able to quickly deploy that character into a war zone as a shock trooper, capture a degraded logistics chain, or assume a mechanical leadership position in a large Settlement, and that will make those characters less valuable to speculators.
OK, I got a Cleric to Level 8.
Like the Fighter build posted elsewhere, the purpose of this build is not to make a necessarily "fun" character, just to get one to Level 8 as quickly as possible.
I will note that there appears to be a bug in that the system is not enforcing the 11 Constitution gate for the 8th level Role achievement for Cleric. We'll look into that.
That's a good thing because right now, using just the Dreadnaught, Temple and Seminary, you can't get 1 point of Constitution. I've mentioned that to Lee and we'll talk about it internally.
A new build will likely be deployed this week which may fix that bug, and thus make this build fail. :( Don't know the exact status of that at this point. My guess is that if you have reached the 8th level Role achievement and the Con gate is reimposed, you'll keep it, but I am not sure about that either.
What I Did
As in the Fighter build, I used Dev commands to give myself a huge amount of XP. That was the only Dev command I used - everything else this character achieved was done by me playing just like you would play.
I purchased Pot Steel armor from the Auction House in Marchmont, but ended up with several sets of Heavy Armor shortly after, so it was a unnecessary purchase. I did not use Dev commands to create any items.
Clerics get Divine Achievement Points by killing things with their Focus, and by killing Skeletons, Ghouls and Cultists. Nobody in the office was sure if all Razimirans counted as Cultists and I didn't bother to try and independently confirm that, and I've never seen a Ghoul. :)
So I killed a lot of Goblins and Skeletons.
It took perhaps a total of 7 hours to do all the killin'. I say "perhaps" because I lost two to three hours of kills in the crash last night and had to re-earn them this morning and that put me in a foul mood and I didn't bother to precisely calculate how much time I had to re-do.
I did the Tutorial, and slotted the Tutorial Feats.
At Level 2, I trained Touch of Darkness 1, and slotted it with the free Focus you get from the NPC next to the Temple trainer.
I then started killing Tutorial Goblins (the Goblins right outside the starting NPC Settlements) to generate Divine points via Focus Expert.
The Cleric trains Focus Weapon Proficiency, Power, Hit Points, Divine Attack Bonus, Willpower Bonus, Base Attack Bonus, and Fortitude Bonus as its Role Feats. I trained these as fast as my Divine points would allow.
After reaching Cleric 4, I trained Heavy Armor Proficiency, Crusader and the Glory Domain, and equipped Pot Steel Plate. I slotted Glory and Crusader. (I was tired of dying and wanted to speed up how fast I could kill goblins).
I killed 250 Goblins to enable me to train all Feats up to Power 10.
That gave me 15 points of Divine.
I trained Hit Points 5 out of sequence because I wanted more Hit Points.
From the middle of Cleric 4, after getting Power 10, I switched to killing Skeletons. There are several active Skeleton escalations running, and I found one and killed most Skeletons there solo. Killing 20 Skeletons allowed me to get 3 points of Divine so I could clear the Cleric 5 gate.
At this time I trained Destructive Smite 1 and Holy Lance 1 because Touch of Darkness does Negative Energy damage and the Skeletons are immune to that. I never actually slotted or used Destructive Smite and killed all the Skeletons with Holy Lance.
I killed up to 50 Skeletons to get the 6 points of Divine I needed to train the Feats for Cleric 6.
I then killed up to 100 Skeletons. Lisa joined me for the end of the Skeleton hunt and helped me dispatch the final 22 Skeletons, mostly by taking on the Red and Yellow skeletons in the escalation mobs with her uber character of uberness. I could have finished this solo, it would have just taken me another hour or so of careful target selection and I wanted to go to lunch.
I trained the rest of the basic Cleric Role Feats to the max of my Divine points, then I started training Orisons. There are 23 Orisons (there should be 24, but we found one that had been omitted from the Trainers - so that's 1 Bug for Ryan!)
I trained them to Level 3 to generate Wisdom. At some point in this process I qualified for Cleric 7. I actually ended up with 12 points of Wisdom so I could have trained fewer of them to 3rd level and saved some XP. The 3rd rank of the Orisons are 1,881 XP each, so training fewer of them would have substantially cut the time required for this build. I think, but am not exactly sure, that I overspent by 9,405 XP (there are some short-term discrepancies between my spreadsheet data and the live game data, but they're close).
I then went out and started killing anything that moved. I killed a total of 500 enemies (Goblins and Skeletons included) to get Focus Expert 6, which gave me enough Divine points to train Power 16 and receive Cleric 8.
Hit Points: 645
XP: 98,395 (40.99 days - would have been 33.9 days if I'd skipped all Orisons to level 3)
Base Attack Bonus 3
All Orisons at Level 3
496 copper (less 100 copper for Pot Steel armor purchased in Auction House)
Loot of Note:
17 recipes including seven +1 and one +2.
Also, I have to say that Attributes are core to the d20 experience from which Pathfinder descends. Not having them would be a bad break with the tabletop game. Making them present but irrelevant would be a bad break with the tabletop game. Making them work like the tabletop game would result in Mandatory Builds, a pressure in MMOs overwhelmingly greater than anything you're likely to see on the tabletop. So using them as gates is a compromise with a good mechanical rationale (as described in my previous post).
Audoucet wants to park characters on an account, let them accumulate XP for a long time, then rapidly train them to be uber as needed. He doesn't want to spend any time at all engaging with the content in the game except for killing other players (apparently).
It looks like an investment strategy to me, not a gaming strategy.
So, how do you find a Dreadnaught 3 trainer so you an advance beyond level 8?
War of Towers.
PC Settlements will get increased training by holding Towers. To get the trainers to advance to Level 3, a Settlement needs to hold 8 towers. Since there are 6 towers around each Settlement, that means they have to hold the Home Ground, plus go looking for 2 more.
When the Settlements succeed in holding towers and their Trainers level up, their characters can then pursue training beyond what the NPC Settlements can provide.
I will do a Cleric build today.
My Fighter build was designed to get the character to level as fast as possible, not to necessarily be a "fun" character to play on a daily basis. The Dreadnaught 2 trainer in the NPC Settlements doesn't know how to teach higher ranks of Base Attack Bonus so I would need to seek out a higher level trainer if I wished to continue to advance (and kill a lot more goblins).
I only used 2 dev commands. I gave myself a lot of XP, and at one point I added a nonmagical Hunter's Longbow to my inventory to save myself a trip to Thornkeep, or a bit more killing of goblins for random loot.
Hitting the Martial Achievement goals took about 4 hours of play and during that time I was working on other things as well not paying 100% attention to the goblin killing. It was pretty "casual".
I spent 29.21 days of XP. Since I reached all the Achievement goals in 4 hours, I think its pretty reasonable to say that at least for this build, the amount of time you would spend pursuing Achievements compared to the amount of time you could spend doing self-directed things is pretty trivial.
What I did:
I trained all my Feats at the Fighter College and the Dreadnaught School in a starting NPC Settlement.
The only things I killed were the Tutorial goblins right on the edge of town.
Heavy Blade Expert 4 (Steel Greatsword): Killed 100 goblins (10 Martial points)
Goblin Slayer 5: Killed 270 goblins (15 Adventure points)
I did the tutorial and took the 3 Tutorial Feats.
At this point I had 22 Martial points (Heavy Weapon 4, Longbow 4, Hammer 2).
At this point I hit the Strength gate. I trained all the Weapon Attack Feats for Heavy Blade and Hammers. I also took all the Master of Opportunity Feats. This cleared the Strength gate.
At this point I hit the Constitution gate. I trained the Armor Feats, the Shield Feats, and the various Feats like Bravery, Bulwark, etc. This cleared the Constitution gate.
Feats (final ranks):
Base Attack Bonus 4
Bull Rush 2
Master of Opportunity: Suffer 2
Heavy Armor Proficiency
Medium Armor Proficiency
Trophy Charm Implement Proficiency 1
Heavy Meele Attack Bonus 2
Hammer Weapon Proficiency 1
Shield Weapon Proficiency 1
I never equipped armor. All goblins were killed with the 2 starting Feats you get with each weapon.
I slotted Toughness with the Tutorial.
No other Feats slotted.
Strength 11, Constitution 11
620 Hit Points
Notable Loot (lots of loot not listed):
270 copper, plus a bit I spent on Greatswords in the Auction House
Recipes for Pot Steel Plate, Quiet Iron Shirt, Steel Ingot +2, Warm Hat, Weak Antiseptic Extract.
I got as loot all the weapons I used except the Greatsword, but I ended up with 2 in inventory so if I had just been patient I wouldn't have needed to go shopping, and the Hunter's Longbow I conjured with dev commands.
We chose Unity because it was the best available client-side solution. The asset store was not a factor.
For the most part, the stuff in the asset store isn't going to work with our tech without so much re-work that we would end up spending about as much time configuring it to work in our toolchain as making it ourselves. We were able to get a large package of weapons which we're using rather than spend time modelling "Pathfinder Specific" stuff (although we'll get to that eventually).
We are, in addition to using Unity, using an off-the-shelf animation package (which we had to integrate with Unity because Unity's built-in system is too slow for an MMO), we're using Havok Physics, we're using Grome, we're using an off-the-shelf content management system for the website, and an off-the-shell forum system that integrates with that CMS.
The two parts of the system where we have invested the most time are the art assets and the server and networking code. Those are essentially "the hard parts" of making an MMO.
We have also spent a lot of time working with Unity itself to determine how to maximize its visual performance. There's a lot of under-the-hood tech in Unity that has to be carefully tweaked and configured and we're using it in a nonstandard way. Our objective is to find a mixture of good quality and a large number of moving entities (usually games are either/or). We have a full-time staff position called a Tech Artist who works almost exclusively on that kind of thing.
We are beginning the Alpha Stress Test today. Over 9,000 emails are going to be delivered to everyone who has a valid goblinworks.com account over the next 24 hours. If you have an Alpha, Alpha invite, Early, Explorer or Open Enrollment account you'll be able to start playing Pathfinder Online TODAY.
To help control the load on the servers, we are staging the notification of the Stress Test and I'm asking that you not log in until you receive your notice if you are not already playing the Alpha.
This should be an awesome weekend to play if you've been waiting for a reason to try the game. There will be lots and lots of new players learning how the systems work and starting their adventures. We can all help them by answering questions they'll have and giving them some tips and pointers as they figure out the game. Making this first weekend of the Stress Test Community Friendly is a big goal of the Goblinworks staff!
The most important thing for PvP in Pathfinder Online is that it be meaningful.
To make it meaningful we need to have things to do other than kill each other.
The critical building block is the economy. The biggest allocation of the effort we have invested into this game has gone into its economic system. That's why there are so many outfits, weapons, implements, etc. and so much richness to the crafting system.
The economy has to exist before there can be meaningful Settlements, but Settlements are the next area where we've spent effort. That's why there are so many structures and so much work has been done on creating modular systems that make it possible to make a lot of different Settlements.
The actual "killing of each other" then becomes a good design sink which is why we're now spending time on things like War of Towers and on targeting and looting.
And of course to make all that work in a huge world that can support thousands of simultaneous players.
We may have a delay with this blog being posted so here is the text:
This update is mostly focused on fixing bugs and improving system stability but there are a few new features of note.
We have been working on many of the animations and animation systems and preparing a major update of the animation tech. That update arrives today. Animations will now appear more fluid and lifelike and you will see fewer strange animation behaviors. Weapons will be attached correctly to hands. We have built these animation suites using the male human avatar and that model will show the most improvement. The other racial avatars and the female models are using the new animation suites but they have been retargeted from the male human model and have not all been hand-adjusted for maximum improvement. That is work that we'll be doing incrementally over time.
* Various player movement animation improvements.
* Selecting different races and genders in character selection has less of a delay.Major overhaul to animations, which fixes much of the combat timing for feats (reducing the number of dropped animations).
* If a character has no target and that character receives damage, the source of that damage is automatically targeted
* Added missing war wizard facility and trainer to Wizard/Fighter template.
* Adjusted reinforcement percentages to make Escalations more manageable at 100% strength.
* Update the auto-generated feats that show up when first selecting a weapon.
* Removed energy resistance from heavy armor. This is the first stage of reducing the desirability of heavy armor for players of lighter-armored roles, as the removal of speed penalties left it way better than intended at low level. Expect further adjustments to armor balance as a few more tech pieces (like spell penalty and encumbrance) are implemented soon. Once those are in, we can get a more updated impression of what else armor needs to have a good continuum of desirability.
* Doubled the (relative) drop chance for common refining recipes at +1 to +3 (i.e., when you get a recipe drop for a refining skill, it's now slightly more likely to be a common recipe and slightly less likely to be an uncommon recipe).
@Loko Loki - the problem with the suggestion of making the act of crafting "more interesting" is that it does not scale.
If you want to be a meaningful economic participant, that is, you intend to make logistics (and crafting is a part of logistics) your thing, you are not going to make one sword. You are going to make thousands of swords. Multiply that by all the items in the database.
In theory spending time to make an item "different" sounds cool. In practice, it turns into a repetitive, easily macro'd or bot'd process that robs the system of value and is despised by those who are forced to use it.
Over time we will iterate on the crafting system and add more interesting things to it. We are not going to add "minigames" or anything that requires the player to interact with the system continuously in order for it to function, but we might add things like occasional work stoppages that require a player intervention to resolve.
The "meaningful choices" in crafting are legion. You have to decide what to craft, where. You have to decide what to pay for the inputs (recipes, raw materials), and how to time your crafting jobs. Once a job is finished you'll have to decide where to sell the output and how to get it to that market.
Just within the question of how to craft an item you have to decide which materials to use - what combination of refined resources vs. scavenged loot, what "+" values to use, etc. Within a given category of item you have choices: Steel, cold iron, silvered, adamantine, etc.
what else are you going to do on your "main" crafter besides manage your queues every few hours at best?
The people in EVE who spend their time focused on economic concerns spend their time searching for pricing arbitrage - looking for places to buy low and sell high.
They also spend their time trying to get as close as possible to total production efficiency. Even slicing tiny percentages off of their costs can give them a pricing advantage in the market that allows them to squeeze out the competition. Those advantages sometimes come from things they can do to their characters like training certain abilities, but most of the time they come from figuring out how to reduce the total cost of the inputs to their economic engine - and time is a big part of that cost structure so they'll be constantly trying to figure out how to increase the rate of production of a material, decrease the time it takes to get that material from place to place, etc.
Because warfare between corporations shifts locations and tactics continuously there is no stead-state in the economy. The thing "everyone needed" last week might fall out of demand, and something new will take its place. So the economic masters are never able to just reach a point of "perfect economic capability" and go on autopilot. They are continuously updating their information and using those updates to change their logistics chains.
In other words, they're playing a game above the combat mechanic. A more holistic game that treats the people fighting as just a small part of the whole. These masters of the economy might spend 10 hours a day "playing the game", and never undock from a station or fly from place to place, or do anything with their ships.
We want that in our game too.
I agree with Doc 100%. Most people with Alpha access who have logged in have probably not spent any material time in the game. Most of them just wanted to see it and take a look around. They're not interested in putting in any serious time because it will all be wiped out.
There's a small number of people who are "playing the Alpha" just because they enjoy it. That's cool. A small number of using the Alpha to test strategies they intend to use in the live game, that's cool too. Every week due to Alpha invites being distributed we welcome a number of new players to the Alpha and they're likely showing up for an hour or two to check it out. We typically send those welcome messages on Thursday or Friday so that has an impact on when those people give the Alpha a spin.
When we start the Stress Test, we'll multiply the number of people who have access by 400%. Population density should be markedly different for a while afterward, but those people won't stay logged in either for the exact same reasons.
Low population at this stage is not concerning to me.
Checking right now:
Apprentice Charged Staff: 15c-24c (lots of auctions)
Right now Auctions last 48 hours. Peak use is going to be on the weekend. So auctions will likely start expiring on Monday and be mostly gone by Tuesday, and won't start to be refreshed until Friday or Saturday.
Also the vast expansion of the map has diluted player activity so people are not concentrating all their Auctions in any one place. This will start to fix itself as the number of players increases but it is likely that a "full market" will appear in one location and concentrate most activity on the server. Until the big market-makers decide where they want to set up shop things on the ground may be a little sparse.
We don't want that to be in Thornkeep either so expect the Thornkeep auction house to either shut its doors or be limited in some other ways.
I think that for the next several iterations the only thing that will be affected by selecting which character gains XP is the XP gain itself.
Eventually I want to go to a system where characters that are not gaining XP also cease active crafting jobs and cancel auctions and want to buy orders.
They probably will not be able to log in, either.
Then we'll think about how to let people buy back in to some of those services at reduced prices via MTX, but that will be dependent on getting the MTX system built.
This is all separate from the issue of having some kind of "trial" account for new players.
Existing accounts should not think of inactive characters as freely playable alts.
If he has a 64 bit version of Windows 7, here is my advice:
1: Verify that the drivers being used are the most current:
2: Verify that the client is running on a partition with a meaningful amount of free space (more than 1GB)
3: Try running the program as Administrator. Find the Pathfinder.exe app, right click, select "Run as administrator". (If this fixes the problem he has some permissions issue with his directory structure).
4: Attempt to run the game in Fast graphics mode:
Make a new shortcut to Pathfinder. (e.g. drag from start menu to desktop)
Unity is not free, btw. Nor is Havok AI, or Granny, or 3DSMax, or Adobe Creative Cloud, etc. We are a Source Code licensee and that's quite expensive. We are not taking the "cheap road" on anything. What we are doing is keeping our staff as small as we can, because salary costs are the biggest part of any software development project. Smaller staff means smaller feature set. You are not seeing the effects of being stingy on spending. You're seeing the effects of being very small in size.
One of the reasons I think we are seeing some of this discussion about gating is that we have never operated the game in the long-term play mode.
We had several short tests with lots of restarts so people played for a couple of days or a week then got restarted. Those short term characters were always gaining new Feats because starter Feats are cheap.
Then we had a couple of tests where characters earned enormous amounts of XP and we removed all the gates so players never had a sense for how long it would take to earn what they got when they were supercharged.
Now we're actually operating the game in its intended mode. We are seeing characters starting to hit 7 and 8 levels in Roles because they've been played for 3 weeks or so. But we're right at the point where the rate of advancement also starts to slow down. Up to this point there's been a rapid accumulation of Feats along a Role's training path but as the characters age into their first month, that starts to become less and less frequent. For the first time players are going to have a sense for what it means to know that if they want that next incremental upgrade they'll have to wait a material amount of time for XP to accumulate or they can go train something else quickly. Now we'll start to see players raising concerns about how rapidly they're gaining XP, not how onerous it is to kill monsters.
Anyway, as I understand from your post, ramping up Character-progression does not necessarily lead to better player retention?
If you don't reward new players with cookies often and quickly they quit really fast.
That's why our character power graph is not a straight line but is instead a curved line. You'll get most of the benefits of a Role within the first 6 months. New characters get significantly better almost continuously.
Psychologically that's how we get you to want to engage quickly and passionately. Once you've had a chance to explore the game and start making social connections and discovering long term goals your incremental character power increases become less significant in retention.
@Pigtails - having an absolute time required to reach certain points means we have total control over when those points are reached and can't be "surprised" by players who find creative ways to overcome in-game challenges faster than we thought possible.
Also, as I've said before, you should be time-gated much more often than you are achievement-gated. The achievements should be won as a natural part of playing the game at a moderate level of engagement. Most of the time you should be thinking "how long will it be before I have enough XP to buy that Feat I want", not "what do I have to do to complete this Achievement so I can get the Feat I want".
Having absolute control over character advancement enables us to roll out features and to iterate on existing features at a rate we can sustain with the resources we have available. Of all the innovative things CCP did with EVE, I think it is the most material innovation they made that created their long-term success.
I do not want to train miscellaneous skills that have nothing to do with my role as a wizard in order to raise attributes which will then allow me to progress.
Confirmed with Lee and Stephen this morning that the above quote is the intent of the system and when the fix to the ability score gate is applied you will be able to achieve the necessary ability score increases simply by training things from the two (or three) trainers for your chosen Role.
Let me reiterate to be clear: The design objective is that if you want to be a better Wizard, all you have to do is learn Feats from Wizard trainers.
(Lee notes that there may be an issue with Constitution right now but that's something that we know about and intend to fix, and that training Hit Points is usually a solution if you find yourself in a dead end on Constitution.)
This is not a change and has always been the design intent. Perhaps we could have been more clear about this so that there wasn't any confusion and for that I apologize.
I consider EVE's system to be flawed. The fact that you can just pay for training, logging in increasingly infrequently as you train skills that have longer and longer windows, and doing nothing when you conduct those periodic logins to reset your skill queue is, in my opinion, bad for the game.
In fact, the reason that EVE did not even have a skill queue until Apocrypha was because the designers understood that forcing players to log in regularly was a necessity to keep them engaged with the game. The compromise for the current skill queue design was hard fought and many people still think it was a mistake.
The risk is that players who are paying, but not truly engaged, are fragile. They're more likely to quit than someone who has something meaningful to do in the game on a regular basis. It's much easier to say to yourself when you look at your monthly credit card bill "why am I paying this $15 for doing nothing?"
Infrequent logins also mean that player's clients become woefully out of date very quickly, which can generate all sorts of customer service headaches. It also means that if your credit card fails to bill for some reason and you ignore emails about the problem (most such emails are ignored) you may in fact not be getting the training you thought you were getting and when you do log in, finally, and discover that due to a payment failure you missed months of skill points, it's pretty easy to imagine that you might just quit right there on the spot.
I believe that you must be playing the game to be a player of the game. I believe in having a system which reinforces that design philosophy. We've been very up front about this from day one - you do not become more powerful simply through the passage of time. You become more powerful through succeeding in doing meaningful things in game, AND the passage of time.
I shouldn't have to say it but apparently I do. You're looking at the minimum viable product for the Achievement feature. It does the minimum required to be viable. In a year we'll have a more robust system. In three years we'll have a system so robust you couldn't imagine the game without it. But today, it's just very simple. Complaining about the design based on the current minimum function is poor criticism technique.
You don't want to do anything but let time pass to become more powerful? That's a legit critique. I disagree, so you lose that argument. But I respect you for making the argument. But saying that because the tiny handful of Achievements we've implemented are boring or grindy means the design paradigm is flawed doesn't move me very much.
You may notice that the server goes off-line at 9am (Pacific) starting later this week. We do not have a notification system built yet, so you will get no in-game notice when this happens, you'll just be disconnected, won't be able to reconnect, and then you'll find the server lets you back in when it has returned to service.
We are going to try and have a staff member in game at about 8:45am (Pacific) to remind people of what is happening but of course many people will not see that message. So do your fellow Alpha Testers a solid and remind them about scheduled downtime when they get back into the game and start asking why the server crashed. :)
(And yes, an in-game notification is something that is on the to-do list)
I'd be happy to spend more time interacting with the community on reddit, tbh. The exposure would be great. Incubating here on the paizo.com forums has been wonderful but we need to start to push out beyond the group of people who are already deeply vested in Pathfinder and talk to MMO enthusiasts in general.