Player Features and Ideas Discussion

Both

1 more day until the release...curious what the feedback is going to be, and if this feature will be lost in all the new UI stuff, or enhanced a bit more by it.

Be sure to post feedback once you have tried it!

Probably the biggest problem with manual control is that there is no sense of constraints/boundaries b3cause there is no clear road, and roads have boundaries.

Idea:

What im suggesting is basically some sort of road which is only viewable on a pilots UI (it doesnt really exist and it is customizable by the pilot depending on his intensions and style of pvp.

A "holographic visual roadmap" will have to exist so that you know you are slowly steering your ship in the right direction (like a big truck on the highway or a bus is constantly checking if he is on course or not while he is slowly adjusting the direction)

Currently you have two main points of comparison, nose of the ship and the destination you are trying to point the nose toward, i think pilots will ultimately need another source of comparison (you need more than two points of comparison, the nose of the ship, the destination and everything in between).

Now that this feature has been in awhile, what are everybody's thoughts on it?
Has anyone been using it really, or was it just simply a "wow, look at how I can move my ship", then back to the same old same old?
Anybody find any novel uses for it?

The second great design decision CCP made is the model for determining how much math has to be done to check for collisions in space. They use a path projection based on your ship's bounding box and speed. Essentially, they use motion prediction (which itself is computationally cheap math) to figure out where your ship could potentially be in a number of ticks dependent on your speed and direction and ship size, then determine if that overlaps with anything else (other ship's predicted paths, structures, etc). This cuts down further the number of calculations needed to figure out who hits what. Instead of the million calculations the WoW servers have to do to figure out if a player makes it out a door, the Eve servers do a couple of geometric equations to determine if a player's ship is near enough to anything to interact with it (a station, another ship, etc). If not (which is 99% of the time) the server just updates the client with a position and state and goes on to serve other players.

Have you ever watched your ship try to auto-pilot around an obstacle, wobbling back and forth at low speed like a blind man looking for the door? That's the bounding box... it only projects forward a little at low speed, so your ship literally does not "see" that it has to turn 90 degrees and thrust away to get clear... it only sees a few feet, beyond that it can't predict.

Even the events that would need collision detection (which again generates load) are handled in ways that avoid it. For example docking in a station. Ever wonder why you just fly up to a station and request to dock, then you get towed in? You'd think it'd be simpler and more fun to just mark a blinking landing pad for everyone and let them fly in, or maybe have robot arms grab your ship as it goes by. The answer is that an automated docking sequence avoids needing collision detection entirely. Once the player is approved to dock, their ship is removed from space and they're docked.

You might also have noticed that ship attitude doesn't matter much except for looks. Even if the blasters on the port (left) side of your ship are facing away from the target, they can still fire and hit. Eve models account for this by putting extra copies of the weapons on either side of the ship, so it doesn't look like they're firing through their own hulls. Not having to take ship attitude and angle into account when figuring out anything also keeps calculation cost way down.

Combat in Eve is likewise kept computationally cheap. For any weapons fire event you calculate hit and damage based on the positions and speed of both ships, the stats of the weapon, skill levels of the player, any environmental items, any effects from modules, etc. and get a couple numbers out. It's essentially solving one medium complexity equation after which numbers go to the client. Just like a GM rolling dice.

Also in order to keep Eve's server and network load to a minimum, Eve doesn't update client state more than a couple times per second. That means slow networks can keep up with the traffic load, the Eve servers can keep up because the only have to update everyone's position a couple times a second, and client PCs get a break, too. While this would make for horrible play in Counter-strike, in Eve it works because the client and server can easily predict the game state and show what's happening in between ticks and they're rarely getting different answers. This is the major reason Eve is playable world wide, so it's a huge win for CCP.. a bigger market means more players.

So... back to the original comment. Hopefully if you read the above you'll see that trying to put basic flight sim mechanics into Eve won't work without a major re-write of the entire game. Sure, you can turn/move your ship with WASD now, but taking the step to allow ships to "line up" on other ships and fire with the press of a trigger would require the server (can't use the client, remember, because it's exploitable) to do calculations for ship attitude and most critically would require a near real-time response level from the server... which would require faster network updates than Eve can do. That right there kills the possibility of twitch anything in Eve.

Plus player input for anything besides ship attitude means a lot of calculations, too. Right now the game saves itself a lot of math by prediction of ship locations (simple equations, remember) which would be impossible if a player could change ship direction like a fighter maneuvering by pressing a button. Think about it - allowing a player to really fly a ship would require the client to receive/send network updates to the servers at 10x (at least) the current speed, and there would still be lag. The server would have to perform ship collision calculations at a proportionally increased rate (lots of 3d geometry), the number of locations a ship could be in in the future few seconds would be much greater and harder to predict (humans are unpredictable) requiring a lot more path to cover all possibilities, and all these calculations would have to be multiplied by the number of players online at a given time, many of whom are interacting with each other in the same system.

Also remember that the way Eve handles extreme load at the moment is TD... slowing down its update rate, giving the servers more time to predict future positions and game state. That's the opposite of what they'd have to do for fighter style combat, you want more updates, not less. The number of people who could be flying in a single system at once would be cut to 10% of the current number or less.



TL, DR; Eve's big design wins that let us have everyone on one shard and good play with slow networks directly oppose the changes needed for twitch gaming or manual flight. Changing this requires a complete re-write and more hardware than players would want to pay for. Effectively, it's impossible to do.

Watch this video then read below: http://youtu.be/4UieJcWI-YM (S.N.S Super Nerdy Slicer (ING) - recent in-game frigate fighting and maneuvering analysis using the double click mechanic we have known for years)

This manual piloting through clicking is a less efficient method, but necessary as you see how far he is necessarily zoomed out, and thus uses the center of the concentric circles to reference where to spam the clicking, and has to constantly reposition the camera. I would argue that implementing some of my OP would mitigate this to some degree and make the kind of PvP experienced in this awesome video, a bit more intuitive.

Even though this video is awesome, and the gameplay is quite intense (to our keenly trained eyes), it is still quite aesthetically unsatisfying, as you are essentially keeping a huge set of concentric circles (might as well be an atom at the center) at some range and in some relative motion to floating brackets.

By focusing on the types of modifications to the experience I mentioned in the OP, we can totally change the feel, immersiveness and translatability (making people actually want to do this) to a much broader audience, while still keeping the core gameplay mechanics. So none of this would be driving eve to a first person shooter, but rather aid in situational awareness and enhancing ones ability to use BFM - Basic Fighter Maneuvering (as he references it) to position your ship to optimize your automated weapons platforms.