Welcome to Jester's Trek.
I'm your host, Jester. I've been an EVE Online player for about six years. One of my four mains is Ripard Teg, pictured at left. Sadly, I've succumbed to "bittervet" disease, but I'm wandering the New Eden landscape (and from time to time, the MMO landscape) in search of a cure.
You can follow along, if you want...

Sunday, October 2, 2011

Incompatible beliefs: TiDi

On 30 September, the Time Dilation blog and associated video came out.  Let's cover a semi-related good thing first, before I get to the meat of the content.  The major devblogs this week have included videos: this one, the CQ one, and the new cyno effect one.  This is a good thing.  Videos do a great job of highlighting new features, showing them off, and getting the player's attention.  Please keep this up, CCP!

Immediately after the "TiDi" devblog came out, Seleene pinged me on Twitter:
@RipardTeg ETA on your blog about this blog? I am saving myself until after the weekend. :)
I responded that it was also going to take me a few days because I was both in favor of TiDi and against TiDi.  Here's where I again demonstrate my ability to believe two incompatible things at once:
  1. I think time dilation is an outstanding idea that will result in a remarkable improvement to EVE Online.
  2. I think time dilation is a pointless exercise in futility and wasted dev time that will backfire massively.
Yes.  I really do simultaneously believe both of those things.

Let's cover them both, starting with why TiDi is a good thing.

The meta-game is a huge part of EVE Online.  Thousands and thousands of people started playing EVE because of things they heard about that were happening in-game.  Lots of times, this involves the enormous corp thefts that make such good reading.  Still, other times, it's all about the massive fleet battles.  The potential to get involved in these was one of the things that drew me personally to EVE.  I was really intrigued about the Influence Map when I first saw it, and wanted to get involved in that part of game-play.

That was about two years ago now.  I struggled for quite a long time to find and join an alliance that wanted to get involved in sov warfare.  Nobody seemed interested in it, and I couldn't understand why.  Why wouldn't people be interested in this?  Isn't this the EVE end-game?  Then I finally caught on with a sov-holding alliance myself... at which point, I learned.

Sov warfare is a sewer.  Nobody can explain to you how awful it is: you must experience it.  I was heavily involved in it for eighteen months and came to hate it more and more and more over that time.  Sov warfare is great for selling EVE Online, but the experience?  Oi.  Waiting hours for tower bashes, mindless following of orders for days at a time, alarm clock ops followed by boredom on a massive scale, getting hot-dropped, not getting hot-dropped (sometimes, this is worse -- a nice hot-drop at least relieves the monotony), alliance logistics and finances, recruiting, internal politics, external politics, training for fleet doctrines... it's truly horrific.

One of the worst things about sov warfare are the massive fights that invariably come when you finally drive an alliance that you're trying to push out of sov to their breaking point.  At that point, you get "one big fight" that usually settles matters decisively as both teams pull out all the stops to bring in everyone they can.  This results in hundreds of people in a single system, fighting it out.

Or trying to, anyway.

More often, what you're left with is a black screen which you get to stare at for a few hours as the game struggles to get you into the system.  Or, if you do succeed in getting in, you're rewarded by your guns spinning and spinning and spinning... only the target isn't taking any damage... and your ammo isn't decreasing.  Lag in an EVE Online fleet fight is a horror.  It truly is.  In null-sec battle reports, you inevitably get a sort of weather report to go with the BR to state how bad lag was for that fight.  ;-)  Lag in a really big fight is so bad that the systems around the lagged-out system also become lagged out.  In a null-sec sov-holding alliance, there are EVE player skills that you have to learn in order to know how to "manage lag" when it strikes.

Let me repeat that so it's clear: In order to survive in 0.0, you have to learn, develop, and practice a whole skill-set around managing deficiencies in the so-called "game" you are playing.

Remember: this shit is fun!

As a result, Time Dilation can only improve the EVE Online experience.  By making large fleet fights possible, manageable, and fun without requiring a lot of additional player skills to do it, and taking away the horrific side effects of these fights... that can only be a positive for the game.  Sov fights come down to "one big fight" and its associated lag today mostly because once you lose such a fight, it's almost impossible to get your team to sign up for another.  With TiDi in place, sov fights should naturally evolve toward requiring several large fights to push someone out of their sov.  TiDi won't be as good as instant gratification, but it will be a thousand percent better than what sov-holders are experiencing today.

...which is why TiDi is also going to be the worst piece of shit to be inflicted on null-sec, ever.

Trebor's Law states that "Fleet size will grow to fill available lag."  Today, it takes about 1200 to 1500 people in a system before EVE Online becomes near-unplayable in not only that system, and the surrounding constellation.  So, I think we can expect that TiDi will increase it to 1700 to 2000 people in a single system before doing the same thing.  TiDi solves the symptoms of lag, not the cause!  And as a result, it's going to encourage yet more lag.  It's a nice stop-gap, but it's going to backfire massively.

In the video, time slows down to 50% with just a few hundred Drakes on the field.  You might recall what null-sec is increasingly looking like lately.  Both sides of "The DCF" can put a few hundred Drakes on the field laughably easily.  Click on that Influence map I linked earlier.  Gaze in wonder at the swaths of red and yellow to east and west, and the pets and vassals they surround.  It's increasingly coming to the point where either alliance could put a few thousand Drakes on the field if they wished to... if the game supported it.

With TiDi, CCP is making it possible for them to do just that.  Who is going to be able to compete with that?  Nobody, that's who.  As I put it in a Tweet I sent to Seleene in reply to his asking when this blog post would come out:
@Seleene_EVE TiDi strikes me as just a lovely way to make sure the entire GSF/DRF blob gets on your fleet's KMs if you dare enter null.
Mittens is usually very smart, but he also has this amusing belief that null-sec changes like this one are going to draw ex-pat sov warriors like myself back.

Yeah, not so much.

9 comments:

  1. Before TiDi, supercapitals were effectively immune from retaliation, since a large enough supercap fleet would prevent reinforcements from loading grid, allowing them to pick off any threats at their leisure. (note that a 'large enough' supercap fleet is still an order of magnitude smaller than a sub-cap fleet 'large enough' to prevent grid loading)

    After TiDi, this tactic becomes considerably less reliable, especially if log-off timers are dilated as well.

    ReplyDelete
  2. It seems fairly obvious that the TiDi video in the devblog was not a demo of lag-induced TiDi on a reinforced node. Had he used such a node I would have expected to see a LOT more drakes.

    ReplyDelete
  3. I suspect that the demonstration video was running on rather old server hardware, if not on a stock computer, rather than on one of EVE's servers, so using it as a benchmark for how fleet fights will be like under time dilation may not be a good idea. It's possible to have a mostly lag free fleet fight involving 300 players on TQ as it is, after all.

    In the fleet-fight portion of the time dilation test on Singularity, there were approximately 350 ships on grid, plus the associated missile, drone and fighter spam. Since turret cycles did not properly work client-side, everyone had to manually cycle their guns, which also no doubt added to server lag. The node held up fine -- the devs were hoping for upwards of 500 people, so they may have even put the system on a dedicated node. So they had to manually invoke time dilation.

    From what I saw in that test, time dilation has huge potential in making big fleet fights of over 2000 players possible, assuming that the number of players the server can handle scales decently with the degree to which time dilated. Currently, a reinforced node can handle somewhere around 1200 players before it starts to freeze up. That number will probably increase as hardware improves and CCP optimizes its server-side hardware. But if slowing time by 50% allows 50% more players onto the grid, that means it would be possible for 1800 players to load grid and shoot someone, instead of the first 800 loading grid, the next 400 having to wait ten minutes, and everyone else simply staring at a black screen for an hour.

    The game was perfectly playable at half-speed, and was okay at quarter-speed. It obviously wasn't playable when the devs froze time completely, but the comments in fleet chat of players being able to see the missile coming to pod them made the wait well worth it.

    ReplyDelete
  4. Veritas' video was likely him running a local server on his machine or something. TiDi will not solve everything, but I can't see how it makes things worse. Right now, people still *try* to have 2000 vs. 2000 fights, they just suck. Clearly there needs to be an eventual solution to blobbing, but it is a hard design problem.

    ReplyDelete
  5. The 50% level was manually selected by the dev, not automatically chosen by the server as a result of a few hundred drakes. In previous discussions the devs said they don't think TiDi will go much beyond 10% slower than real-time.

    ReplyDelete
  6. To me it looked like that was a stand-alone sol node running on the developers own box (along with all the thin clients), thus the mighty UI lag and stop-start motion you see through the video.

    Given what appeared to be a hard limit on time dilation (0% for non-laggy, 100% for … half speed?), we should expect to be able to squeeze about 141% of the ships onto grid without blackscreen-inducing lag.

    Of course time dilation doesn't address session change issues, so people jumping into a whopping huge gate camp are still SOL.

    ReplyDelete
  7. Hey, quote me correctly... it's "Fleets expand to fill the lag available", coined in my CSM5 election manifesto.

    TiDi just buys CCP time -- and probably less than expected -- to fix the core issue with EVE fleet combat mechanics, which is that the power of a fleet is the number of ships raised to an exponent > 1. As long as it is always better to use one large fleet instead of two smaller ones, the blobs will grow.

    My guess is that TiDi will scale at about 1.5 - 1.7 power, which means that a TiDi factor of 10 gets you ~ 4x larger fleets.

    ReplyDelete
  8. @Trebor: sorry. ;-) Couldn't find the exact quote. I concur with your estimate, though. Feels about right.

    ReplyDelete
  9. I was strolling around your blog, and was wondering if you could bring up your current opinions on TiDi in a new post, perhaps a Junk Drawer?

    ReplyDelete

Note: Only a member of this blog may post a comment.