NA relic weapon

[Elexia]

So if anything, NA/EU relic owners are likely to be better players than japanese players! Go us!

(Notice how i turned something bad into something cool? Wasn't that awesome?)

And yet most NA shouts "THM/BLMx3+ needed!" for everything.

so:

Now as I said, this doesn't make the fight impossible for us, but it does raise the amount of skill required for us to beat it.

No.

es, sucky players will make you wipe to ifrit, but the limit on what defines a "too sucky player" is stricter for people outside of japan.

No.

A player in japan doesn't need to be as good at dodging and predicting as a player

Yes they do.

So if anything, NA/EU relic owners are likely to be better players than japanese players! Go us!

Once again, no.

[Mirage]

Mirage is wrong.

Please explain how the data travels from the server to europe in the same 70 ms that it needs to reach japanese players. If you can't, I'm sorry, you're wrong.

Let me take an example to clear up why it is impossible that japanese players are not experiencing less lag:

First, let's assume that there is a lot of inherent lag in the server. Most people seem to agree with this, even Yoshida has said this.

I don't know exactly how much this inherent server lag is, but it doesn't really matter. I'll use 150 ms for now. There, now every player in the game has a base of 150 ms lag.

Now let's say the server makes Ultramonster uses one of his fancy AoE attacks This particular attack has a charge time of 1 seconds. Because of the server's inherent lag that we've put at 150 ms, it will take 150 ms from the server decides that ifrit will use this attack until this data exits the server and enters the internet.

This data is now underway to the clients in this particular instance. after 70 ms, it reaches a japanese client. The client runs at 60 fps and is set to a buffer of 3 pre-rendered frames, equaling a local client input lag of 50 ms.

This adds up to a total lag of 270 ms, which means that 270 ms after the server initiated the attack, the japanese player sees the animation start. The japanese player is pretty good, and manages to interpret this data and respond to it within 250 ms. The signal now travels back to the server, and uses the same 270 ms to get back. All this added together comes to a total of 790 ms. Barely within the time limit of 1 second (1000 ms). The japanese player is able to evade the attack with a margin of 210 ms, and felt that that was a close call.

Now we move to a player from the Netherlands. A central location in Europe's internet infrastructure.

The server still has 150 ms of inherent lag, but 70 seconds after the data was sent from the server, it still has not gotten out of Japan. Finally, after 250 ms, the Dutch client receives the data. His client also has 3 frames buffered at 60 fps and therefore gets 50 additional ms of input lag there. we're now up to 450 ms lag. The dutch player is as good as the japanese players, and reacts in 250 ms. His client now sends his action back to the server, and the server receives this 450 ms later, 1150 ms after the attack was initiated by the server, and 150 ms after the time limit was over. He is hit by the Ultramonster's attack, and takes 1500 points of damage.

To counteract this, SE decides "hey let's increase the charge time for that attack to 1.5 seconds instead of 1 second". Result: The attack is now dodgeable by the good dutch player by a margin of 350 ms. The japanese player can now dodge it with a margin of 770 ms.

Conclusion: The japanese player has roughly twice as much time to react to an attack than the dutch player. Also known as "much easier". You can't avoid this unless you want the server to add artificial lag to japanese clients only to make it an even playing field for everyone, and that is never going to happen.



So what can SE do to reduce lag? They can do two things:

1. Reduce the lag inherent to the server. If they manage to get this down from our "assumed" 150 ms to 10 ms, they'll reduce the lag by 140 ms for all players all over the world.

2. Create a client that can supply a stable 60 fps frame rate with 0 pre-rendered frames buffered, reducing input lag on the client side by 50 ms.

The total reduction from this is 190 ms. Apply this to the previous numbers, and we get that the Japanese player will get a lag of 80 ms, while the dutch player will get a lag of 250 ms. If we assume the same 1 second charge time for Ultramonster's AoE, the japanese player will now have a total reaction window of 920 ms, and the dutch player will have a total reaction window of 750 ms. This is obviously better, and makes the effective difference between the two smaller, but the fact that the japanese player still has a slight advantage is crystal clear.

Until SE wants to make an even harder fight, where you have an AoE/gaze attack with a charge time of half a second. Then we're again in a situation where an attack is hard to dodge for a japanese player, and extremely hard/impossible to dodge for an european.

There is a 3rd thing SE can do, of course, but I'm sure you know what it is, and know that SE will probably not do it.

-edit-
And again, I am not saying that Ifrit Extreme is impossible for EU and NA players.