805 replies to this topic

[Ritualist]

Scorpion, on 24 February 2013 - 05:55 PM, said:

Since we'll get to choose the which dailies we want to do which would be the best combination? I was thinking:

Aquatic
Dodger
Crafting
Laurel vendor
Events??? If you get the maw you can take care all of it in a few minutes.

Keep in mind that once we get to choose our own, we'll also get new achievements - and they mentioned that we'll most likely see area-specific ones among them, for instance, kill a guy that only shows up in one area or visit a specific vista or something.
So based on this, I am guessing that the ones we currently have available each day will also be the best option for those of us that do not want to visit specific locations.

[Scorpion]

Ritualist, on 24 February 2013 - 07:24 PM, said:

Keep in mind that once we get to choose our own, we'll also get new achievements - and they mentioned that we'll most likely see area-specific ones among them, for instance, kill a guy that only shows up in one area or visit a specific vista or something.
So based on this, I am guessing that the ones we currently have available each day will also be the best option for those of us that do not want to visit specific locations.

Oh I didn't know about that. Yeah that definitively changes things up. The less zones I have to jump through the better.

[XPhiler]

Kymeric, on 22 February 2013 - 04:29 PM, said:

Just to throw another viewpoint into this, it's not just a question of it taking longer for someone who plays less often.

There was a great post in the mirror thread to this on the official forums that summed it up well.  A casual player can choose to work toward ascended gear or not.  They may take longer than someone who plays more.  For some, those things aren't a problem at all.  What is a problem for some casuals is that if they choose to work toward ascended gear by dailies, it becomes the majority of the game for them.

The hardcore player logs on for a three or four hour session, knocks out dailies in the first 45 minutes, then goes on to do whatever else they want.

The casual player logs on for a 90 minute play session and spends an hour of that time making sure dailies are done, and only has a little time left to do anything else.

If dailies are meant, like legendary weapons, for the hardcore, grind-enjoying player, then that's fine, and the rest of us can shrug and walk away, like we do with the legendaries.  Of course, those players are probably already staying up to date with ascended gear by doing fractals.  If, however, it's meant as an alternative, but slower path to BiS gear for casuals, it needs some tweaking.

Which it's going to get on the 26th with 5/9.


So this thread has kind of outlived its usefulness.




But somehow I can't resist continuing to post, anyway.

Thing is they dont need to focus on the daily at all to achievement the daily, they only have to focus on it if they need to log before its complete so they finish it off. This varies from player to player of course but nearly all tasks get you kills, kill variety, gathering, dodges, heals and to a degree and depending on what you like to play also veteran kills. Crafting and laurel vendor takes literally 10 seconds to complete.

[XPhiler]

Kymeric, on 23 February 2013 - 10:01 PM, said:

Apparently results vary.

What time do you normally play?

There are times when I've spent a lot  longer than 10 minutes looking for the last couple of DEs to finish out the list.

I'm getting pretty good about hitting certain areas to complete the tasks.  Hop to Gendarran Fields and knock out Oakhearts for Vets.  Hop to Thaumanova to get DEs and Healer (unless it's bugged, or on the Champ elemental and no one else wants to join in... got it to 1/3 health with one other player recently before we died and it's health went back to full, tried to find others in map chat with no results).

But did ANet really intend this to be something we speed-farm each day to get out of the way?

If I just do normal zone exploration, it can take much longer than 45 minutes to get DEs done.  I've hit 2 hours actively moving around a zone and not had them complete, before.

There are plenty of places in the game where you can guarantee you'll find 2 DE in 10 minutes. There are event places where you can get DE on demand even.

No Anet didnt intend for people to do Speed-farms everyday. Thing is people dont need to speed farm at all but it seems that for some players missing the daily becomes a tragedy so they have the option to speed farm the remainder of the achievement to avoid said personal tragedy. It may sound incredulous but its also quite possible to just play the game the way you like playing. If that doesnt hit the daily so what? In my play session yesterday I didnt do any WvW, dungeons or fractals either. People dont have to do everything.

[raspberry jam]

XPhiler, on 22 February 2013 - 03:01 PM, said:

I never said a 5% increase in damage means you have a 5% chance of killing a mob with one less hit. Its impossible to say what is your likeliness of killing a mob with an extra hit or not. I only talked about 2 extra hits because thats possible to calculate the point when it becomes a possiblity. In my opinion saving 2 hits is impossible on a regular mob (entirely possible on veterans and above of course)

It's not impossible, it is a matter of calculating the probability.

Volkon, on 22 February 2013 - 03:01 PM, said:

Don't be ridiculous. Of course it does. Besides... you're not talking number of hits originally, remember? You were talking about time to clear mobs, then drifted into number of hits. But... well, let's look at some numbers, because numbers are fun.

First off, we're talking about an increase in damage, not an increase in dps. Yes, they're different. The formula for damage is this:


Damage done = (weapon damage) * Power * (skill-specific coefficient) / (target's Armor)

So, plugging in numbers using the amulet on my mesmer as an example:

Before:  1100*1982*.5/1000 = 1090.1 (Assuming a coefficient of .5 and an armor of 1000 to simplify the math)
After: 1100*1993*.5/1000= 1096.15

(1096.15 - 1090.1)/ 1096.15 = .00552 or a single attack damage increase (not DPS) of approx. 0.55%.

The only way to consider this equivalent to a dps increase would be, of course, if all attacks happened like clockwork over a uniform period of time. We know for a fact that that's not the case in GW2. Many of the skills you use may do little damage but have a stronger controlling or supporting component to them. You often need to dodge, move, etc. all interfering with "optimal" dps. As a result, it's clear that actual dps increase will be well below one half of one percent.

I see no distributions in your calculations... How is weapon damage calculated? It is random, right? Is it an even distribution on the interval? In either case, taking the average (or any specific value) is of course not enough.

Number of hits is just a part of the clearing time, but the dodges and CC and such needs to be done anyway. If the number of hits go down the amount of dodges and CC and so on will go down as well, but such things usually need to be distributed across the DPS time anyway, so they can be ignored: calculating the number of hits is enough to determine if the clearing time goes down or not.

[Volkon]

raspberry jam, on 25 February 2013 - 09:32 AM, said:

It's not impossible, it is a matter of calculating the probability.

I see no distributions in your calculations... How is weapon damage calculated? It is random, right? Is it an even distribution on the interval? In either case, taking the average (or any specific value) is of course not enough.

Number of hits is just a part of the clearing time, but the dodges and CC and such needs to be done anyway. If the number of hits go down the amount of dodges and CC and so on will go down as well, but such things usually need to be distributed across the DPS time anyway, so they can be ignored: calculating the number of hits is enough to determine if the clearing time goes down or not.

You don't need a distribution. The actual weapon damage is irrelevant; everything else remains the same. If I used a weapon damage of 1 instead of 1100 you'd see this:

(1982 * .5) / 1000 = .991
(1993 * .5) / 100 = .9965

(.9965 - .991) / .9965 = .00552 or the same approx. 0.55%.

Regardless of weapon damage we're looking at a 0.55% increase in damage per hit. Now, regarding the dodges, CC, etc. that's all relevant in the sense that it's not completely predictable. The mobs don't all fight identically, so each one may require a dodge at a different time, a utility use at a different time, etc. A good example is the little Karka on Southsun. Now, my mesmer eats them up... they're highly susceptible to reflects when they do their tail attacks. The only problem is that you're never entirely sure when they'll actually use it. Sometimes they'll lead with it, sometimes there's a notable delay (meaning more hits spent until the coup de'grace). Sometimes they'll fire it at a clone, in which case Feedback is a lovely choice. Sometimes they'll fire it at me, in which case I'm using the heal that reflects projectiles offensively instead of defensively. Then of course you need to add the variables such as clones up when you shatter (sometimes they kill the clones quickly, others not so much), other players, other aggro, etc.

[Pyrea]

Will look forward to the changes implemented today where you can customize your achievements. Anet are certainly going in the right direction with this

[Lordkrall]

Pyrea, on 25 February 2013 - 01:00 PM, said:

Will look forward to the changes implemented today where you can customize your achievements. Anet are certainly going in the right direction with this

Tomorrow you mean? ^^

[raspberry jam]

Volkon, on 25 February 2013 - 12:51 PM, said:

You don't need a distribution. The actual weapon damage is irrelevant; everything else remains the same. If I used a weapon damage of 1 instead of 1100 you'd see this:

(1982 * .5) / 1000 = .991
(1993 * .5) / 100 = .9965

(.9965 - .991) / .9965 = .00552 or the same approx. 0.55%.

Regardless of weapon damage we're looking at a 0.55% increase in damage per hit. Now, regarding the dodges, CC, etc. that's all relevant in the sense that it's not completely predictable. The mobs don't all fight identically, so each one may require a dodge at a different time, a utility use at a different time, etc. A good example is the little Karka on Southsun. Now, my mesmer eats them up... they're highly susceptible to reflects when they do their tail attacks. The only problem is that you're never entirely sure when they'll actually use it. Sometimes they'll lead with it, sometimes there's a notable delay (meaning more hits spent until the coup de'grace). Sometimes they'll fire it at a clone, in which case Feedback is a lovely choice. Sometimes they'll fire it at me, in which case I'm using the heal that reflects projectiles offensively instead of defensively. Then of course you need to add the variables such as clones up when you shatter (sometimes they kill the clones quickly, others not so much), other players, other aggro, etc.

No, you need a distribution. You need to calculate the amounts of hits that you'd expect to score a kill with, and then you need to calculate the probability of getting a kill with X hits less than usual given a certain alteration of stats. What you calculate is the increase in damage per hit, which is all fine but not the number that we want.

[Volkon]

raspberry jam, on 25 February 2013 - 02:43 PM, said:

No, you need a distribution. You need to calculate the amounts of hits that you'd expect to score a kill with, and then you need to calculate the probability of getting a kill with X hits less than usual given a certain alteration of stats. What you calculate is the increase in damage per hit, which is all fine but not the number that we want.

Too many variables and too much overlap. You'd have to discard the RNG element entirely (Greatsword, for example, does 995 - 1100 damage per hit, hence the choice of 1100), you'd have to know actual hit points of your mob, the armor of your mob, etc. and you'd have to make the assumption of two fights being identical which strays from realistic at best.

But, if we go with the aforementioned calculations as a reasonable example, we can play with them a little bit. Above you see the difference in hypothetical data where damage goes from 1090 to 1096 per hit. With this scenario, starting with a "one-shot", something is clear. A mob with 1090 or less health will be one-shot before and after. A mob with 1090 to 1096 health would take two shots before, one after. What you have is an increase of 0.55% in hypothetical low health mobs being taken down in one hit instead of two. This, of course, includes the caveat that you're using the exact same attack every time with no additional boosts of any kind, no crits, no blocks or misses, etc... just a single mob standing there taking one for the team in a fully controlled experiment.

[raspberry jam]

Volkon, on 25 February 2013 - 03:05 PM, said:

Too many variables and too much overlap. You'd have to discard the RNG element entirely (Greatsword, for example, does 995 - 1100 damage per hit, hence the choice of 1100), you'd have to know actual hit points of your mob, the armor of your mob, etc. and you'd have to make the assumption of two fights being identical which strays from realistic at best.

But, if we go with the aforementioned calculations as a reasonable example, we can play with them a little bit. Above you see the difference in hypothetical data where damage goes from 1090 to 1096 per hit. With this scenario, starting with a "one-shot", something is clear. A mob with 1090 or less health will be one-shot before and after. A mob with 1090 to 1096 health would take two shots before, one after. What you have is an increase of 0.55% in hypothetical low health mobs being taken down in one hit instead of two. This, of course, includes the caveat that you're using the exact same attack every time with no additional boosts of any kind, no crits, no blocks or misses, etc... just a single mob standing there taking one for the team in a fully controlled experiment.

But no mob that is worth fighting has that low health. You need to compare to high health mobs, and you need the distribution because you need to make probability-theory calculations.

[Volkon]

raspberry jam, on 25 February 2013 - 03:07 PM, said:

But no mob that is worth fighting has that low health. You need to compare to high health mobs, and you need the distribution because you need to make probability-theory calculations.

That's what I'm saying... too many variables (to me at least) to get a viable scenario. Fights against even the same mobs aren't exactly the same. Fights against high level mobs could very well result in you needing less hits overall as variables come into play, such as crits, mob avoidance, etc. Because of these variables you can't extrapolate that, for example, for mobs with 10x the health above it will be 5.5% of the mobs that you need less hits on. Now, maybe someone with a much deeper well of mathematical training can come up with something, but even then there will have to be assumptions made that aren't likely in "the real world".

[Kymeric]

XPhiler, on 25 February 2013 - 08:31 AM, said:

It may sound incredulous but its also quite possible to just play the game the way you like playing. If that doesnt hit the daily so what? In my play session yesterday I didnt do any WvW, dungeons or fractals either. People dont have to do everything.

You're absolutely right.  I used to shrug off the daily if I didn't get it done with the old system.  Meh, what's one jug of karma more or less?

The new dailies, for a handful of reasons I've listed before, just feel more present in my game play to the point that I don't really like the changes.  I'm looking forward to the updated system tomorrow.

Edited by Kymeric, 25 February 2013 - 04:34 PM.

[raspberry jam]

Volkon, on 25 February 2013 - 03:14 PM, said:

That's what I'm saying... too many variables (to me at least) to get a viable scenario. Fights against even the same mobs aren't exactly the same. Fights against high level mobs could very well result in you needing less hits overall as variables come into play, such as crits, mob avoidance, etc. Because of these variables you can't extrapolate that, for example, for mobs with 10x the health above it will be 5.5% of the mobs that you need less hits on. Now, maybe someone with a much deeper well of mathematical training can come up with something, but even then there will have to be assumptions made that aren't likely in "the real world".

True, you can't extrapolate, you need to make a probabilistic calculation. And for that you need to know the distribution of the random variables (like weapon damage).
Avoidance isn't affected by Power... I guess if you boost Precision instead you'll hit better, but let's assume that Power is the best choice. If not, then any result we get for Power is a lower limit to the efficiency increase.

I'll try to find the right formulas when I get home.

[Resolve]

Lordkrall, on 25 February 2013 - 01:55 PM, said:

Tomorrow you mean? ^^

Is this a big update?

[raspberry jam]

Resolve, on 25 February 2013 - 05:44 PM, said:

Is this a big update?

There is a big update coming that will solve any problems that you can possibly have with the game. And when that doesn't happen, there will be an update next month. It will solve everything. And not cause any new problems. At all. And those will be solved in the update after that. Promise.

[Lordkrall]

Resolve, on 25 February 2013 - 05:44 PM, said:

Is this a big update?

Define big.

There will be selectable dailies.
New Dailes.
A new event system (Guild Missions)
Next part of Flame and Frost.
New sPvP map
New sPvP tournament type.

[XPhiler]

Kymeric, on 25 February 2013 - 04:34 PM, said:

You're absolutely right.  I used to shrug off the daily if I didn't get it done with the old system.  Meh, what's one jug of karma more or less?

The new dailies, for a handful of reasons I've listed before, just feel more present in my game play to the point that I don't really like the changes.  I'm looking forward to the updated system tomorrow.

Thats quite a dilemma dont you think? Something that was useless to you before is now useful (thats a good thing, I suppose we agree on that)  You were totally happy with missing it before but now that doesnt seem to be an option for you any longer just cause of the useful rewards. Thing is by missing it now and then you're not giving up on the rewards it will just take a bit longer. Why is that a problem?

[raspberry jam]

raspberry jam, on 25 February 2013 - 05:41 PM, said:

I'll try to find the right formulas when I get home.

The correct values seem to be more than 12% chance of killing in one hit less, around 4% of killing in two hits less, and below 1% killing in three or more hits less, for a 100k health enemy. I'd definitely take those odds.

[XPhiler]

raspberry jam, on 26 February 2013 - 10:27 AM, said:

The correct values seem to be more than 12% chance of killing in one hit less, around 4% of killing in two hits less, and below 1% killing in three or more hits less, for a 100k health enemy. I'd definitely take those odds.

And which mobs have 100k health?  Only champion mobs+ which you're never going to solo anyway introducing yet another variable . In any case we're talking of rare fights that take minutes to resolve 3 less hits is not going make a big difference at all. Better to have the advantage then not to have the advantage of course so that is not to say one should ignore dailies but a few days extra before you earn an ascended amulet is not going to set up back much.

[raspberry jam]

XPhiler, on 26 February 2013 - 10:54 AM, said:

Better to have the advantage then not to have the advantage of course

Yeah... that's true.

[Kymeric]

XPhiler, on 26 February 2013 - 08:34 AM, said:

Thats quite a dilemma dont you think? Something that was useless to you before is now useful (thats a good thing, I suppose we agree on that)  You were totally happy with missing it before but now that doesnt seem to be an option for you any longer just cause of the useful rewards. Thing is by missing it now and then you're not giving up on the rewards it will just take a bit longer. Why is that a problem?

We seem to be stuck in a circle at this point, so I'm not sure there's any point in continuing around it one more time.  You're right that I can choose to ignore the dailies.  I don't care for them in their current implementation.  We get a patch later today that will change them, hopefully in a way that I will enjoy.

Thanks for the discussion.

[Volkon]

raspberry jam, on 26 February 2013 - 10:27 AM, said:

The correct values seem to be more than 12% chance of killing in one hit less, around 4% of killing in two hits less, and below 1% killing in three or more hits less, for a 100k health enemy. I'd definitely take those odds.

Something seems... off... there. The larger the pool of health you're attacking the more likely should be the odds of reducing the number of hits you need to take it down. This looks like you're stating the reverse. I'm going to need to see your work!

[raspberry jam]

Volkon, on 26 February 2013 - 12:49 PM, said:

Something seems... off... there. The larger the pool of health you're attacking the more likely should be the odds of reducing the number of hits you need to take it down. This looks like you're stating the reverse. I'm going to need to see your work!

Mm, I'll try to write it down. I did it online, on WolframAlpha, and I got two different (but almost the same) results, very annoying.

A larger health pool means that the chance is bigger of course, but the effect is strange, it is large at first and smaller as the health pool grows.
Another surprising result is that even though most kills is done with overdamage, the most likely kill is the one that reduces the health to exactly 0.

[Resolve]

Lordkrall, on 25 February 2013 - 06:03 PM, said:

Define big.

There will be selectable dailies.
New Dailes.
A new event system (Guild Missions)
Next part of Flame and Frost.
New sPvP map
New sPvP tournament type.

Looking forward to the patch notes then.

raspberry jam, on 25 February 2013 - 05:56 PM, said:

There is a big update coming that will solve any problems that you can possibly have with the game. And when that doesn't happen, there will be an update next month. It will solve everything. And not cause any new problems. At all. And those will be solved in the update after that. Promise.

Meh, as long as they actually start making positive progress and stop with the "well there's some good here but also a lot of bad" updates.

[raspberry jam]

Resolve, on 26 February 2013 - 01:59 PM, said:

Meh, as long as they actually start making positive progress and stop with the "well there's some good here but also a lot of bad" updates.

We are talking about ANet here.