194 réponses à ce sujet

[ZeroDivision]

Terraflare wrote...
Snip...

Still think so?

1st post - Assuming animation canceling is not intended as part of the game, yes. Even if it is intended, the Javelin still better for the uses of a structured team with adepts, engineers, etc skilled enough to handle the small targets more efficiently than a sniper.

2nd post - Yes, though on further testing the Widow propper is indeed outclassed by the Black Widow.

Edit - I love iPhone autocorrect.

Modifié par ZeroDivision, 03 avril 2012 - 06:03 .

[deadpixel92]

Khevan77 wrote...

There is a large flaw in the logic of relying solely on mathematics. Mathematics can only produce answers from a given set of conditions, and it's prohibitively difficult to account for every variance or variable that can occur.

In other words, real life events, such as needing to run because a Phantom is keen on sticking her sword up your pooper, or suddenly realizing a Geth Hunter is shooting you in the back of the head, or simple enemy movements (like rolling just as you pull the trigger...) all of this affects the damage per second, and fudges up the numbers that are so neatly laid out according to pure mathematics.

Math is theory, using the weapons in game is reality. Reality will always trump theory when comparing effectiveness of any particular system.

^THIS

[Sp3c7eR]

@ZeroDivision: The free trial of Fraps allows you to record for 30 seconds without paying. More than enough to kill an Atlas with Warp ammo III, Sniper rail III and Power Efficiency Module III. With Javelin III we wouldn't see any differences in terms of damage output - it would still take 6 cloaked cycles to kill an Atlas on Gold. The only difference would be firing speed/delay which I am curious to see how I could improve on. So please record it and show us. If you have trouble with Fraps, there is a program called MSI afterburner - it is absolutely free, no time limit on it and does an even better job at recording because it doesn't drop your frames per second in game. Honestly, I'd do another run with Javelin but I don't have unlimited supply of equipment and I don't see a point after observing a difference of 30%.

The Javelin has no real synergy with reload cancelling unfortunately so it would be nearly impossible to pull off what you described in your last post. It would still take at least 4 if not 5 cloak cycles with your method and if anything, it will probably delay killing an Atlas even further. But again, I'm curious to see the video which will be the most definitive way to prove your point. Without a video those numbers you're producing mean nothing... And you can also involve your team's sniper - if you can kill an Atlas in less than 18 seconds with Javelin, no matter what rank it is, I'll admit it's fair to compare it to Black Widow in terms of DPS.

@Canasian81 Got a laugh out of me Thanks.

Modifié par Sp3c7eR, 03 avril 2012 - 06:06 .

[Terraflare]

ZeroDivision wrote...

Terraflare wrote...
Snip...

Still think so?

1st post - Assuming animation canceling is not intended as part of the game, yes. Even if it is intended, the Javelin still better for the uses of a structured team with adepts, engineers, etc skilled enough to handle the small targets more efficiently than a sniper.

2nd post - Yes, though on further testing the Widow propper is indeed outclassed by the Black Widow.

Edit - I love iPhone autocorrect.

Wait my brain seized for a bit. A Jav shot does 1030.5 damage a shot and a Widow X does 1083.3 damage a shot. 

Both fire single shots. Both take same reload time. Javelin takes 0.25s to charge a shot. I'll give you the benefit of the doubt and just say that you didnt test the Widow X at all. Might want to stop claiming you test stuff when you dont actually do (as you did when claiming the BW took 3 seconds to fire 3 shots). 


So you would translate (by your calculations, assuming you can pull this off in a video) a 7% faster kill time (roughly 20 vs 21 seconds a real game scenario) on an Atlas (compared to every single of the BW's strengths), to the BW being a "terrible gun" and being only "suitable for silver and bronze" and that people who use it well are simply doing so on "silver and bronze" and on gold "it gets destroyed"?

Modifié par Terraflare, 03 avril 2012 - 06:12 .

[ZeroDivision]

Terraflare wrote...
Wait my brain seized for a bit. A Jav shot does 1030.5 damage a shot and a Widow X does 1083.3 damage a shot. 

Both fire single shots. Both take same reload time. Javelin takes 0.25s to charge a shot. I'll give you the benefit of the doubt and just say that you didnt test the Widow X at all. Might want to stop claiming you test stuff when you dont actually do (as you did when claiming the BW took 3 seconds to fire 3 shots). 


So you would translate (by your calculations, assuming you can pull this off in a video) a 7% faster kill time (roughly 20 vs 21 seconds a real game scenario) on an Atlas (compared to every single of the BW's strengths), to the BW being a "terrible gun" and being only "suitable for silver and bronze" and that people who use it well are simply doing so on "silver and bronze" and on gold "it gets destroyed"?

1 - You're assuming that every individual has a Widow 10 before they get a Javelin 2. For the record my Javelin is 3 and my Widow is only 5.

2 - Comparing a Widow to a Javelin of equal rank isn't even close.

3 - For a structured team using a strategy even close to the one my team runs, yes.

[probablynot]

I don't know whats more troubling about this thread, the fact that the OP uses bad math to loosely prove his point (anyone who has ever played an mmo knows how easily theory crafting at this level decays, especially in a game considerably more entropic than an mmo), or that so many morons are coming out of the woodwork to claim how they "distrust math". I'm sure some people really mean "I believe this math is not thorough enough to warrant my attention" but for everyone else, shame on you. This is blatant anti-intellectualism and it makes me sick. There is absolutely nothing in this universe more trustworthy than good math.

[ZeroDivision]

probablynot wrote...

I don't know whats more troubling about this thread, the fact that the OP uses bad math to loosely prove his point (anyone who has ever played an mmo knows how easily theory crafting at this level decays, especially in a game considerably more entropic than an mmo), or that so many morons are coming out of the woodwork to claim how they "distrust math". I'm sure some people really mean "I believe this math is not thorough enough to warrant my attention" but for everyone else, shame on you. This is blatant anti-intellectualism and it makes me sick. There is absolutely nothing in this universe more trustworthy than good math.

Please explain to me how the math itself is wrong, not the debatable data used in the math, and i'll be more than happy to fix it.

[Terraflare]

ZeroDivision wrote...


1 - You're assuming that every individual has a Widow 10 before they get a Javelin 2. For the record my Javelin is 3 and my Widow is only 5.

2 - Comparing a Widow to a Javelin of equal rank isn't even close.

3 - For a structured team using a strategy even close to the one my team runs, yes.

Ahhhhh, might want to think about prefacing your claims of 'fact' with all these little bitty qualifications.... 

Summary time: 

Terraflare wrote...

So you would translate (by your calculations, assuming you can pull this off in a video) a 7% faster kill time (roughly 20 vs 21 seconds a real game scenario) on an Atlas (compared to every single of the BW's strengths), to the BW being a "terrible gun" and being only "suitable for silver and bronze" and that people who use it well are simply doing so on "silver and bronze" and on gold "it gets destroyed"?

 

Is answered with :

ZeroDivision wrote... 

3 - For a structured team using a strategy even close to the one my team runs, yes.

Harder than squeezing water out of a rock. I'm glad I'm not in your team, you must truly be at the top of this game. because only your team seems to know how to do gold. Have fun making that video that you really should to prove your points.

Modifié par Terraflare, 03 avril 2012 - 06:43 .

[ZeroDivision]

Terraflare wrote...

Snip...

Harder than squeezing water out of a rock. I'm glad I'm not in your team, you must truly be at the top of this game. because only your team seems to know how to do gold. Have fun making that video that you really should to prove your points.

Nice straw-man.
Nice passive-aggressive comments.
Nice logical fallacy ( Arguement ad Ignorantiam ).

If this is what you've fallen to I really will start ignoring your posts.

Edit - As a final aside, I'd like to point out that I preface the original post with precisely the setup and the reasons for the setup that this comparison is run from. Which is exactly what you asked me to do in the rest of the above post.

Modifié par ZeroDivision, 03 avril 2012 - 07:03 .

[PluralAces]

you
need
a
life

that is all

[ZeroDivision]

PluralAces wrote...

you
need
a
life

that is all

Yes, yes I do. I can't speak for anyone else in this thread though.

[probablynot]

It isn't a matter of being wrong or not. If you think the math is totally right then you have bigger problems to worry about. When you simulate something like this, it's more about how much "wrong" is acceptable.

I don't even need to go into specifics before something immediately jumps out to me. You don't take latency into account in your calculations. Even at the absolute most basic levels of dps simulations on wow, latency is a part of the calculations. That's on a game with a spell queue and significantly better lag.

Okay, so you are laboring under the assumption that you are always hosting. I can live with that. But then you totally ignore ammo/sniper rail amps in your calculations? These are non-trivial factors. It massively changes how many shots it takes to do a great deal of things.

These are just a couple things, there's more of course. Now you may say "Oh, that's a slippery slope argument as there is no real upper bound to how specific you can get." No, not really. These are the basics of theory crafting for a game like, say, WoW. So anything less amounts to, you are being less precise in a game that is more entropic. Especially the rail/ammo stuff, I mean come on. Some of these factors might even help your argument, idk, I don't care to theory craft about this game.

Point is, if you are going to use objective mathematics to quantify your argument, try to be a little more specific with it. You are at that absolute worst level where the anti-intellectual morons will distrust you, and the people who know how to theory craft dismiss your numbers as too basic an analysis.

That's all assuming everything you said is factually accurate. I won't contest that as it has nothing to do with my argument, but I would just like to point out that plenty of people have called that into question.

[Shahadem]

How can you theory craft without using advanced statistics to simulate what actually occurs in a match?

Your rote math shows nothing more than which gun hits harder, a fact which is already known by all.

[Nickle]

Have you guys factored in what happens if you're too busy muzzing out to a peaking track and your hand starts flying all over the place, or you're too busy fist-bumping to aim?

[shadowkinz]

i want a valiant so bad, i really hope i don't get a duplicate eagle =( i love my eagle but i've been dying for a valiant, it makes me so upset lol

[UK Wildcat]

I rate this a 5/10 on trolling scale.

[ZeroDivision]

probablynot wrote...

It isn't a matter of being wrong or not. If you think the math is totally right then you have bigger problems to worry about. When you simulate something like this, it's more about how much "wrong" is acceptable.

I don't even need to go into specifics before something immediately jumps out to me. You don't take latency into account in your calculations. Even at the absolute most basic levels of dps simulations on wow, latency is a part of the calculations. That's on a game with a spell queue and significantly better lag.

Okay, so you are laboring under the assumption that you are always hosting. I can live with that. But then you totally ignore ammo/sniper rail amps in your calculations? These are non-trivial factors. It massively changes how many shots it takes to do a great deal of things.

These are just a couple things, there's more of course. Now you may say "Oh, that's a slippery slope argument as there is no real upper bound to how specific you can get." No, not really. These are the basics of theory crafting for a game like, say, WoW. So anything less amounts to, you are being less precise in a game that is more entropic. Especially the rail/ammo stuff, I mean come on. Some of these factors might even help your argument, idk, I don't care to theory craft about this game.

Point is, if you are going to use objective mathematics to quantify your argument, try to be a little more specific with it. You are at that absolute worst level where the anti-intellectual morons will distrust you, and the people who know how to theory craft dismiss your numbers as too basic an analysis.

That's all assuming everything you said is factually accurate. I won't contest that as it has nothing to do with my argument, but I would just like to point out that plenty of people have called that into question.

Thanks for the reply.

The reason that I don't include rail amps, ammo, and other damage bonuses is simply because, to my knowledge, there has been no difinitive answer on whether they are additive or multiplicative. The results would be similar either way, but these kind of details matter.

As to latency there is no numerical value available currently to base an average on or even an idea of what it might be, so one cannot use it in a quantified comparison.

From the details perspective you are entirely correct. I realized this quite quickly and made an attempt to correct this problem with my second and third comparisons. Is the most recent comparison still not detailed enough? If so I could run a comparison assuming additivity, multiplicity, or both when dealing with other damage bonuses.

Concerning the factual accuracy of the data I have used, my testing and assumption that the reload-cancel bug is indeed a bug, and therefore should not be relied on for legitimate comparisons, apparently differ from some others.  That being stated, I based the most recent comparison on their data, when available and accurate, and the math still supported my original claim within the parameters provided by my first post.

[Phaelducan]

The real question is whether people with bad aim should use a Viper X. Now that's some math that needs exploring.

[ZeroDivision]

Phaelducan wrote...

The real question is whether people with bad aim should use a Viper X. Now that's some math that needs exploring.

If they're really bad at aiming the best option, as far as snipers go, is probably the Raptor.  Outside of that a shotgun might do them well.

[Terraflare]

ZeroDivision wrote...

Terraflare wrote...

Snip...

Harder than squeezing water out of a rock. I'm glad I'm not in your team, you must truly be at the top of this game. because only your team seems to know how to do gold. Have fun making that video that you really should to prove your points.

Nice straw-man.
Nice passive-aggressive comments.
Nice logical fallacy ( Arguement ad Ignorantiam ).

If this is what you've fallen to I really will start ignoring your posts.

Edit - As a final aside, I'd like to point out that I preface the original post with precisely the setup and the reasons for the setup that this comparison is run from. Which is exactly what you asked me to do in the rest of the above post.

Ah, so you are quick to point out fallacies and are not stupid.  Good.

And yet you fail to see that the blatent fallacy you make when you generalise a highly specific, situational, hypothetical, yet to be proven (where's the video) opinion into a statement that the BW is a "terrible weapon" and only suited for "bronze and silver". 

[RTS suDnd3th]

ZeroDivision wrote...

4 - Math is much more telling and balanced.

Ppl like you that rely 100% on numbers which only work in a perfect world and no world is perfect.

Yes Math is great for number crunching and rough estimates but it never works out exactly to what comes out on paper no matter what when put into practical use. Games or RL its the same damn thing.

You remind me of a person I had a run in with who absolutly drove me nuts. Saying that his numbers were right and so forth, yes they were but when put into practical use I had to spend 3 days fixing it so it would work correctly after it was built.

[Keldaurz]

probablynot wrote...

I don't know whats more troubling about this thread, the fact that the OP uses bad math to loosely prove his point (anyone who has ever played an mmo knows how easily theory crafting at this level decays, especially in a game considerably more entropic than an mmo), or that so many morons are coming out of the woodwork to claim how they "distrust math". I'm sure some people really mean "I believe this math is not thorough enough to warrant my attention" but for everyone else, shame on you. This is blatant anti-intellectualism and it makes me sick. There is absolutely nothing in this universe more trustworthy than good math.

There are only a few things i really asked myself for weapons.

Exposure time
Cycle time
Damage per cycle/reload
Damage/cycles needed to kill given enemies with/outh ED.

If you beyond that with "considering you have a DPS of 1 gazillions you can do on a match 100 gazillions damage" when you are not taking into account overkill, headshots and whatever else you really want to add, it's already telling you there's some places where math can't bring you, atleast not with only a few basic, perfect, and static asumptions.

So yes, that about "only skills matter" and whatever, it's just... well the other fanatic side of the coin.

Modifié par Keldaurz, 03 avril 2012 - 07:44 .

[Trix-Rabbit]

Still waiting for a realistic comparison between the 4 main sniper rifles.(mantis/widow/bw/valiant) hell you can even throw the jav in there if you want.

which includes Prox Mine and Energy Drain. And being able to spam them more often+not having to take the SR weight reduction certain weapons.

[ZeroDivision]

Trix-Rabbit wrote...

Still waiting for a realistic comparison between the 4 main sniper rifles.(mantis/widow/bw/valiant) hell you can even throw the jav in there if you want.

which includes Prox Mine and Energy Drain. And being able to spam them more often+not having to take the SR weight reduction certain weapons.

Unfortunately not every Infiltrator race or spec has Proxy Mine or Energy Drain. The purpose of a class based weapon comparison is to compare weapons with only the factors common between all variations of the class, not to decide which weapon is better for which spec or which race does more or less damage with its powers based on the weapon used.

[Sp3c7eR]

 So... I rummaged through some of my footage from yesterday and found something curious, and before ZeroDivision declares it invalid I hasten to add - it is not a perfect kill. Yet notice what happens to the 6th shot of the Black Widow:

Shield Gating? Where is it?

And yes, this test scored a time of 15.25 seconds. Could be better as I was a little thrown by the body jerk animation of the Atlas, but hey these things happen. Only people with link will be able to see video.

Modifié par Sp3c7eR, 03 avril 2012 - 09:07 .