Please upgrade your browser for the best possible experience.

Chrome Firefox Internet Explorer
×

R/E frustrations


Danylia's Avatar


Danylia
01.24.2013 , 07:03 AM | #31
Quote: Originally Posted by DarthScruffy View Post
Depending on how they set up their RNG it may not be random.
This. In fact, it indeed is not "purely" random, it is pseudo-random, because the RNG must be based on something - it must have a "seed".

The seed in Neverwinter Nights, another Bioware game, was for example based on the area you were in (among other things), and it did reset each time you changed an area. Also, once the seed was set, the "random" sequence of numbers was always the same, meaning that if with seed A you got rolls like 45, 89, 76, 3, 15 etc., you would get exactly the same rolls if you would manage to get seed A again.

The consequence of this was that a) some areas were more "likely" to produce a successful roll, and b) changing an area after a series of unsuccessful rolls was a bad idea since the RNG had to start "from the scratch", while in the original area you would be approaching the "good numbers". I don't know what kind of RNG is used in SWTOR, but from my observations, the point b) seems to hold true here, too - when I fail RE'ing something five times in a row, I try, if possible, to make the next five attempts in the same area - so far it has paid off.

Though, of course, it may also be just a "Las Vegas casino manual".
Quote: Originally Posted by IronSalamander View Post
Scorpio is just a glorified Microsoft product ("time to upgrade", "I must upgrade",etc.) and I really want this thing off my ship. Kaliyo went from obnoxious to worthless really fast and I would love to dump her out the airlock.

DarthScruffy's Avatar


DarthScruffy
01.24.2013 , 01:56 PM | #32
Quote: Originally Posted by Danylia View Post
This. In fact, it indeed is not "purely" random, it is pseudo-random, because the RNG must be based on something - it must have a "seed".

The seed in Neverwinter Nights, another Bioware game, was for example based on the area you were in (among other things), and it did reset each time you changed an area. Also, once the seed was set, the "random" sequence of numbers was always the same, meaning that if with seed A you got rolls like 45, 89, 76, 3, 15 etc., you would get exactly the same rolls if you would manage to get seed A again.

The consequence of this was that a) some areas were more "likely" to produce a successful roll, and b) changing an area after a series of unsuccessful rolls was a bad idea since the RNG had to start "from the scratch", while in the original area you would be approaching the "good numbers". I don't know what kind of RNG is used in SWTOR, but from my observations, the point b) seems to hold true here, too - when I fail RE'ing something five times in a row, I try, if possible, to make the next five attempts in the same area - so far it has paid off.

Though, of course, it may also be just a "Las Vegas casino manual".
If the seed is based on a "normal clock" certain times of days will produce the same result. But there are better ways you can even use a clock a seed and you could do some real advanced logic stuff if you really wanted to get closer to random.

Kaskali's Avatar


Kaskali
01.24.2013 , 06:12 PM | #33
We have computers that win at Jeopardy! and calculate pi to ten trillion digits. Do you honestly think no one has figured out how to approximate a die roll that is sufficiently random for video game crafting?

This is one of those situations where a little bit of knowledge does more harm than good. Unless you are a cryptographer or a mathematician, computer randomization is more than sufficient for your needs.
Quote: Originally Posted by Danylia View Post
The seed in Neverwinter Nights, another Bioware game, was for example based on the area you were in (among other things), and it did reset each time you changed an area. Also, once the seed was set, the "random" sequence of numbers was always the same, meaning that if with seed A you got rolls like 45, 89, 76, 3, 15 etc., you would get exactly the same rolls if you would manage to get seed A again.
I never played Neverwinter Nights, but single-player games are sometimes designed in ways that appear (and may, in fact, be) not entirely random in order to discourage players from cheating by saving the game and reloading if they don't get the results they want. This is most noticeable in turn-based games like XCOM or Civilization; if you reload a saved game and do the same things you did first time, you will get the same combat results. You can reload your saved game a hundred times, but if you have the same unit move the same way and attack the same enemy unit, the results will be the same each time. It may be that Neverwinter Nights was designed as it was not because of a limitation of computer randomization, but to discourage cheating.

Khevar's Avatar


Khevar
01.24.2013 , 09:58 PM | #34
Quote: Originally Posted by Kaskali View Post
...

I never played Neverwinter Nights, but single-player games are sometimes designed in ways that appear (and may, in fact, be) not entirely random in order to discourage players from cheating by saving the game and reloading if they don't get the results they want. This is most noticeable in turn-based games like XCOM or Civilization; if you reload a saved game and do the same things you did first time, you will get the same combat results. You can reload your saved game a hundred times, but if you have the same unit move the same way and attack the same enemy unit, the results will be the same each time. It may be that Neverwinter Nights was designed as it was not because of a limitation of computer randomization, but to discourage cheating.
This is an excellent point.

When playing XCOM I noticed that exact same behavior. It's almost as if each move or action has a value assigned to it, is somehow combined with the previous RNG result, and is then used as a new seed.

I agree that single player games are likely to have a completely different philosophy about RNG than an MMO.

NLxAROSA's Avatar


NLxAROSA
01.26.2013 , 12:45 PM | #35
It's true that the Firaxis games (latest Civ and XCOM) deliberately store the seed on save, to prevent people 'reloadcheating'. Developers confirmed that it was deliberate. Still easily defeatable if you have a few other characters with moves left though.

Zhiroc's Avatar


Zhiroc
01.28.2013 , 04:22 PM | #36
Almost a year ago I did some analysis on this. It turned out that for a 20% schematic the 90th percentile of trials is at 10.5 rolls, and the 98th is around 20. In other words, 1 out of 10 times it will take you more than 10 rolls to succeed, and about 1 in 50 times (actually, it was 1 in 45), it will take you 20 or more.

If you roll the dice a million times, you will almost always find at least one streak of > 50 failures, and it's about 50/50 that you'll find one taking over 60.

ebado's Avatar


ebado
01.28.2013 , 04:46 PM | #37
Quote: Originally Posted by Facime View Post
As for those of you saying its a re-roll each time. I understand that, but Im wondering how many of you might not understand odds and statistics. The way it works is like this: Statistically I have a one in 5 chance of hitting a schematic. That means that the more times I attempt and fail, the greater my odds are that I will hit on the next attempt...statistically speaking of course. When you dont hit after 30 or 40 attempts, either the mechanic is broken or you are deep nito a statistical anomally. .
Sigh.

Your odds don't change. Even if we were talking about a statistically significant sample size in the millions. It's 1/5 each time. Your odds each time are 1/5. Over the course of a statistically significant sample size of rolls, you don't increase your odds at any time.. No, you don't. Not statistically speaking. Rather, over the course of a statistically significant sample size of rolls, you have enough indepedent rolls for the actual chance to show through and eliminate and / or reduce the effect of outliers.

That is completely different from saying "I've failed 20 times, that means I have a better chance now than I did the first time."

psandak's Avatar


psandak
01.28.2013 , 04:58 PM | #38
Every time I read arguments over statistics and odds I am reminded of the Don Henley song Garden of Allah

"Today I made an appearance downtown
I am an expert witness, because I say I am
And I said, 'Gentleman....and I use that word loosely....
I will testify for you
I'm a gun for hire, I'm a saint, I'm a liar
Because there are no facts, there is no truth
Just a data to be manipulated
I can get any result you like
What's it worth to ya?..."

Zhiroc's Avatar


Zhiroc
01.28.2013 , 10:47 PM | #39
By the way, you generally don't want to "re-seed" your RNG from something that you "think" is random, like the time of day. If you're using a good RNG (and there are plenty in various runtime libraries), nowadays they generally are designed on sound statistical principles, and provide a uniform pseudo-random distribution. If you think you "know better" and try to throw what you think is more randomness into it, you are probably making it worse, not better.

And to reiterate what I said a few posts back, if you roll the dice a million times, you will almost always find a string of 50 or more failures in a row. So... think about it. Let's say there are 100,000 players RE'ing each day. If they each roll about 10 times, that's a million. So roughly speaking, that means EVERY DAY, someone in SWTOR fails on 50 tries to succeed on a RE'ed item.

Yes, the odds of failing 50 times in a row is miniscule. You'd never bet on it... but when you have a sample size the size of SWTOR, it happens---REGULARLY.

ScarletBlaze's Avatar


ScarletBlaze
01.28.2013 , 11:09 PM | #40
I actually find RE in Swtor easier than I did when RE was first introduced in SWG. That was a headache there was no ratio. You could get something all the way up to one point where you wanted it and the next time you re the item hoping to get it to max it would go all the way back down.

It was until they made re tools (which were good for one use only) that RE got easier but then you had to have:

the food, drink, a buff from an entertainer and an entire day to set there and get them to where you wanted them.
A southern girl is a girl who knows full well that she can open a door for herself but prefers for the gentleman to do it because it demonstrates a sense of respect. Current Guild Roleplay: Redemption: The battle of Xa Fel https://groups.google.com/forum/?hl=...ps/r-1rAjOuo30