Incarnate Trials April - Dec 18
By
Bubbawheat,
Posted on: 2011-12-19 02:16 in CoH and CoV General Discussions
Quote:
Well that hardly makes a fair chart does it | And yeah, [dripping sarcasm] Underground, TPN, and MoM are really being run often on Triumph [/dripping sarcasm]. The one TPN trial, as well as one BAF & one Lambda, were on Virtue. Oh, and I got an uncommon on that TPN... Again so much for the "improved chance of a Rare or Very Rare". |
ONE TPN run and you dont get the rare/very rare? Well cant do a sarcastic post when you run just one TPN and state that the increased chance isnt there . . . . . . its called an increased chance and you need to run multiples of them before you can do the whole Quote:
| Again so much for the "improved chance of a Rare or Very Rare". |
TPN/MoM/UG are all run now on Union quite often, slightly less than baf/lam/keyes but still quite often so the improvements DID work [at least for our server].
@Damz Find me on the global channel Union Chat. One of the best "chat channels" ingame!
Sample size of data seems too small to draw conclusive information.
Let him cry his eyes out and complain about the drop rate with insufficient data.
"while some people would say fish, cow, ambush!"-Ice9
The Paladin
Steel Canyon, Virtue
Exalted
@Paladin
Dude, lemme just clarify:
Did you seriously just post a chart with a sample size of one?
Come back when you have cumulative numbers from a couple hundred.
Okay, using this revolutionary methodology, I've finally proven that computer RNGs are not as good as people like to think.
I had a computer generate a random number between 1 and 6. As you're aware, in theory the probability of each number should be even, like this:
1 | ********
2 | ********
3 | ********
4 | ********
5 | ********
6 | ********
----------------
What I actually got was this:
1 |
2 |
3 |
4 | ************************************************
5 |
6 |
----------------
So much for the fair distribution of results, or the alleged "one in six" chance of getting a 5!
Quote:
You can't make a sweeping statement about the probability of an event being skewed based on a sample size of one. That's not even statistics... it's common sense. If the original percentage of R/VR drops was 10% and on a TPN it's 15% (completely hypothetical numbers), that's an improvement, but still an 85% of not getting a R/VR. There is nothing shocking about this at all. Even your sample size of 516 is fairly small, given how many purported adjustments have been made to trial drop rates since trials went live.|
Originally Posted by Snow Globe
The one TPN trial, as well as one BAF & one Lambda, were on Virtue. Oh, and I got an uncommon on that TPN... Again so much for the "improved chance of a Rare or Very Rare".
|
If you flip a coin once and get tails at the start of a hockey match, would you accuse the ref of having a trick coin that flips tails 100% of the time? Dear lord.
There are no hard iTrials, only hard peo-... er... wait.
Quote:
Until they give out the percentages, no one actually knows if they did what they said. Yes, I expected that I would have got a rare/very rare from TPN. A good portion of the players on that TPN run said they got a common or uncommon, with only a few saying that they got a rare or very rare.|
Originally Posted by EU_Damz
Well that hardly makes a fair chart does it
ONE TPN run and you dont get the rare/very rare? Well cant do a sarcastic post when you run just one TPN and state that the increased chance isnt there . . . . . . its called an increased chance and you need to run multiples of them before you can do the whole |
Quote:
|
Originally Posted by EU_Damz
That alone made the thread from a good/factual thread with actual figures to one that is once again hating on certain parts with no data
TPN/MoM/UG are all run now on Union quite often, slightly less than baf/lam/keyes but still quite often so the improvements DID work [at least for our server]. |
As to the commentary, take it or leave it. I could care less either way. The numbers are clear. If you have the records to show otherwise, or the developers, I encourage that the numbers get posted.
And Keyes? It is showing more Commons and Very Rares, not Rares and Very Rares. I noticed that you, and the other people that felt the need to post in an inflammatory way, glossed over that fact.
Quote:
|
Originally Posted by seebs
Dude, lemme just clarify:
Did you seriously just post a chart with a sample size of one? Come back when you have cumulative numbers from a couple hundred. |
As to the reason some trials are under-represented, the answer is simple: on the server I run on, they are being avoided. That alone should be enough for the developers to take notice and to run their OWN checks to see. They have far better access to the data than any player, including all the posters in this thread.
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters
Quote:
Okay so I've never seen that picture before and I must say I "LOL'ed" in real life |
Originally Posted by Paladin
|
Leader of The LEGION/Fallen LEGION on the Liberty server!
SSBB FC: 2062-8881-3944
MKW FC: 4167-4891-5991
Quote:
Also irrelevant in confirming a probability bias. Since each player's random roll represents an independent event, even if every person on a 24-person league got C/U drops, that would still be insufficient evidence to show that there wasn't an increase in the frequency of R/VR drops.|
Originally Posted by Snow Globe
A good portion of the players on that TPN run said they got a common or uncommon, with only a few saying that they got a rare or very rare.
|
Again, if ten different refs flip ten different coins and each of them gets tails, this is not evidence that any number of the coins in the sample is biased toward tails. If the hypothetical chance of getting a C/U is 85%, every person on the league could have just rolled 85 or less.
There are no hard iTrials, only hard peo-... er... wait.
Quote:
Then the developers can publish the actual percentages so we can have accurate expectations. I think that they are scared to do so as players would slow down or stop doing the later trials.
|
Originally Posted by Iron_General
If the hypothetical chance of getting a C/U is 85%, every person on the league could have just rolled 85 or less.
|
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters
Double post much?
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters
Quote:
If you find that you can't trust the devs when they make claims about how they modify the game mechanics, then you should also have no faith that the mechanics they post, numbers or otherwise, would be any more believable. |
Originally Posted by Snow Globe
Then the developers can publish the actual percentages so we can have accurate expectations. I think that they are scared to do so as players would slow down or stop doing the later trials.
|
Quote:
|
Originally Posted by Snow Globe
Double post much?.
|
There are no hard iTrials, only hard peo-... er... wait.
Quote:
|
Originally Posted by Iron_General
Also irrelevant in confirming a probability bias. Since each player's random roll represents an independent event, even if every person on a 24-person league got C/U drops, that would still be insufficient evidence to show that there wasn't an increase in the frequency of R/VR drops.
Again, if ten different refs flip ten different coins and each of them gets tails, this is not evidence that any number of the coins in the sample is biased toward tails. If the hypothetical chance of getting a C/U is 85%, every person on the league could have just rolled 85 or less. |
This is true. However I think we need to remember one caveat--you cannot build incarnate powers with "chance." You build them with the salvage that is the result of chance.
In other words, a player who has run 500 trials may not be in a position to make a claim about overall drop chances. However, the player is able to make a solid statement about the results of those invisible odds, and how he or she feels about them.
I don't completely back Snow Globe's conclusions that the odds of getting a Rare or Very Rare aren't much higher in some trials than others. However, I can look at Snow Globe's data and draw conclusions about the position this actual player is in in terms of advancement. I can do the same thing for any of my characters.
FWIW I have one character I have been waiting for a V Rare drop on for a while. I think have gotten two or so total. Meanwhile I currently have three or more of every Rare available. But these drops are mostly useless to me; rolling a Rare isn't "almost" as good as rolling Very Rare, and rolling 20 Rares you don't need isn't as good rolling Very Rare one time. I just hoarde them in case they are useful later, and continue waiting on the one big drop I need. I don't really care what the overall drop chances are, only what I personally receive. It might also be useful to note I was against a random drop system from the beginning.
100% agreed Tex. However, I wasn't disputing Snowglobe's feelings. I was disputing his conclusion that the TPN does not have an increased chance for a R/VR, based on having run one.
There are no hard iTrials, only hard peo-... er... wait.
Increased chance to drop rare/very rare in Keyes does not state that the rare/very drop chance would be higher than common/uncommon drop rate. It simply says that the rare/vr drop rate would be higher than the rare/vr drop rate was previously, which I have noticed to be true. (No, I don't have a pie graph to show you my statistical refrence, I'm not ocd enough for that)
Also, sample size is indeed too small. If you got say 5-10 people together (because running 4000 trials yourself is obviously a stretch) and had them all record their drop rates as you did and compile the results, then that would work to prove a point.
Quote:
To be blunt: the only people with access to that kind of data are not sharing it. The only group with that data are the developers, and you know it.
|
Originally Posted by DreadShinobi
Also, sample size is indeed too small. If you got say 5-10 people together (because running 4000 trials yourself is obviously a stretch) and had them all record their drop rates as you did and compile the results, then that would work to prove a point.
|
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters
Quote:
Technically, he could, but his conclusion wouldn't be considered to be as reliable as a test done with a larger sample size. Say, n = 2.
|
Originally Posted by Iron_General
No, just once, in error, and I deleted it when I realized it. Again, you can't conclude anything on a sample size of one.
|
61866 - A Series of Unfortunate Kidnappings - More than a coincidence?
2260 - The Burning of Hearts - A green-eyed monster holds the match.
379248 - The Spider Without Fangs - NEW - Some lessons learned (more or less.)
Come visit Pinnacle, we've been running TPNs with some regularity lately.
I'll keep track of my trial results and provide them for your analysis, Snow Globe. I disagree almost completely with your opinions on the incarnate system, but I'd like to contribute instead of posting snarky non-sense like I did above.
Virtue's a hotbed of incarnate activity, and I usually get around 4 trials a night in my quest to claim every T4 ability out there. This includes Underground, Keyes, and the new trials - it's usually active in the evening PST after 8 pm, and if no one's forming anything it's not because they don't want to join, it's because they don't have anyone leading. At that point it's easy enough to form up a league yourself in about 20 minutes for the Underground, or 5-10 minutes for the others.
The Paladin
Steel Canyon, Virtue
Exalted
@Paladin
Your data is unreliable. You are not controlling for any variables that determine drop rates outside of completing the trial. Any conclusions you may draw based on this fundamentally flawed data are wrong or right by accident.
Edit: Unless the devs divulge drop rate determination, this exercise on any data set, dev provided or no, is just that, an exercise.
Badge-Hunter: @Dogma
CIT: @Dogma
XFire: BrckaLo
Playfire: BrckaLo
GamerDNA: CoHDogma
Quote:
Again, the only people with that data aren't sharing. If you are that convinced my methods don't have controls, then get the developers to share their data.
|
Originally Posted by Dogma
Your data is unreliable. You are not controlling for any variables that determine drop rates outside of completing the trial. Any conclusions you may draw based on this fundamentally flawed data are wrong or right by accident.
|
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters
Quote:
Well, I don't believe they are going to give you that information nor do they have to.
|
Originally Posted by Snow Globe
Again, the only people with that data aren't sharing. If you are that convinced my methods don't have controls, then get the developers to share their data.
|
"while some people would say fish, cow, ambush!"-Ice9
Quote:
Neither do I and I don't expect them to. However asking me to provide clinical trial data to that extreme is going to get the exact same type result as asking the developers to release that data.
|
Originally Posted by Elric
Well, I don't believe they are going to give you that information nor do they have to.
|
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters
Hi all, time for another update in the wonderful world of trial rewards.
First off, straight values for the trials (by major patch):
Click for the LARGE version
Next we have the percentages of fail/threads/common/uncommon/rare/very rare for the trials (by major patch):
Click for the LARGE version
Finally we have the percentages without fails or threads included (by major patch):
Click for the LARGE version
I suggest looking at Keyes for the last two periods. The supposed increase of Rares and Very Rares. Instead, more Commons and Very Rares are showing up, with a decreased chance at Uncommons and Rares.
And yeah, [dripping sarcasm] Underground, TPN, and MoM are really being run often on Triumph [/dripping sarcasm]. The one TPN trial, as well as one BAF & one Lambda, were on Virtue.
Edit: ---------------------------------------------------
Even if you don't agree with the commentary, take a look at the numbers.
The results of the TPN/MoM/UG were included to show where the overall totals came from. If you think a particular trial lacks significant data, then feel free to draw your own conclusions.
As to the reason some trials are under-represented, the answer is simple: on the server I run on, they are being avoided. That alone should be enough for the developers to take notice and to run their OWN checks to see. They have far better access to the data than any player, including all the posters in this thread. I'm usually on between 6-7pm to around Midnight Pacific time every night. This is not exactly late, even on a server like Triumph.
Edit 2:
I guess that I should point out that these trials are across 10 characters, not just one.
Bedazzlement: Mind/Psi Dominator
Blastbot: Bots/Dark Mastermind
Dr. Cobalt: Rad/Rad Defender
Little Imp: Fire/Fire Tanker
Mechanized Maiden: Bots/Pain Mastermind
Pearl Angel: Broadsword/Shield Scrapper
Shi Otome: Ninja/Ninja Stalker
Snow Globe: Ice/FF Controller
Strobe: Peacebringer
White Succubus: Illusion/Empathy Controller
Triumph: White Succubus: 50 Ill/Emp/PF Snow Globe: 50 Ice/FF/Ice Strobe: 50 PB Shi Otomi: 50 Ninja/Ninjistu/GW Stalker My other characters