Neverwinter "RNG" Database

trinity706

In an ambitious attempt to provide tangible, verified "RNG" drop/reward information players can send screenshots of their dungeon rewards/skirmish rewards, lockbox rewards, Heroic Encounter rewards, etc. directly to "DatabaseCenter" (without the quotes) on PS4 and Xbox (same name on both) via the console's message system or email them in compressed archived format (such as WinRar) to thedatabasecenter@gmail.com

- take a screenshot
- start a message
- attach the screenshot
- send
- use the previous message to send additional screenshots

After viewing a number of "community" edited drop tables none were seen to confirm the information provided, while this doesn't necessarily make them invalid human error (whether intentional or unintentional) can and does happen. A drop/reward table with verified information to support it's entries not only provides validity but that validity can be the basis to support critiques, responses, etc. of the "RNG" system in Neverwinter.

This will not be an easy task and will take quite some time (months) of consistent information to even begin to draw from the information compiled. It will take 100's of runs of a single dungeon/skirmish before even a tiny glimpse of what drop rates "might" be can form. Things have to start somewhere though.

Spread the word, the more players actively contributing the more information that can be compiled and the sooner things can take shape.

- Currently it is simpler to use the PS4 and Xbox apps to retrieve/view screenshots and to verify their legitimacy.
- Messages will be viewed for the screenshots they contain, any unnecessary information, questions, comments are likely to be ignored.
- Edited screenshots will NOT be accepted.
- Screenshots from all participants of a dungeon/skirmish run are GOLDEN.
- Older screenshot are invaluable.
- thedatabasecenter@gmail.com for players that want to get fancy and send bulk screenshots in compressed archived format (such as WinRar) transferred from their consoles to a PC via a USB flashdrive (possibly videos as well)(you know the same PC many users use to type their posts LoL).
- The DatabaseCentral's PC storage is in the process of being upgraded above it's current 6TB dedicated for this information.

randomdigits

Sorry, but:
1) You can't have a verified RNG stats database unless you can verify that all the results are submitted without exclusions.
2) Some of the things I'd like to see stats for (like HE rewards and DH/Fey/QM utility enchants) are seemingly affected by some internal cooldown mechanism which is fine tuned to give the most rewards within the first hour or so of playing per day. Just knowing the drop rates would be insufficient in this case.

omegarealities

trinity706#8838 said:

This will not be an easy task and will take quite some time (months to years) of consistent information to even begin to draw from the information compiled. It will take 100's of runs of a single dungeon/skirmish before even a tiny glimpse of what drop rates "might" be can form.

Spread the word, the more players actively contributing the more information that can be compiled and the sooner things can take shape.

I really do not see much point to this as each time there is a patch these drop rate can change radically. Next month the drop rates could be completely different from the current ones.

Add to the fact that I think that most people will not want to take the time to take a screenshot, download the screenshot and then upload the screenshot of every dropbox/chest from every dungeon, skirmish and heroic encounter they participate in.

trinity706

@randomdigits#2166
Unfortunately you have mistaken what the term "verified" means in this case, it means that there will be a screenshot/video evidence to support the information in the database.

Since this is a "RNG" database all things "RNG" related will be contained within, apologies for using "etc." instead of listing each and every example of "RNG" related things in the game.

HE rewards, utility enchantment procs, Wanderer's Fortune, etc. are already being tested.

@omegarealities#7219
The point is changes to rates will be visible between patches/drop rate changes (if some are made) and with previous data there is something to compare things to.

Players don't have to download/upload screenshots via a website, they can be attached directly to PSN/XBL messages and sent to the profile "DatabaseCenter", if players want to send them in bulk they can transfer them via USB to a PC (that a number of users utilize to post on the forums anyway), compress them and send them via email which will only take a few moments depending on how many screenshots are involved (compression/archiving would be completed long before a number of long forum posts can be typed out).

If some players don't want to take the time to send screenshots that's fine and they will continue to gripe about drop rates with nothing to support their statements. The players that are contributing are doing so because they want to, regardless of the effort it takes which isn't much. The database will still take shape.

randomdigits

Sorry for not being more clear, my first point was about what's known a https://en.wikipedia.org/wiki/Survivorship_bias.
In this case it means that people far more likely to submit proofs for good drops than of bad drops. Thus the calculated drop rate would be artificially inflated (biased). (I'm a math major with specialization in statistics, if that's not already obvious )

My second point was about the cooldown variable. It would take a lot of testing to understand the inner mechanics of cooldowns - what triggers them, how long do their last, what's the impact on the drop rate. I was thinking myself about doing this analysis for lanolin drop rate in Lonelywood, but I never managed to do this. I image that in such analysis one would record all drop times and results - for a few weeks and then would fit a model of cool down logic - which may have several variables in it. Taking lanolin as an example, my starting hypothesis would be that the rate is about 33% in the first hour but drops to around 10% after that. Similar observations for Evidence of Evil drops in ship heroics.

But then again, what if there are other factors affecting that drop rate (such as your damage contribution). Before starting that work I'd very much like to get some feedback from devs about the contributing factors.

On the other hand, if the main goal is to detect RNG nerfs/boosts - it might be sufficient to standardize the test conditions and repeat them periodically comparing the results.

trinity706

@randomdigits#2166 players aren't forced to supply each and every drop/reward, if they only provide "good" screenshots then so be it, the data will be noted accordingly. The group of players that are sending in all dungeon/skirmish runs, etc. are part of the control group, even then it's not guaranteed that all will be submitted which is ok, the data will not be handled as 100% accurate.

The only people who have the ability to access the drop/reward information of ALL players is Cryptic and since they more than likely won't allow outside people access to that information (even for something like this) it is being done from the outside and consequently with limited participation. It would be great to have large groups of players that contribute 100% of their drop/reward info, etc. but that is highly unlikely though the database is growing slowly but surely regardless.

A person doesn't have to be a math major to create a database such as this or to know which variables are more suitable to be used as criteria in analysis, those on this end aren't new to this. This isn't a spur of the moment idea for Neverwinter. Damage other than in dungeons/skirmishes would be a nightmare to try and calculate in open world. With limited resources bias has been embedded from the start, the control group is nearly the closet thing to unbiased information provided though still won't be considered as completely unbiased, to go further the closest thing to pure unbiased information would be from those on this end that meticulously log EVERYTHING "RNG" related (no exaggeration).

A number on this end compile detailed information for at least some of the games they play if not all.

kclowe

It seems the only way the data collected will be accurate is if every single drop is sent in to be marked down. If half the people send in all there rewards than only half the data is being collected. Also whoever does send in their screenshots will have to be sure they send in eveyone of them. Otherwise the data is incomplete on a number of levels. Correct me if im wrong.

trinity706

kclowe said:

It seems the only way the data collected will be accurate is if every single drop is sent in to be marked down. Correct me if im wrong.

*Incoming correction* @kclowe

The consistent information from the control group alone is "accurate" (to a degree), more accurate than the scattered "unverified" information presently available from various sources. Information from 100 players is more accurate 50, the more information that comes in the more accurate the database becomes. Stating or insinuating that for something like this to be "accurate" it would require 100% input from all players, in a light tone that is silly, "accurate" is relative. It would pretty much be impossible at this point for the database to be 100% accurate, it is highly unlikely for it to happen anyway knowing that people will be people and not all will contribute for whatever reason(s) though it's accuracy is increasing which is fair enough.

Even if some players outside of the control group only send in some of their screenshots the information from them can still be held up in light of the information provided by the control group. The lowest amount of runs, lockboxes, etc. for a specific item/rarity and the highest number within a specific time frame based on information provided by the control group will set the scope(s) (will probably create an image(s) depicting this).

--+-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+ +-~* *~-+--

"RNG" in Neverwinter is not cumulative for the most part even though some players treat it as such which tends to garner frustration. Opening 100 lockboxes does not guarantee a desired result, neither does opening 1k though knowing/having evidence of 100's/1000's of lockbox rewards helps others draw logical/reasonable conclusions rather than conclusions pretty much solely based on personal opinions, bias, hearsay, etc. Bias in this regard is how players often disregard the cumulative value of dungeon/skirmish/lockbox rewards. Players opening lockboxs are guaranteed something of "value" (often quantified by market value), if lockboxs had the chance to be empty it would be a different story altogether for the most part.