Failing 16 out of 24 times at 60%+?!?!

General Discussion
Here's my theory on the real works of all this:

0% to 20% =0% no dice!
30% to 50% =25% small chance but don't hold your breath!
60% to 70% = it's a 50/50 thing!
80% to 90% = almost a sure thing with a dash of RNG for the right contender!
and 100% = winning!

...For most, I think it's as close as it gets for realism.
06/15/2016 11:12 AMPosted by CyberDoc
^^ I think the translation works out to 'Days containing an I character' (cap i not lower case L)
Thanks, now I'm going to start failing my 60% rolls! /maudit verrat!
HaHa..happenes all the time. And yur surprised at this? O.o
06/14/2016 08:42 AMPosted by DustInDaWind
Law of Total Probability:
If B1,B2,B3,⋯B1,B2,B3,⋯ is a partition of the sample space S, then for any event A we have
P(A)=∑iP(A∩Bi)=∑iP(A|Bi)P(Bi).


Now that we have cleared that up... ... I'll hold out for my Heart of Gold to appear out of nowhere, along with a sprem whale and a bowl of petunias.
This topic comes up over and over and there are so many theory's out there but until someone takes the time to do a controlled test with a huge sample size its all assumptions period.

Its possible to do 100 gem upgrade attempts and fail all 100 as its calculated per try, not did you get the upgrade last time. Now is this likely to happen, hell no, but it is in the realm of possibilities.

Is the gem upgrading chance skewed a bit? I don't know.

Dose anyone on the forums know? Not likely.

Blizzard is the only ones that would know and since there is no reason to have a calculation error and not fix it I tend to believe its calculating correctly and people are just thrown off by there limited sample size results.
06/15/2016 03:44 PMPosted by Kriton
Blizzard is the only ones that would know and since there is no reason to have a calculation error and not fix it I tend to believe its calculating correctly and people are just thrown off by there limited sample size results.


I'll plus one your post and offer this in reply:

It's not that people are thrown off by their limited results.

It's that the game is thrown off by the devs endless and comprehensive results.

The devs think: So long as they can glimpse towards the horizon and see that the target 60% is being hit, after a countless many number of scientifically accurate tests involving the game code and everything controlled....they think their work is done.

Then you get a hand-full of players venting their disgust over failing a string of boring as hell 60% upgrades as fun as flipping a coin...

Then you get players who support the devs above and the game as it is, unflinchingly, all puffed up like a pack of retarded pigeons claiming to know "oh so much" about statistics,

They mock the players who are bored and/or frustrated,
and they stand in support of the game and everything it currently has to offer,

And as a result the game's growth and evolution is perpetually stunted.

if people look at a few 60% fails and say "derp the math checks out" the game sits idle...things are as they should be.

If they look at it and question it and criticise it, it has a chance to be improved.

This is why people who recite the line RNG is RNG love to think they are so wicked smart, but in reality they are worthless; They don't care about the game getting any more interesting than flipping a coin with Urshi every few minutes...all they care about is RNG being RNG...
06/14/2016 03:26 AMPosted by MikeAkaJB
Each roll is independent of one another. Fail / success has no influence over your chance of success in following tries. Look at a roulette wheel, the lottery, or flipping a coin. Just because the ball hit black 15 times in a roll doesn't mean the next is going to be red. You could literally flip a coin 100 times and come out with heads 80% of the time even though it's a 50% chance for either to show up. Just because the lottery is now up to $500 mil and no one has hit out of the millions of tickets purchased in previous drawings doesn't mean a winner will be drawn.
Very typical response.

But players don't form opinions objectively, based on hard, cold facts and statistics; we form opinions subjectively based on how the game "feels."

That's why I continue to assert that the game needs parametric randomization; what is coming is based in part on what has come before.

The game already does this with Legendary gems, for example. Start doing GRs, and the game will give you one of each until you have them all, then stop. Use one for Caldesann's Despair, and that one will drop on the next GR you do. The "Pity Timer" in days of old used to do a similar thing, but based on time elapsed, not item in inventory.

This game already tracks and adjusts for hundreds of things in real time, as you are playing. Urshi, Kadala, Kulle, and Myriam could all do the same thing.

One component is sadism on the part of the devs. Ever tried to get a pair of Ancient UE gloves from Kadala? Any UE gloves at all? The devs know that the UE set is the most powerful set for the majority of DH players, so above all items, they make the drop rate on gloves the lowest of all.

Another is the brainless RNG system this game has, which could be made much smarter, with a few instructions.
06/14/2016 08:42 AMPosted by DustInDaWind
Law of Total Probability:
If B1,B2,B3,⋯B1,B2,B3,⋯ is a partition of the sample space S, then for any event A we have
P(A)=∑iP(A∩Bi)=∑iP(A|Bi)P(Bi).


I thought that a game should be fun? That certainly does not look like fun. And One shouldn't need to have a masters in math to enjoy a game either.
06/14/2016 09:51 AMPosted by UnifiedEarth
are you keeping the GR level 13 levels above the gem? if you do then the chance is 100% for all 3 or 4 tries.


and this is not always possible. If I'm pushing the limit griftwise, then I'm going to be stuck with those 13 levels. Not all of us are running these micro builds and pushing high 90 grifts you know (and neither should we have to in order to ENJOY the game).

Personally, I think Blizzard has it wrong with the current upgrade chance - you should have 10% upgrade chance up to and including the grift level that you're running. If you're trying to upgrade a lgem to a higher level than the grift that you're currently running, then odds should diminish. But hey, Blizzard has a long history of not being able to balance games imho.
I once went 18/18 at 60% success. Ended up 22/24 that day.

RNG...
06/14/2016 10:33 AMPosted by adm0ni
Don't listen to the clowns telling you to take a statistics course. It is they that need the refresher.

The best bet you can do is maintain a log of your pass/fails and submit them to the bug forum. I would suggest you don't submit anything with less than 30 attempts. More would be better.


Yeah, but Blizzard doesn't care. Especially if you're on a console. *^_^*

Sadly, these forums are full of people who just are well, for want of better words, nasty. They exist to attack others posts, ridicule other players (aka bullying) and call other players names (troll is a commonly used example). Of course, Blizzards mods don't care and don't do anything about this rot.

I too have had this lgem - One success, then 8 in a row failures @ 60%, then a success, then 2 more fails. 2 from 12. I've never seen this happen in reverse!

The "maths wizs" all come up with "too small a size sample!" or "you're a troll" or "you're lying", but as I said above, that's par for the course with these forums and the mostly low life users that inhabit them. Yeah, I call a spade a spade. I find it amusing (but not unsurprising) that Blizzard bans users such as myself for saying this, but does nothing against those who are name calling (troll!) or bullying other players and making fun of them. It just reinforces my opinion that Blizzard does not care for anyone who dares criticise their game.

I bought DS:ROS (ps4) for a mate 2 months ago and he's never played it. When I gave it to him, his first comment was "are item drops still crap?". He then went onto make a comment to the effect that he dislikes Blizzard's RNG implementation - and for exactly the same reasons why I do. It makes no sense to see an "unlucky" player miss out, but a lucky player get "loads of hits" when they both put in the same effort. The game should be rewarding players for effort, not luck.
06/14/2016 09:44 AMPosted by Shaolin
I think you're crazy to do empowered rifts. Just run a couple of rifts instead. About the silliest thing diablo introduced.
Because gold has any sort of meaning, right? ...oh wait.
06/16/2016 03:05 AMPosted by mrtheoden
I thought that a game should be fun? That certainly does not look like fun. And One shouldn't need to have a masters in math to enjoy a game either.
Then don't worry about the 60% and complete your next rift and try again!
I've had 3 failures at 60% several times. My bad is assuming 60% means 'I should get this' where as 50% I'd think "total coinflip". Only a 10% increase but it changed my mentality more than 10%. Obviously each roll is independent of last everyone knows that.

I'm sure you've all done this, I accidentally clicked the wrong gem to upgrade to hurry up to please a random group to be nice, it was at 3% and success lol. Bane of Trapped of course, because Trapped is used in almost every build if not literally every single one right? Different thread topic right there though.

Maybe devs should create Bane of the Upgrade Chance where you increase your upgrade gem chance through a gem? Just a dart at the board.

GL OP change your mindset to 50% chance works for me.
Looks bad because nobody makes threads about

"I got 6 success in a row at 60%! The game is broken, please fix Blizzard!"
60% is illuminati confirmed. 30% works more often than 60%.
06/16/2016 07:36 AMPosted by Shanso
Looks bad because nobody makes threads about

"I got 6 success in a row at 60%! The game is broken, please fix Blizzard!"


well, I'm yet to ever see Six straight successful upgrades @ 60%. I've had Six straight fails at 60% many times though. Now, I'm not a lucky person, in fact, many of my friends consider me to be the unluckiest person that they've EVER met. But - a game should reward you on effort, not luck, so my bad luck shouldn't even come into a game. This is where RNG that is intelligent and looks at how many fails you've had and prompts for a success if you've been unusually unlucky is a great idea and a far better way of implementing meaningful RNG imho. I guess if you're a naturally lucky person, and generally have good luck with RNG, then you don't care. For us unlucky sods like me, we do care about it. And it limits our enjoyment of the game as you can appreciate. Not everyone is born with the luck of the Irish...

edit: to give you an idea of my "luck", I once played a game of Two up (ANZAC game in Australia, played only on ANZAC day) where you flip a coin and pick heads or tails. 27 straight goes before I guessed correctly...and only One beer was involved (you typically play the game on ANZAC day at your local RSL having a few beers with mates and remembering those who fought in WW1 and WW2 and the Vietnam wars etc).
Sounds about right.

Join the Conversation

Return to Forum