Seasons are technically a substitute for power creep. D2 has not seen a power creep for many years yet those who even play it regularly have the feeling of progression.
you dont have to test anything - theres plenty of sources / guides / simulators already for the next patch. google it.
No, you see, if you make a claim "here's the reason why those 2 sets are so much more OP than others, and why they should recieve another nerf:" and then tell me to go off and search for the reason myself on Google, then I can't really give your argument much merit.
That x1000. OP is just a troll and this whole thread should be reported for lacking substance. Thanks for taking the effort, but I fear OP just wants us to get frustrated.
My point is: I really don't see the problem with the paragon system. If you don't wanna grind levels - then don't. Noone is forcing you.
Most of the people discussing the paragon system belong to either of two sides : 1- the side who hates the system , 2- the side who tries to argue that side 1 is not legitimate. I've never really heard a good argument in favor of the paragon system.Truth is, it is plain bad: it encourages repetitive and boring gameplay, it promotes the use of bots, it rewards time spent in-game instead of proper gameplay achievements. People have to realize that the alternative to parangon is NOT simply the same game without parangon. The latter would suck, too! There are plenty of alternatives! But of all the ways to provide a "carrot-stick" to the players, the paragon system ought to be about the worst. Changing the carrot-stick for a better one would be to the benefit of all.
You need about 1-2 million tries to determine ~ mathematical expectation. I know it because I worked in casino and bots on our servers spin each game from 1 till 2 million spins to determine mathematical expectation and then we verified results.
There is no sense to do it on 100 or 500 or 1000 tries.
You can estimate the probability of success with any number of tries. 1 is already enough. I know this because I am a mathematical statistician. More tries gives you more precision in your estimate, in that you are correct. So what can we tell with only 30 tries? let's see:
Each trial follows a Bernoulli(p) distribution.
Number of success S for N=30 trials follows a Binomial(N, p) distribution. The probability mass function of binomial is:
(N choose S) p^S (1-p)^(N-S)
We thus have S ~ Binomial (30, p). The data collected is S1 = 13. OP says it seems a bit low; he questions whether p, the probability of success, is really as high as 0.6, given that he observed S1.
We consider testing H0 : p >= 0.6 versus H1: p<0.6. That is, we want to test whether our data has sufficient evidence to rule out a probability of success of 60%, or more.
Several test statistic are available to answer this question, each with their own power functions for a given level. Here we consider the p-value test. We ask ourselves: given that p is 60%, what is the probability that I have 13 successes - or less - in 30 trials? If this probability is smaller than a fixed level (typically e.g. 0.05), then we decide to reject H0, that is, we reject the hypothesis that p is indeed greater than or equal to 0.6.
This probability of having such a bad luck as you did is straightforward to derive:
Here, you barely reject H0 at a 95% confidence level. So your data, at 95% confidence level, goes against the hypothesis that the probability of success is indeed 60%. It would suggest a lower probability.
What does 95% confidence level mean?
It means that were you to decide in this way everytime you make such a test, you would incorrectly reject the hypothesis (that p=0.6) 5 % of the time (this is an exact result). Giving the benefit of the doubt to Blizzard's developers, it seems that your case belongs to this 5%. That is, if everyone complains whenever they obtain a p-value < 0.05, 1/20th of the people who do this test will end up complaining, even when the probability of success is 60%.
And in everyday English?
It means that if indeed the probability is 0.6 as advertised, on average only 4.8% of the players who make your test (make 30 upgrades, observe S1) have a result that is as low as you or worse.
In other words, you belong to the 5% most unlucky players. It is rare, and some people in science would already consider your data as enough evidence to doubt that p=0.6, but it is not extremely rare. Indeed,it cannot constitute a proof. More than anything, it calls to make another such test... and you'll have only 1 in 20 chance to repeat such a bad luck.
Frankly, I can't really think of any other game where this would be different. There're always top-builds. Tried being in the tops. Boring. Staying in the 'gray zone' ever since for fun and embracing all varieties of builds :3 Quite new to d3, but not gonna change my attitude here. And tops seem to be occupied by bots in d3 anyway. Don't think they care about variety. Casual players and new-comers have got much choice here so far.
OK so. None of this is a personal attack against you....but here it is:
MANY games do not suffer from this disease of "best build"
Starcraft did not have best units......Hearthstone does not have best cards........and Wow does not have a best arena or raid composition.
You can't compare these with diablo. Starcraft is a strategy game: you adapt to your opponent. If your opponent does strictly the same thing all the time, you will find one best way to beat him (e.g. in the campaign).
Hearthstone's "campaign" mode suffers from the very same thing: for every boss, you have one best deck to defeat it.
It's like chess: if someone plays always in a pre-determined way with always the same moves and answers, there will be a best way to beat him.
So once you re-center your argument on something that is comparable, it falls appart. If you spent time reading the answers here, you would note that the real issue is that diablo end-game is focused on a single objective. If there is just one objective, and it does not change or adapt itself, there will be one best way to beat it.
Wyatt Cheng goes one step further than your (and ultimately quinn's) analysis. He realizes that having a very large choice set means you can identify the best solution, and this means you really only have one choice. The idea is then to create multiple choice sets, not just one. These items, while not the solution by any means, are a step towards this idea. It allows to balance things in such a way that different skills can have pretty much the same overall efficiency at clearing GR. Then it is up to the players to choose the skill, and the items that go along it. You end up with different gameplays to clear the GR as efficiently -- in short, you increase variety.
We should not complain about them reducing variety while all they are doing is just the opposite.
I welcome this change of mindset. But what we need I think (and we are many to share this thinking) is multiple end-game objectives, not just GR.
There is 0 item diversity in this game. And the new change, doesn't change a squat.
Every single one of you who played over paragon 1 knows, that for end-game builds, you have spesific items you must own (obviously with good rolls) to successfuly run high GRs.
Take a good look on the builds page, do you see alot of items diversity? No, you don't!
You can't run with w/e weapons you like on WW Barb. No, you have Bul Kathos for the best WW performance. Sure you can use IK weapon. but it won't be nearly as good.
Same goes for DH current weapon of choice for MS build.
You are in denial, if you think you can choose w/e the fuck you want to wear and get away with the same/better performance on current builds in-game.
itemization only works on Torments.
You cant do much tweaking once we talk GR60+.
This entire thread, along with Quin's, is a waste of time. Unless you're a casual who don't run high GRs and care about minimal gear swapping for gameplay diversity.
I just have to agree with this.
Diablo 3's itemization has been an illusion all along. Whatever choices you think you have, there will always be 1) one build that preforms best, and 2) corresponding best-in-slots items for every slot.
The difference is, now we don't need to look out for which build/item choice works best, Blizzard gives a straightforward answer.
For the less casual players, it means you don't have to figure out things. But if you look a tthe build diversity, I'd say less than 0.1% actually had to figure out anything.
For the very casual players, finding one such item will give them a surge in power, with the excitment that usually accompanies that.
In short, this kind of items grants casual players exciting moments in the game, while they just remove "high-lvl" players' illusion of choice. They just don't realize it was just that all along-- an illusion.
Some, like me, like to test different builds and it is true that the possibilities are now more limited. But whoever, like me, tried other things, must have realized how boring it quickly became when these other builds just sucked.
On the other hand, for Blizzard, blocking these item slots makes the game way easier to balance, also with respect to the set dungeons. That means more fun in general.
The general idea is to make diablo 3 endgame and leaderboard competition more attractive for the more casual / non-botter player. As mentioned, the mere time you spend on the game makes your character more powerful, and gives you a tremendous advantage for the leadboard which some consider unfair.
Leaderboards are meant for competition, and competition is often meant to be skill-based (tactical, strategic, execution, ...) rather than time-based. More often than not, however, the amount of time you spend affects your skill, but this is still considered fair.
In diablo 3, time played gives you at least these advantages:
better items
better rift levels (Grift fishing)
improved skill and techniques
more dps/survival from parangon and XP
An ideal competitive system would eliminate all advantages but improved skill and techniques.
Your solution proposes to get rid of the parangon problem (although the same problem still persists, to a lesser extent, within categories). But the remaining problems, which you would not fail to notice and write about once your proposition goes live, still remain.
In short, someone who invests more time in the game will still have an advantage over you, even within the parangon leagues.
There is no way to address in a straightforward manner all the above-mentioned problems. Parangon leagues (or having a separate league with parangon bonus capped at 800) somehow dampen the problem but doesn't solve it by any means.
One idea that has surfaced could be the following:
Similar to the set dungeons idea from Blizzard, seasons could propose a static dungeon withincreased difficulty per dungeon level and "saves" at each level once the previous has been completed. Upon arriving into the dungeon, all items's stats currently worn are standardized to their maximum possible value. Parangon bonus is capped at 800.
Each class will have to figure out how to beat the current level, and leaderboards can keep track of who has advanced the most inside the dungeon, by level, and then % of level completed.
I don't want to diverse into discussing the obvious advantages of this idea (which isn't mine by any means): true, fair competitive setting, increased community activity, almost uniquely skill-based (still need to access the ancient items), more excitment in the game, meaningful in-game goals, leading to better gameplay experience, leading in the community at large better streams, increased popularity of the game, etc...
TL;DR: What you truly want is not to tweak the GRIFT experience to make xp grinding less relevant but to add another game mode that is truly based on skill/strategy/tactics/execution and where time played does not give any advantage at all after a certain point (required to have the proper items). If we had such a mode, chances are that you wouldn't want to make any change on the current GRIFT leaderboard paradigm, as the only leaderboard that would really count would be the skill-based one. Instead of asking for half measure, we should investigate more into the best competitive mode, which the GRIFT can never be.
There is a reason why you, sir, are not a game developer for Diablo 3. Well, plenty of reasons in fact, as many as your bad good ideas above. Don't take it personal, but this is no solution you are proposing, this is just plain wishful thinking.
I don't want to be controversial here -- and don't get me wrong, I usually like your content. But these builds you have been posting as of late are a bit lackluster in the creativity and research realm. I understand you want to have views and followers, so posting new content is important. But pretty much what it boils down to in your build videos is your having an awesome gear, making pretty much all builds with a semblance of thought viable.
The build you are showcasing in this particular video would have a lot of trouble going past GR30, even with your incredible gear. Spin to win is not back, at least not in this form. Consider a raekor-WW build instead, which has awesome movement, is great at packing mobs together (for teamplay, for AOE maximization, etc...), likely deals more damage and has more survivability.
What's more, it is impossible to get perma wrath. This would require 77% of effective CDR, which no barb gear can achieve, even with new leoric, gogok and belt. Part of the reason is that CDR and the passive boon of bul kathos don't work well together: the base on which CDR is calculated already includes boon's bonus. Too bad. That being said, 50% WOTB uptime is possible at a fairly low cost to dps, which makes it OK as part of a late-game build I guess.
I like your choice of Auguild, which people might have discarded as sub-par itemization, but which really is not. That being said, it is not that good with the new Furnace, because its bonus damage is additive, not multiplicative to other elite dmg bonuses.
Overall I think you started with a good idea, and made it worth the viewer's time because you have a truly amazing gear. But the build itself, the research behind it and the itemization is a bit lackluster. That being said, I still enjoy your content and am impressed by your productivity when it comes to sharing content with your viewers. If it can allow other people to refine these builds and think them through, it would be all for the better, so keep it up!
1
Seasons are technically a substitute for power creep. D2 has not seen a power creep for many years yet those who even play it regularly have the feeling of progression.
2
1
1
Too short answer.
You can estimate the probability of success with any number of tries. 1 is already enough. I know this because I am a mathematical statistician. More tries gives you more precision in your estimate, in that you are correct. So what can we tell with only 30 tries? let's see:
Each trial follows a Bernoulli(p) distribution.
Number of success S for N=30 trials follows a Binomial(N, p) distribution. The probability mass function of binomial is:
(N choose S) p^S (1-p)^(N-S)
We thus have S ~ Binomial (30, p). The data collected is S1 = 13. OP says it seems a bit low; he questions whether p, the probability of success, is really as high as 0.6, given that he observed S1.
We consider testing H0 : p >= 0.6 versus H1: p<0.6. That is, we want to test whether our data has sufficient evidence to rule out a probability of success of 60%, or more.
Several test statistic are available to answer this question, each with their own power functions for a given level. Here we consider the p-value test. We ask ourselves: given that p is 60%, what is the probability that I have 13 successes - or less - in 30 trials? If this probability is smaller than a fixed level (typically e.g. 0.05), then we decide to reject H0, that is, we reject the hypothesis that p is indeed greater than or equal to 0.6.
This probability of having such a bad luck as you did is straightforward to derive:
p-value = Probability_under_H0 (S <= 13) = SUM_i^13 (30 choose i) 0.6^13 (1-0.6)^17 = 0.048
Here, you barely reject H0 at a 95% confidence level. So your data, at 95% confidence level, goes against the hypothesis that the probability of success is indeed 60%. It would suggest a lower probability.
What does 95% confidence level mean?
It means that were you to decide in this way everytime you make such a test, you would incorrectly reject the hypothesis (that p=0.6) 5 % of the time (this is an exact result). Giving the benefit of the doubt to Blizzard's developers, it seems that your case belongs to this 5%. That is, if everyone complains whenever they obtain a p-value < 0.05, 1/20th of the people who do this test will end up complaining, even when the probability of success is 60%.
And in everyday English?
It means that if indeed the probability is 0.6 as advertised, on average only 4.8% of the players who make your test (make 30 upgrades, observe S1) have a result that is as low as you or worse.
In other words, you belong to the 5% most unlucky players. It is rare, and some people in science would already consider your data as enough evidence to doubt that p=0.6, but it is not extremely rare. Indeed,it cannot constitute a proof. More than anything, it calls to make another such test... and you'll have only 1 in 20 chance to repeat such a bad luck.
2
You can't compare these with diablo. Starcraft is a strategy game: you adapt to your opponent. If your opponent does strictly the same thing all the time, you will find one best way to beat him (e.g. in the campaign).
Hearthstone's "campaign" mode suffers from the very same thing: for every boss, you have one best deck to defeat it.
It's like chess: if someone plays always in a pre-determined way with always the same moves and answers, there will be a best way to beat him.
So once you re-center your argument on something that is comparable, it falls appart. If you spent time reading the answers here, you would note that the real issue is that diablo end-game is focused on a single objective. If there is just one objective, and it does not change or adapt itself, there will be one best way to beat it.
Wyatt Cheng goes one step further than your (and ultimately quinn's) analysis. He realizes that having a very large choice set means you can identify the best solution, and this means you really only have one choice. The idea is then to create multiple choice sets, not just one. These items, while not the solution by any means, are a step towards this idea. It allows to balance things in such a way that different skills can have pretty much the same overall efficiency at clearing GR. Then it is up to the players to choose the skill, and the items that go along it. You end up with different gameplays to clear the GR as efficiently -- in short, you increase variety.
We should not complain about them reducing variety while all they are doing is just the opposite.
I welcome this change of mindset. But what we need I think (and we are many to share this thinking) is multiple end-game objectives, not just GR.
3
I just have to agree with this.
Diablo 3's itemization has been an illusion all along. Whatever choices you think you have, there will always be 1) one build that preforms best, and 2) corresponding best-in-slots items for every slot.
The difference is, now we don't need to look out for which build/item choice works best, Blizzard gives a straightforward answer.
For the less casual players, it means you don't have to figure out things. But if you look a tthe build diversity, I'd say less than 0.1% actually had to figure out anything.
For the very casual players, finding one such item will give them a surge in power, with the excitment that usually accompanies that.
In short, this kind of items grants casual players exciting moments in the game, while they just remove "high-lvl" players' illusion of choice. They just don't realize it was just that all along-- an illusion.
Some, like me, like to test different builds and it is true that the possibilities are now more limited. But whoever, like me, tried other things, must have realized how boring it quickly became when these other builds just sucked.
On the other hand, for Blizzard, blocking these item slots makes the game way easier to balance, also with respect to the set dungeons. That means more fun in general.
2
Hi,
Nice post, good to read!
The general idea is to make diablo 3 endgame and leaderboard competition more attractive for the more casual / non-botter player. As mentioned, the mere time you spend on the game makes your character more powerful, and gives you a tremendous advantage for the leadboard which some consider unfair.
Leaderboards are meant for competition, and competition is often meant to be skill-based (tactical, strategic, execution, ...) rather than time-based. More often than not, however, the amount of time you spend affects your skill, but this is still considered fair.
In diablo 3, time played gives you at least these advantages:
An ideal competitive system would eliminate all advantages but improved skill and techniques.
Your solution proposes to get rid of the parangon problem (although the same problem still persists, to a lesser extent, within categories). But the remaining problems, which you would not fail to notice and write about once your proposition goes live, still remain.
In short, someone who invests more time in the game will still have an advantage over you, even within the parangon leagues.
There is no way to address in a straightforward manner all the above-mentioned problems. Parangon leagues (or having a separate league with parangon bonus capped at 800) somehow dampen the problem but doesn't solve it by any means.
One idea that has surfaced could be the following:
Similar to the set dungeons idea from Blizzard, seasons could propose a static dungeon withincreased difficulty per dungeon level and "saves" at each level once the previous has been completed. Upon arriving into the dungeon, all items's stats currently worn are standardized to their maximum possible value. Parangon bonus is capped at 800.
Each class will have to figure out how to beat the current level, and leaderboards can keep track of who has advanced the most inside the dungeon, by level, and then % of level completed.
I don't want to diverse into discussing the obvious advantages of this idea (which isn't mine by any means): true, fair competitive setting, increased community activity, almost uniquely skill-based (still need to access the ancient items), more excitment in the game, meaningful in-game goals, leading to better gameplay experience, leading in the community at large better streams, increased popularity of the game, etc...
TL;DR: What you truly want is not to tweak the GRIFT experience to make xp grinding less relevant but to add another game mode that is truly based on skill/strategy/tactics/execution and where time played does not give any advantage at all after a certain point (required to have the proper items). If we had such a mode, chances are that you wouldn't want to make any change on the current GRIFT leaderboard paradigm, as the only leaderboard that would really count would be the skill-based one. Instead of asking for half measure, we should investigate more into the best competitive mode, which the GRIFT can never be.
8
1
The build you are showcasing in this particular video would have a lot of trouble going past GR30, even with your incredible gear. Spin to win is not back, at least not in this form. Consider a raekor-WW build instead, which has awesome movement, is great at packing mobs together (for teamplay, for AOE maximization, etc...), likely deals more damage and has more survivability.
What's more, it is impossible to get perma wrath. This would require 77% of effective CDR, which no barb gear can achieve, even with new leoric, gogok and belt. Part of the reason is that CDR and the passive boon of bul kathos don't work well together: the base on which CDR is calculated already includes boon's bonus. Too bad. That being said, 50% WOTB uptime is possible at a fairly low cost to dps, which makes it OK as part of a late-game build I guess.
I like your choice of Auguild, which people might have discarded as sub-par itemization, but which really is not. That being said, it is not that good with the new Furnace, because its bonus damage is additive, not multiplicative to other elite dmg bonuses.
Overall I think you started with a good idea, and made it worth the viewer's time because you have a truly amazing gear. But the build itself, the research behind it and the itemization is a bit lackluster. That being said, I still enjoy your content and am impressed by your productivity when it comes to sharing content with your viewers. If it can allow other people to refine these builds and think them through, it would be all for the better, so keep it up!