Great work!
Thank you for sharing! I think this survey method works.
I love the way you communicate with us, and also love your game. With this kind of attitude towards your players I expect a great future for this game and many new players to come!
So, great communication. The results of the survey are definitely not surprising.
I am definitely not happy with the leaderboard solution though. I donât understand why you donât just purge the entries for bugged builds that you fixed.
Using timestamps is sketchy because many people legitimately that got their high score during but they are still competing with the bugged ones.
Ideally the bugs caught before a cycle are released are more highly prioritized as these recent bug builds had been reported for quite some time.
You are really great this is how a gaming studio should communicate and listen to their players. Every other gaming studio could learn something from you. I believe you are on a right direction with this altitude to become top ARPG in a genre (for me you already areâŚ)
The official site used to show at least 100 rankings in beta, iirc.
So the data is there, it just currently isnât being shown in-game. Them choosing not to show a giant-ass list doesnât mean they donât record your Rank if you would have gotten 51st, there is no proof of that. Again, nothing to do with design philosophy, just how data is shown ingame (or on 3rd party sites)
If you donât understand that data may exist without it currently being visible, I canât comprehend it for you.
I wonât even dare read the comments because there surely be the usual very small vocal minority (as the pool showed) that will complain like babies.
Good stuff, great outcome from the community. Itâs a balanced decision and overall makes sense considering that even the mildly over-performing might get patched anyway at the end of a cycle, so itâs a matter of patches in 1-2 weeks vs 3-4 months, not âpatches vs non-patched everâ.
Keep being open and listen to community feedback EHG. You are doing great, and this shows also what kind of feedback can be safely ignored that might be erroneously appeared as majority when reading Reddit or the forums.
Bro there are people in this thread who have been quoted the post that says âOur stance on balance changes isnât set in stoneâ minimum 3 different times
Still saying that they feel misled because EGHâs stance on balance changes seemed set in stone.
Nothing will ever convince these people they were wrong for thinking that, or that the majority disagree with them, itâll always be âthe loud minorityâ even when theyâre outnumbered 3:1.
Please consider not using google for your surveys in the future. This prevents privacy conscious people from participating. Thank you.
I am once again asking for an official stance or even a similar survey or something like that to address the issue of severely underperforming skills. Nerfs being this heavily discussed and adjusted with so much direct community involvement is amazing, but I feel like the discussion about what to do with skills, items and specs that are severely underperforming compared to even the average should be happening as well.
Pure speculation says that developing, balancing and testing such fixes takes too much time, which would be perfectly understandable, but Iâd like to hear that officially, if itâs the case.
I kind of wish we had some balance changes during Cycle iâll be real, iâm fine for future cycles if the balance is more finely tuned prior but for this one the newer masteries are so vastly superior i feel the older ones could at least use some numerical help.
Uhm wasnât the survey only for highly overperforming bugs? This reads like you want to work on everything that is highly overperforming even if it isnât a bug.
Can we also have an identical survey with just the word underperforming in place of the word overperforming?
Thanks.
hehe this is similar to my answers. Fixing bugs is nobrainer. Not changing non bug skills in the middle is also my prefernce and i dont care about ladder so i checkd, dont know
Read it again, I donât think you parsed it right the first time.
I am not sure if this was discussed yet, but I think the purposed solution has a problem when you add a filter to it.
Legit players that didnât have played after the date of the fix wonât appear in the leaderboards after the filter is applied. You might end up in a situation where people might manipulate the leaderboards to claim positions they donât actually have.
In the era of streaming, bad information can make a game look bad. It is also very easy to spread false allegations, etc⌠For now this is not the case, but the competitive scene tends to be a popular topic in these mediums.
Personally, I donât care about leaderboards, but pretty sure competitive players do. Besides, because of that future features to leaderboards, like cosmetic rewards or any other stuff that could be related with how far you go up may become a problem really quick⌠Or always needs to be time gated or something like that.
I donât have a solution for this, but I advise reviewing how this will happen and how to mitigate above problems.
How about fixing the detailed search function of the exchange before patching the balance?
Are you doing localization?
Please donât keep looking at other parts.
Please keep your promise first
Yup you are right I completely missed the plot here
Absolutely, Iâd be as equally happy to say it the other way, it just depends on how one wishes to message the statistics to fit oneâs preconceptions. Statistics donât lie but liars use statistics, after allâŚ
People who donât mind it either way should simply be taken out of the equation because they donât care about either. So it would boil down to what the 1-2 and 4-5 people want for this survey. If someone donât care about a specific point and says so itâs obvious from my point of view that their input isnât meaningfull to the topic at hand and can be diregarded because they donât tend to either extreme viewpoint.