Dev Data vs. Player Feel?

General Discussion
Prev 1 2 3 4 9 Next
01/23/2017 09:31 AMPosted by Ceratisa
Isn't D.Va being adjusted cause of how people felt according to the patch notes?


No, she's being adjusted because of the data that showed she was op af
01/24/2017 07:59 AMPosted by MortystMorty
01/23/2017 09:31 AMPosted by Ceratisa
Isn't D.Va being adjusted cause of how people felt according to the patch notes?


No, she's being adjusted because of the data that showed she was op af
Hey being played 3rd most across ALL modes doesn't show us she's OP. Reinhardt must need a buff if he's 12th so faarrr behind D.Va.. wait..
01/24/2017 07:59 AMPosted by MortystMorty
No, she's being adjusted because of the data that showed she was op af
Statistics can tell you things like "character X on average does very well in duels against character Y at such and such ranges" etc.

What they don't necessarily tell you is why. Is it the character that is the cause, or is it the player base?

Ana is a great example of this. She's a typical example of a character that was overly buffed, because the statistics showed that she performed poorly, which a lot of people interpreted as meaning that she was weak. And then, after a while, people started getting used to her. And suddenly she's the root cause of one of the most overall effective comps in the game.

Oh, and regarding the post you're replying to:
01/11/2017 05:23 PMPosted by Geoff Goodman
As she was, it was common to hear people ask 'What do I do about D.Va?'.
01/24/2017 08:00 AMPosted by Ceratisa
01/24/2017 07:59 AMPosted by MortystMorty
...

No, she's being adjusted because of the data that showed she was op af
Hey being played 3rd most across ALL modes doesn't show us she's OP. Reinhardt must need a buff if he's 12th so faarrr behind D.Va.. wait..
Nobody is even remotely close to Dva's Data far as K/D Ratio goes.

Im thinking...
- Booster 5 sec cooldown huntdown shutdown outrun outgun almost everyone superiority
You're thinking...
- tags targets w/ wide spread

Something has got to give.
I really hope they use relevant data only (which means competitive matchs only at a certain level of play) and absolutely not people opinion.

Imagine a hero is totally OP and easy to play. He will be played by a huge number of the community. And if someone complains about him on the forum he will get : Git gud, L2P, L2switch (maybe for the only character that will soft counter him). Every post related to a nerf will be downvoted to the ground.

You should never listen to players for balance things imo.
I think people need to learn the crucial difference between balancing based on "data" and balancing based on "Jeff From The Overwatch Team's interpretation of data".

They're not taking these objective numbers and plugging it into an equally objective Optimal Balancing Machine, they're looking at numbers, possibly with little to no context attached, and forming their own very subjective opinion based on them. Any gaps or unanswered questions that the numbers leave are filled in by (subjective) player experience, left to the developer's (subjective) imagination, or due to human error are just never asked in the first place.

Data analysis is just as much asking the right question as it is having the right answer.
01/23/2017 09:22 AMPosted by Bullgrit
The developers with their unshown data behind the scenes


Unshown, that's the issue. Add to that recent comments about protecting riptires and "not playing Sombra correctly", and it paints a situation where the developers are trying to force a particular style of gameplay onto us, in which case none of our comments here would matter because their minds would already be made up.

That aside, in these forums (or in any discussion, really), the most vocal minority will always be overrepresented, so we should be careful about considering "player perspective" as well.
This is an interesting discussion and we appreciate the discussion on the topic.

To provide some insight to how we make decisions, we weigh a lot of factors. There is no one single thing that forces our hand with decision-making regarding the game.

I would say the three things we look to the most are:

  • Player Feedback
  • Statistics
  • Our own instincts as players and designers


There isn't a formula that tells us when or how to make changes. We carefully use the three guiding categories above to help lead us to the right decision. Those three categories don't all weigh equally in every decision as well. For example, according to Statistics, Symmetra was fine and did not need to be touched at all. But our instincts and player feedback was contrary to what the stats were showing. Obviously, we made changes in spite of the statistics, not because of them.

Another example of how this works in practice is Competitive Play. Our instincts as designers tell us the most fair, most competitive and most balanced version of Competitive Play would require 6-person teams only. It's a great design.... on paper. If people were Omnics. But they aren't. And Player Feedback has overruled our instincts in this case so we allow for any group size players see fit (including solo) to queue for Competitive Play.

More often than not, we use Statistics as a guideline or a tool. Most of our decisions are based more off of Player Feedback or the OW team's instincts for gameplay. Stats are just good to tell us how far off those instincts are. Of course, if something is truly egregious in the stats, we will take a look at it more closely.

I'm over-simplifying the process here but trying to provide some insight into how decisions are made. Time, resources and technology all play a large role in this as well. But hopefully this gives a better understanding of why we do the things we do.
Highly Rated
Considering everyone exaggerates what they say on the Internet, I'd have more trust in the dev stats.
Whats up jeff
01/24/2017 08:26 AMPosted by kilojuliett
Considering everyone exaggerates what they say on the Internet, I'd have more trust in the dev stats.
God you like exaggerating don't you. EVERYONE huh?
Jeff's awake right now, inb4 people ask for numbers to be shared which Blizzard never really does.
Edit: Downvotes but I'm right, people are asking.
01/24/2017 08:29 AMPosted by Ceratisa
Jeff's awake right now, inb4 people ask for numbers to be shared which Blizzard never really does.


The only number I want is his
This might sound stupid, but I listen a lot more to emotion. HOWEVER you need to look at the purpose of the emotions. An example would be me. I personally think symmetra is super powerful, but then again my main is countered by her (dva) so my opinion isn't the most useful in this case.
01/24/2017 08:28 AMPosted by Ðr0wsŷLőuiš
Whats up jeff
Highly unexpected. Thank you Jeff for the clarification.
01/24/2017 08:25 AMPosted by Jeff Kaplan
This is an interesting discussion and we appreciate the discussion on the topic.

To provide some insight to how we make decisions, we weigh a lot of factors. There is no one single thing that forces our hand with decision-making regarding the game.

I would say the three things we look to the most are:

  • Player Feedback
  • Statistics
  • Our own instincts as players and designers


There isn't a formula that tells us when or how to make changes. We carefully use the three guiding categories above to help lead us to the right decision. Those three categories don't all weigh equally in every decision as well. For example, according to Statistics, Symmetra was fine and did not need to be touched at all. But our instincts and player feedback was contrary to what the stats were showing. Obviously, we made changes in spite of the statistics, not because of them.

Another example of how this works in practice is Competitive Play. Our instincts as designers tell us the most fair, most competitive and most balanced version of Competitive Play would require 6-person teams only. It's a great design.... on paper. If people were Omnics. But they aren't. And Player Feedback has overruled our instincts in this case so we allow for any group size players see fit (including solo) to queue for Competitive Play.

More often than not, we use Statistics as a guideline or a tool. Most of our decisions are based more off of Player Feedback or the OW team's instincts for gameplay. Stats are just good to tell us how far off those instincts are. Of course, if something is truly egregious in the stats, we will take a look at it more closely.

I'm over-simplifying the process here but trying to provide some insight into how decisions are made. Time, resources and technology all play a large role in this as well. But hopefully this gives a better understanding of why we do the things we do.


Hi Jeff we appreciate your insight on the matter. Is there way to give the community more insight on your data?
So we could be more on the same ground regarding balance?
devs, please make the game you set out to make. btw - the amount of events you have been rolling out has made your game viable... i would have been completely bored 6 months ago... without maps and new characters and events..
important to note is, while feedback and player feels are something that has to be taken into consideration... remember that those people who post here in the general forums are still the minority of players. This may sound harsh to most and I will get downvotes for it, but think...

most players don't go onto the forums and post/rage/troll/flame/give feedback around like we do when balance changes come... most either just look at them and then keep going or don't even know that things change...

Of course this doesn't mean player feedback should be ignored, however statistics are a bigger factor and give more reliable feedback than a main who got pissed because his favorite hero got nerfed.
01/24/2017 08:25 AMPosted by Jeff Kaplan
This is an interesting discussion and we appreciate the discussion on the topic.

To provide some insight to how we make decisions, we weigh a lot of factors. There is no one single thing that forces our hand with decision-making regarding the game.

I would say the three things we look to the most are:

  • Player Feedback
  • Statistics
  • Our own instincts as players and designers


There isn't a formula that tells us when or how to make changes. We carefully use the three guiding categories above to help lead us to the right decision. Those three categories don't all weigh equally in every decision as well. For example, according to Statistics, Symmetra was fine and did not need to be touched at all. But our instincts and player feedback was contrary to what the stats were showing. Obviously, we made changes in spite of the statistics, not because of them.

Another example of how this works in practice is Competitive Play. Our instincts as designers tell us the most fair, most competitive and most balanced version of Competitive Play would require 6-person teams only. It's a great design.... on paper. If people were Omnics. But they aren't. And Player Feedback has overruled our instincts in this case so we allow for any group size players see fit (including solo) to queue for Competitive Play.

More often than not, we use Statistics as a guideline or a tool. Most of our decisions are based more off of Player Feedback or the OW team's instincts for gameplay. Stats are just good to tell us how far off those instincts are. Of course, if something is truly egregious in the stats, we will take a look at it more closely.

I'm over-simplifying the process here but trying to provide some insight into how decisions are made. Time, resources and technology all play a large role in this as well. But hopefully this gives a better understanding of why we do the things we do.

Hello Papa
Jeff, can I politely ask you how is being able to hook someone, shoot them(I'm fine with this for now) and in that same split second, combo it with a melee attack against the victim to ensure a kill good design? It's also bothered me about roadhog how is he allowed to sync both a shotgun blast with a hook and then shoot you again when the hook ends, I am just asking(politely) because in both this game, and WoW, you have integrated internal cooldowns for abilities in the form of delays in this case, its seems to me that roadhog breaks some of those rules in terms of that philosophy used over the years.
I have a lot more faith in the devs than in the crybabies in this game that say that Genji's nerf would make him useless [and they were wrong], that Widowmaker and her ult are useless and should be reworked [and they're constantly being proven wrong], that Roadhog is going to be useless because you actually need to AIM his hook [and again, they're getting proven wrong].

Join the Conversation

Return to Forum