The statistics Smokescreen.
by Preekness | 17/02/2010 16:39:07![]() How can blizzard deny that some classes like druids and Dks are underpowered every metric points that their representation is extremely low compared to Paladins, and Warriors. So what happens we get this smokescreen saying that the numbers are useless. I wont point you to proof that druid and DK representation is stupid low because some blue will be in here saying that those numbers don't matter. I wish the developers would balance the game better. I wish they would make it so that specs like resto druids/Shaman could have a place in pvp. Rather than saying the statistics are a lie. PVP is balanced around retarded burst with nerfed healing. Druids have no dispel protection for their hots. DKs have no dispel protection for their diseases. Are we seeing a trend here. The worst represented classes in high rated 3s and 5s rely on something that can be dispelled. |
by Ghostcrawler | 17/02/2010 20:13:11![]()
Sound advice from Believe. Ghostcrawler Lead Systems Designer |
by Ghostcrawler | 18/02/2010 18:48:03![]()
There are some risks with being too open. One is that you kill the discussion. If we immediately answer every question, then there isn’t much reason for players to go back and forth, and we lose the potential to learn something about what our players are thinking or feeling. WoW has a healthy theorycrafting community. If we over explain things, then we reduce or remove the need to experiment. We really want the forums to be a place for discussion, principally among players. If we want to get a message out, we have better ways to do it that have a chance to reach more players. I know I personally slip into politician speak when I’m worried about players misinterpreting something or trying to use it as a big club to get their way. If we say “There is a bug,” then players might respond with “OMG Blizzard admits releasing unpolished content.” If I admit where we made a mistake, players can respond with “OMG Blizzard is just flopping around helplessly and needs to let the community make all the big decisions.” I don’t like slipping into that mode and I try to catch myself, but it has definitely helped me understand why politicians do it so often. :)
Representation is a data point. It’s not the only data point. Another data point is what the best players of that class / spec can manage. Another is what the worst players can manage. Another is whether players think that class / spec is fun to play. When we’re dealing with who plays what class, we are trying to interpret a very noisy data set that probably has literally hundreds of variables that are difficult to measure or model. We can’t always tell you why some classes have 8% representation and others have 15% (though some players will try to boil it down to very simple and probably inaccurate explanations). Since it’s hard to determine cause and effect, the most useful information comes when there is a big shift. If you see representation jump or fall over a short period of time then it’s likely that something happened. Maybe one of the variable suddenly overwhelmed all the others and caused a shift. If we introduced a big PvP buff or nerf it wouldn’t surprise me at all to see PvP representation suddenly change (though I’m sure there is also an effect where representation changes just because players heard about a buff or nerf and not because power actually changed that much). [ Post edited by Ghostcrawler ] Ghostcrawler Lead Systems Designer |
by Ghostcrawler | 18/02/2010 18:48:08![]()
Statistics are very powerful, but only when coupled with interpretation. If you look at say a medical study, what you will see is a bunch of numbers and graphs that are probably hard to interpret unless you are very familiar with the questions being asked. This nearly raw data is called “Results.” What comes after that is a part of the study called “Discussion.” This is the section that is the most important part of the study. In it the researchers try and interpret what is going on in the data. They are trying to answer questions, or at least explore whether they can reject a hypothesis. The data are really just there in case someone disagrees with their interpretation. When the mainstream media report on the study, they typically don’t do much with the data. They go straight to the discussion or the conclusion of the study. Too many of the posts in the WoW forums (and not just the Blizzard ones) are results with no discussion. Linking to arena (or whatever) representation is step one of your post, not the entire post. Numbers can rarely ever just speak for themselves unless the situation is as simple as reading the temperature off of a thermometer -- a relatively simple measurement generally held to be accurate enough for answering the question being posed: what is the current temperature? If you’re asking a question more complex than that, and whether Resto druids are underpowered certainly qualifies, then you need to explore things a little more than just linking easily available numbers. That is the point I am always trying to make with responses like this. I realize you might rather see me post “Here is how we’re going to buff druids,” but if we wanted to say that, we would (see my reply above). On the other hand, directing players a little on how to make more useful posts can potentially pay big dividends over the course of the entire game. I plan on reading these forums for years to come, so it’s in my interest to do what little I can to try to elevate the quality of the discussion. :)
Yes. [ Post edited by Ghostcrawler ] Ghostcrawler Lead Systems Designer |
by Ghostcrawler | 19/02/2010 08:06:03![]()
Personal experience is very informative. It's probably one of the most valuable things we can get from these forums. You will get more credibility if you phrase your points in terms of "My experience is..." instead of "My experience is... and therefore all players' experience is...". Ghostcrawler Lead Systems Designer |


Recent Blizzard Announcements