line0042

Why the erebus nerf is foolish, and what to do instead

Why the erebus nerf is foolish, and what to do instead

I know lots of people have been whining for Erebus to get nerfed lately, and this has been done. But statistics don't really lie, at least not ones this simple. Erebus is middle of the field in the pantheon. If her were truly as overpowered as all the whiners have been saying, he'd be higher up, if not on the top. That is all there is to it.

So the patch came through to placate people who were getting raped by him. Sure, I agree he's obnoxious and powerful (in my opinion he should be an assasin, and swap places with TB, but that's beside the point) but not OP, or he'd be higher up in k/d ratio.

 

If anything needs to be changed, it's the demigods on the bottom who should be getting a buff. The only safe thing right now to say is that Queen of Thorns needs a buff. She's been on the bottom since release, and by a wider margin than any other demigod. That is statistics, and that is the answer. Nerfing a mid-range demigod makes no sense whatsoever.

356,960 views 120 replies
Reply #101 Top

Quoting line0042, reply 25
My patience is at a razor's edge for this thread. So I'm going to just throw this out to the slobbering masses of retards commenting here.

Despite the fact that the statistics aren't prefect, they even out. Everyone is getting poorly recorded wins/losses, not just erebus. People CLEARLY don't understand how statistics work. Everything evens out with errors.

Fucking duh.

You resurrected a post that hasn't seen any replies in almost two days to say that you're losing patience with it?

Reply #102 Top

Quoting JagerJack, reply 1

You resurrected a post that hasn't seen any replies in almost two days to say that you're losing patience with it?

Pretty much.. cause he didn't add anything to the discussion either.

Reply #103 Top

Quoting BeardKing, reply 10
I share a similar sentiment that I had expressed in another thread. I am not sure if Erebus is OP, I play almost all the demigods, less rook and torch bearer, and holding them in juxtaposition I can honestly say I never really had a problem with him (and in another thread posted some methods for dealing with him). I worry a great deal about a nerf so early in the games development when a competitive seen has even yet to developed. Based purely on my own past experiences an action such as this sets a bad precedence and leads to greater calls for nerfs, balances and buffs, usually with no backing, leading to a vicious cycle of balances and rebalances such that no player ever really get theirs bearings.

            There is an interesting chapter in Jonathan Baron’s book ‘Thinking & Deciding’ called ‘Quantitative Judgment’ that hits precisely on the matter at issue, peoples subjective opinions and statistics. The standard model for quantitative judgments is a statistically derived model called multiple linear regression. Basically the goal is to construct a statistical prediction rule (SPR) that measures whatever particular case, in regards to a number of different dimensions and cue’s. Students that go to graduate school for example are evaluated by an SPR that takes into account GPA, GRE scores, letters of recommendation, and other such things, which are each given a weight in accordance with their relevancy to the issue at hand. SPR’s are in common use now, and studies show time and time again that they are much more accurate than the opinions of experts, that is, they are far more effective than expert human judgment, (and in this case I do not believe any of us can even claim the status of expert). Stats don’t show bias, they don’t omit information, they express a value of whatever their function might be.

            You can argue against the above all you like, but studies have shown time and time again [please see Baron’s book] that you are wrong, statistical analysis trumps human judgment every time, and should therefore be the method both preferred, and deferred to, when trying to determine balancing issue’s; and if we at current do not posses enough data, that is all the more reason to withhold major game changes, we need the data if we are to make informed decisions, PERIOD. What’s more, if we have this data the chances are high that the balancing put into place as a result, will be the only balancing needed.

 

 

I so dearly wish this was true about game design. If it was we could just crunch a bunch of numbers and make the best game ever. But until you tell me how to quantify fun, game design will remain a fuzzy domain and one that can use numbers, but can't rely on them.

The data suggests that QoT needs to be buffed. Which buff do the numbers suggest? Increasing her health, mana, auto attack damage, spell cooldowns? Which spell do you buff? Do you need to nerf others to account for the increase in this one area? You can't tweak just one value and know what will happen. Changing on spell might completely change that character as well as change how all the other characters relate to that character!

 

Poke around here for a while to read more about why you can't balance by numbers, especially in a asymmetrical game: http://www.sirlin.net/.

 

Reply #104 Top

No offense, but I think some of you are really overthinking this.

Reply #105 Top

erebus isn't op if u die to him u got out played simple as that

Reply #106 Top

Quoting Xinoxlx, reply 2



Quoting JagerJack,
reply 1

You resurrected a post that hasn't seen any replies in almost two days to say that you're losing patience with it?


Pretty much.. cause he didn't add anything to the discussion either.

 

I didn't add anything because I've made the point over and over again, and have yet to have someone actually make a valid argument against what I originally said.

Reply #107 Top

heres some more proof that erebus wasnt op, look at the stats after the nerf already he dropped several spots an is jus above sedna and QoT. now if stats didnt matter then shouldnt erebus have stayed roughly the same? ppl must not realize that its based off win % if a demigod was op he would a higher win % than the other dg's now to start off with he was ranked 4th pre nerf, now after nerf hes dropped down to jus aboce sedna an QoT 2 majorly support chars so he is now the lowest ranked none support dg congrats the nerf wasnt needed. another interesting note pre nerf dark side was winning by roughly 1% now after nerf not only has light side caught up but they are now ahead by roughly 1%.

 

really dont care what people were saying they had no facts that erebus was op beyond "omgz he killdz me so ez" whereas the stats show differently. guess maybe dark side players need to take a tip from light side and whine till they can win easier. sigh guess im done, thanks again for reading my post.

 

Reply #108 Top

I so dearly wish this was true about game design. If it was we could just crunch a bunch of numbers and make the best game ever. But until you tell me how to quantify fun, game design will remain a fuzzy domain and one that can use numbers, but can't rely on them.

The data suggests that QoT needs to be buffed. Which buff do the numbers suggest? Increasing her health, mana, auto attack damage, spell cooldowns? Which spell do you buff? Do you need to nerf others to account for the increase in this one area? You can't tweak just one value and know what will happen. Changing on spell might completely change that character as well as change how all the other characters relate to that character!



Poke around here for a while to read more about why you can't balance by numbers, especially in a asymmetrical game: http://www.sirlin.net/.

I wanted to open with how cool the link you posted is, i bookmarked it, and it looks like i have a new read. That being said, game design does subsist in a fuzzy domain, im a philosophy major and games of abstract strategy are one of my main interests. What makes a game fun? All of us can point to examples, or sight instances, but what is the 'fun making property'? That is a scary hard question to answer and one i have put a great deal of time, thought, and study into; and i as of yet have no answer.

How to balance a pre-existing game is a different sort of question, and statistics are the answer. Pure win loss ratios can tell us where balancing needs to be done, and they can do a much better job than or own intuitions; but, as you said, not how to go about it; to do that requires a much more comprehensive set of metrics, but i believe that it is possible and that SPR's are the way to do it.

What would we need? All variable data for each demigod relative to their win/loss percentage, and a large sample. This is a very general answer; and now that you have me thinking about it i hope you will grant me some time to come up with a more precise response.

I think my [/thread closed] command is broken, but im happy that it is as now i have a new project. :D

 

Reply #109 Top

Quoting Lionheart1118, reply 7
heres some more proof that erebus wasnt op, look at the stats after the nerf already he dropped several spots an is jus above sedna and QoT. now if stats didnt matter then shouldnt erebus have stayed roughly the same? ppl must not realize that its based off win % if a demigod was op he would a higher win % than the other dg's now to start off with he was ranked 4th pre nerf, now after nerf hes dropped down to jus aboce sedna an QoT 2 majorly support chars so he is now the lowest ranked none support dg congrats the nerf wasnt needed. another interesting note pre nerf dark side was winning by roughly 1% now after nerf not only has light side caught up but they are now ahead by roughly 1%.

 

really dont care what people were saying they had no facts that erebus was op beyond "omgz he killdz me so ez" whereas the stats show differently. guess maybe dark side players need to take a tip from light side and whine till they can win easier. sigh guess im done, thanks again for reading my post.

 

 

Yeah, pretty much. I'm unsurprised that he dropped. Although to play devil's advocate, some of that might be from less people playing him due to the nerf.

Reply #110 Top

Lol statitics in the pantheon? No one plays pantheon, and if they did, there are a lot more funner Demigods. Erebus doesn't take any thinking to play as.

Reply #111 Top

Quoting LookOverThere, reply 10
Lol statitics in the pantheon? No one plays pantheon, and if they did, there are a lot more funner Demigods. Erebus doesn't take any thinking to play as.

 

Um, you clearly haven't noticed the thousands of people playing pantheon. All you have to do to see them is to click pantheon and then look at the big, big numbers.

Also, I'm not trying to say which demigod is "more funner", or if he takes skill. Only if he needed a nerf, which he didn't.

Reply #112 Top

Quoting LookOverThere, reply 10
Lol statitics in the pantheon? No one plays pantheon, and if they did, there are a lot more funner Demigods. Erebus doesn't take any thinking to play as.

Considering that he's in the bottom three...

Reply #113 Top

Our most accurate metric at the moment is reasonable projections, tested and found to be true. Things like his ability to permanately stay in Mist form if a player invests towards it, his movement advantage given by Bat Swarm over impassable terrain and the amount and extent of the utility co-efficients stacked on his primary damage ability. We know he can do the first two; feats not as easily replicable by other DGs without significant trade-offs and the third is a more complicated matter but seems such a 'must have' that it's drawn suspicion.

QFT. This was not a knee-jerk reaction. Erebus simply had far too many options available - damage dealing, mobility, lifesteal, and disabling his opponent... plus minions. He needed a nerf. To what degree that nerf should be taken, we will see. Balancing is a constant process; just wait.


I also agree that QoT needs a buff. Although her shield is very strong, she completely fails to meet her "original intention". If you notice, her strengths listed on the hero select screen are "heavy siege damage" and "powerful area damage". There is no mention of healing at all. Consider then that her AoE/siege damage is pitiful compared to TB/Rook and this makes me wonder why those things are even listed on the hero select screen.

Reply #114 Top

If Erebus needed a nerf, I'd argue that Oak may need one too. But above all, QoT needs a buff.

Reply #115 Top

Quoting TheGuildfordStrangler, reply 14
If Erebus needed a nerf, I'd argue that Oak may need one too. But above all, QoT needs a buff.

Oak? Now this I gotta hear. Considering how pathetic his shield is i'd like to hear a reason.

Reply #116 Top

Quoting Xinoxlx, reply 15
Oak? Now this I gotta hear. Considering how pathetic his shield is i'd like to hear a reason.

I have to say, I agree with TGS here.  I considered Oak the most powerful DG before the Erebus nerf, now it's not very close.  Oak takes a high level of skill to play successfully because he's weak up to level 5 (at which point the SoF/DJ combination greatly increases his staying power and creep farming ability), but there is no demigod that can hang with a skillful, well-built Oak in the mid-to-late game.  His spirits are by far the most useful minions in the game, and his skill tree synergizes extremely well with a heavy minion build - SP, DJ and SoF combined with 3/4 levels in Raise Dead ward is just sick...not to mention the awesomeness of Penitence in the hands of a skillful player.  Then there's Last Stand kamikaze Oak if it comes to that late game.

I actually agree with you that Shield is pretty weak, overall.  I often buy only the first level when playing as Oak and use it exactly once per combat.  The first level is quite useful and the 3rd level is also worth the investment for the debuff remover, though the 2nd level is basically a waste and the 4th level/Purity is a joke.

There aren't a ton of really good Oak players out there because his best build has more moving parts (micromanaging the minions can be hard) than that of other DGs, but a skillful Oak player is damned near unstoppable unless you can gank him very early in the game.

Reply #117 Top

none of the dudes  need nerfs there all easy and fun to play  so stop  with all this needs a nerf crap and counter it

Reply #118 Top

Quoting InfiniteVengeance, reply 22

This doesn't make sense to me. Do you really believe that should be a factor in deciding which skills get nerfed?

Uh, yes, it should.  All builds should be viable in different situations.  If every build has one ability then that ability is obviously too strong (or the others are just too weak)

All right, since I've heard this argument before, I'd like to hear your opinion on Sedna and heal.

Reply #119 Top

The biggest downside in panth/skirmish to a no-heal sedna is your allies screaming at you for never healing. :)

Reply #120 Top

a sedna without any heal is stupid  and is a shit build  sedna needs atleast lvl 1 heal it gives her that little boost  that is needed