WoW is P2W - Do you agree?

I think that’s an important thing to underscore. The game is almost made so that you succeed more easily and efficiently if you play in an organized guild.

But increasingly people don’t. Ergo another reason why boosting thrives.

Hmmm this is arguable. Raiding in a guild is more enjoyable and more “guaranteed to succeed eventually” but most guilds do take longer to get AotC. There are a lot of PuG groups that can clear AotC in week one.

Also joining a guild and forming relationships has always been the key to success in this game.

Because you have to play on a schedule if you play in a raiding guild, most of the time.

I can raid when I want to, by joining pugs. I’ll clear aotc no problem if I want it.

I can play m+ when I want to, by joining pugs. I can do keys well beyond max reward by pugging, I did +10 like 2 hours after season 4 launched, with pugs.

I’m not gonna set aside time to start a raid at 19:30 or whatever time a raid starts and then play until 23:00 or whatever the end time is. I play when I want to play and I don’t play when I don’t want to play.

There’s also the other part… Unless it’s like a very high guild in regards to world ranking, chances are you’re gonna be better off just pugging if you just want to clear stuff.

I personally wouldnt play wow if it werent for the ppl i play with, but ofc alot of ppl dont wanna do that. But if ppl wanna achieve those things (ksm, curve etc) without buying boosts there’s plenty of options out there (guilds, communities and so on). Ppl just seem to prefer to play it as a solo game and ofc it will always be easier to just buy that boost then to invest the time into making and maintaining friendships and connections. I could probably achieve the same stuff in pugs, maybe even more who knows. I just dont want to.

1 Like

Common… what is so hard to understand :

The bots YOU see are only the ones that did not get banned.

The ones that did get banned… you dont see.

But you dont know how many. You dont know if its 1M bots that got banned. Or none at all. There is no way to tell, unless you work at blizzard. Which you dont.

Are you somehow claiming that from the bots you count in the open you can extrapolate that value to the ones that got banned? You cant !!!

Why do you give me these coments ?

You cant be serious here. All I said is : you dont know what blizz is doing or not doing. You only see the results !

Lets say blizz is doing a bad job. And out of 100 bots 90 get away. OR… they could be doing a fantastic job and out of 1 million bots 90 get away.

You only see the 90. But we are talking about what can be done about those 90 that got away.

If its 100/90, im 100% certain blizz can do better.

If its 1M/90 then I doubt blizz can do better. Because “there will always be some that get away”.

But you cant tell the difference. Neither can I. We only see the 90.

But I give blizz the benefit of the doubt. In the sense that there is a chance that its the 1M/90 case.

And if those 90 bots that get away can generate so much pain. Then its better to change strategies to stop them.

That is all I said.

What is “pretty dumb” about that?

I made that number up. I dont know how many AH post the bots put.

But still. Lets change 20000 AH posts for “90 bots that got away” from above. Now it makes sense.

Who said it was small ? It still pales in comparison to wow though. I dont even need to see the stats. I am convinced that there are atleast 150k (10x more) people right now. MINIMUM.

Or not. Maybe its the 1M/90 case. Maybe its the “there will always be some that get away case”.

To ban them, you have to identify them first. And avoiding false positives like this guy that Amonet pointed out :

If it was so easy as to “ban them” there would be ZERO botters in any game ever made. Even in AE2.

They would have already done it dont you think? That is my point of view. There is some information that WE are missing from this story. If it was that easy to just “ban them” then they would already be banned by now. So why not? They are missing “motivation”? They are missing “intention”?

OR maybe there is some technical reason that we dont know… THAT is what I am arguing…

They haven’t. Ores are still rocks you click that give you tradable items. Herbs are still things you click that give you tradable items. Skins are still killing beasts and skining them. And they give you tradable items.

Tradable == not bind on pick up.

That has been the case since 2004. Nothing has changed.

So all I am saying is : I somehow believe that “banning them” is not working. They ban them and they keep coming back. And to leave this stupid “back and fourth” situation something has to change.

Now. Imagine for a minute that all the items you farm are bind on pick up. What would the bots do in the end? NOTHING. Then all mobs from SL and back give 1 copper as loot. Not more. Gold comes from quest rewards only.

The drama would be over in a day. No need to ban any bot. They would be out of business. No need for fancy AIs or GMs spying on us…

Tell at what do u progress faster in your vieuw? Am curious?

Ahh, you’re talking about selection bias.

Well, uhm… how the hell does that matter? I’m not talking about statistics.

Are there too many still? Yes there are. They could be banning 5 or they could be banning 5 million - it makes no difference, there are too many.

Because sometimes I literally can’t tell what you’re talking about, if you wanna know. Language can be a bit finnicky sometimes so I have to make a guess and understand it as best I can, and the above… ? Not in my wildest dreams had I expected that to be what you meant. It’s completely and utterly irrelevant to us in every way possible.

I mean they told us at several points how many they banned, plus Thor on YouTube told us how they banned, plus we have information from the gold sellers. So… I think that’s reaching a point where we can extrapolate quite a lot, actually.

But banning the bots is never going to work. As I said for the last several weeks, they have to kill the market. It’s the fact that people buy gold that causes ALL OF THIS to happen.

The amount of market distortion does not depend on the ratio between banned and unbanned sellers, it depends on the ratio between sellers (+their bots) vs the amount of regular players. That’s the only thing that matters; the amount of bans they are handing out simply doesn’t matter.

Nono, I know - so did I. The point is if they can keep the amount low the amount of market distortion is smaller - it doesn’t matter how many they have to ban to do that though.

You did. The number you gave would make it rather small.

Yes, that’s true - which is why I have suggested machine learning. But if you really wanna do this manually and hit them where it hurts, you ban their customers. The bots are replaceable, but the people who buy the gold have INSANE levels of investments. For context, my character is old enough to vote. Just think about that; think about the attachment, the history. To lose her would be absolutely devastating, so if they tell me I shouldn’t buy gold, I won’t!

And the thing is, the amount of people in my shoes is very large. It’s not all of us of course, not even close, but if Blizzard actually goes after these buyers then the fear of breaking the rules will spread, which is what we want.

They need to make it hurt. Banning these bots doesn’t really hurt very much and therein lies the real problem.

If they pull a Reaper of Souls on WoW they will destroy it. MMORPG’s build on the idea of a market economy. It is one of their most defining features by far. To destroy the auction house and cooperative crafting would be a huge and very negative decision.

I will not allow you or Blizzard to use botting as an excuse to destroy a quarter of the game, because that’s what you’re suggesting. Ban the bots, don’t destroy the games.

The consequences of your suggestion would be cataclysmic.

Hmmm it’s a very hard question to answer. I don’t think the ban waves are the way to do it, or at least not with them being 6 months apart. I think that no matter what Blizzard do there will be someone smarted than them that will figure out how to bypass their efforts and start botting again, and if one person figures it out then that will become the “meta” for doing it shortly after. It’s an arms race. I think they should continue to improve, but again I also don’t think innocent people should get banned. The thread I linked mentioned that that person has filed nearly 20 appeals, you have a limit of 20 appeals before it’s game over for good. Now idk if that person did use bots or didn’t, but I can say for a fact you can get cought in the cross fire even if you didn’t.

And my hypothesis is that it will change nothing. The botters will make an AI of their own and bypass the security measures once again.

Like they have always done. Like they always will do.

We will never get rid of the “90 bots out of 1M that got away”.

OK. If taken at face value yes. But if we agree that “there will always be 90 that got away” we can begin a debate about how to kill the bots farming capabilities with out creating a cataclysm.

Let me begin :

The “work order” system that blizz made is perfect IMO. In the sense that you dont need crafting materials that are tradable like before this system.

Many craftable items require “crests” which are bind on pick up. Cant trade them, not even with alts. But you can still transfer them to an individual to craft an item, even if they are binded to you.

So we can expand on that.

First : The reason people dont want to farm in the first place is because its boring. So, give alternative means of aquiring raw materials (which would be bind on pick up).

For example : A WoD style garrison farm. Except that you have to imput “dragonflight expedition resources” that are quired by doing activities. For example, WQ, Delves, M+ and Raids.

They would all drop this “expedition resources” and that can be traded for ores and herbs.

That you then use to craft items.

You can also farm these ores in the Open World, but they would be “bind on pick up” and account bound.

And crafters would have the option of supplying the materials for you (for a price).

IMO a system like that would be similar to what we have now, with a cooperative crafting.

Then the AH would only exist for refined items. Like potions and items.

Bots cant make money off refined items. Because to make them they would neceserely have to play the content like a regular player does. You cant automate “raiding”… :slight_smile:

Machines are flawed and can be outsmarted, even if they “learn”.

This will 1000% get exploited and people who are not “their customers” will get the short end of the stick. Like I said before, I will gladly sit back and enjoy the poop show from the sidelines if that happens.

The poopshow this created in WotlK when they actually did rain the hammer on gold BUYERS was epic…

As you said.

It was so epic that they haven’t done anything close to that ever since…

1 Like

Exactly.

You brought the crime example. IRL we have several departments and infrastructures that are specifically designed to combat crime and yet we have not solved crime. We have police officers, we have investigators, forensics teams, lawyers and judges and we still cannot solve this. We have all these people that are specifically trained to do their jobs, have the governments backing and can get access to pretty much everything and yet it is not enough. What chance does a gaming company which can only operate within their own product do to solve fraud/framing (I’m really bad with legal terms), or laundering money.

Ishayo is right if you brake the customers trust the botters will lose, but that goes both ways if Blizzard rank up enough false positives they will be the ones braking their customers trust and they will end up losing. This is something that can and will kill this game faster than any army of bots or goldselers can.

1 Like

And my hypothesis is that you’re completely wrong, so I suppose we could butt heads over that for all eternity.

How much do you know about machine learning? I’m just curious. It has some particularly interesting strengths in regards to this, but of course it is based on linear algebra and probability at the end of the day, so precision will be an issue. But, the more you ban, the better it gets.

Maybe not, but 90 bots is a trivial problem. I’d say it’s not even a problem at all. I can live with 90. I can’t leave with tens of thousands though.

We can discuss ideas for enhancing professions, especially ones that involve effort from the crafter leading to greater yields. The fact that some of these mechanics will make life terrible for bots can be a positive coincidence, HOWEVER:

Allowing cheating to inform design decisions is not up for discussion. At least not with me. I consider it a terrible thing to do.

Unfortunately any discussion of how to actually do tradeskills well is therefore off-topic to me, but I will nevertheless give a few comments based on yours - but this is not an invitation to debate shaping crafting around bots and gold buyers. That clear?

There are potential reasons one might want to limit trades - putting a cap on how many crafted items a player can get by using crests and the work order system is a fantastic example of a good use of bind-on-pickup, because it means people don’t get to buy tons of gear from everybody else and short-circuit the gear progression curve, while simultaneously allowing crafted items to be really good - even BiS. This is a good idea regardless of the ability to buy gold - it just solves two problems: There is now good crafted gear in the game, and you can’t use that good crafted gear to boost yourself to near BiS really fast. That’s a win-win.

Your other idea about using dungeons and raids for dropping important items and recipes is another great idea and an idea they’ve implemented in the past. The idea of asking for effort and skill on the part of the crafter has serious negative implications for botting because they don’t have skill, yes - but to me that’s a coincidence. What I really want is to avoid a situation where you cannot invest by playing but must instead invest with large capital only. That’s a problem. Investing with large capital is of course not a problem, it’s only a problem if it’s the only choice.

It should still be very possible to contribute without running mythic raids or whatever. There are a lot of ordinary players in this game who just enjoy hopping around the world and picking up stuff, and going from “it’s dead because of bots” to “it’s dead because we killed it because of bots” doesn’t really help these people. It might mean we need to buy less gold, but the bots will find other things to farm and wreck our day anyway. Farming gold directly from dungeons and causing inflation is a common tactic, and it’s not just something they can do for old content.

The only crafted items that are worthwhile are the ones with embelishments and you can only equip 2 (3 if you count flavor pocket), the rest are just stat fixes.

Theirs a lot of people out there who will pay for those carries even though they’re more than capable of achieving it themselves.

I’ve known many people who would rather pay gold to get it done and dusted so they can do something that’s more enjoyable overall.

Boosting will never vanish and rightfully so blizzard should probably stop catering to the 1% and cater to longevity and social aspect of their game calling people noobs for valuing their time more is sad because those noobs could have been veterans in the day.

Bring back server community and remove sharding

1 Like

If I’d get banned for something I didn’t do, I would just quit the game it happened in permanently. I’m not staying in a game where I get banned for something I didn’t do. Don’t care if they’d reverse it or not, it’s the fact that it even happened if it happens.

2 Likes

If my account got perma banned I would quit (in my case it got restored after 1 appeal and it was only for a couple of hours). I would also quit if I saw more people getting banned out of the blue for no reason.

3 Likes

I am not at the forefront of AI development. I just code AI for industrial applications.

So I know about AI. Atleast the ones that can be commercially applied.

False. That is just the ONE of the myriad of optimization algorithms used to train the AI.

If you want more info, ask for it. But I will give you a TLDR of what machine learning is :

There are many types of AI. But they all have 3 basic elements :

  • (1) A set of parameters. They can be Hours Played. Movement ingame. Particular key strokes… Any parameter you can put a number to. These set of parameters are set by the programmer. Lets call these parameters P1, P2, P3… up to Pn.

  • (2) Some weights. It just means “how important is each parameter”. Lets call this W1, W2, W3… up to Wn.

  • (3) A win condition. A number between 0 and 1. Lets call this variable “WIN”. 1 == a bot. 0 == not a bot.

So. An AI at its core is the following formula :

  • P1W1 + P2W2 + P3W3 + … + PnWn == WIN.

So how do you make it work ? You measure some P variables, what is the Weight of each one such that WIN approaches 1 ?

You give it training material. A set of examples with a known WIN (WIN == 1, its a bot) and known P values. The optimization algorithm figures out what W values correspond to that set.

And then you can use it.

But you need millions of cases of “this is a bot” and “this is not a bot” to make it work properly. So who decides this? The coder of course. He has to manually give that set.

And when you apply the AI in real world cases, you will never get exactly 1 as a win condition. You have to also define a threshold. For example, that 0.8 constitutes a bot, but 0.79 wont.

So as you can see… it does not “learn” just like that as you claim. You have to train it.

This is false. It will be just as good as it got trained. And it its enough to change the value of a P value by the boters to escape detection. And then, you have to go through the training process once again.

IT… WONT… WORK… Not better than the systems already in place…

And its definitely not a silver bullet that solves everything. Its a program, like any other. With exploitable weakness. And any mediocre coder can figure them out.

Dont get fixated on the number itself. You know what im talking about when I say “90”… :slight_smile:

You misinterpret me. I don’t want to allow cheating. And I will not justify any discussions as “allowing cheating”.

I just want to make cheating a waste of time. In such a way that nobody will do it.

The only “loose cog” of this whole bot issue is raw materials. That is the only thing left to fix.

You fix that… there will be no more cheating… it will be IMPOSSIBLE to cheat…

It is the opposite of what you claim, that changing the gathering system is somehow “allowing” or “justifying” cheating…

In the example I proposed, people can still do that. Its just that you cant sell it in the AH directly. You have to transform it into something else first. Like a potion.

I think its orders of magnitude less profitable than farming stuff and puting it in the AH.

Also, atleast you don’t SEE the bots. So they are not in the open world stealing your herbs, nor are they wreaking the AH. They just farm gold directly and sell it in their webpage for $$$.

And because they are in instanced content, you can police those dungeons much better. Its easier to police than the whole open world.

Let’s look at that “False” once again:

Firstly, that
“* P1W1 + P2W2 + P3W3 + … + PnWn == WIN.” thing is literally linear algebra. This fact is one of the major reasons NVIDIA were so fast at winning the AI processor race - all they needed was FP8 compute in addition, and they were already masters of FP16, FP32, and FP64 compute at incredibly large scale and parallelism. They already had the most optimized libraries for linear algebra on the planet.

So yes, it is linear algebra. As for the probability part of my statement, unless you use something like decision trees or K-Means clustering you’re probably going to use probabilistic methods for the training - and not only are you using probabilities for the training, for classification algorithms you even get a list of possible results with attached probabilities as output from the trained model - which is what we need here - we want to classify is bot vs. not bot on a basic level. So you’ll get something like “98% sure this is a bot”.

In order to train you would indeed need to feed it tons of data parameters, such as indeed hours played, how long the account has been active, its movement patterns (does it repeat the same movement for a long time?), spells cast, gold transactions, etc.

You would indeed need this.

Now as it happens I happen to know that World of Warcraft has sold over 150 million copies and that during WotLK classic Blizzard were saying every month how many accounts they had banned - it was usually between 200k and 300k. And they claimed to have been doing that since Classic came out.

Now, I have a hard time believing that but if we assume it to be true then we do indeed actually have millions and millions of data points, and if we don’t have it yet, assuming they’ve continued, we’ll have it by the end of the year.

Yes, so you retrain it every month or so. Sorry, I assumed that would be obvious but from a strictly technical point of view it of course is not.

Yes I do, and the point you seem to be making by picking such a small number is that it’s a very small and trivial number of bots - which doesn’t bother me. I’m not the one fixated on those numbers.

No. If there’s a mechanic, you can cheat on it. That’s just… fundamental to game design. Cheating means nothing more than not doing what the rules say you must do.

Well, well then farm materials, craft into pots, flood the market, and now you’ve not only destroyed materials farming for people who just wanted to do that, you’ve also destroyed the crafting professions even more than they were already by making the base ingredients that can actually be posted harder to get that what you craft with them, thus ensuring that crafting anything with what you get is actually always a net loss. Now everybody is going to need to play 1 gathering 1 crafting or they can’t get anything done.

True, but now they’re causing inflation, meaning all the prices go up and world quests become hopelessly incapable of keeping up (since they don’t inflate)

Sure. You do you. Give me lessons on AI.

That is not linear algebra. Its just algebra. You can make it linear if you want to work with vectors.

As I said. There is more than 1 AI out there. Go and wikipedia it.

NVIDIA worked fine for SOME applications of convolutional neural networks. But not all of them. And specifically because using kernels on images is much more efficient if you use matrix computation (linera algebra) rather than scalar values.

Dont bust your head arround it. Its an optimization algorithm. And there are may, many types.

Its literally answering the question : Given some fixed set of P, what values of W do I need such that WIN == 1.

That is what training is.

You can use probabilistic algorithms if you want. But its just 1 tiny subset of the available options out there.

And none of those training sets would have anything in common. A player in 2004 does not play like in 2024. And neither does a bot.

Your set of “trainable” examples is quite small. Especially if it lasts you what? 1 or 2 months before the botters figure out how to cheat it? Then you will have wait for 1 or 2 months, gather that data to train your AI. And then ban them again.

There is not enough data to train a robust AI that wont make mistakes.

This is what I mean. You retrain it with what? 1 or 2 months worth of data? What data? Data hand picked by a GM saying “this is a bot” and “this is not a bot”.

Skip the middle-man. You still need someone to classify that data. Might as well just ban them on the spot if you see them. Why wait to train an AI…

See what I mean by “its not a silver bullet”?

AI is not the solution to all our problems. Its just another tool. Like a Wrench or a hammer. With advantages and limitations.

Fine. It was just a proposition. Obviously I will never farm anything because I find it horrifically boring. But I respect those who do.

I just wanted to say : There are other options other than working on banning bots.

I think that a 20 year long battle with 0 results… is gonna drag another 20 years if we dont change how we approach it.

That is what I beleive. But you dont.

So lets leave it at that.