This -will- be nerfed. (Hollow Hound)

Multiplayer Discussion
62

Top 1k is not everywhere, but that’s apparently how you and blizzard are defining everywhere when it comes to balance.

And we’re going to disagree with “reasonable” here, too. The fact of the matter is that “reasonable” means miracle rogue, DH, or control priest wins against it at top 1k or it’s getting nerfed. It’s fine if all of the popular things people want to play at other ranks get dunked on as long as top 1k is happy, the deck is “fine.”

This is crap. This is what you are advocating as “balance” and it’s not balance in any real way.

What is a win rate? What does that top line number tell you about a deck? how does that number translate to different rank brackets?

What makes it go up and down?

Saying the average wins went down, so it’s fine is not how stats work. It only works this way if all the other variables stay the same, and they don’t.

Edit: PREFERENCES account for more of the change than anything related to power. PREFERENCES. That does not somehow negate the power of a deck.

And it’s wrong because win rate doesn’t measure skill. To measure “skill” he would have to look at the same player against the same decks over time at different brackets. Aggregate data doesn’t tell us much about skill.

Win rate is not a proxy for skill. It may be the best data we have, but it does not represent skill in the way that you are telling me it does.

Also, it’s okay not to take ZachO as gospel. VS does lots of good things, but one thing they don’t do well is stick to conclusions based on their actual data. THere’s a whole bunch of opinion and conjecture at times in what they put out there. They are (usually) good about pointing this out (“ZachO thinks…”) but the community is not as good at understanding the difference.

63

It’s not… Entirely wrong.

At the highest skill levels of the games, they play decks that have no issue with hunter. As hunter players reach those top skill brackets, their deck is doing less well for them than it would against worse opponents.

Some of that is from different matchup spreads, but, for pure pally, for example, you saw it drop off within specific matchups as higher skilled players were just better at dealing with the power turns the deck could do. (Obviously horn changed that and closed off the weaknesses the deck had)

It’s ok for decks to be strong, so long as there are ways to counter them. Decks tend to get nerfed more from deck frequency than absolute power anyway. An overpowered deck you only see 1 in 20 games doesn’t ruin your experience in the same way that a slightly too good deck that you see 1 in 5 games does.

But that slightly too strong deck is probably not in any need of nerfs if it drops off the better players are.

64

You just contradicted yourself in your post.

This sentence contradicts the next sentence.

It is not, it’s doing less well against DIFFERENT DECKS. The opponents are not the point, the decks are. You are looking at decks in winrates, not opponents.

Again, you didn’t show this with data, you showed me that the match ups changed enough that the top line number was okay to you at that level. It never negated the fact that the same deck was NOT okay at lower ranks.

Further, this only applies if you look at individual players in longitudinal data.

No, it’s not okay for a deck to with 60+% of games because this one or two decks that top 1k players prefer beats it.

That’s your whole argument - that if it’s okay at top 1k there is not a problem. That’s bullcrap and not a way to balance a game.

Again, this is a huge double standard if you look at it by rank.

AT TOP ONE K it changed. AT THE TOP. IT HARDLY MOVED AT LOWER RANKS BECAUSE IT WAS ALREADY BROKEN.

Seriously, you’re clinging to nothing. The horn didn’t break paladin, it just made top 1k players mad that they were losing to a “smooth brain” deck that was beating their self-appointed god decks.

Rogue sucks. It’s a terrible game experience. If I never saw a thief rogue again, it would be too soon.

But any time it’s tangentially viable it will see insane rates of play at top 1k meta. Same for DH and control priest. This is not “normal” if normal is the whole player base.

The paladin nerfs exposed the dev team as an absolute clown show beholden to a niche group of players who do not represent the whole game but get outsized influence in many areas. This fact is more responsible for alienating customers than anything else they have done.

65

I honestly can’t verify this. I don’t pay for HSR premium and never will.

Paladin decks prior to horn did lose win rates BOTH from the matchup spread and the individual decks lost less to it in the top end. I have no idea what hunter is doing with those trends.

It’s likely not seeing that kind of win rate. You need to have paid stats on HSR to actually see win rate %s that make any amount of sense, because the front page splash has hound hunter and pure pally at 68% win and enrage warrior at 64% which are aggregate win percentages that have basically never existed overall.

The most recent VS report has hound hunter on basically the same level as enrage warrior and pure paladin in D4 - L. That’s hardly looking like a power outlier.

It’s definitely not a deck that needs intervention before the Titans launch. Who knows what things are going to look like afterwards.

What are you talking about? It literally flipped every bad matchup pure paladin had… That’s a textbook definition of breaking a deck.

66

I literally linked the hunter deck’s winrates in diamond to legend in this thread.

The bad match ups weren’t played that much at lower ranks is what I have been saying through this whole discussion.

The change at lower ranks wasn’t that big compared to the change at top 1k, where those decks saw more play. It literally was only a huge change for the top 1k meta, and that’s the only thing that got it nerfed. If it didn’t’ move at top 1k, blizzard would have told the rest of the game to get bent like they always do.

You have to actually look through the data, not just take the top number summary and assume everything.

67

This is objectively false. It is not a questionable opinion, it is an outright falsehood.

One way to calculate the winrate of a deck is the weighted average method. Let’s say there are n archetypes in the format. If so, the winrate of any one deck is
sigma(1,n) w(n) × p(n)
where w is matchup winrate specific to an opposing deck archetype and p is the popularity of that opposing deck.

If we want to get real results, we would pull w and p from the same ranks. But if we want to get hypothetical, we can pull w and p from different ranks. For example, we can use the matchup winrate table from top Legend, and we can use the popularity table from Diamond. This would show us what it would be like if Diamond players played with top Legend piloting skill while keeping the meta proportions completely unchanged.

Quite some time ago — admittedly, prior to the release of Horn — I ran these numbers using actual winrate and population tables. And in the vast majority of decks that lost winrate going from Diamond to top Legend had a lower winrate with top Legend matchups and Diamond popularity than they did with Diamond matchups and top Legend popularity; and the vast majority of decks that gained winrate going from Diamond to top Legend had a higher winrate with top Legend matchups and Diamond popularity than they did with Diamond matchups and top Legend popularity.

So no, it’s not just that top Legend players like some decks more than others. They like some decks more than others because the skill difference between them and Diamond players actively changes deck power levels. The “pocket meta” is not the cause of but an adjustment to these changes in power.

Again, this is the kind of math that anyone can do (it’s very tedious, but not particularly difficult) and the data is out there. No one needs to take my word for it.

3 Likes
68

Yep, decks don’t just stop being good because everyone arbitrarily changes what they are playing at top legend. The better skill players play the decks that win more with that skill. Then the decks (like hound hunter) drop off in win and play rate as a result of better decks being around.

The only time that doesn’t eventually trickle down to diamond is when those decks are particularly tough to play well.

1 Like
69

Although I mostly agree with your position, I can’t endorse this point. I’d say a good rule of thumb is that what the top Legend players will figure out in a week, Diamond players will figure out in a month. So very early in a meta, yes, what you see in top Legend will trickle down. But for the most part, metas don’t last long enough for the trickle down effect to actually work, and the vast majority (75%ish) of top Legend development that we see is simply never going to reach Diamond before a balance patch or a new expansion.

In short, yes, it trickles down, but it’s so slow that “always” doesn’t describe it well.

1 Like
70

Yeah, that’s fair.

Given unlimited time, those metas will usually slowly align, but we don’t have that without meddling of balance patches or new set releases.

1 Like
71

Both of you fail to grasp what a win rate consists of, what longitudinal data measures in contrast to cross sectional data, or why you’re both flat wrong in your assessment.

I haven’t read all of scr0otum other than what you’ve quoted, Smeet, because have have that person on ignore.

No, but that’s EXACTLY why the win rate changes - it’s a different meta.

No, they play the decks they like to play, which aren’t the same as other levels. History is full of examples of decks being fantastic at top meta that can barely scratch out 30% at lower ranks.

When you go and really look at the data, you see that the decks that eat that deck are barely played at top 1k, which is why you can play that deck. Zok druid is a relatively recent example of this, where it was a meme outside of top 1k where there was little to no aggro to stop it. It’s not about power at all.

Balance should be based on the common meta, not the niche. Until they learn this at blizzard they will continue to lose players and alienate gamers.

72

Nope, fully grasp that. You are just putting too much stock in what those win rates are telling you.

Yep, and the meta changes because they do better with different decks, not just because they don’t like hunter up there.

They find out “oh, outcast DH is the strongest deck, we are going to play that instead, because hunter isn’t.” Then hunter’s win rate drops, because it isn’t the best deck. And the meta shifts away from hunter.

It’s pretty damn rare that players at top legend that will avoid playing the actual best deck in the format for long. Hunter isn’t secret OP up there. If a different set of matchups makes hunter much weaker up there, that same matchup spread will make it crappy in power tiers too.

So, hunter is over performing in lower brackets because players aren’t using the actual best decks… So… Nerf hunter?

The actual best decks in this case aren’t weird anomalies that can only beat a tiny fragment of decks that are only seen at top legend.

But really, that’d just bring back the face burn meta that we had prior to hollow hound and healing combos with horn of the windlords pushed them aside.

The hunter deck isn’t the best deck in the game. The top legend format is proof of that. It’s a strong deck, and easy to pilot, so it’s strong in lower ranks.

Outcast DH is a stronger deck than hound hunter.

1 Like
73

How do you figure? Explain exactly what you think they are and I will show you line by line how you are exactly wrong.

You know what they don’t tell you? Skill. Because if you are saying skill changes, then you need to show me how the same player loses to the same deck at a different rate as rank increases.

Not that deck wins change, players.

I’ll wait for you to provide it. Skill is not attached to a decks, it’s attached to players.

Do you understand that you are actually, in this sentence, completely conceding my point?

You can’t say that the average player must play X deck to be competitive, but that isn’t meta warping. You can’t tell me that hunter isn’t strong because everyone is playing demon hunter and in the same breath tell me that it’s a skill issue.

You don’t understand how meta works, though. There are a list of decks that are strong that are significantly underplayed at top competitive levels because they are unfavored styles of play. You can’t deny this and expect me to take anything else you say seriously.

That isn’t what I said.

No, it’s not proof of anything other than top 1k plays what they like to play and it’s not related to the rest of us.

74

Skill is attached to the players that are playing the decks.

The players that are using hunter at top 1k are in general better than the players using the deck at diamond.

The people playing the other decks are also better at higher ranks.

Let’s look at the most recent VS report…

https://www.vicioussyndicate.com/vs-data-reaper-report-267/

The win/loss matchup matrix for D4-L shows outcast DH with a losing matchup to hound hunter.

Filter it for just legend (not even top 1k), and suddenly the matchup is a tossup.

That’s showing that when the players of both decks got better, hunter became less likely to win that matchup.

At a quick glance the others are staying about the same when just filtering legend, but the report doesn’t show the spreads for top 1k only, and I’m not going to buy HSR premium just to do so to find other examples.

Skill plays a noticeable part in what decks are best. For a long while pirate questline warrior was a great deck to get to legend, but utterly worthless once you got there because people knew how to play around the power turns.

No, I’m not. You are asserting that the meta at high legend is arbitrarily different, which causes the win rates of decks to get better/worse.

It’s the other way around… Skill and new discoveries push other decks to higher win rates, which then shifts the meta to account for the better decks. Hunter didn’t become a worse deck at top legend, there are just better decks there being played better, which leaves no room for hunter to look “overpowered” because other decks are performing better there.

They aren’t going to swap to the Diamond to legend breakdown because the decks hunter farms are not good enough to be played there, so they don’t.

So yes, a meta shift caused a change in the win rate, but the meta only shifted because there are better decks than hunter to play. It wasn’t just a personal preference.

75

Yes, it does.

It 100% is arbitrarily different.

To assert that it isn’t is to assert that the top 1k doesn’t have a preference for specific decks/deck types. There are years of data to support these trends. Years.

You cannot deny this.

You cannot separate the two, though. There is significant evidence to show you what types of deck are played at top 1k, an it’s not hunter and paladin.

I didn’t say that it did. I said it LOOKS worse because of the other decks that are more popular. This is how a weighted average works, though. When bad match ups are played 3X or 5x more often, the win rate goes down. It did not change the power of the deck. What changed was deck preference.

You’re literally conceding my point then contradicting yourself.

Demon Hunter (all types) are TWENTY FIVE PERCENT of the top 1k meta. One of every four games is DH match. It is about 7% at D4-1. That’s a HUGE change, and the winrate changes are a function of changes like this.

Same thing happens with enrage warrior. Same thing happens due to the lack of better match ups because top 1k prefers different decks.

76

Because you believe there is some way to build an utopian meta where no deck will dominate lower ladder forcing the learn curve of the players to adapt around it and building the top meta with decks that win against it.

Reality is that learning curve has a direct relationship with those “low rank meta tyrants”.

Why would someone learn to play the game BETTER?
Because they have fun trying to climb the ladder.

And to climb the ladder you gonna have to win against that type of deck.

And there are 2 ways people usually try to do that:

1.They either learn more of the game to pass the barrier those decks present.
2. They play that deck in an almost mindless grind. And no. The term mindless isn’t trying to offend anyone but i also gonna call it what it is.

There are characteristics that form those decks that you already heard from me nonstop like a inflated performance against other decks when both players are BAD.

What can and should be done isn’t on game balancing level but on design level.
Have a better control on the learning curve of the decks you let get online.

It will not “stop” but less it by a significant amount.

But if you prefer believe that people in bronze are as good as people playing the game in legend ranks and the “different meta” isn’t fruit of people getting better go ahead.

77

What an utterly ridiculous demand.

Fact: Out of 285 Vicious Syndicate (VS) recorded top 1000 Legend games, Pure Paladin won against Outcast DH 80 of those games, to earn a winrate of 28.1%. As published in the most recent Data Reaper, #267.
Fact: Out of 262 VS recorded Diamond 4-1 games, Pure Paladin won against Outcast DH 100 of those games, to earn a winrate of 38.2%.* Also DR#267.

There’s no “deck popularity varies” argument to be made here. It’s pure Archetype X vs Archetype Y data. And yet the winrate shifts (just over) 10%. How is it that the great Neon explains this difference then?

The only way that one can defend the argument that the power level of decks doesn’t fluctuate with rank (by which we mean fluctuate with skill) is by refusing to acknowledge reality. And that’s exactly what Neon is doing here.

* Calculation note

VS publishes the matchup winrates here as 33.35% over 911 games for D4 and up (implying 304 games won) and as 31.47% over 649 games for Legend (implying 204 games won). By subtracting the latter from the former, I got the result for just D4-1.

78

I deny this.

The key word is “arbitrarily.” Of course they’re different, but those differences are not arbitrary.

79

No, there is no top 1k legend breakdown in the VS matchup matrix (there is in power rankings, and deck frequency sections). They don’t usually include that range due to a lack of enough data.

As you go from D-L, to Legend, to top legend, the deck’s power score consistently drops though.

They do, but it’s pretty rare that they will flatly ignore the best deck in the format because of those preferences. If they can’t play around it, they play it, even when it’s paladin.

It’s pretty exceptionally rare that a deck is undoubtedly best, but barely played at top legend.

Yes, but those decks also generally become worse as skill improves due to being one dimensional, so it’s not JUST personal preference. Personal preference doesn’t drop a deck from tier 1 performance to nearly tier 3.

If there are decks up there that handle hunter just fine, without actually building around hunter (because it’s a tiny part of the meta), that means there are decks you can use in lower brackets that don’t have major issues with hunter.

So your complaint that hunter needs a nerf falls flat when there is an entire meta’s worth of decks that make it barely a positive win rate overall.

There are a few good options to take it down, or at least play evenly against it.

Pure pally, enrage warrior, totem shaman, unholy DK, outcast DH are all either favored or break even against it.

80

I’m literally looking at it. It’s in the drop down menu. Yes, it has top 1k.

How many times will it take before you understand that the power score only exists relative to a meta? That as the meta changes, the score changes, too.

Explain how control priest is trash at one tier and tier 1 at top 1k? does it suddenly have better cards or does it somehow get better match ups more consistently? This works both ways - the meta affects weighted averages. The win rate of a deck is only meaningful in themeta that it is calculated and it doesn’t translate to look at them across play levels. You can’t just look at top 1k and assume X deck will rake at diamond. It does not work this way because the meta isn’t the same.

Let’s recap:

Top 1k is a unique niche place that should not be used to make balance choices.

The performance of a deck at top 1k isn’t generalizable to the rest of the game.

Choosing top 1k to balance the game is killing interest from the wide player base because you are ignoring the factual experiences of literally 100’s of thousands of players who are not playing top 1k pocket meta. This is BAD game balance and poor financial choices.

They don’t have to ignore those decks for those decks to look worse at that level, all they have to do is prefer to play the bad match ups more frequently and the win rate looks worse. It’s literally definitionally how weighted averages work.

You understand that win rate is based on how a deck does against all other decks, that some decks are better match ups than others, and changing the mix of decks changes the overall win rate? THis is the central point here - looking at “top 1k” meta win rate only tells you about the top 1k meta, not the overall game or overall power.

You keep going back to skill, but that’s not the point or even the main reason for most of the change. If you really wanted to understand you would need to make some kind of an average if all decks were played equally to understand some sort of absolute power rather than relative power (weighted averages are a relative measure).

Whoa, so the fact that DH is 25% of the meta isn’t because of hunter, it’s because of preference? And I’ve been saying all along that the preferences of top 1k change how the power of many decks looks? I mean, you’re still making my points while refusing to agree with them.

I said hound needs a nerf, I stand by it. When it happens, I expect you to apologize, lol.

So we do have to change decks to play around it? I thought we didn’t. Can you pick one and stick with it, please?

Huh.

Only two of those sees high play rates at top 1k. The play rate of enrage warrior triples from D4-1 to top 1k and the play rate of outcast DH goes from basically zero to more than 10%.

But tell me again that the meta isn’t based on arbitrary preferences at top 1k.

Edit:

Watch this and you will understand my point:

https://www.vicioussyndicate.com/drr/vs-meta-score/

Start at platinum, then go through each rank. Notice how amazingly everything changes at top 1k? It’s unrelated to the vast majority of the meta but is somehow revered. It’s bullcrap and bad game design/balance.

(Hound hunter goes from 100 meta score at all ranks below legend to 98 at legend, and suddenly 50 at top 1k. This is not a “skill change” it’s a meta change based on arbitrary preferences.)

81

I love coming back to this thread now and seeing these wonderful statements.

1 Like