Why doesn't HS consider deck composition?

Which are determined by the actual meta at the time of the decks being played. Not something that can be playtested.

So let’s go with that system and a fresh new meta

Let’s say we have 10 decks being played all starting at 50% winrate.
2 Decks are really better and reach 55% winrate when the others are spread between 45 and 52%
The 2 top tier decks almost only face eachother and rarely the 50%+ other decks
One new decklist is created to counter those specific 2 decks. This is a brand new deck, how do you insert it in the meta to begin with ?
Let’s imagine you put it at 50% by default. It will primarily face the 8 decks it is not made to be good against, and will have a mediocre 40% winrate.
You made a specific deck to counter the top of the meta and you will never have the possibility to play against them… because you asked for a rigged matchmaking

Just provide a source for that instead of claiming it, it will close the debate

3 Likes

Implies the extreme scenario that Smeet, Derkan, and now you provided which is obviously not meant to exist.

Variance still exists, you won’t face only the people playing X.

You won’t.

The whole idea is: If you a good enough player you’ll push through the added counter-matchmaking and reach the higher realm. Counter-matching still has variance, and if you are in the rank you don’t deserve, the other player won’t be as good as you. Slowly, you’ll come out positive and climb.

Spot on.

Exactly what would happen xD

Your skill advantage is not as high as the winrate differential between two decks which counter each other. 33-67 is normal for countered matchups, and you’re definitely not 18% better than your opponent, and the higher your rank is, the more that becomes an obstacle.

For example:

I’m stuck at rank 200, playing mainly Insanity warlock. My winrate on the deck? 52.2%

Its’ winrate in top 1000? 52,28%

What can I conclude? I can only conclude that I absolutely belong to top 1000. That means that noone in top 1000 has over 2.28% higher skill than anyone else, because if they had, they would have a higher winrate on insanity warlock (more than 52,28%)

This is the optimal matchmaking, and it’s literally the proof it’s not rigged.

Did you miss the part where I said specific win rate is to be considered?
The deck’s 40% win rate won’t matter if its win rate against the top two decks suffices to matchmake it against them.

They would still face this counter deck. Where did you get it not being possible to face them?

It’s on every public statement about how the matchmaker works that Blizzard has ever put out, including the rollout of the rank system and how bonus stars work. It’s on the wiki if you want to look it up.

Which would be easier for you do do by googling “how does hearthstone matchmaker work” instead of accusing me of lying.

I play 50-60 games per day and it still wasn’t enough couple of months ago to get me where I belong

You’re underestimating the variance. Variance can be huge, much bigger than % difference in skills between ranks.

Enough, victim.
Your lie was claiming MMR and ranking is the same.

They do work exactly the same, no matter how often you whine otherwise.

See how this didn’t end the debate, Derkan?

My whole point is diminishing that variance.

No they don’t xD

You were right up to this point.

Rank gain is fixed. Your MMR dictates who you play against, and your star gain per win, but ranks are discrete (change in countable steps) while MMR is continuous (depends on exact decimal number)

The extreme scenario not meant to exist being the creation of a brand new deck in the meta ?
Dump the idea that it has been created to be a specific counter.
If your deck starts by only facing decks it’s good against, then crashes on the top tier decks it can’t win against, you obtain a deck that has had its winrate artificially inflated by all that time it has not faced its weakness

What would be easier for you is to provide a soure instead of wirtting a sterile back and forth with someone else on the forum.
Please quote the moment I accused you of lying. Asking for a source is not a lie accusation.

What was to be seen ? I said a source would close the debate, no source has been provided
“google it yourself, it’s everywhere” is not a source

But you don’t do it successfully, because you match players who are better than their ranks with players who are worse, but play stronger decks.

You can’t discriminate between “deck strenght” and “player skill” in your model. Both enter the formula to determine who you play against.

That means that both the one and the other will decide who you play against. If it’s only player skill, you’ll play against average number of counters and average number of decks you counter, while if it’s both, you’ll be playing your counters more because your skill goes up and deck strenght stays the same.

If I can’t play against people who play only tier 1 decks because I play only tier 3 decks, then it’s different categories all together.

It’s like having a light category boxer (tier 3 deck player) competing on the worlds (MT, since we’re in the MT qualification season right now) against a heavy weight boxer (tier 1 deck player), even though they never faced each other and one clearly has deck advantage.

It’s just not fair, not balanced at all. You would benefit from playing tier 3 decks only if you’re skilled, or tier 1 decks only if you’re bad.

And when you don’t have bonus stars, your rank is the only input to the matchmaker, and the matching algorithm is identical.

Do you deny this?

I told you, it’s on the Hearthstone wiki, which archives Blizzard posts with direct links to the original.

Your laziness is not my problem. But snipping the part where I answered you out of your quotes will become my problem if you continue.

I match players who are better than their ranks, against players who are worse, with same power level decks. That regards matchup specific win-rate etc. to determine them being a “fair” matchup. The better player will win more here, and he will climb until he reaches the rank where the matchmaker is meant to keep him at 50% at. Right there, it’s their job to breakthrough that by being better than the other, more innovative etc etc.

That’s why I am not. The “player skill” doesn’t change the matchmaking. It is actually meant to bypass it and face people of different skill to allow the better person to win more.

No, that part is true, but irrelevant for the most part

People who are new to the game won’t reach high ranks where the difference matters anyway

Only when your account is new on a server will you be able to exploit it to your advantage, and even then it’s only temporary.

When I was still a 1-2k player on EU, I started fresh on NA, got 3 stars to work with, and hit rank 70 on NA because of the difference between MMR and ranks.

Needless to say, now my ranks are ~200 on both, so it balanced out after a while.

Legend uses MMR, and doesn’t care about the number in the hexagon whatsoever.

It’s the same algorithm.

I don’t know how to be clearer.

You can’t bypass it, or you lose a variable to determine who you get matched against.

Your winrate has to be taken into account, or else you’d be matched against random tier X players from any rank possible. You’d play one game against a Chicken 10 and another against a top 10 player.

It determines by your rank.

It does, and it matches you more against counters. Still, a good player will come positive from this and climb, because the handlers of said counters are players of the undeserving rank the good player is in.

It’s not. There’s 1 algorithm and multiple categories upon which the algorithm gets applied. That means, yes, the algorithm is the same, but the output is different, so no, it’s not the same, the algorithm DOES care.

You can hit top 20 with 49% winrate if your TheoHS. Also, you can hit Legend 10k with 70% winrate if you’re a dumpster legend player.