I never said it proved anything, you’re the one trying to find patterns in chaos.
Because it isn’t random. Not because it is chaos.
This game has as much in common with elo matching as I do with Taylor Swift.
You wouldn’t have a clue what random looks like.
Before you get confused thinking that’s an insult, there are about a million papers written on the fact that humans find it very difficult to detect randomness.
But I’m pretty sure we’ve had this conversation already, so what is the point.
But I do have an excellent understanding of what an elo system looks and operates like.
Team 5 needs to stop repeating that fairy tale.
No, but unfortunately, we simply have differing opinions on this topic. You believe that this happens, and that it happens regularly and intentionally, I do not, so I can’t factor it into my response. All I can offer the OP is a rational explanation as to why suffering a 13 loss streak isn’t as unreasonable a series of events as they believe. I have yet to see anyone complain about going on a massive win streak. That’s just intelligent deck choice and consistent good play, but somehow a massive losing streak is always RNG manipulation and fudged matchmaking, never a poor deck choice and inconsistent/bad play.
I am aware that a matchmaking algorithm accounting for deck strategy is a popular theory, but still no one has been able to explain who is targeted with bad RNG or why.
2 Likes
I would modify that: I observe it happening regularly.
As for intentionally, I cannot prove that.
I believe it to be, as it happens too often to be coincidence imo.
I am not someone who discounts facts or Science.
But; I have been playing the game a very long time, and am convinced that there are more variables to the matching than Team 5 cops to.
1 Like
But…
Good old but.
I wish that Team 5 would just go back to the old ranking system.
It was harder, but it was also more straight forward in how it worked.
Back then I had no doubts at all that if you weren’t good enough then that was that.
We have a new month.
And I’m sure this is what is gonna happen: Smooth sailing bronze through Plat5 week 1.
Week 2: Loss streaks on D10 break point. Week 2 - Week 3: Grind D10 with 50/50 results, then sudden win streaks that push me to D5 breakpoint, then loss streak back to D10.
Week 3: Algorithm allows to go through into D5. Then grind D5 with 50/50 results, when suddenly by end of Week 4 algorithm allows to climb, then loss streak back to D5.
This exact same scenario happened in August, September and now in October. I want to try and record this in November.
How can I record my monthly stats?
I don’t know. In order to prove what We see it takes an enormous amount of statistics, and the players who poo poo everything you say are keenly aware of that.
I have tried in the past to compile matches, but I am no statistician.
2 Likes
Random is Random
The human mind is trying to find pattern and meaning when there are none. Can you be matched against 13 unfavorable match-ups in a row? yes, it’s possible.
Plus, this thread mentions 13 losses, not 13 unwinnable match-ups. Is it possible the OP made a few misplays that lose him a few winnable games? Yes, it’s possible. No need to reach to some voodoo explanation here.
Also, one needs 15 stars to get from diamond 5 to legend. with a winrate of 55% (which is quite high for the average player), you earn 1 star every 11 games, meaning you would need at least 161 games. I had 52% winrate, so i needed near 300 games myself.
Going to legend is a marathon, not a sprint.
4 Likes
Use HSreplay and it will record your games.
This is true, but being given a run of statistically unfavorable matchups is pretty common at tier breakpoints.
I have observed it for years, but never thirteen in a row.
The problem is people. People are attempting to see a pattern where there is no pattern.
2 Likes
Well; like I said above: I think there were far less complaints when We had the old matching system.
We should just go back to it. This one sux.
Op being a bad player is an factor he himself tosses out of his initial post, and which we can’t quantify.
The matchmaking becoming wonky is explicitly quantifiable, if true.
That’s like saying that if a guy crashes every car he had before six months, a conspiracy theory involving car dealerships and him being a bad driver were of equal probability.
They aren’t. Chances are OP is just a bad driver.
2 Likes
Exactly this. Within random distributions there are always clusters of similar events which we normally assume not to be random. It gives us the illusion of some conspiracy. But there is none.
3 Likes
You are right. I would imagine He is an average or below average player.
The matching is still crummy.
this matching system clumps in peculiar ways with remarkable regularity.
I liked the old ranking system way better. I don’t believe I ever questioned it.
I just took my lumps.
Yes clustering is an important aspect of random distributions. And it is also an aspect that is generally underestimated by most humans. I am aware of this yet i still would like to claim that hs rng events do cluster more then natural random events.
There is ways to check this btw,though it would be a very rough aproach that would not really prove all that much in the end.
You take the average winrate over x amount of games. And based on that average winrate you can calculate the amount of clusters expected. In this case we can look at the very simple clustering of wins vs loses. I am sure if you would perform this simple analyzes that it would give proof of more clustering then when flipping a coin.
Unfortunatly there is more factors at play which makes that this aproach would be up for discussion ,or at the minimum require an extraordinary amount of data to give a good indication. But mathematically this aproach would be sound.
Deck choice is a pretty big factor in this as well. People will often target the most popular deck because they are under the illusion that this will give them an edge on ladder, but this is rarely the case. This is because the most popular decks don’t generally achieve a play rate over 20% (usually somewhere between 10% and 15%), meaning >80% of your games will not be against the deck you’re targeting. Further to this, decks that are designed to target a specific strategy often stack up quite poorly against decks that are not the one they are designed to beat. If your deck crushes Imp Warlock but has a bad matchup against Druid, Paladin, Mage and slower Priests and Rogues (POV you’re playing Boon Priest) you’re going to have a pretty bad time on ladder.
You could say something similar for home brew builds. Generally speaking, most home brew lists are prone to loss streaks as although they often contain powerful plays, they lack the refinement or available card pool to make them consistent and will be preyed upon by decks that can achieve their power spikes more consistently.
Switching up decks regularly is another problem for a lot of players. It is very rare for a casual player to have equal ability across a lot of strategies. Every class and archetype plays differently, and even decks within the same archetype or class can have major differences in both their own play styles and their play patterns against different decks. In essence, if there are 12 commonly played decks in the meta and you’re playing 1 deck yourself, you have 12 matchups to learn. Now say for arguments sake you are rotating 4 or 5 decks, suddenly you have 48-60 matchups that all have different mulligans and strategies. Suddenly, you’re going to lose a lot more games as you run into more matchups that you don’t know how to mulligan and play correctly.
Lastly on the list is decks that are in the meta but known poor performers. OP has advised that they were playing Silver Hand Paladin. The data shows this decks good matchups at Diamond ranks (mostly Rogue lists) are rarely seen at the D10-L rank bracket (except for Ramp Druid) and its bad matchups cover some of the most widely played decks on ladder (Imp Warlock and Spooky Mage to name a couple).
Add that to the list of things that influence win/loss streaks, and the possibility that 13 may have been an exaggeration (although I’m happy to take it at face value because I think its a reasonable number of games for a player who’s at Diamond rank on the last day of the month to lose in a row) and I think we have a logical explanation.
1 Like