Rigged competitive system is the reason for toxicity

Competitive Discussion
147

Rank is your classification. If it’s redundant it’s a nonsense label. It’s also your payout and reward. It’s your flaunt. If it’s redundant, the laddering process is non-genuine (a synonym of fake, btw).

Again, wrong. You have the wrong level of scale. You’re forcing a bunch of heads and tails onto the same team so it balances out fairly. That’s rigging.

We can’t be sure if diamond and gold are redundant metrics, if sr doesn’t properly classify your skill. If you want to make this about rank, then SR (with pbsr and other sr offsets) is all you ever need to bring people to their true rank label. MMR is redundant and inapropriate. It exists only to rig matches for 50-50 outcomes. That’s not fair competition.

It literally is. Nice try tho. Maybe use the dictionary term to settle it:

So rigging for 50-50 is 100% rigging. By definition. That was too easy. Next?

Why would someone with a good, well-founded, logical argument bring up hardstucks? Sorry for being hardstuck in the top 5%. Your lack of logical reasoning is deteriorating and you’re bringing in ad hominem attacks lmao. Again, too easy.

Random within a strict SR bracket is absolutely fair. As it should be. That’s the soccer league analogy you keep boring us with. If I’m 3300sr, I want to be matched RANDOMLY with and against people at 3300 +/- 250sr, say. MMR is totally unecessary - you still haven’t proved why it’s needed anywhere, at all.

Nice. Guess I win again. No logic or math just ad hominems. When sheer truth and intellect break down your arguments, it’s back to hardstuck and offtopic nonsense about climbing. Try again.

Do you want your ladder to be unfair? You can’t have both. c.f. NFL theorem.

Actually I probably do because I don’t ad hominem and make cash teaching this subject to people who try to look smart talking about z-tests lmao. I already reminded you of the virial theorem (for why a reset is good), which is based on entropy in line with thermodynamic and information theoretic definitions.

That would be an interesting data source if you can provide it.

Our last data reveal from Blizz was in 2018 and it was by players not accounts iirc. I’ve already explained how the two are entirely different and ruin ladder integrity.

Again, if you’re going to “hypothesis test” anything, test the trajectory length and time-to-rank (including wait times and cancel rate and leaver rate etc.). Rigging + noise = unfair laddering. People don’t get to where they belong with any efficiency, and more people get to where they belong faster if you remove mmr rigging - making sr better metric when no mmr handicapping is present.

I cook data people for a living. I’m not ur fam, and a structural or algebraic argument will defeat a data argument for breakfast, because it argues around a bounded set of all inputs rather than specific cases.

Or you could disprove the trajectory data with your supposed “hypothesis testing” math. Let’s see it.

In a stationary, wide-sense-stationary, or non-stationary sense? It’s a very non-smooth trajectory under 1st order conditions.

1 Like
148

MMR = points you scored on an exam. SR = how you rank against your classmates

Not redundant.

It forces you to play with a fair coin–as you would expect from any game. The dice are fair. The coin is fair. The cards are fair.

In the system you’re proposing, SR is meaningless. SR is not meaningless under the current system. It’s a simple question: Diamond players are more skillful than Gold players. Why you wouldn’t answer that question is beyond me.

PBSR and other SR offsets allow players to converge onto their true rank faster.

MMR is still used for matching balancing teams. Not redundant.

Sorry, but a fair coin will never ever be rigged.

Show the data. If the system is working against players, demonstrate that a more skillful player cannot climb.

Unless a Gold player has been riding the stochastic win streak train with 5 teammates that are hundreds of SR above the enemies.

Well, it is fair. So yes, I like fair ladders.

Oh no! How dare someone test your sacred belief with a perfectly applicable methodology!

You’re equating alt accounts to signal noise and that hasn’t been demonstrated nor is it logically tractable. Again, feel free to show the math, the data, or literally anything aside from anecdotal testimony.

Again, it doesn’t matter. A percentile is a percentile every day of the week. It doesn’t matter if there are 10000000 accounts belonging to 100 players. They will always be ranked against one another and they must necessarily follow a normal distribution.

Your hypothesis, your data. Walk the talk, fam.

Walk. The. Talk. Let’s see it.

2 Likes
149

If by rigged you mean you don’t belong in an Elo and the game is trying to correct for you being out of place then, yes reciepts, it’s rigged.

I always knew you were :joy:

No it finds players/stacks with MMR similar to yours to determine if you’re a smurf playing below your rank or a legitimate player who’s having a few good games in soloqueue/stacked queue.

You must be a popular guy

Elo Hell discord notwithstanding since it’s got a lot of players in there.

It’s not rigging, but then again you’ve failed to provide a reference for what you consider rigging so perhaps we’re at a crossroads.

It’s trying to prevent smurfs from ruining games while ensuring it doesn’t accidentally boost the wrong player out of low Elo. Hence why you get skill checks every couple hundred SR or so while climbing low accounts.

Yeah yeah, the same line I’ve been seeing you write out for years now.

No more than you. The difference is I truly understand the system and you pretend to.

1 Like
150

No balls :joy:

Remember I told you, you were going to lose brain cells in this conversation?

I’ve put him on ignore because I just can’t listen to it anymore :joy::joy:

At this point I’ve seen you provide

  • academic paper on match making
  • unverified (though probable) math
  • inductive/deductive reasoning
  • decent arguments

What has he provided?

  • speculative arguments
  • unfounded rhetoric
  • no math
  • no academic paper on match making.

This is why I generally ignore these conversations other than to find a few players I feel would benefit from some guidance from higher Elo players and are willing to put to rear in efforts to realize their competitive aspirations.

1 Like
151

How you rank relative to others is a neither SR nor MMR. It’s a distribution of % players by rank (or % accounts by rank if that’s your thing). It’s still redudant, because people are being hosted around MMR, and get to take “the test” multiple times (multiple matches, multiple alt accounts). So even with that corny analogy, at the end of the day, SR is more “points” accumulated or scored, and mmr is more “how you rank”.

So your poor analgoy is inverted (if at all valid). And since they never show MMR, it makes SR meaningless or replaceable with a single distribution. They can just show everyone how the rank with SR dist.

And why would you matchmake along hidden skill criteria that isn’t relevant to final payout and classification? MMR makes no sense, and it needs to go.

Exactly. You can get people where they belong without ever calling on MMR. And once they are where they belong, why would you ever rig with MMR ?

Wrong scale. A bunch of fair coins cherry picked on both sides of a match to balance heads and tails equally will always be rigged.

Wrong interpretation. Rigging isn’t about end result climb, it’s about taxing over-performers and buffering under-performers in what is supposed to be unadulterated competition. You don’t pervert the contest with 50-50 rigging. The 50-50 is supposed to happen once you reach SR, via random shipping around current SR.

Please explain how this system defeats the NoFreeLunch Theorem in all of optimization and systems desing. If you over-fit the micro you fail the macro, and vice-versa. Which is why random matchmaking around your SR is the correct way. It doesn’t deceive global results at the cost of local fixing.

Imagine being called out on z-test methodology and refunting it with “lol hardstuck”.

Well it has been shown, by Receipts in the thread i linked. Do you disagree with that math? If so, where? Again, it’s data-free so you need to use some algebra.

Where is the evidence that shows they force a (symmetric) normal pdf? They absolutely won’t distribute that way, because by their very definition a person has more alts than mains and the alts are on average lower rated then their mains. And again, the thread I linked shows how every alt disrupts the ladder integirty, because of sampling with replacement for some but not others, leading to constant-sum effects and geometric absorption costs.

You are free to debate that math if you want.

So you’re going to disprove algebraic and structural (i.e. data-free) results without any math or data of your own? That makes no sense, and no one is going to do work you should have done before enterring this thread, for free.

1 Like
152

I linked the Receipt’s thread on how every alt disrupts the ladder and lowers integrity. Has anyone even debated that using some math, data, or peer reviewed journals? I think not. Was it just over every single ad hominem gamer’s head? You know, the kind who +1 each other, and throw around “hardstuck” but not “Cauchy”.

153

So you deny the Meriam-Webster definition for rigging?

So rigging for 50-50 is 100% rigging. By definition.

Have they introduced you to online dictionaries yet? Bro that’s like the lowest bronze you gotta make it through that before we can go further. Just click accept it’s not that hard.

2 Likes
154

MMR is a performance metric against all other players. It necessarily follows a normal distribution by design.

MMR is the raw point performance. SR is the translation of those raw points into a rank. Performance metrics and ranking metrics are well detailed in the academic papers on the subject. Feel free to read up. I’m not going to belabor this further.

MMR is more granular and therefore more suitable for match making over percentile ranking. It also allows for faster convergence.

Giving both teams 6 fair coins each isn’t rigged and it never will be.

Are the outcomes rigged or not? What you’re describing is a system that must discern volatility from actual signal. Over performers are moved up the ladder. Under performers are moved down the ladder. Everyone has good and bad games. The system doesn’t reward nor punish you for RNG [volatility] in a few good/bad games. It rewards and punishes you for your consistent performance. There is nothing unfair about that.

  1. It’s a bounded search space
  2. Not all players perform equally within the search space

Further, the match makers are tested with real data and simulated players.

Imagine failing to reject the null at p < 0.00001 with an alpha of 0.05 on the basis of emotions rather than data. Pretty wild.

I didn’t see any math or the .py simulations in the post you link? Can you post the exact comment?

MMR is translated into standard deviations from the average player.

Feel free to solve the problem I posed earlier: 122414 accounts with a mean of 2266 SR. How many alt accounts at 3000 SR do you need to shift the mean 234 SR to 2500?

Are you serious right now? You proposed a hypothesis, refused to support it, and are now claiming I have the burden to disprove your hypothesis?

Get real.

1 Like
155

Having fun yet?

:joy::joy::joy:

2 Likes
156

Right, but you said:

Which is why there is confusion. You don’t score points on an exam based on an a prior pdf from 5 years ago, or any pdf (normally). So no, your analogy doesn’t even work. Stop trying to force it.

Still wrong. If you’re giving them 6 fair coins you’re shipping a random match. And that’s what we’re advocating for in SR-only system. But with hidden MMR, you’re perversely rigging the odds. By adaptively holding player data against them, by analyzing every player or “coin” to decide if it’s “due” to show H or T (even with fair coins you can predict if it’s due to H or T based on streaks and assumption of 50-50 bias). So load-balancing around that, you’re rigging the match.

There is no way deny it’s hard rigged, every. single. match.

Do you not agree forcing 50-50 odds is a kind of rigging? Or are you still caught up on basic terminology?

With mmr 50-50, they are stochastically rigged. The outcome (in many measures like expectation and variance) is rigged in advance - towards 50-50. The devs have said this and the patents seem to support this. That’s rigging as per definition, word for word.

If every game is parametrically rigged, there really isn’t anything else to say.

That has little to do with NFL in this context. NFL says you can’t constrain match odds (by forcing outcomes) and have a natural ladder progression. You can’t force micro results and expect macro to fall in place, and vice versa. But in random SR only system, the micro and macro are equal, (mmr, the micro manipulation - doesn’t exist and goes away). So you avoid NFL because everyone is on same criteria: where matches, ranks, and payouts are using the same transparent metric.

Eh? What does that have to do with rigged matches? Either they force outcomes or not. No fair competition forces the outcome ahead of time. They might force your overall bracket or league or team, but never match on match. Their are fairplay orgs and sports governing bodies that validate this mantra. You don’t rig matches and call your contest fair lmao.

A fool’s errand. Where is that data posted? Are the alt accounts winning and losing at 50% winrate or is there some negative bias because they’re alts and games are disposable? Still no evidence they force a normal pdf, articially sending you to places to preserve the shape of their curve from 2015. And you haven’t answered my question: how many games does it take a GM to go from gold to GM in a random SR only system vs. system with mmr rigging that arrests mobility? Assume no duplicate accounts per player and matches formed randomly for an SR range of +/- 250sr.

But you haven’t done an ounce of real math. And you haven’t shown a single source of data. Most of the debate is structural and algebraic, not data-based. We’re aruing about a model, so it needs to validate for all inputs with maybe some a priori knowledge of those inputs.

1 Like
157

But it’s not statistics, it’s sorting. What is a rank? How many operational steps does it take to get you to your proper rank? Does the sorting alg take more or less steps when you rig every comparison for 50-50 outcomes, instead of naturally (randomly) binning the sort/selection? Laddering is about sorting players to their proper rank on the curve.

Imagine thinking that stats, a 1st order uncertainty metric, is not outclassed by structural methods like sigma algebras and algorithmic complexity.

And that’s just the ranking and classification issue. Still have all the alts corruption issue. See the thread Receipts posted for more info on that subject. Feel free to refute with better (data-free) math btw.

Forum Mod Edit: This post has been edited by a moderator due to language. Blizzard Support - Forum Code of Conduct

1 Like
158

Has anyone pointed out yet that the lopsided population distribution of the ranks where gold, silver and bronze hold the majority of the player base is the reason why a dps only has a 7-10 min que…but at higher elos you can wait upwards of 30 mins?

Wouldn’t this A) incentivize higher ranked players to create smurf accounts to play dps role in ranks with lower wait times, and B) disincentivize dps player from wanting to climb into higher ranks where they will experience longer que times?

They nerfed tank shields, and tank dmg (zarya beam dmg ramp being the most recent…yet syms remains unaffected…or wrecking balls pile driver dmg radius ). They nerfed support healing output and dmg output (bap and anna most recently).

If the idea is to reduce que times for dps…why would you strip power from supports and tanks…especially tanks…when increasing those populations would naturally lower que times for dps? Tanking and supporting in lower elos would certainly be less frustrating if you weren’t so reliant on dps.

Do i need a research paper to make that obvious?

2 Likes
159

No you don’t. You only need people to accept the premise. The conclusion follows:

  1. Higher ranked players are incentivized to create smurf accounts for lower queues.
  2. Higher ranked players exist, they number several, and play several more matches per day or month than casuals.
  3. That 1 and 2, leads to additional disruption in lower ranks, increasing entropy, lowering signal/noise, and taxing someone’s trajectory through the ladder.

Most of this was clarified in the thread on alts:

Every alt disrupts the ladder by sheer counting and existence. It’s more basic than a stats argument. Receipts treated alts as clones of their mains, i.e. zero smurfing ratio for all of them (which we know is false), and showed that even a 0 delta discrepancy was enough to wash away the meaning of ranks. You can still climb the ladder, but are heavily taxed (he even quantified the amount of extra effort, approximately).

Alts ruin ranked integrity. So does rigged matchmaking. And so does never resetting things 5+ years into the same season. Then you have all the human elements like toxic psych and disposable gameplay, as a byproduct of the above.

1 Like
160

No. A random system gives everyone a coin that may have entirely different probabilities. That results in more asymmetric and unfair matches.

You need to demonstrate why being matched against similarly skilled opponents is synonymous with “rigging.”

If the system is adaptively working against the player, you need to show that higher elo players cannot climb to their appropriate rank.

It’s not “forced” 50-50 odds. It’s giving you 6 fair coins against another 6 fair coins.

Can you demonstrate that the trajectory is statistically different from the system you’re proposing?

Yeah–just like defining the parameters of a fair chess board.

Given a number of optimization algos over an unbounded search space with near identical solutions…

None of that applies here.

Meaning, the optimization is tested empirically and through simulation i.e.–it makes accurate predictions.

Fair competitions pit equally or near equally skilled opponents against each other. Fair competitions don’t randomly assign players to matches.

Why doesn’t a league/bracket qualify as forced match on match? It’s analogous to the MM. You don’t have a predefined team unless you 6-stack, but you’re always playing matches against equal or near equal opponents within your bracket.

57,289 accounts at 3000 SR are required to move the mean from 2266 to 2500. That’s 46% of the total accounts sampled initially. You then have to make the assumption that every single one of those accounts is played during each season. That is, you need a massive number of alt accounts playing at the same elo during the same season for negligible mean shift.

Can you demonstrate that these requirements are met?

Answered multiple times. Matching and ranking against your peers results in a normal distribution, by design.

[Current] MMR rigging system that “arrests mobility”: 23 games in 5 hours and 50 minutes. Awkward’s unranked to GM on Ana. Starts in Gold ends in GM with a 96% winrate.

[Proposed] Random SR only system: No real-world examples from Overwatch. Assuming an average SR of 25 per win, starting at 2250 and ending at 4000, you need 1750 SR. At a minimum that’s 70 games, but with a 96% winrate, you need 73 games total. That takes roughly 18 hours and 30 minutes using Awkward’s rate.

Your system should be about 3x slower than the current system.

Sorry, I forgot basic statistics doesn’t qualify as math. Again, let me know if you’re actually going to present anything other than personal anecdotes.

There’s a post on Medium entitled, “Attempting to Collect Unbiased Data About the Player Base of Overwatch (PC)” by Mörkenbörken. Unfortunately the forums won’t allow me to post the direct link, but it should be easy to find. The author pulled Overwatch open profiles at the end of season 9 and compiled the ranking data from 122414 profiles.

Kaplan’s post which coincides nicely with the mined data:

What are you talking about? It is a statistical hypothesis test. The rank is top 1% at 4000 SR. What is your contention here? Are you complaining about “rigged” outcomes or how grindy the ladder is?

Imagine being all bark and no bite. Walk the talk.

Yep, just seeing complaining and anecdotes. No proofs. No methods. No data. No math. Just erroneous assertions based on normalized distributions (IQ), no evidence of entropy, and lots of complaining.

3 Likes
161

Because it’s forcing the odds, as per definition of rigging.

No. You just need to show it takes fewer matches in a rigged system vs. a random one to claim everything is fine. It needs to be shown that rigging gets people to where they belong more efficiently than random SR only system. But the Virial theorem says for large populations (distributed normally by skill), random is best possible. So current laddering is handicapped and not fair.

It is as per dev statements and patents. See for example:

That’s forcing the odds “for match fairness”, therefore it’s rigging as per definition.

Trying to determine the outcome of the game, and rigging for it, before it’s even played - is nothing like any other ranked environment. You can assign pbsr and sr bonuses for deltas incurred by randomly shipping two teams around an SR margin. No place for MMR except to rig “for fairness” - which makes the overall laddering process anti-competitive and unfair.

That would be random SR naive matchmaking. With MMR rigging, it’s cherry picking coins that are biased to fall one way or the other, and shipping 6 of them together to balance the other 6. It’s rigging the match outcome to be 50-50 by picking biased coins.

Chess is a two player zero sum sequential combinatorial game. It’s state space doesn’t change with balance patches, and it will soon be weakly solved (for example, a proof by induction on base games that white can always force a win, which is greatly suspected).
Chess is in a simpler complexity class than Overwatch. OW is an N-player stochastic differntial game you can transform to a fixed base game with perturbations and adapt epsilon Nash strategies for via e.g. pointer networks or quantum gans, like alpha.star.

They randomly seed or have you qualify for a bracket, then let you (and/or your team) progress naturally through the ladder, brackets, tournaments, etc. Even for single player ranking systems (like Chess), they use only 1 metric (elo), rank you accordingly, pay you out accordingly, and matchmake you naively around your current elo projection. If Hou Yifan is climbing a new account in Chess, the matchmaker doesn’t analyze her performance and cherry pick her subsequent opponents. Who she faces on her climb is entirely random around her elo and winrate. And there is no adaptation to how she is doing or what her gameplay was recently like.

I did a rough calc and also got 57k accounts to move the population mean 250 sr. Where are those numbers from? Who cares…because that’s not the issue. If you are attempting to climb through a bracket of 7% of the population (by player % skill), but it actually contains 11% of the accounts, you’re playing a lot more matches to sieve through. You want to play against diamond players and outrank them, not diamond accounts, which could be over-represented by master players parking alts over and over, each one having a constant-sum amount of SR to distribute, imparting geometric amounts of additional effort for anyone attempting to ladder through.

Have you shown math to dispute the following?:

As well as:

As well as:

And he goes on to give example using race car drivers and race times:

So you really can’t have alts + rigging + 0 resets and call it “fair” laddering.
I have to side with Receipts because the logic holds. His examples are extremely conservative (they argue around a data-free base case). Which is the best you’re going to get unless you can show actual data.

1 Like
162

That’s a great wake up call, I am happy to hear that!

Believe me, if I can do it, you can do it too! I used to be Low Gold and absolutely terrible: https:/ /streamable.com/nx36sp

VOD Reviews and Aim Training has helped me climb ALL the way to GM

3 Likes
163

You are Master, not GM. =]

164

You’re considering fair game pieces to be rigged.

I reject that worldview. Fair coins, cards, dice, and chess boards will never be classified as “rigged.”

You and others are complaining that “limited mobility” is a form of “rigging” and holds players back. Quickly converging on the player’s skill level demonstrates the opposite of “limited mobility.” If anything, it’s hyper mobile by comparison to the system you’re proposing.

The system quickly converges on the player’s skill relative to their peers. That isn’t “rigged.” That’s an efficient system.

Ensuring everyone has a fair coin to play the game with isn’t forcing odds.

That’s how you match players by skill. Can you think of a way to match players by skill without considering the player’s skill?

No and I already explained why it isn’t. Substantively respond if you want.

The likelihood of both teams having equal coin flips (same number of heads/tails) is cumulatively 22.5% under your system. Meaning, the remaining 77.5% of games will not have equal coin flips and be unfair.

Frankly, forcing 77% of games to be asymmetric seems “rigged” to me.

Engage with the point: giving both players a fair set of pieces doesn’t constitute “rigging” an outcome

If you want to talk analogous elo systems in chess, you need to account for differences in players autonomously seeking out tournaments versus the match maker’s need to automatically match players to a “tournament.” High elo FIDE tournaments are categorized by only 25 elo differences. Limiting the SR window like that would greatly exasperate queue times–if a match could even be made.

Additionally, team chess elo systems and pairing are also more applicable here.

That is exactly the issue. You’re suggesting that alt accounts are endangering the fidelity of the ladder. You need a 47% increase in alt accounts at the same elo being played in the same season for minor SR changes to the ladder. Not only is that extremely unlikely, the consequences are negligible for the integrity of the ladder. It also hasn’t been demonstrated using real data.

You’re not playing more matches as you don’t play every person in your bracket. You still require the same amount of SR to rank up or down–you’d still be playing with and against players of your skill level. Only 1 alt account can be played at a time.

You’re assuming a) each alt actually distributes SR and b) that the amount of SR distributed causes SR requirements to change.

Example: You need $100 to purchase a phone. 1000 people walk into the store with $100. The average amount of money in the store doesn’t move and there are an unlimited number of phones that be purchased. The requirement for purchasing the phone remains the same.

That’s not how the burden of proof works. If you’re claiming the ladder has deteriorated due to an influx of alt accounts at the same elo being played every season, you need to demonstrate–not speculate–that’s the case.

It’s not even clear where the ~7200 SR is coming from. Every player starts in Gold around 2250 and may “distribute” SR from there. The distribution of that SR is still measured relative to all players.

SR ranks are hard coded brackets. They don’t change. You always require the same amount of activation energy regardless of the potential energy in the system.

There’s exactly zero reason to believe that. Players at all ranks have alt accounts. You need to demonstrate a massive influx of alt accounts at the same elo being played during the same season.

You need a massive influx of accounts to shift the average player performance. Demonstrate that’s happening.

4 Likes
165
  1. People have been saying the system is rigged since it cam out, spoiler alert: ITS NOT. Why would a company rig a system so the players have no fun and stop playing? Answer: they wouldn’t.

Actually I don’t agree with you. The system in casino is rigged. Why? They don’t want to loose money but also is a way to trap people who refuse to accept their looses. Is simple human psychology. Some people get addicted to the looses and it’s a way to ensure they will have a fan base. So yeah, the system can easily be rigged to hook those type of players.

They are a business, they don’t care that you have fun. They care you are hooked and they can suck your money dry.

So yeah, they system is beyond rigged.

3 Likes
166

Toxicity is in nature of competitive games and it will always be a part of those games, things like saying ‘just don’t be toxic’ and trying to erdicate is in just people being naive.

The system is not rigged, Its just bad, outdated and not working properly. But hey appearently its the best blizzard can offer, maybe still not much but best they can.