ProtossCraft 2 cemented SC2's fate in stone

General Discussion
83

The pareto principle is the 80-20 principle (80% of a country’s wealth is generated by 20% of its people, you can achieve 80% of a performance with 20% of efforts and the remaining 20% would be way more time consuming …)

I do not know what notion you confused it with but it is not and in this discussion Pareto has no place so please stop throwing it as a random buzz word.

Next: you confuse popularity and strength. A race very popular is more played, a race more powerful will win more matches at similar abstract skill level. Zerg is not popular (fewer people overall) yet people with it wins more and therefore rank up (hence even fewer zergs in lower leagues). Being strong does not make a race necessarily popular (terrans have been the most played race for years and I’d argue they are neither strong or especially easy especially below master).

Zerg are over represented up to master (included), it is only among the top 600 sc2 account worldwide (including smurfs and alt account) that they are fewer.

Now again if you believe comparison is unfair with other races do take zergs only for example. You can even do your own custom leagues by comparing yourself only with other players of your own race and then see how you fare. Typically a platinum zerg player has as many zerg players lower than him (in ratio) as a gold terran player would. If zergs were to rank only against other zerg players then they would rank lower than their current mmr in all leagues but GM, the terrans would rank higher everywhere and the toss would be mostly unaffected currently except some of the GM would be master 1.

This is not merely a theory since you can do the distribution yourself going on rankedftw and taking each race separately to compare how many players are below or above other players of the same race at any given league. It is easy to do and the experience can be done and proven by any one who wants to check it.

2 Likes
84

You have my respect Sir NoobIsHere.

TheBatz, can you provide me a link to your May’s work please?

1 Like
85

It sounds like you are theorizing that mirror matchups are a better indicator of skill, and if that is the case you’d be correct on your theory. You would be wrong, however, on the relevant facts. I did an analysis on Aligulac rankings of players on a per matchup basis. Mirror matchups are a better indicator of skill, but unfortunately for you it was Terran that did worse in their mirror matchup.

FYI, every chart on rankedftw shows Zerg decreasing with time. Zerg went from 35% in Diamond to 31%. Were the zergs moving up into Masters league? Nope, Zerg’s numbers in Masters went down across the same time period from 33.9% (remarkably balanced) to 28%.

Your theory that Zergs are being pushed to higher leagues from balance is very clearly wrong. It contradicts just about every pertinent fact about the ladder. Zergs are being pushed out of the higher leagues and shoveled into lower leagues.

Thanks for confirming your degree is from Google university. The pareto principle is the emergent behavior from a random process that has a minimum threshold. One of those emergent behaviors is the 80/20 principle which is highly relevant to this scenario.

Are you really going to argue that how difficult a race is has no bearing on how popular it is? I don’t need to refute that point. It’s patently absurd.

Let me get this right, the least common race is over-represented, but the most common race isn’t? What year is it? I thought society left sophism behind in the 5th century. Inversion of reality is not a valid point.

You’re right. A theory has a plausible chance of being correct.

Again, fact denial is not a valid argument. Zerg is under-represented in every league except platinum (where it is equal) and diamond (where it has a lead). Your theory is that Zerg is being pushed to higher leagues, yet it’s the higher leagues where Zerg is most severely under-represented and where, over time, Zerg’s representation has been continually going down (meaning zergs are being pushed to lower leagues).

When your theory contradicts the facts, your theory is wrong. I am very sorry that I even need to say that but clearly I do.

86

Nope.

I could keep answering but I have made the same points that I said in different ways several times now hoping you would understand them and you did not.
I never mentioned mirror matchup, only your skills or results compared to other players playing the same race (including how well they fare against other races).

I’ll insist on this again because that’s the thing I tried to show you and you keep missing and hence shooting left and right and everywhere except at the target.

To try bypass all your biases and emotional blindness I’ll use a scenario exempt from any direct comparison to anything you are emotionally invested to. Let’s say we have a competition among three groups: the Spoon, the Orange and the Shoes. Each group has exactly 100 people and fight with each other and also against other groups. We have leagues, worst 10% bronze, then 20% silver then 20% gold then 20% plat then 25% diamond then top 5% are masters so the distribution overall is: 10; 20; 20; 20; 25; 5

Now imagine at first we wanted to compare all three groups together but some in each groups thought it was not fair because the other groups or one of the other group had it easier.
In the initial league distribution we had Shoes were represented as follow (still bronze; silver; gold; plat; diamond; master):
Shoes: 15; 23; 24; 20; 16; 2

If Shoes wanted to compare their skills to each other (in all matchups) because they felt the comparison is just unfair with the other groups (other groups don’t have to fight the same way they do) then they could try to see what league they’d be in if the distribution was as intended but for Shoes.
So to get from
Shoes: 15; 23; 24; 20; 16; 2
to overall: 10; 20; 20; 20; 25; 5
you have 5 bronzes (15 -10) who rank up in silver then you have 8 (23+5-20) silvers who rank up in gold, 12 gold who rank up in plat, 12 plat who rank up in diamond and top 3 diamond who’d be in master league. Alligulac here would give more info about the top 0.1% which here is a fraction of a person hence we can’t really use it here. Same issue with the GM, among a population of 100 people they don’t exist hence I did not put the league.

Now I’ll need to add some complexity for it to be closer to the ladder and before closing the analogy.

Now the distribution is the same as before but Shoes are twice as many as each of the other group, they are 200.
So shoes in each league:
Shoes: 30; 46; 48; 40; 32; 4
What it would give on rankedftw in distributions is this
Race : Total, Shoes, …
master: 20, 4 (20%)
diamond: 100, 32 (32%)
plat: 80, 40 (50%)
gold: 80, 48 (60%)
silver: 80, 46 (57.5%)
bronze: 40, 30 (75%)

If you want to redistribute among the league with the same proportions as before you need to account for the total population (here Shoes 200) before you do the rebalancing.

The result would not give you who is better in mirror matchups but only how a Shoes in a given league compare with other Shoes and what would be his league if and when compared to any other Shoes in all matchups.

Coming back to sc2, if we do that same thing then Zergs league would be lower when compared only to other zergs, Terrans would be higher when compared to only other terrans and toss are mostly untouched currently. (Again, completely unrelated to mirror machups).

As for your mention of evolution in between seasons … that would be the impact of a patch and a change of the balance. We do see that zergs have been nerfed recently both against protoss and against terrans hence they are doing worse now than they did before the last few patches and yet even then they are over represented at master, diamond and platinum. By over represented I mean compared to the number of players: they are 26% overall so any league where when you exclude the random players zergs are above 26% of the remaining players they are over represented. As such they are over represented in all top leagues (except GM now) and under represented (fewer than 26% when you exclude randoms) in bronze, silver and gold.

“Are you really going to argue that how difficult a race is has no bearing on how popular it is? I don’t need to refute that point. It’s patently absurd.”

It does have some impacts but does not mean a more powerful race is necessarily more popular. You’re proof of it: you seem absolutely certain your race is not the most powerful and yet you play it instead of switching. There is a way people like to play and as long as they can still win playing as they like then plenty rather keep at it even if it is a sub optimal choice (nod to the silver mass void ray since WOL, the rush BC strat before it was viable, etc).

2 Likes
87

Lying is not a valid argument. I stopped reading your post there.

88

You clearly have no clue what you are talking about but you seem to be interested. So, I’ve taken the time to write a simulation that shows how the pareto principle affects the SC2 ladder. Doing some very basic tuning of the parameters of the simulation shows almost identical behaviour to the SC2 ladder. If you are interested, feel free to read the following post where I explain it in detail:

89

Funny, that’s exactly what I was thinking about you. You are brave to engage people with actual knowledge in the matter (@TyrionThelmp and me) because you clearly don’t understand much. I am impressed you know of Monte carlo though so maybe you do have some knowledge in informatics . I answered there and pointed some of the things you did wrong but the idea behind it could be interesting.

Yet since it starts with bad hypothesis I doubt you’d go far there. I suggest you reread what I mentioned (and the post you stopped to read because you fail to understand it or its logic and therefore called it a lie). It is quite simple and if you have any question I’ll dumb it down for you as much as necessary.

2 Likes
90

Words mean nothing in the face of evidence, kiddo. Your opinions don’t invert the truth as so defined by the facts. You denied over and over again that the pareto principle was relevant to this scenario and then I proved beyond doubt that the pareto principle is capable of explaining the current ladder scenario. This proves I am extremely knowledgeable on the topic, meanwhile all you’ve got to offer are some words in a post which are, more often than not, incorrect.

91

What you proved is that you don’t understand what the Pareto principle is and have little grasp on language.

Nope but that you try to use authority argument instead of actual evidence shows you are either trolling or wholly unaware of what arguments (or facts) actually are. It also explains why every user who tried to talk with you ended up facing a wall of incomprehension from you and were eventually accused of being alt accounts of some other users I can only assumed you disagreed with at the start.

I am amazed a guy like you managed to go far enough to know of Monte Carlo simulations and even run one and yet manage to not know even the basics in how to show results, what information need to be in graphs you share in order to be usable (such as the scales used, the distribution used, etc).

Yet, and that made me smile, you eventually did read what I wrote since you mentioned somewhere that a favored race would be less represented in the lower league (contrary to what you first stated here) . You mentioned protoss at the time I believe but one more step and you’ll get it for zergs as well (zergs being even much less present there).

Finally you can like to pretend you are “winning the argument” but fact in point almost no one agrees with you. So not only are you wrong you are also loosing the argument in all your debates.

2 Likes
92

No actually it proves you don’t. The pareto principle is the emergent behavior of a random process with a minimum threshold, which produces a certain distribution where a small sub-set of the full distribution eats up the majority of the correlation. Closely related to this is the 80/20 rule, which is an application of the pareto principle to very specific parameters. On twitter for example it’s 2% of people make 80% of tweets. You’re getting confused between the 80/20 rule, and the actual pareto principle.

The pareto principle is THE KEY to understanding the SC2 ladder distribution. A race that is underpowered will be under-represented in all leagues, so why do they peak in the middle leagues? Well, that’s where new players start out. If you were to chart the number of games played vs MMR, you’d see that the upper & lower mmr regions play the vast majority of games. In other words, zerg’s representation is good in the middle leagues because new players are put there and then don’t play much.

Argument ad populum is a fallacy, sweetie. It’s highly likely that most people won’t agree because complex mathematics are difficult to understand.

93

Does your model depends on this? Is that critical to your explanation? Because if it is I am happy to tell you you can once and for all forget everything you for some reason believed for that above is false.

If you do look at MMR and map it to the number of games played you see a positive correlation between number of games played and current mmr. In particular if you were to look at the bronze who played the most (as I am writing this) he is barely at the 45th spot, after all leagues but GM (which only has a few hundreds players) appeared several times.
Next Bronze player is 130th in number of games played.

If on the other hand we look at those who played the least we have dozens of pages of Bronze and silver players who played exactly one game.
If you collect the data and draw it the graph it gives directly contradicts the non sense you are spreading.

As for Pareto

Try to understand that part, it is critical in understanding why here it does not even remotely applied. A bit like gravity you keep referring to for some reason. No, gravity does not explain ladder performance and no Newton did not write the rules of gravity thinking of you. You also don’t understand how science works (I am more and more convinced you are some computer grad students or stopped your education there and believe you know something except you don’t and you argue with people who do but you just lack the education to understand).

2 Likes
94
  • Claims model is false because of X.
  • Confirms X the next paragraph.

I didn’t read the rest of your post :laughing:.

95

Case in point, disproved you.

gg, better luck next time.

1 Like
96

No, you said I was wrong because I rely on X which you claim false. Then you do a 180 in the next paragraph and admit X is true. You were arguing my point for me on accident because you didn’t understand my point to begin with, which is remarkably similar to how another forum troll tends to face-plant in debates. Flack was right about you.

97

I don’t know what you smoke but that sure seems good.
You said lowest and highest mmr played the most, I show you it is not the case and you congratulate yourself because I prove you right?

And who is Flack, your imaginary boyfriend? Your left hand?

Anyhow I’ve wasted enough of my time already speaking to a wall. Maybe one day you’ll get wiser or at least smart enough to be worth chatting with. Until then … good bye.

2 Likes
98

This bird. Does not understand. Own words he types… or is crazy…

1 Like
99

I don’t know how to say it any simpler.
Players with lowest mmr on average played the least (positive correlation between mmr and number of games played) which goes directly counter to them playing the most.

What word do you not understand?

1 Like
100

To get big change. Of rank. Players play many games… are ok small bird?

1 Like
101

Lol.

Change of rank is unrelated to what was mentioned here (current mmr). I forgive you though, you don’t seem to speak English.

1 Like
102

New player. Must change rank. To get high or low league… spread wings small bird… and you can fly to…

1 Like