A poor argument regarding hidden pool discussion

Discussion in 'Dota 2 Matchmaking and Communication' started by DMR_Chane, Aug 30, 2016.

  1. aberjams

    aberjams Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    How recent are you thinking? Past 25 games? 100?

    The algorithm is there. They choose not to share with the rest of us. We are left to wondering and the #hiddenpool. Don't bother losing sleep over it. They could be tweaking the matchmaking algo and we wouldn't know.

    :insert theorycraft:
    I think that most people are assuming that behavior score, matchmaker rating, what ever else is involved(p, q, z, etc..) I would imagine a diminishing target value is applied in relation to time . This would open the variables for behavior score and elo as more time passes. The more time that passes by, the more varied of pool you would have in relation to your personal numbers.
  2. Animus2280

    Animus2280 Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    Chances are, the university I went to is far superior to the one you went to, (unless you attended Oxford or Cambridge.) However, I won't bother arguing with your "160 IQ" Mr. Genius because you haven't given a single shred of evidence (let alone proof) that that Valve is matching people based on behavior scores.

    I repeat, not a shred of evidence. You haven't even made your case before preemptively attacking the other side and saying that it's "too difficult" (??) to set up an "experiment" because there are too many "factors." (rofl)

    Furthermore, there is no reason to believe someone at 7000 behavior score is going to be "stuck" at a ranking below what s/he deserves because you've heard people with 9k behavior score on this forum reporting that they get matched with players who call mid or feed at the start. Maybe that ought to be your focus instead bragging about how you learned "p-value" at the "basic high school level" without taking a course.

    By the way, your "high IQ" was conspicuously absent when I insinuated that statistical concepts you brought up are not sophisticated, but you misinterpreted that as "I had to take a college level course to figure it out?" Does the fact that I have taken 2-3 statistics courses and am aware of the meaning of p-value indicate that I would not know/be capable of understanding without those courses?

    Wait, let me help you out:


    As for my "argument," I merely pointed out that the vast majority of people making this claim are 3k-low 5k shitstains who couldn't carry a game out of a paper bag, and are routinely making excuses as to why they cannot climb, so it is unsurprising they would latch onto any new conspiracy theory like a moth is attracted to a flame.

    (That is an observation, not an argument; hopefully you can distinguish.)

    Anyway there's nothing further to discuss, and all these responses to a trivial topic are mind-numbingly absurd. Plz jack off to your intellectual prowess more.

    Edit: I should probably adopt the nadota response technique in the future, this entire ordeal was a poor usage of my time.
    Last edited: Sep 1, 2016
  3. DMR_Chane

    DMR_Chane Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    I am not attending University yet, and judging by your reading ability I would've guessed you didn't either. I don't quite understand why you keep harping on about things like 'Mr Genius' considering I've not once praised myself in this regard. I've only commented on your apparent abilities, or the lack thereof.

    Now to get on topic. I've not once said the experiment is 'Too difficult' to conduct, I've only said that it is difficult to do it well, as you want to reduce the amount of changing variables and the degree in which they change as much as possible to obtain a more sound result. If such hurdles as obtaining two almost identical MMR accounts with vastly different but measurable behaviour scores and the exact same report/commend record, a player that is very close in skill level to the used MMR, is consistent in role played in the experimental games as well as in communication, large enough sample size etc. etc. etc. It's doable, I've never said it wasn't, only that the results will never be as conclusive as we'd like them to be.

    The evidence that points towards factors like behaviour score and the report/commend record being used is their very existence. It is illogical for a company to create such a system to then not utilise it. This is why such experiments should aim to DISPROVE the hypothesis that these factors matter. The assumption that a system would be used for it's purpose is more logical than the assumption that it wouldn't be. This is why we assume that it is used and attempt to disprove it.

    ---------- Post added at 07:37 AM ---------- Previous post was at 07:32 AM ----------

    There is no way to say for certain how recent, the question would also be if it declines at a steady rate or if the most recent games are relatively exponentially important compared to past games.

    I do not subscribe to the theory of a sepperate hidden pool, only that factors such as behaviour-score are used alongside your ELO to find matches.

    We know for a fact that the MM system starts narrowly and expands it's search-range the longer you keep searching, Source1 had this bar which expanded outwards as you kept searching and said it was expanding search range. It is not unreasonable to assume other factors would also be taken into account.
  4. Trew

    Trew Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    Not a shred of evidence?
    You can take MMR and Behaviour score. Low Behaviour score = Low MMR. The trend is clear. By these data you can atleast gain 500 MMR by simply changing attitude and not improving your skill at all.
    Feel free to use the data yourself if you don't believe me.

    ---------- Post added at 09:12 PM ---------- Previous post was at 09:10 PM ----------

    Also if a 7k player can barely keep 50% winrate on a 4k account with low behaviour score that should more than prove it. What are you even argueing about?
  5. mitochy

    mitochy Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    Yeah Dota is a team game that requiress teamwork, so it should be a given that attitude is part of skill, would be strange to consider it otherwise. E.g. not tiltable is a must for mid, and being cool header is critical to preserve teamwork. In basketball if durant screams like a 5 years old every time hid team fail to pull a move his team will never win. Or get angry so much he plays badly. Team morale is important to win
  6. Trew

    Trew Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    You are not responding to my 2nd point
  7. mitochy

    mitochy Well-Known Member

  8. Trew

    Trew Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    Interesting article
  9. GoLD-ReaVeR

    GoLD-ReaVeR Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    I know it may sound ironic, but can I get a TL;DR on that?
  10. henpara

    henpara Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    TL;DR of the article:

    To get to the best team you necessarily need the best of the best, nor do you necessarily need ppl who are similar in interest or habits.

    A thing that every great team has in common was a bond between team members: They
    kind of cared for each other, nobody was scared of anyone else.

    But they also say that except for this there was no pattern or too many patterns on how to build the perfect team.
  11. Teremus

    Teremus Well-Known Member

    Re: A poor argument regarding hidden pool discussi

    Bad player complaining about getting bad teammates and thinks he's good all though he is just as bad as them.