usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Thank you! This post though wasn't about climbing - I'm happy to never-ever-climb again and whatever division I am in to be named "Basalt IV", as long as the skill of the people IN THE GAMES I PLAY is such that the games are fair and competitive, not skewed heavily 3/4 of the time (in my experience). I think the matchmaking system could be improved to foster that. Not more, not less.
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Yes yes, I get your point, "get good" - well said! Let's maybe briefly go back to my original point, which was: "unfair games 3/4 of the time due to sub-optimal matchmaking" and not "I should be Challenger and am not"... but hey, who cares, right? You have a view and that must be the truth! But since you are so bored at work, here's some reading for you (although since you do recognize the "eloquence" part, maybe there's hope?): https://www.researchgate.net/publication/12688660_Unskilled_and_Unaware_of_It_How_Difficulties_in_Recognizing_One's_Own_Incompetence_Lead_to_Inflated_Self-Assessments Bye now!
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
You proved to me that my gut-estimate about the 100x100k coinflips average head-occurance not converging to two zeros as the first two digits after the "." was wrong - nothing else. The initial question wasn't whether that's true (although yes, I did make a mistake in my gut-feeling-guess - feel free to get hung up on it in every post from now on), but how to know what your average winrate SHOULD be. Anyway, listen - I don't need to evangelize you. Believe, do and say what you want, and good luck with your life if you base your decisions on "average probabilities for which you have data" and if you "put your stock in things you can show, math out, and prove". Maybe if you grow a bit older and gather some experience outside of case studies, you will see my point about "reality" vs. "averages". Maybe not. As a guy who got the massage (you may have heard of him) once said (roughly, quoted from memory, I'm too lazy to look it up and you won't care either way): "Ships may sail around the world and people will still believe in the flat earth theory"... or, more to your "prove and show" point (quote from me... kidding ;)): "Einstein may have already developed Special Relativity and people will still prove things with Newtonian mechanics". Now take care - I am not interested in discussing further once my discussion partner, however interesting he or she appeared at first, proves that they are close-minded and have a very narrow perspective on the world :)
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
No, we don't :) Sorry, that seems like a silly discussion now - if all the system does is "you have an MMR, and the team gets put together with others who have the same", then it's NOT a working matchmaking system and all my points are even more valid. There is so much more you need to account for and I am certain Riot is not doing it. In any case, Blizzard or any other company talking about "MMR systems" has zero implication for what Riot is or isn't doing. That point seems fairly obvious to me. Why is the basic elo system adaptation (described in your link) not valid... I don't wanna type for the next day or so explaining this - there are many reasons, but you have to be a bit open minded for that. The main idea is: if you play chess, EVERYTHING depends on you - NOTHING is left to chance. If you play a team game, there are COUNTLESS interactions and interrelations that you have to model - we talked about this before: auto-fill, impact of each role, champion knowledge, even RNG (triple-infernal with Ryze? Great!! triple-ocean... not so much). If you are theoretically interested, read the book by Elo and tell me that THAT thinking even remotely compares with the dynamics of LOL... hint: it does not. And again: the idea that "everything just averages out" is one of the fundamental, underlying, theoretical problems of people who understand statistics "just well enough"... every heard of "efficient financial markets"? Was really a thing back in the days... people BELIEVED in this with all their heart (my running partner at university, for example). More simplistic: the chance that I die in a car accident is x%, on average, in North America... except, it's not, because I live somewhere where I don't need a car! Now, here is the more important problem: IF I nevertheless die in a car accident the one time of the year when I use a cab, I will NOT CARE about averages - that's why "things that are on average good" can mean nothing in individual circumstances, such as "a lol match with 10 real players" - THERE they should work correctly, and THERE they should "balance". Not "on average across everybody over 100 games". That's like the hunter who misses the dear by a foot on the left with his first shot, then by a foot to the right with his second, goes home thinking "statistically, the deer is now shot" and then dies of hunger... But I have the feeling it's a "religious" issue with you - just like my friend from university... he could not understand (and he was absolutely smart and highly intelligent) that this could be a concept that may be faulty. And, by the way, it's with Riot, too, that is PRECISELY their thinking - so we will just have to live with the fact that "on average, everything is fine". Last thing: fine with the 20 champ requirement - then lax it a bit... make it a combination of number of champs owned, make more champs free to play, make it a fixed roster of champs that are always free to play, ... - there are countless options. And honestly, getting mastery 5, that's like 30-ish games with the champ, not an issue if you want to play it competitively... be open minded! It helps expand YOUR horizon as well (since that is what YOU preach) ;)
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
P.P.S. I think the guy who gets to level 10 playing only Garen, Tryndamere and Ashe is WAY better than the guy who is level 30 and plays first time [INSERT CHAMP NAME]... the game isn't THAT complex (at least, nobody IN GOLD knows how to play it anyway, so...)
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
P.S. re. "punishment" - the guy who plays first time Irelia into my Trynda in ranked punishes MOSTLY his team, because he himself will LOVE the fact that he can learn that champ in a "real" environment and wasn't expecting to win anyway... that's why so many players get to G4 and then switch roles, because they think nothing can happen and now they are where they want to be. Also, as aside note, current rank 1 player in Korea believes that the best role to climb solo-queue is... SUPPORT... for exactly the same reason(s) I am outlining here :)
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Haha naw, to not write another "novel" here: 1) no, "MMR systems" are not publicly known - nobody knows if Riot uses a "k factor"... you talk about ELO systems, which are designed for a completely different problem (for 1-v-1s, and for tournament outcome assessment where the tournament performance itself determined the matchup, not the elo - after the initial requirement). How to implement an "Elo-like system" for a multiplayer game to (a) represent the team size and (b) to use it as a means to CREATE fair games in the first place is an interesting scientific exercise and my no means "public knowledge" 2) sorry to say, but you play support main (as you say) - that means I will have to discount the mid tips. The fact that it gives your team 90% of the time an "autofilled advantage" is probably enough to get out of gold and plat 3) I agree that the main thing is "knowing where to be when" and to "know the matchups inside out" - those things are way more important than cs, roam etc. - the latter are "fundamentals", and the fact that many Gold players have no idea how to shows already that "luck" is quite a factor here 4) I really cannot see why my proposals are not crystal clear - duo-ing will always distort the competitive balance, because you cannot know up-front what the impact is going to be. I have lost GAZILLION of games where "my duo" was so dramatically bad and the matchmaking system assigned them a higher MMR value "together" that was just bogus (it's often enough two mid-laner mains playing botlane together, and the system doesnt even realize it's NOT their main role) 5) mastery 5 is a requirement? What about getting to level 30? That takes MONTHS
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Hey - so first of all, let me say a few things: 1) thank you very much for the work in this post - greatly appreciated!! It seems you are a very smart, super decent and friendly guy. So please be as aggressive as you want if THIS is also what you post :) 2) regarding math (now that I care what you think) - while my coinflip example isn't ideal, it is actually also not as "wrong" as you made it sound... if you like this kind of stuff, expand your code so it reflects players put into teams of 5 and the TEAM has a winrate of 50% every single time, but the "wins" are counted by player. I have built this in Python (seriously, who uses Matlab in 2019...?! ;)) Just kidding) and the results are what I say they are (you get a distribution that is very similar to the one we have in ranked, EXCEPT that the players' rank-positions change more often - but if you add an "actual skill value" and adjust the percentage of winning based on that "actual skill value", you can get away from that OR you simply make it very difficult to drop at certain intervals (from G4 into S1) and you will remove some of the change in rank-positions... Now everything depends on the assumption of HOW MUCH the actual skill value changes the percentage of winning, obviously... and HERE is where I think Riot has made a few critical mistakes in their system. If you REALLY wanted it to be competitive, there are a few changes you could make easily, see below. By the way, if you are bored at work, I posted a lot in the (distant) past around that topic, when we were trying to built a much better matchmaking system (which isn't that difficult, except Riot never tells us how theirs works really, so it's not possible to compare). Now- also again THANK YOU VERY MUCH for all the thoughts you put into the tips. They are generally excellent and super helpful. I have to say though that: a) I am aware of almost all of them (and as you say, sometimes you can implement them, sometimes you can't) - that's the "tips and tricks" that many (really good) players give and they are not "wrong", but knowing and implementing them is not always gonna be enough, because... b) the implementation by itself is already "team dependent" - you seem to know very well that the no.1 objective to "win" is to make sure the enemy team tilts first and / or more than your own team - that requires A LOT of work, also in your team... for example, you know you shouldn't team fight 4v5 right now with Baron, but just siege (especially while your Renekton is botlane splitpushing with Ignite)... but you HAVE TO because the Volibear goes in (you wanted to watch a Replay? my forum's account last game is an ideal example... more on that later) and he has already question-mark-pinged you before because you didn't die with him in another stupid engage... if you run / stand back, your whole team will tilt and losing that teamfight may be better than losing the game with ff-vote-spam because "you didn't help the team" c) not all match-ups are "correct", but that's fine - Ryze cannot "just shove" early against a decent Ziggs, his mana problems pre MFB-stacked + Lost Chapter are much less than Ryze's (as are probably all other champs as well ;))... but even if they weren't, they still depend ON THE TEAM - if your botlane always mindlessly pushes although they are 0-3, your roams will come to late to safe them, etc. - you know what I mean. You still have to have a halfway decent team that's "not yet tilted". These things are what I mean with "luck" and, to be honest, that's why I think your tips will not work by themselves most of the time (although they are absolutely and without doubt right!). Instead, the reliable way is to play a "one vs. five" champion (Jax, Yi, Darius, maybe Yasuo, ...) or a "very hard to not get ahead very early, with the objective being to tilt the enemy team quickly" champ (Udyr, Pantheon, Jayce, sometimes Wukong, maybe TF if you are really good, sometimes Zed, ...). There are some others that work reliably in games that you expect to last 30+ minutes like Mumu, Annie, ...) but I think those are already "risky" unless you play them very well (because of tilt, again). THESE are the champs that "boosters" do play for a reason (although I understand that they are good enough to win with Soraka mid, fine - they STILL pick those aforementioned champs, and still for a reason). Go back (if you like only of course) to the aforementioned last replay - why did we win this? Because Vayne gets a triple-kill bot... did we enable that? Yeah, maybe.. were there mistakes (from me? At least 3 - get caught twice, one bad ult.. .probably a lot more that I don't even realize I made) ...? Of course there were... but THAT was the game-breaking moment and without it, Vayne tilts / is done / the enemy gets excited, and it will not be won. Did I contribute to that specific instance? Not at all. That's just "lucky". And yes, maybe over time, the winrate will get you somewhere, but the way right now is just "too much luck for too long" to "feel" really competitive... and I think it could be better IF people would stop defending the current system, especially people who are in Diamond and above, because THOSE guys will ALWAYS be listened to by the rest of the community :) Also, if the "climb" takes too long, then YOU (as in "me") will actually not be able to "learn" anymore - I wholeheartedly agree that learning should be the number one objective (I play Ryze - it's one of the things he constantly says in game ;)). BUT if these "luck" moments determine the game most of the time, I will not get into the situation often enough (2 of my big mistakes in the last replay? They are made AFTER the triple-kill Vayne... if he doesn't get them, I won't get the chance to make these mistakes in the first place, and I will learn nothing from that game and it will re-confirm my belief that it's "just luck"...). Yes, cs-ing with Ryze is a pain in the b... early, that I could improve... but seriously, that's not gonna be it (no need to get Faker's cs in Gold - as long as I have more than my opponent, I'm doing fine)... But again, you are awesome and if all people here and in the game would discuss with your depth of thought and desire to help, maybe Riot WOULD listen eventually, but in any case, the whole experience would be SO much better and more pleasant. So thank you for that :) P.S. - "Suggested changes to matchmaking in solo queue to increase competitiveness": i) no more duos in "solo"queue - ever.The adjustment factor just CANNOT work (if I wanted to pay the duo-booster to get me into D1, he is not gonna play on a D1 account, he's gonna play on a S1 account with horrible winrate, thanks to bots playing that account... more likely, he's gonna play 3 different S1 accounts over time with me) ii) match auto-filled positions across both teams - jungler A auto-filled means jungler B also auto-filled. NO discrepancy here, EVEN if that means slightly longer wait times iii) no champs are playable unless you have mastery 5 AND played the champ at least 5 times over the past 30 days (across ranked and normal SR, not ARAM / TT / etc.) - possible exception: you are auto-filled
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
My last 5 games for you - in a ROW: 1) 27 min game, me: 6-3-8, 245 cs - top GP, 1-10 vs. Yasuo... loss 2) 28 min game, me: 2-2-4. 221 cs - bot combined 1-14 Cait / Soraka vs. Ashe / Zyra... loss 3) 33 min game, me: 5-6-6, 285 cs (vs. a f... Neeko as Ryze) - jungler Kay 0-10 vs. Amumu... loss 4) 21 min game, me (as Kat this time): 6-3-7, 125 cs... - jungler Lee 0-10 vs. Xiin... loss 5) 28 min game, me 4-5-5, 235 cs - bot combined 3-16 as Twitch / Lux vs. EZ / Yuumi... loss Any questions about "luck"...?
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Good, you are insulting already but that's fine - let's be constructive then: explain to me the following - assume I play consistently better than my opponent as, say, Ryze-mid. My opponent is, say, Malzahar, Anivia, Ziggs or any other easy-farm-champ onto which you will NOT get a kill as Ryze-mid if the opponent is halfway decent (we are talking Gold - they are not idiots and they are not autofilled mid). HOW do I win a game where my autofilled-jungler is 2 levels and 5 kills behind the enemy K6 / Xin booster at minute 8? HOW do I win a game where the botlane loses first tower at 12 minutes and the enemy adc is 3 kills up? HOW do I win a game where the top-Tryndamere gets first tower at 9 minutes? I know what you are going to say - "you don't, but hey, if YOU play consistently better than your opponent, it will not matter!! Over time, you will STILL have a positive winrate, because all the things above could also happen to the other team!!". But it's NOT - because ONE game that I am supposed to win, I am unlucky and give a stupid kill away because I misclick the root on the wrong target or I flash instead of cleansing vs. LB. And another game, my opponent actually doesn't play Malzahar, but he plays Zed and is a booster and DOES get 2 kills on me.... That gets back to what my OP was - it's LUCK to get out of Gold, unless you are A LOT better than Gold, OR you play with a friend all the time AND dodge consistently. That's how every streamer does it FOR A REASON. Lucky for you (Diamond 2 - your forum account at least), you are in group 1. But I am not. But I am better than Gold 4, and getting out of it requires luck or hundreds of games. Period.
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Again, why so aggressive? Listen, I am not a Matlab expert, but I believe you if you say so. Then let the coinflips in your experiment converge to 50.00%, my apologies and maybe I have talked me into a corner here with that example... So that means you DO do a good job! The point I was initially making (see OP) would still stand, but you have managed to discredit it.
usul1202 (NA)
: They explicitly define boosting as logging into someone's account. Otherwise you get interesting lines like, if I duo with a friend in danger of demotion, am I boosting? In some situations it becomes quite close, so they set a hard line a bit further back, at 1 account multiple people.
Since you know Riot so well, can you please confirm: if I pay a dude (he's a challenger player) money so he duos with me until I reach Diamond 1 - is that boosting? Or am I allowed to do that? And if I'm not allowed to do that, why not? What if the challenger player is "my friend" and I'm "in danger of, say, not getting into Diamond 1"...?
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
It's funny, going through your posts... what I say below is EXACTLY what you are doing - everything Riot does, says, plans, ... is fine, great, working as intended. Everybody who has any sort of criticism gets shot down by you, trying to imply that you are "just the average dude" so that the other posters think that they are the stupid minority. Not a particularly "ethical" approach if you ask me... just accept the fact that things don't always work as intended, that stuff could be improved, and that you don't know everything... that would be a start (and since we talk about probabilities: a "highly unlikely" occurrence, for sure).
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Well, saying "your lack of knowledge" or "I know statistics better than you", then posting numbers (which are not correct - for example, there is an extremely low probability that the "average winrate" is going to be 50.00% in your "experiment")... is exactly what Riot does - it's still a flawed understanding of the difference between statistics "on average" and what that means for individual players. And it's the reason why the current iteration of the matchmaking system does not work as well as it could. Yes, you get a bell curve if you run a coinflip experiment often enough... but that's not what's happening here, right? We have a mechanism that does the following: if you win, the next time you win, you get more points for the win and lose less for the loss... but if you start with a loss, then [...] - I'm sure you know the system. Also, there is a starting point that's somewhere between Bronze and Silver. These are (together with game-dependent factors - different champions are differently well suited for solo-queue, as we all know) what creates the "bottom-weighted"-ness (which is not a term, by the way, but it's clear what you mean anyway). I seriously think that there are people here (maybe you are one of them) who work for Riot and are paid to make these statements here that are intended to "discredit" everybody who does criticize the system. Because wouldn't it suck if people would understand that - ON AVERAGE - it's just luck that makes them climb faster or slower, and that ON AVERAGE the number of games they will need to climb is determined by math? And as per your "proof" of non-luck - no, that's not what happens to me. I get exactly one account higher up over many hundred games each season, and the others I get to Gold - because I need too many games to climb. Not enough time to get the rest out as well.
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
...well, I'm not sure whether or not you are saying I may be the typing monkey... If not, then this is the same idea, but certainly not in the same ballpark. The probability to get n wins over k games is much higher for higher numbers of "n" than people commonly believe... even if the probability each time is 0.5. But anyway, seems the discussion died...
usul1202 (NA)
: No, coin flip means you have no effect on it because it's 50/50.i was better than golds so I climbed, then got stuck. Then I got better than plats, so I climbed again. That's not a coin flip, it's a slow climb. Grind is a decent word for it. And luck definitely happens, I have a high plat level friend who lucked into low día, and it's hard to fall out of a 4 division. But luck doesn't take you too far, and people will get within a division or so of their true elo real consistently. Remember, the idiots are on both teams, and so are the good people. Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! At the end of the day, you have control over your character and how you influence the game, and that gives you at least some control over your winrate, and what's you you climb/grind/crawl your way up the ladder.
Dude, I am really really knowledgeable in statistics - but you don't have to be to know that if everybody of a certain population has a 50-50 chance of winning each game, then there WILL be the distribution that we have in ranked. It's not gonna be all in the middle - that's what people (who have no clue about statistics) THINK, but that's just NOT how it is, whether you believe it or not doesn't change that fact (that's actually "a fact"). This: "[...] Luck becomes irrelevant over a decent enough number of games, and the people who cheat that with low numbers you'll rarely see; they don't play! ]...]" is also WRONG, but again, you need to understand statistics for this. Luck is irrelevant ON AVERAGE, that's very different from the relevance of luck for small minorities who got consistently lucky... if you let 100,000 people do 500 coinflips in a row and you give them 1 point for each time it's heads, I GUARANTEE you that you will get a distribution of points that looks eerily similar to the rank distribution (minus the people who are ACTUALLY good enough to really get there all the time... I am not arguing they don't exist, I'm just saying there is a middle ground - it seems somewhere in Gold - where "luck" is much more important than not-substantial skill differences, as in Gold 4 vs. Plat 4). Also, where is your evidence for this statement: "[...] people will get within a division or so of their true elo real consistently [...]" - what's the data? Why are you saying this? I can tell you: there is no data, you have no idea whether that's the case or not, but YOU got lucky and you are happy where YOU are, so it must be right. That's why this discussion here is ALWAYS the same - the lucky ones argue for the system, the unlucky ones argue against it, and nobody knows for sure. Example? You even argue against basic statistics - not my opinion, STATISTICS - just to "confirm" that YOU were not lucky, but YOU are within a division of "your true elo". It's make-believe, not facts.
usul1202 (NA)
: I got out of gold and hardstucked myself in plat for 2 seasons, never spammed that many games. Pick any hardstucked plat account with a medium number of games, they didn't magically get gold. I've never seen someone making videos of their climb although I don't doubt it exists (low view numbers probably). Disguised toast is probably the biggest gold streamer (was last time I saw, although improving at the time), might want to check him out. I don't mean to be too aggressive, but I coach people from time to time, and I honestly find the mindset you have to be the absolute largest mental roadblock to improvement. I do wish you the best on your climb though, hitting a goal is always very very satisfying.
...that's what happens when you have a coinflip - effects of probability, not "the individual" - that's what I'm saying. Again, I understand that people who don't cs well, die all the time, don't do damage if they are on a champ that should, don't roam, don't focus objectives etc. etc. don't climb - maybe for them it's a mental roadblock. I do all those things more than well according to all metrics - I am climbing, but it's a GRIND not a "climb" that I find to be disproportional to the state of ability to play the game I am in. My opinion is that many people who are lucky to get to Plat etc. quickly reinforce the believe that "you just have to believe in you / don't have a mental roadblock"... EVERYBODY and their mum believes they are at least Challenger, so that's simply basic psychology, mixed with probability / statistics. And the lucky ones talking just makes Riot look the other way and assume "all is good", instead of fixing the matchmaking and finally accounting for things such as boosters, duos, champion mastery etc. - THAT would make the game competitive and NOT a conflip (until you reach D1+ proficiency and can beat 'em all).
usul1202 (NA)
: Except that the %'s of different elos are readily viewable, and that 0.5% is of ranked only accounts. It's well known that bots lose a vast majority of games, and wouldn't fill out the numbers to makeup the ranked ladder that we have. If they never play those accounts they'd sit in iron/bronze where new accounts place. Irons pretty empty for that to be a thing. You're making ridiculous justifications for why people/the system prevent you from climbing, when plenty of people do. There's plenty of facts (and peoples experiences, watched a friend slog through gold last season and he's doing it again now!). It's doable, but you have to improve and be high gold to beat low gold, and low plat to beat high gold.
Again, I don't think any of that is "observable" or "readily viewable", and therefore my assumptions (I am not "justifying" anything...) are just as ridiculous as yours. And I also disagree with the statement that there are "plenty of facts and peoples' experiences" - every person online, every streamer, every booster I know, anybody who does an "iron to challenger" video series does the things I mention in my above post: plays with somebody else, dodges all the time, or is already so good that it really doesn't matter to them (my "D1+ dude", whether that's accurate or not). If that's not true, show me ONE video series of a guy who plays all by himself and gets out of Gold (from an 18-18lp account) alone in under 250 games and is NOT a D1+ level player... There is also generally no need to start being aggressive towards me - it's great if you can do all these things, it's also great if you don't have negative experiences. I can't, and I play "better by a margin" than the average player (see above) consistently. And I have these experiences all the time - I get to 10 wins in a row, then I get teams that lose the game even before it really starts until I'm back at 55% winrate... that's how it is for me. I AM also climbing, steadily, and I will be in low Diamond again by the end of the year with ONE account, because it's going to take me 1,250 games (I can't even do 750 as I mentioned above)...
: For 3 game promos Chance to win 1 game is 55% So you have to win 2 out of the 3 So thats 55% and 55% and 45% (if you have to play all 3 games. Since there is 2 combinations to for those results We have 55% x 55% x 45% twice Which is 27.225% Now let's see the probability of winning both games in a row Since its only 2 we just have 55% x 55% Which is 30.25% Now since you can get through by winning 2 games jn a row or 2 wins and 1 loss We add them together Giving a grand total of 57.475% Meaning there is only a 42.525% chance you loose promos (Im kinda a geek ok{{champion:74}} )
Haha first of all, thanks, I appreciate the effort :) I am a geek, too, and I am certainly able to do the math... there remain two problems with your logic (not your math): 1) IF a 55% winrate would IMPLY a 55% chance of winning the game (which it, in theory but also ONLY in theory in this case, according to my argumentation line, WOULD - see no.2), I would still say that something which happens with a 42.5% chance isn't "unlucky"... maybe "a tiny bit unlucky", but certainly not "unlucky"... maybe that's me, I think I'm "unlucky" when something happens that really wasn't supposed to happen (for me, that's like a sub-10%-ish probability) 2) now, more important is that I am arguing that the individual game is NOT a 55% chance - it's very one-sided very often, especially when you are in the realm of "above 50% winrate", because the game is not trying to give you a 55% chance of winning a game, it's trying to keep your overall winrate at 50%... so, the second you get too high, it gives you players that are more likely to lose you the game (e.g., 2 autofilled vs. none) - and MUCH more likely than with a 45% chance. That's the point of what I experience. To be clear: the IDEA that you slowly improve and that you have to carry a "larger share" etc. etc. is all fine with me. I am saying the implementation is not working out well - maybe because of what one of the above posters said (1 bad player is WAY worse than 1 good player for your team - unless you are "so good" that you can compensate that, but that's just not "gold good"...). I think you have to become a "challenger-level" player (or whatever, D1, I dont know...) to get out of gold... but by being "only better on average" than your opponents, you REMAIN in the coinflip thingy... your improvement has to be disproportionately dramatic (or: you get lucky - hence the distribution - see the example of what happens when you give 100 people 100 dollar and they exchange it randomly with each other.. or you simulate that in a computer, I'm sure a geek like you knows what's the outcome... if you don't: it very much resembles the rank distribution over time) in order to climb out "on your own". And all the other factors I mentioned (boosters, smurfs, boosted accounts) just make it worse.
usul1202 (NA)
: No but the person I was originally responding to said that was the big one, hence my reasoning. Correct, so there's even less people there to smurf. Making it less likely, since people at high elo will consistently climb back to that elo.
What if every D1 player had 1,000 accounts that he plays? What if 60% of all "active" LOL accounts were bot-level-farming accounts, intended to give the D1+ players the accounts to boost with? What if every player has 2 additional accounts that he / she doesn't ever play with? Are they? I don't know... are they not? You don't know. So don't come up with probability judgements about these things that are heavily based on assumptions. Instead, let's stop making assumptions and stick to what we experience. If you don't have my experiences, great, good for you!
: Why do you have to be atleast D1? Surely if you play in lets say... Gold 3 Surely if you are gold 1 or even just gold 2 standard you will be better than the average player in your game
EDIT: Actually, never mind, I may not have explained it: So, first of all, D1 is an estimate - I don't know if that's correct, but all booster services say they "only hire D1 or above", so maybe there's some truth to it. Second, I play better "on average" than my enemy. Not only on average actually, I do that consistently in, say, 90% of my games. How do I measure that: cs, damage dealt, vision score, damage to objectives, kda, what op.gg says regarding my "rank" in the game - you name it. STILL my winrate is 50%, because the second it goes over it, I will get three games in a row where my team has 2 autofilled (jungle and adc), and the enemy team has NONE autofilled - guess who wins. Yes, if I were "D1" (or whatever), maybe I could still carry - but I think the matchmaking system INTENDS to NOT allow for that, UNLESS you are so fucking significantly better than your rank that it doesn't matter. Whether that is intentional or just a side-effect of that winrate-based matchmaking, I dont know...
: You say games are unfair and still a coinflip? A coinflip IS fair So there is an even chance that your team gets the smurf or booster. So to have a positive winrate you simply need to be able to have more of an impact than your opponent when you are both, as you put it "gold good"
Haha ok, I give you that "fair" vs. "coinflip" thing :) Correct you are. However, what I meant (and I'm fairly certain you did understand me pretty well) is: what I do is relatively irrelevant (as long as I'm doing my job, according to my champion pick, decently well). The "coinflip" for me is: do I get the booster? Do they get the feeder(s)? And also yes, I did not say that I wouldnt climb - I said with THAT if I am doing "gold good", I will maybe (with a bit of luck still, see above) get out of Gold in 750+ games. Which, obviously, is pretty much in Riot's interest (since it means I play all the time, which I do) so we can assume it's how it's "supposed" to be. Is that a competitive game? No. It's a "game". Like playing D3 online vs. others.
usul1202 (NA)
: 5 feeder chances on enemy, 4 on yours. 5/9 is 0.55, 55% win rate. If you're average on lp, (+15 -15) then over 100 games that's +150. That doesn't include the fact that you'll climb in mmr and start gaining 16-20 and losing less lp for a loss. And that is the bare fucking minimum, if the only thing that matters is not feeding. If you're actively contributing to wins it'll be even quicker. And, also to the original post, d1+ is less than 0.5% of players. And they have to play their own games (at least 1 a day for master+). You are not constantly getting smurfed on by that level.
It's not, as I have explained in the OP. Please read. Feeders is ONE element. And yes, you constantly get smurfed. And no, it's not 0.5% of players. It's 0.5% of accounts.
: A 55% winrate will mean that loosing promos would be unlucky.
: But that one player is only doing bad because the enemy player is good. > If one person is doing well, sure he is basically 1.5 players. Sure by himself he counts as 1.5 players But he also takes another bad player out with him on the enemy team And now the rest of your team can get fed off that? At the very least its a 2.5 player swing Which makes up half a team. Now since its the coinflip You will usually get 2 or 3 good players Thats 5 - 7.5 player advantage? What if you now get good enough that when you are the bad player against a good. You can make it onpy a 1.5 player swing? Or when you are the good player against a good player you make it a 1 player swing for you? Assuming you get 2 good and 2 bad.... That means your team has the advantage
That's just not true - there is a difference between the other player being "gold good" which will make his opponent 0-2 and that's fine the team can still win. And most certainly, one REAL feeder will make more than up for that. With 2-0, "gold good" will not consistently win (nor will "plat good"... you have to be D1 and above). And again: you are assuming that the games are fair on average - I disagree! There are so many smurfs and boosters in these games (and yes, they are easy to spot imho), that you still need luck - hence, coinflip.
usul1202 (NA)
: A good player will consistently make their opponent the feeder on the enemy team though. Even if not, if you're a hyper passive laner that scales, there's now 5 chances for the enemy to feed and 4 for yours. That's a 55% win rate right there if the only thing that matters is not having the feeder on your team.
That math unfortunately makes no sense... and even IF it did: 55% winrate, that meansd 50+ games a week AND you have to be lucky that the games you win are when you are in promos and the ones you lose are just after when it doesn't matter. If it's the other way round, 55% winrate will NOT get you out of Gold anytime soon (yes, play 750 games, then it will).
: > [{quoted}](name=Cind3rkick,realm=EUW,application-id=3ErqAdtq,discussion-id=vOmM1hIU,comment-id=0000,timestamp=2019-05-21T14:36:22.980+0000) > > Or you just become good for your elo and are able to get it to a 60/40 flip as you can make up for a teammate that doesnt play as well One bad teammate is much more impactful than a single good teammate, and that’s the problem now. If one person is doing well, sure he is basically 1.5 players. But one person who plays badly and feeds, they don’t just feed one person, the enemy team is going to capitalize and use that one person to get multiples of them fed. TLDR. One person can easily feed 3-5 people on the enemy team. But one person can only carry with the strength of 2.
: Or you just become good for your elo and are able to get it to a 60/40 flip as you can make up for a teammate that doesnt play as well
That's a good idea, see my option 1. But not realistic, since you have to be at least D1 to "consistently make your opponent the feeder" - if you really knew the game, you'd know. Say you play mid, enemy picks Malzahar... or Anivia... and plays super-safe. One feeder, game over.
Rioter Comments
: > [{quoted}](name=rito are scambgs,realm=EUW,application-id=yrc23zHg,discussion-id=EF6AgGkM,comment-id=00060000,timestamp=2019-05-16T00:20:46.675+0000) > > you might learn few things from diamond players if you are bronze or iron but you won't learn anything from golds and plats they are just slightly better.. nope golds and plats still make tons of mistakes that diamonds can point out. I'm in a discord with high diamonds and a few masters that could teach you things you don't even know.
> [{quoted}](name=GreenTeaAndPoro,realm=NA,application-id=yrc23zHg,discussion-id=EF6AgGkM,comment-id=000600000000,timestamp=2019-05-16T01:28:37.590+0000) > > nope golds and plats still make tons of mistakes that diamonds can point out. I'm in a discord with high diamonds and a few masters that could teach you things you don't even know. "Teach you things you don't [...] know"... wow... you and your logic buddy...
rujitra (NA)
: > [{quoted}](name=GreenTeaAndPoro,realm=NA,application-id=yrc23zHg,discussion-id=A6tgpLWd,comment-id=00000001,timestamp=2019-05-16T01:32:16.990+0000) > > whats your ingame name. I'm trying to op.gg you "I don't believe that you could be right, so I'm going to stalk you". Not a healthy attitude dude.
> [{quoted}](name=rujitra,realm=NA,application-id=yrc23zHg,discussion-id=A6tgpLWd,comment-id=000000010000,timestamp=2019-05-16T01:34:29.442+0000) > > "I don't believe that you could be right, so I'm going to stalk you". Not a healthy attitude dude. Don't be mad at him... imagine you had HIS life... you'd do the same... or... I dunno, something much worse. As long as he is here talking, that's better than the alternative :)
: > [{quoted}](name=rujitra,realm=NA,application-id=yrc23zHg,discussion-id=A6tgpLWd,comment-id=000000010000,timestamp=2019-05-16T01:34:29.442+0000) > > "I don't believe that you could be right, so I'm going to stalk you". Not a healthy attitude dude. it's not stalking. Op,gg is a public website with public stats. The fact that you are afraid to show your stats is pretty suspicious.
> [{quoted}](name=GreenTeaAndPoro,realm=NA,application-id=yrc23zHg,discussion-id=A6tgpLWd,comment-id=0000000100000000,timestamp=2019-05-16T01:41:32.361+0000) > > it's not stalking. Op,gg is a public website with public stats. The fact that you are afraid to show your stats is pretty suspicious. Trying to annoy others? Not doing a good job buddy... more like, "beginner troll" or something? Take lessons, learn, grow up, then come back <3
: better question why do losses upset you so much? its a game.
> [{quoted}](name=GreenTeaAndPoro,realm=NA,application-id=yrc23zHg,discussion-id=A6tgpLWd,comment-id=0002,timestamp=2019-05-16T01:31:19.582+0000) > > better question why do losses upset you so much? its a game. Logical flaw in the question...
rujitra (NA)
: Except there's people like me who have an 83% winrate with Yuumi with over 25 games played over multiple live accounts and my PBE account. Why are you punishing me and making your chance of win smaller by not letting me destroy the enemy team with a champion I'm good at?
> [{quoted}](name=rujitra,realm=NA,application-id=yrc23zHg,discussion-id=A6tgpLWd,comment-id=0000,timestamp=2019-05-16T01:27:59.297+0000) > > Except there's people like me who have an 83% winrate with Yuumi with over 25 games played over multiple live accounts and my PBE account. > > Why are you punishing me and making your chance of win smaller by not letting me destroy the enemy team with a champion I'm good at? You fully understand what he's saying so don't pretend you don't. And he's certainly correct (99% of the time).
WrÆth (NA)
: I'm sorry however
You have to be level 30 to play ranked (which takes forever), but you can play any stupid champ you've never tried before in ranked anytime... makes no sense. Require mastery X (in my mind: 5... but fine, require 4 if you must) and one of the biggest issues with LOL is solved (in ranked)... BUT that means less money, so... not gonna happen. "We want an eSport"... roflmao... 25% of sales go into the pricepool... roflmao....
: I think Yuumi's win rate is heavily suffering from people genuinely not knowing how to use her.
That's like saying "I think I get wet after it rains because rain is wet"... obviously, the experience is missing. That's why I DO flame my teammate if they pick her in ranked. That's just being a dick to all others who are actually trying and spend long hours learning their champs. Play her in normals until you have mastery 5, then go into ranked.
Rioter Comments
zPOOPz (NA)
: or even more likely someone is new to the internet and don't know that bitching about downvotes will get you even more downvotes... You got 4 downvotes out of 344 page views. Geez, cry moar. Also this is a RANT and belongs here: https://boards.na.leagueoflegends.com/en/c/rant
...or STILL more likely, I really don't care about "number of downvotes" - I was trying to get people to open their mouths and say something instead of just clicking "downvote"... but hey, if you'd got some balls, you wouldn't hang out here and distribute insults, right? And no, it's not a rant - it's an observation. Also: I upvote you because you seem to be cute and because I just LOVE how you distribute your super-awesome advise around all the posts... you just know it all, don't ya? :)
: Be good enough to BE the booster.
YES! Please... how? Can you help / provide some insights / strategies etc.? And don't say "farm up", that's not it...
: When i played it was the same in Dia, but not with the "who gets the booster", but "who gets the boosted" :D
Haha well... :) That sounds like a nicer problem, but is still the same (and related...)
Gojirha (EUW)
: New player (1 month into this game) permanently banned discussion
You are doing it again - you are flaming this board. And by "again", I mean: every single one of your ranked games is 10+ deaths... that has nothing to do with "I'm a new player", that's intentional feeding. Period. And I am certain you know this and just enjoy this discussion here now...
: Internet-culture of 2019... some dude reads a post, downvotes without leaving a comment... little afraid I guess...
OR more likely: this whole game by now consists of 50% new players who dont get why they can never climb, and 50% booster / smurf accounts from streamers etc. - who also make up this board... I get it
: It's literally "who gets the booster on their team" in Gold IV...
Internet-culture of 2019... some dude reads a post, downvotes without leaving a comment... little afraid I guess...
Rioter Comments
: MMR was messed up at start of season, people were getting the wrong lp gains and losses, allowing a lot of people to rank up faster than they should have. That's why they reset the top of the ladder. Consider yourself lucky, because you were gaining 22 lp in games you should've only gotten 11 in until it was fixed.
P.S. Game right after... win... +22lp... so, I think we can agree on "bug" :)
: servers are dying rn thats why.
: MMR was messed up at start of season, people were getting the wrong lp gains and losses, allowing a lot of people to rank up faster than they should have. That's why they reset the top of the ladder. Consider yourself lucky, because you were gaining 22 lp in games you should've only gotten 11 in until it was fixed.
That's an explanation, but like not correct... we are talking Silver III here buddy... around 40 games into the season, up from Bronze III where I was placed...
Rioter Comments
: I won 2 of my placements because my team always had a afker or a player who took smite top and was 0/4 at 8m It was a total shit fest
Yeah, and now the same thing will happen and every time, you will LOSE LP :) Sorry to hear nevertheless...
: So this is how low elo ppl that belongs in low elo see dodges? I will dodge anyone who is showing any signals of toxicity/trolling, really bad teamcomps and players I played with before and they were terrible... Penalty is OK to hold me from dodging every possible bad things(e.g. L9 in name, lol), while I'll will rather wait 5 mins and lose 3 lp, than playing with Janna jng.
And because of that, I want the penalty to be higher - you should not be allowed to dodge what YOU decide is "a signal of toxicity / trolling"... how can you even tell? And from the initial sentence "this is how low elo ppl that belongs in low elo" I can tell two things: (1) English isn't a language you have mastered, (2) your opinion on this matter is childish, actually "toxic", and irrelevant :)
Show more

Bademeister

Level 42 (NA)
Lifetime Upvotes
Create a Discussion