Boards›Gameplay

Uninstalled after 4 years of LOL

samcamo··5 votes·2,648 views

Its been a while since I started playing league of legends. I remember the old days, where item didt get the unique passive. I remember time playing amumu with tears on and 6 sunfire cape. Insta death. Ha, old times. Dodge was also a pretty nice feature. Anyway, just uninstalled league because of cumulated frustration on ranked games. Yesterday I was promote to Silver 3 (which is were I started last year, and because of matchup with teamates who were having trouble playing steadily lost 8 outta 10 pre ranked match and fall to bronze 3) , and I was demoted today because I lost a game at 0 LP. I got very frustrated and I red the demotion system. It is based on MMR, if Riot think you dont belong in this league because of to low MMR, you are prompted to be demoted. The problem is that the MMR imho is a failed system. In my average games, I have a positive KD ratio, I focus on objectives ward and call the shot for the team (for who ever wants to hear me). I must throw 1 game outta 20, which I think it is pretty good for silver 4-3. The problem is that I keep losing the other 19 games because of:

1- Teamates who don't communicate. They dont speak in champ select, they oinly speak to blame other people. You call the shot they don't listen, they don't have map awarness, and they think they know everything.

2- People who always FF no matter what happen. Ho its 4/7 its gg we lose, jng fault blablabla. I never vote yes on a surrender vote. At silver elo, anybody will make a mistake that can cost the game, and as the enemy team we can take advantage out of it. How many times I have won my lane top and every other lane was failing and they surrender, while I think I was able to carry. In ranked, surrender vote should take all the team to vote yes, or people who said no have no LP penalty or the half.

3- Trolls. There is always some trolls. In champ select, I have to dodge last week in my promos because on guy called adc and took it. After, the troll duo took trist and lucian. I dodge not to lose my time because so far we had 3 adc and 1 support and we couldnt win the meta. I was out of my promos. 4- AFK and problem with connection at start of the game. How can you not get frustrated when somebody is AFK in ranked, saying like I was on the phone, or gotta go and stuff... man its ranked, allow a time to play it properly or play normal. Worst it's player who do it on purpose. So finally it's just another game that will lower my mmr -.-

5- People who really don't know how to play. Bad positionning, bad timing, bad call, wrong item, don't ward. There is a lot of ressources on the internet to know how to play LOL. Plz read it or at least listen when your teamates are trying to help you in your play. I always listen to teamates who are trying to help me in pointing out my mistake. Sometimes it hurt the ego, but if it's true then better to improve it right now than latter.

6- People blaming the jungler and they never asked for help or pinged. I know this is awesome when the jungler come gank at perfect timing, but you can't blame it if you feed. You are the only responsable for your 0/6 vlad.

7- Being matched with lower rank player. It is okay to be matched with lower ranked player, but not with one with 3 or more lower rank than you. I kept losing because of those lower ranked player, while facin a team with an higher MMR. Some people are good in lower rank, they got stuck like me because of bad matching. But bronze elo in your team while having a gld 5 in the other it is almost gg at start.

I am not saying that I am a perfect player. I did have a lot to understand and learn. But the MMR system is a failed one. It should be based on multiple factors like wins, KD ration, Average CS/game while not support, How many gold you did make/team avg gold, how many objectives you took or helped to took down, did you vote yes or no at surrender, did you wwin your lane. It is all idea that I have about remaking the MMR system. I know some of them aren't realisable, but it's just suggestions. To conclude, I have to say that I am sad to see where leagues is going. In first season, people were awesome. When someone bought a skin, we were all like damm man nice skin gl hf and all. Now there is still good player and nice people, but there is a lot of negative people, trolls and kids (not like young people but like kids in attitude) who ruin the game. On that good luck to you all, I think I'm not the only one who feel this way, feel free to continu this thread.

It's been a long run over 3000 normal games (with loses). Thank you for all the fun LoL, Adieu

84 Comments

Daedalus8714 votes

Time and time again, I see people complain about 1, 3, 4, 5, and 7. And time and time again, I see a logical response that says "Assuming you're not the troll/AFKer/whatever problem your having, then the enemy team will have more of them in the long run because the have 5 spots for undesirables while you only have 4. So that's not keeping you from climbing."

And you know what? I still haven't seen a good counterargument.

AMYS GRAVE5 votes

statistics.

If there is a 4/9 chance on your team and a 5/9 chance on the other team there is a significant fraction of the lol community that has more trolls/afks on their team than the opposing team. It has to balance out for the community but it does not have to balance out for an individual. With 27 million players daily some strange things start happening.

If you have played enough to have 100 games with trolls/afks then there is a ~11% chance to have more of them on your team than the enemy team. So that is nearly 3 million people who legitimately are being held back by this.

There is one poor bastard who, if you count his games that had a troll/leaver somewhere, has had that troll/leaver on his team for the last 17 of those games in a row....

The average player is average, but if the odds of being imbalanced are 1 in a million then you would expect there to be a few dozen people for whom that happens.

Kazaashi1 votes

{quoted}

statistics.

If there is a 4/9 chance on your team and a 5/9 chance on the other team there is a significant fraction of the lol community that has more trolls/afks on their team than the opposing team. It has to balance out for the community but it does not have to balance out for an individual. With 27 million players daily some strange things start happening.

If you have played enough to have 100 games with trolls/afks then there is a ~11% chance to have more of them on your team than the enemy team. So that is nearly 3 million people who legitimately are being held back by this.

There is one poor bastard who, if you count his games that had a troll/leaver somewhere, has had that troll/leaver on his team for the last 17 of those games in a row....

The average player is average, but if the odds of being imbalanced are 1 in a million then you would expect there to be a few dozen people for whom that happens.

Its acutally the other way around. Since you have to not be the rager/DC/Troll/AFK/feeder for this statistic to work, there are less slots on your team than the other team for one to appear; therefore, over a long enough series of games, you will win 11% more than you will lose due to AFKs.

AMYS GRAVE5 votes

You've misunderstood. I'll do the math explicitly.

Lets say you have 100 games with trolls in them. The odds that you have 51 on your team and 49 on the other are,

100 choose 49 * 4/9 ^ 51 * 5/9 ^ 49 = 3.3%

then you can add 52/48,

100 choose 48 *4/9 ^ 52 * 5/9 ^ 48 = 2.5%

and so on and so on. It comes out at ~11 % of people have more afks/trolls on their team than the enemy team even if you are never the afk/troll. There are more people climbing compared to falling due to this, but the existence of one group does not negate the other. It has to balance out for the community, not for an individual.

EDIT: for formatting, math is hard to show with this formatting syntax... :)

Kazaashi2 votes

{quoted}

You've misunderstood. I'll do the math explicitly.

Lets say you have 100 games with trolls in them. The odds that you have 51 on your team and 49 on the other are,

100 choose 49 * 4/9 ^ 51 * 5/9 ^ 49 = 3.3%

then you can add 52/48,

100 choose 48 *4/9 ^ 52 * 5/9 ^ 48 = 2.5%

and so on and so on. It comes out at ~11 % of people have more afks/trolls on their team than the enemy team even if you are never the afk/troll. There are more people climbing compared to falling due to this, but the existence of one group does not negate the other. It has to balance out for the community, not for an individual.

EDIT: for formatting, math is hard to show with this formatting syntax... :)

Oh I see now, you are manipulating the numbers to prove your point; unfortuatly, that is not correct.

The real odds are as follows: As long as you are not the troll, there is a 45% chance to get to get one on your team and a 55% chance the other team will get one. This means that over a long enough series of games (200 or so), you will have more AKF's/Feeders/Trolls/DC's/Ragers on the enemy team than on yours.

AMYS GRAVE5 votes

Its called a binomial distribution. The mean is in the favor of the non-afk/troll, but there is a variance on that.

http://en.wikipedia.org/wiki/Binomial_distribution

in an n=100 game series with p=4/9

mean = np = 44.44 games (this is under 50 and is as far as your arguments have gotten, please keep reading...) variance = np(1-p) = 24.69 games

so in a 100 game series you can expect with 68% confidence that there will be between 19.75 and 69.13 trolls on your team. In no particular series are you expected to get exactly 44-45 trolls, you can vary from that by quite a bit.

If you double the number of games you will double the average (np=88.88), but you also double the variance (np(1-p)=49.38). In order to ensure that your claim must be true for every league player you have to ensure that they have played at least ~2400 games with a troll or afk in them.

I'm manipulating numbers consistently with the field of mathematics. I've shown my work and it is consistent with my claims. If you wish to disprove me you will have to show your work rather than just saying that i'm wrong. That's how math and science work. It would behove you to study a bit more before claiming that i am incorrect.

doublelift fg6 votes

people who didn't pass 2nd year of highschool can't understand that

UP2006 votes

wow, the ignorance is strong here.

usul12021 votes

That's...... not how probabilities are used properly. nor how you add them over multiple iterations. Which is massively swinging your numbers.

AMYS GRAVE4 votes

That's exactly how that's done. You integrate the area under the probability distribution function(pdf) from one extreme to the point of interest to get the cumulative distribution function(cdf). In the case of discrete pdfs (as is the case here) you sum instead of integrate. That's how numbers work.

usul12022 votes

Really? you have a 4/9 chance of the troll being on your team, and that's 51 on your team, and the 5/9 on the other team is 49? That's good math?

AMYS GRAVE5 votes

100 games total. 51 games with trolls/afks on your team with a 4/9 probability and 49 games with a troll/afk on their team with a 5/9 probability.

nCr p^(r) q^(n-r) returns the probability of exactly r on your team with n-r on the other team. so if you want the probability of 51 on your team with 49 on the other team...

n=100 r=51 p=4/9 q=5/9

the p^r and q^n-r is the probability of event^number of times event occurs. The 100 choose 49 could have been 100 choose 51 but they are equal because nCr = nC(n-r).

so to calculate the odds of having exactly 51 on your team to 49 on the other team you do this once, to find the odds of having at least 51/49 you start r at 51 and repeat this procedure to 100 and add the results.

P(r>50 for 100 trials) = sum from r=51 to 100 of ( (100 Choose r) * (4/9)^(r) * (5/9)^(100-r) )

http://www.wolframalpha.com/input/?i=sum+from+r%3D51+to+100+of+%28+%28100+Choose+r%29+*+%284%2F9%29^%28r%29+*+%285%2F9%29^%28100-r%29+%29+

Yes, that is good math.

Kazaashi1 votes

{quoted}

100 games total. 51 games with trolls/afks on your team with a 4/9 probability and 49 games with a troll/afk on their team with a 5/9 probability.

nCr p^(r) q^(n-r)

n=100 r=51 p=4/9 q=5/9

the p^r and q^n-r is the probability of event^number of times event occurs. The 100 choose 49 could have been 100 choose 51 but they are equal because nCr = nC(n-r).

so to calculate the odds of having exactly 51 on your team to 49 on the other team you do this once, to find the odds of having at least 51/49 you start r at 51 and sum to 100.

Yes, that is good math.

But is it useful to find the probability of having 51 vs 49 undesirable players on your team in a series of 100 games? Short answer is no.

Ultimatly, I think about Ranked this way and it has helped me tremendously:

40% of your games are going to be lost due to DC/AFK/Feeder/Troll/Undesirable Player on your team.

40% of your games are going to be won due to a DC/AFK/Feeder/Troll/Undesirable Player on the enemy team.

Only about 20% of the games you play in Ranked will you be able to have any meaningful impact in. It is those games that you have to identify and make sure your contribution is positive and leads your team to a win.

AMYS GRAVE4 votes

But is it useful to find the probability of having 51 vs 49 undesirable players on your team in a series of 100 games? Short answer is no.

Its actually useful to find the probability of having at least 51 vs 49 on your team in games that already include one "undesirable player". It tells us that there are people who get the short end of the stick with teammates. Otherwise this is sage advice.

Have the serenity to accept the things I can't change, the courage to change the things I can, and the wisdom to know the difference.

But i like to do the math.

If 1% of the community is afk/troll as riot claims then there will be an average rate of 8.64% of afk/trolls in the game assuming you are not one. So 91+% of your games your contribution will have a significant impact on the game. 20% vs 91%. Have the wisdom to know the difference.

Kazaashi1 votes

{quoted}

If 1% of the community is afk/troll as riot claims then there will be an average rate of 8.64% of afk/trolls in the game assuming you are not one. So 91+% of your games your contribution will have a significant impact on the game. 20% vs 91%. Have the wisdom to know the difference.

I feel that is just too optimistic of an estimation. There are a lot more factors than just an AFK/Troll. You can have feeders, unskilled players, people who elo boosted.

I use the 40/40/20 statistic as a baseline. I totally believe that you can do things to bump all of those numbers into your favor, with the proper training, skill, and luck. Hopefully people will be able to look at that and see if they can increase thier chances of winning a 4v5 by 5%, for example, that they acctually have helped themselves climb by a lot.

AMYS GRAVE5 votes

Very true, i use the 1% because it's what Riot has officially put out as toxicity rates, but I've always felt it's a little low. Unfortunately my feelings on the matter are subjective so any other number would just be made up....

usul12021 votes

Its bad math if you're pulling that r value out of your ass. it makes the entire thing meaningless. I can guarantee you that my number is maybe 10/100, at most. Putting random numbers into probability shows nothing, and is therefore meaningless math.

AMYS GRAVE4 votes

r= 51 is the threshold for 50%+1. The sum from r = 51 to 100 gives the total probability for people to have more afks on their team compared to the enemy. r=51 is anything but arbitrary and its interpretation when summed r=51 to 100 is that the resulting 11% of players have more afks on their team than the enemy team. There are 27 million players daily so that works out to just under 3 million players.

If you chose 10/100 then that is arbitrary.

usul12021 votes

You're assuming that every single one of the games has a troll or feeder though. This is mathematical masturbation, its nice in general (and correctly done), but it means nothing because its picking at values. It's math for the sake of math.

Even the biggest complainers on the boards don't get an idiot every game. 11% is a incredibly heavy weighted number due to this fact, and makes the statistic useless, because its not true. You're running on a false base assumption.

AMYS GRAVE4 votes

I stated from the beginning that we were only considering the subset of games that included an afk/troll, so that is not an assumption that was made. You don't have to have one every game for this to hold true.

usul12021 votes

No, but it does massively shift the number down from 11% of people having a >50% number of games where there is a troll/AFK on their side.

AMYS GRAVE1 votes

But that doesn't change the ratio of trolls on your team vs. the enemy team so that's irrelevant. I never claimed that >50% of all of their games had a troll/afk on their side, only that >50% of games that included a troll/afk would have said troll/afk on their side.

usul12021 votes

We're disagreeing on what numbers are significant here. Your math is correct, yes. I simply disagree that an analysis of troll only games has any importance, because it weights the numbers by ignoring balanced games. Its a worst case scenario viewpoint that I don't find applicable when calculating percentages of people as large as 11% of 27 million people.

Daedalus8712 votes

The math itself is done correctly.

The assumptions behind it make it meaningless.

AMYS GRAVE4 votes

The only assumptions being made here are that you aren't the troll/afk and that there is an equal chance for the troll/afk to be one of the other 9 players. Which do you disagree with?

Daedalus8711 votes

No you are only looking at games where there is one troll. You're completely ignoring games where there are 0 trolls, 2 trolls or any other number of trolls. You're only telling part of the story, which isn't good enough.

AMYS GRAVE1 votes

If there are 0 then that doesn't change the ratio of trolls on your team vs. the enemy team so that's irrelevant. Given that the rate of negative behaviour is ~1% then the odds of having two said people in a game is negligible in comparison.

Daedalus8711 votes

Using your logic and numbers, I'm pretty sure I can deem the the odds of having 1 troll in the game neglibible because #of games with no trolls in them/ # of games with 1 troll in them ~ # of games with 1 troll in them/# of games with 2 trolls in them.

Furthermore, you can't ignore games with 0 in them because then you fail to convince me that a meaningful amount of games are ruined by trolls. If 10% of games have trolls in them, who cares? The other 90% of games are still good.

UP2004 votes

wow, learn math dude.

Daedalus8711 votes

Don't make me look up how to do a binomial distribution. Yes I know there are some people who "win" the team lottery and get stuck with undesirables 90 of the time, but chances are, OP isn't one of those people.

AMYS GRAVE4 votes

The threshold for one person is around 70% on their team with 30% on the other, but that's neither here nor there. If you like i did the math in response to another post just below.

An 11% chance that the OP is in this group is not negligible odds, so a definitive statement like "that's not keeping you from climbing." isn't really a fair assessment of the situation and the statistics is the counterargument. I'm not saying the OP isn't salty or even that he's right, just that we can't know if he is wrong or just a statistical fluctuation.

Daedalus8711 votes

Why do you only consider the cases when there is one troll in the game? What if both teams get trolls. OP also complains about afkers and otherwise bad players (after this point all, trolls/ragers/afkers/otherwise undesirable players are referred to as trolls).

So with that in mind, here is some math.

Assumptions:

  1. 30% of people are trolls.
  2. OP never trolls.
  3. Matching making gives trolls randomly and independently.
  4. All trolls have the same influence.

OP can have between 0 and 4 trolls on his team, the percent of that happening is: 0 trolls: 24% 1: 41% 2: 26% 3: 8% 4: 1%

Enemy team can have between 0 and 5 trolls. The percent of that happening is: 0 trolls: 17% 1: 36% 2: 31% 3: 13% 4: 3% 5: 0.24%, calling it 1%.

So now let's look at the games where OP's team has more trolls than the enemy. When OP's team has: 0 trolls, the enemy team has 0% chance of getting less trolls. 1 troll, the enemy team has a 17% chance of getting less trolls. 2: 53% 3: 84% 4: 97% chance of getting less trolls than OP's team.

So now we need to weight this to find out what is the total chance of OP's team getting more trolls than the enemy team: 0 trolls: (24%x0%) = 0%. In 0% of OP's total games, he will have 0 trolls while the enemy team has less. 1 troll: (41%x17%) = 7%. In 7% of OP's total games, he will have 1 troll, while the enemy team has less. 2: 14% 3: 7% 4: 1%

Now if we take the sum, we get the percentage of games where the enemy has less trolls than the OP, 29%. Thus OP's team has a 71% chance of having the same or less trolls/ragers/feeders/etc than the enemy team. Now using this, we go to a binomial calculator and we define success (X) as the enemy team having less trolls than OPs, which has a 0.29 probability, set the trials to 100, and we want to what probability that the enemy team gets 50 or more games where they have less trolls than the OP P(X >= 50).

Thus we get that over the course of 100 games, that OP has a 0.000798% chance the enemy team getting 50 or more games with less trolls than OP.

TLDR: I'm not buying that OP is stuck because of other trolls/bad players/afkers.

AMYS GRAVE4 votes

Assumptions:

  1. 30% of people are trolls.

Where is this coming from? Riot has come out stating that approximatly 1% of players exibit this kind of behaviour so the cross terms where there are multiple afk/trolls is much smaller than you state, seriously inflating the odds to prove your point.

Daedalus8711 votes

My mistake.

By troll I meant trolls/ragers/afkers/otherwise undesirable players that OP also complains about (I have since edited for clarity).

30% seems reasonable for that.

Also if I wanted to inflate the odds for proving my point, I would either assume a low number of trolls (~1%) or a high number of trolls (~99%).

AMYS GRAVE4 votes

30% seems reasonable for that.

It really doesn't. Where is that coming from?

Also if I wanted to inflate the odds for proving my point, I would either assume a low number of trolls (~1%) or a high number of trolls (~99%).

Given that riots behavioural research team has given us 1% as a number to work with 30% is a high number of trolls. 1% is actually quite reasonable and the difference will change the 0.000798% figure you claim by many orders of magnitude.

Daedalus8711 votes

Okay then, using the "Official 1%" that you love so dearly and the same methods as above:

of trolls on OPs team %chance

0 96.1% 1 3.8% 2 0.05% 3&4 calling these 0.

And since I'm feeling lazy, I'm going to call the chance that enemy team gets less trolls than OP 3.9%, IE the chance that he get's at least one troll.

Again 100 trials, the chance the enemy team gets less trolls than OP teams in at least 50 games, P(X>=50), is equal to 3*10^-13%.

The chance of the enemy team getting less trolls than OP in at least 15 games? 8 in a million.

So you were right. Changing from 30% to 1% did change the results in orders of magnitude. It just made my argument a whole lot stronger.

AMYS GRAVE5 votes

N on your team - chance of that many happening.

0 - 96% 1 - 3.8% 2 - 0.05% 3 - ~0 4 - ~0

N on enemy team - chance of that many happening.

0 - 95% 1 - 4.8% (note this is not 3.9% as you claim.) 2 - 0.097% 3 - ~0 4 - ~0 5 - ~0

When enemy team has X trolls - your team has X% chance of getting fewer trolls 0 - 0% 1 - 96%. 2 - 99.8% 3 - ~100% 4 - ~100% 5 - ~100%

Weighting the odds that you get fewer trolls than the enemy team (the 0 case and the 3, 4, and 5 cases are 0)

0.96*0.048 + 0.998*0.00097 = 4.7%

Across a 100 game series leaves you with a 10^-40 chance of getting fewer trolls on your team than the enemy.

I hope you are seeing why i took the subset of games that had one troll in them and analysed that. The number of games with 0 trolls dominates and with 2+ trolls is pretty negligible. If you consider all games then a 100 game series does not have a statistically significant number of trolls in it. Having more trolls on your team than the enemy in 100 games is not very probable, but neither is having it the other way around. When you constrict it to just games with at least one troll then this accentuates the effects to see that even though you have way more games without any trolls whatsoever, it is still possible to have more games with trolls on your team than the enemy team. Any series of games is dominated by a sea of games where there were no trolls at all.

This is why you have to take it to 1%, because otherwise the number of games with 0-0 trolls is downplayed to not reflect reality, and the number of games with 2+ is inflated in the same manner.

Daedalus8712 votes

{quoted} If you consider all games then a 100 game series does not have a statistically significant number of trolls in it. Having more trolls on your team than the enemy in 100 games is not very probable, but neither is having it the other way around. This is why you have to take it to 1%, because otherwise the number of games with 0-0 trolls is downplayed to not reflect reality, and the number of games with 2+ is inflated in the same manner.

Very well then.

I will concede that my original argument of the enemy team getting more trolls is not necessarily correct, but the spirit of the argument is still true, ie OP is not getting fucked over by trolls.

AtheosisX1 votes

{quoted}

Time and time again, I see people complain about 1, 3, 4, 5, and 7. And time and time again, I see a logical response that says "Assuming you're not the troll/AFKer/whatever problem your having, then the enemy team will have more of them in the long run because the have 5 spots for undesirables while you only have 4. So that's not keeping you from climbing."

And you know what? I still haven't seen a good counterargument.

That's an oversimplified take statistically.

AMYS GRAVE5 votes

The only assumptions being made here are that you aren't the troll/afk and that there is an equal chance for the troll/afk to be one of the other 9 players. Which do you disagree with?

Konidias1 votes

{quoted}

Time and time again, I see people complain about 1, 3, 4, 5, and 7. And time and time again, I see a logical response that says "Assuming you're not the troll/AFKer/whatever problem your having, then the enemy team will have more of them in the long run because the have 5 spots for undesirables while you only have 4. So that's not keeping you from climbing."

And you know what? I still haven't seen a good counterargument.

Counterargument:

Even if the math is correct in assuming you will get less trolls/afk than the opposing team, this math only matters after a great amount of games... like tens of thousands.

You could play 20 games in a row with afker's or trolls on your side. The point is that the average person isn't playing 5000+ games a season, so the math is useless.

Also people forget the most important factor: Promos and Series games. These have far more impact... meaning you could potentially get unlucky and get those AFK/trolls during every promo or series game and lose those, while winning the non promo/series games. Therefore, you will never climb elo because the odds are against you at the most inopportune times.

You also have to consider that the enemy team has MORE chance to get a smurf than yours. Assuming you aren't a smurf. Thus, it balances out.

Daedalus8711 votes

  1. I mentioned in the long run, it should balance out. While it is hard to say what the long run is, I doubt it's more than 200 games for most people.

  2. Again, you go through promos enough times, you should be able to get through.

  3. The number of undesirables vastly outnumbers the number of actual smurfs to the point of insignificance.

Konidias1 votes

{quoted}

  1. I mentioned in the long run, it should balance out. While it is hard to say what the long run is, I doubt it's more than 200 games for most people.

  2. Again, you go through promos enough times, you should be able to get through.

  3. The number of undesirables vastly outnumbers the number of actual smurfs to the point of insignificance.

  1. We need actual numbers. Why spit out all these numbers and equations when proving your odds and then forget the part where variance comes in to play? Again, the math does not matter if you leave out crucial pieces of information. "I doubt it's more than 200 games" is not mathematical evidence

  2. How many times? Few enough to be reasonable? I don't want to hear "should" I want to hear "will"

  3. Proof? Records showing number of smurfs per game vs number of trolls/afks/unskilled?

All I keep seeing is "well this basic math says you should win more than lose if you are good enough" but I fail to see anyone explaining the variance and estimated number of games you have to play for the math to work.

There is MUCH more to factor into each match-up. You need to not only look at the odds of your team getting undesirable players vs the opposing team, but also your team vs their team getting a smurf, your team vs their team getting someone around your own skill level, etc.

If I get matched up against another team and both sides don't have any undesirable players, but I'm leagues better than the rest of my team... and they have a single person who is leagues better than the rest of their team, then the match boils down to which one of us gets fed harder... and that can be tough considering we don't know which of us is actually most skilled until it's too late.

Maybe that Nidalee jungle on the opposing team is an equal match to me, but my team fed her 5 kills before I realized how good she was... Now I have to get 5 kills to catch up, or their team is at a significant advantage. I was just dominating top lane and maybe killed the top laner once or twice, but Nidalee roamed around and got multiple successful ganks. So I lose this game because of dumb luck, and it was actually a game I could have potentially won had I known who to look out for.

Then maybe next game I get an AFK.. 2 losses in a row. Then next game I get 2 people arguing all match and they do poorly and we lose. 3 losses in a row. The way people talk, you would think the next 4 games I should easily win... but that is never the case.

Maybe I easy win 5-6 games in a row and get to series, and then I get the AFK/trolls 2 games straight and lose my series. Yay, now I have to win 5-6 more games just to potentially have the same thing happen. This does not prove ANYTHING about my skill level at all. It just proves I'm unlucky. It's an unnecessary grind.

Put me with players at my level please. That's all I want. Because to me, most every game feels like the game is trying to handicap me by giving me the worst teammates while the enemy team seems to be at least semi-organized and competent. The few games I win are just total stomps where the enemy team gets steamrolled.

Daedalus8711 votes

I did math above, showing that you have a 71% chance that your team will have more undesirables than the enemy team assuming that 30% of people are undesirable. I also show that over a 100 games, you have a 8 in a million chance of getting 50 or more games of having more undesirables than the enemy.

If you want something else, then do it yourself. I've notice you haven't shown us any math or stats.

Randomonium5 votes

I think people are getting confused with all this math.

I'll try to make it simple. Let's just say that getting a troll on your team or the other team is essentially a coin flip scenario (yes, i realize the odds are slightly in your favor but bear with me.) Heads you get a troll, tails the other team gets a troll. While in theory you should get an equal number of heads and tails it is completely possible to get 10, 20, even 30 heads in a row (though it is a low probability.) However, if you have millions of people flipping coins (i.e. playing games) it becomes a near certainty than some of them will encounter these improbable number of heads.

While it is true that if you flip the coin an infinite number of times you should be able to regain that 50/50 balance, most people are only going to play a couple hundred games max in a season and they may never recover from a string of "bad luck."

This is what is called variance in statistics. Yes, the theoretical odds are in your favor with trolls but that doesn't mean you are guaranteed to get less trolls on your team than the opposing team.

If you don't believe me flip a coin 10 times every day and keep track of the results. I guarantee you will not get 5 heads and 5 tails every day.

Daedalus8711 votes

I did the math above. I assumed that every non OP player had a 30% chance of being trolls or otherwise undesirable. I came up with that in each individual game, that OP has a 71% chance of getting the same or less undesirable players on there team.

Futhermore: We get that over the course of 100 games, that OP has a 0.000798% chance the enemy team getting 50 or more games with less trolls than OP. (Just replace the numbers in order): 50, 0.141%, 25 20, 3.85%, 10

UP2004 votes

Holy shit 30% of people are trolls! What the hell are you smoking?

Randomonium3 votes

Sorry it took me so long to respond, I wanted to make sure I fully understood your thought process before commenting. While I agree with other people that 30% is high your math is sound and would continue to be sound regardless what troll percentage you used.

My issue is that you didn't take your conclusion further. I used your math and made the assumption that 1,000,000 play league. Using your math this results in over 40,000 players averaging six or more games per every ten played where they have more trolls on their team than the other team. While I agree that 40,000 into 1,000,000 is a very small percentage that still is a lot of people. That's the point I was trying to make with the coin flip. No matter how impossible the odds if you have enough people they become probable.

I do a agree with you that most people who complain on these boards though are not in that small minority.

Daedalus8711 votes

I like you. You may disagree with me, but you do it intelligently.

I used 30% of people being trolls/undesirables because OP also complains about poor players and the like. And while I haven't done the math, and I really don't want to, I'm pretty sure that lowering the % of undesirable teammates would only make my argument that OP doesn't get more trolls/undesirables than the enemy team stronger.

Anyways going back to my conclusions, 40,000 out of 1,000,000 will play in at least 6 games out of 10 where they have more trolls than the enemy. But after 50 games played, about 500 will had 26 or more games where their team has more trolls than the enemy. After 100 games only 8 people (and I know that I am really abusing the law of large numbers here) out of a million will have 50 or more games where they had more trolls than the enemy (assuming everything is independent and stuff).

So it seems fair to me to say "after 100 games, the enemy team should have gotten more trolls than yours". So then the question is "How reasonable is it for someone to play 100 games, at which point the effect of trolls should be negligible?" At a rate of 3 games a day, about a month. Seeing as how the ranked season goes from Jan-Oct (9 Months), I'd say that playing 100 games a season is reasonable.

Randomonium2 votes

I agree, it should. However, from personal experience after having over a dozen games of trolls I got so fed up I just stopped doing soloQ.

I started my season with over 20 games lost in a row. While I know I made mistakes and could have done better in nearly every game I had a rager, a troll or an afk.

Since then I've started to turn the corner and claw back up but its such a grueling and frustrating process that I wonder if its even worth it.

This is the reality of a lot of players. They get one really bad string of bad luck or a couple small strings of bad luck back to back (which are all very possible with the statistics) and it completely tanks their rating and you have to spend the rest of the season clawing you way back up to where you started. Statistics provide little comfort at that point.

Your math has given me a little hope though. I'll keep Qing up hoping the stats will eventually even out.

More from Gameplay

3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay764 comments
7,167
votes
3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay156 comments
3,510
votes
UPVOTE FOR URF
Gameplay606 comments
2,440
votes
Yes plz.
Gameplay456 comments
1,721
votes
3 Refunds per season, up vote if you want 3 refunds per season.
Gameplay114 comments
1,716
votes
When is ANYONE at riot going to address the awful new player experience?
Gameplay634 comments
1,383
votes