Greedy Goblin

Thursday, December 8, 2016

Riot and the total lack of AFK punishment

Even the most casual games have AFK punishment. If you AFK in a WoW dungeon, you can be kicked. If you are AFK in a battleground, you automatically get kicked and get deserter debuff. This is to protect the other players from obvious leeching. While morons and slackers will leech anyway, the game strongly differentiate between those who are just horrible and those who can't even be bothered to press a key.

Not Riot. If you go AFK in a ranked game, you get exactly what your teammates do: usually a defeat, in rare cases a win. I just had a defeat because of Lucian decided at level 3 that he just stays home until everyone else reached lvl 10 - and of course we lost bot towers and they got fed. No matchmaking rig analysis can save me from such defeats. But maybe there is more here than bad luck. If even Blizzard punishes AFK, there must be a reason why Riot doesn't. There isn't a rational - even a cashgrab - reason to cater to AFK-ers. Those who don't care enough about the game to press keys will likely won't care to buy skins, while their teammates will be very frustrated. Not like it would be hard to get rid of AFK-ers from ranked games: if a player is AFK for 2 minutes, those who were not in pre-made with him can vote for remake, the other team gets a win and its rewards, the innocent teammates get nothing (remake), the pre-made members of the AFK-er get a defeat and the AFK-er gets all the defeat punishment for his innocent teammates (so if he queued solo, his AFK is equal to 5 defeats in a row). This would push the AFK-ers to bronze in no time.

But for some reason Riot isn't even giving a slap on the wrist for AFK-ers. This reason has to do something with my totally unscientific experience that I have much more AFK-er teammates than enemies. It's unscientific because I see micro-AFK teammates but only completely AFK enemies. However it's a fact that the enemy initiated /remake 5 times since I play the game and I initiated over 20. Remake is a no loss game end when someone is AFK from the very start. So I claim with certainity that I had 4x more AFK-at-start teammates than opponents, while the proper number should be 0.8x. This is quite a difference.

I believe notorious AFK-ers are used by the matchmaker to rig games. They are the ultimate weapon, as they can't be dodged. I was very sure that we win that game we lost because of AFK Lucian. I had another game where Fizz was AFK at start, but his buddy talked people out of remake because he'll be back. He was, by the time Talon was lvl 3 with 20 minions and subsequently fed hard. We won, but only because the enemy had a great late game plan: "Nunu has Frozen Heart, Spirit Visage, Warmog's Armor and Guardian Angel, FOCUS HIM!"

While the evidence isn't smoking gun, we must ask why a "cutthroat E-sport" is more tolerant to AFK-ers than "everyone is special snowflake deserving a legendary" WoW.

PS: no need to tell me that I sucked in that game. I'm fully aware. It was my very first time as support and only took this game because it was a sure win. I expected myself to suck, give a few kills in laning phase and the Jungler will carry us. And it almost happened 4v5 too. It would have happened if I expected it to be 4v5 and open with a Dorian's shield and a top lane talent/mastery. But there I was first time support, standing 1v2 against the ADC and an offensive, ranged support in support talent with a melee champion with nothing but a support item.


Skeddar said...

"if a player is AFK for 2 minutes, those who were not in pre-made with him can vote for remake, the other team gets a win and its rewards ..."

Actually there was a system like that back in Dota (the origin of LoL). The main problem was that as soon as a player goes 0/3/0 his teammates will urge him to go afk as he 'ruined the game and should take responsibility'. Obviously that led to a chain-reaction of blaming 'I am only 0/3/0 cause we have an afk jungler' and in the end everyone was more busy typing than playing. Which is quite a negative experience Riot wants to avoid.
I think that's the ORIGINAL thought behind not allowing people to let others shoulder their loss. In the last years Riot added the Remake functionality and, according to some statements they published, they punish afklers harder should they go afk in frequent games.

"While the evidence isn't smoking gun, we must ask why a "cutthroat E-sport" is more tolerant to AFK-ers than "everyone is special snowflake deserving a legendary" WoW."
As far as I know being afk in battlegrounds would still give you all the rewards. Even if your team lost you got a basic reward, so without any (instant) punishment there would be 80% afk players. In LoL you don't get any rewards when flagged as afk, even if your team wins you don't get a win (as far as I know).

"However it's a fact that the enemy initiated /remake 5 times since I play the game and I initiated over 20."
You do remember your teammates didn't want to remake either at first, only when you told them there is no punishment for them for 'giving up' they voted remake? Especially when you are the one who is initiating all these remakes that might be the reason.

I am kinda amused how many different reasons you find for rigging, brings the 'blaming teammates for your loss' to a whole new level. But maybe one needs that kind of mindset to find rigging if there is any.

Gevlon said...

This is a quite weird fear. No one would accept 5x loss punishment and no sane man would suggest anyone to do so.

Also, Riot could just remove chat and that would greatly increase game quality at the first place.

maxim said...

I see a much simpler explanation of you as an individual being 5 times more likely to initate a /remake than any non-goblin
Think of it as "goblin ratio". Your team has a 100% chance of having a goblin in it (you), while all the other teams have 20% chance

The matchmaker is not out to get you. The only reason it feels it is out to get you is because you are not great at the game yet

Gevlon said...

@maxim: If I'd lose because of my lack of skill, I should be feeding. I should have 0/6/0 by 10 mins and it's always someone else (OK, when Lucian was AFK, it was me but as Jungler or Top I never feed and can count on one hand giving first blood)

Rohan said...

I remember proposing the inverse of this for WoW: Identify the players capable of "carrying" a dungeon group, and use them to fill up a group which has wiped and had people quit on the group.

It's an interesting idea for PvP. Would a team of 5 players with the same skill level be a better experience than [1 high-skill player, 3 average players, and 1 low-skill player]? It's sometimes nice to see a better player on your team, because you can learn from them. And sometimes you'll be the high-skill player on your team, and sometimes the low-skill player, giving you a wider experience of games.

99smite said...

Wow, removing chat should improve a team-based game?
Why bother with multiplayer games at all?
There is a game called "Dungeon Master", which has been rewritten to work on modern machines and the author added a feature that Gevlin should love. It is a step counter. Players can submit their record on a forum. The least steps to the final boss and you are the winner.... This is pure competition for Gevlon.
No social interference, no rigging, no toxic environment....

Anonymous said...

> as Jungler or Top I never feed and can count on one hand giving first blood

Yeah, but that may just be because you've adopted a very conservative playstyle.

As others have pointed out in previous comments, your CS is unusually low. You take fewer risks and suffer fewer deaths; that's natural. We're saying that a skilled player *would* be taking greater risks -- because he has enough map awareness and champion knowledge to understand ward spots, lane pressure, baiting, escape routes, dragon timing, etc.

In other words: you're choosing matchups wherein you can expect to win with mediocre play, and then you're delivering mediocre play. If you keep doing this then you may be able to increase your MMR, but your actual skill level won't improve much.

Gevlon said...

@Rohan: in LoL it would mean the low skilled would be told to stay AFK in the base

@99smite: chat in LoL is only used for flaming

@Anon: While my CS is low, my gold is not. Mostly because killing red, blue or dragon counts as 1 CS.

"We're saying that a skilled player *would* be taking greater risks" is stupid, since (assuming normal matchmaker) he'll be facing equally skilled opponent who will punish his risk-taking.

Anonymous said...

I feel like it confirmation bias about AFKs. You feel them more on your team compared to the opponent team, so it seems that every AFK gets put on your team. Blaming your team for losing is the kinda of thing reddit trolls do.

In the meantime I still feel like you are only winning because you stack your team and then get carried by Yasuo and Jinx with 26 kills. Consider this game:


You did less damage than Janna, you almost participate in fights and barely hit any turrets. Instead you farm the jungle all game, but without using the farm to attack the enemy its going to waste. Being lowest performer on your team is a consistent trend on your, while you have a teammate or two on your team who goes crazy and 1v5 the enemy and win for you.

Because of your low performance I have trouble believing that its the main cause of your loses.

Anonymous said...

Why not 5 losses?

A dc has to equal to an afk (feels bad) but more importantly if you dont adjust his mmr its just a temporary icon change, if you do you screw over the mm.

Better if you afk you get removed from ranked for x days (same with dodging), x exp increasing.

Anonymous said...

> "We're saying that a skilled player *would* be taking greater risks" is stupid, since (assuming normal matchmaker) he'll be facing equally skilled opponent who will punish his risk-taking.

Chasing an enemy into the fog of war is a risk. Tower diving is a risk. There are lots of risks.

But you can mitigate these risks by knowing where the other enemy champions are located. By playing those champions and getting a feel for their preferred engagement ranges, combo timings, and ambush spots. By warding and counter-warding. By remembering the cooldowns on enemy summoner abilities. By quickly estimating whether the other guy has enough mana for a root or gap-closer.

A skilled player can take risks *without* necessarily being punished, because he has already "done the math" and he *knows* that the enemy has insufficient resources to reliably punish him. He knows that he can cut down that weakened enemy champion and take exactly two tower hits before escaping. He knows when he can walk away and let an Ignited champion burn to death, instead of over-committing to the kill.

You may be neglecting to learn this sort of thing, because your attention is focused on the performance of your teammates (and your study of the metagame).

Gevlon said...

@Anon: what damage you expect me to do with -6% damage taken -4% damage from allies keystone and every item defensive? Or Janna should have tanked? But this is besides the point. The point is that I CAN stack my team, because matchmaking is rigged. Why should I bother to learn to play when I can uphold 2:1 win:loss for over 100 games?

@Anon: but his equally skilled enemy will also watch mana and wards and counterwards and whatnot. The only way out is accepting that:
- one or two members of your team is stronger than the others and carry the game
- in 60-80% of the cases it will not be you
- in these cases your job is to not feed the enemy and support the carrier in teamfights

dobablo said...

My own confirmation bias has led me to conclude that, because I take occasionally 30s to 1min AFKs and my phone has a habit of disconnecting during games, Hearthstone seems to match-make me up against other ropers and AFKers (probably mobile gamers).

Anonymous said...

Ill looked replay of that match.
First 8 minutes, you had no influence on the game at all. Just clearing jungle camps. Then overcommitment to blue buff forced supp to help you and you both died. On 9 minute mark, there where 10 kills and you are involved on 2 of them, by dieng. No assists. You got first kill 11:45, chasing opponents jungle and thanks to help to opponents top lane, barely. You died right after, so you got kill, but you gived kill and assist. Opponents still win that trade.
Then on 13:20, you stumble on overpushing opponents mid. Champ was cornered on 3 sides, without a spell cs and low health, but you use ult asap. Even if mid used flash, she was trapped anyway. Basicly ult wasted. Then you go dragon, Where botlane outplays enemy jungler and you had no influence at all, exept warwick w, what gived bit attack speed.

on 15 minute teamfight you choose to run to mid what pokes under turret, while same time your botlane where low and get chased by enemy graves with ult up. Exeptional support play saves your team from disaster. again, you ult support who literally has no way to escape.

then you force rift herald, and not let 3 players to cash in from succesful teamfight. Missed teamfight on mid and force again rift herald. This is example why you should not ult with warwick asap, 2 times opponents flash out and escape, just ulting after flash could net either of the kills. And then again, when you are sieging turret, where your attack speed would help alot, you recall.

That shows in the match statistics too. From 34 kills, you are involved in 7. Thats very very low participation value. Maybe thats the reason why you are matched with afk players, because you influence is comparable with afk players.

Gevlon said...

@Anon: you gave much more effort to my personal performance than I ever did.
You are wasting time on something irrelevant. I have no reason to improve. Maybe I could increase my 2:1 win:loss to 3:1 or 4:1.

But it doesn't really matter because the game is usually won before we leave the start platform, since the matchmaker is rigged and further understanding how it works helps me to more wins than being better in pushing buttons.

This is why I only play two easy champions and don't really bother to improve. (Maybe I should, just to increase my own enjoyment). The very point is that the game is SO rigged that even the pathetic performance you describe is enough to win 2:1

Anonymous said...

gevlon, just to make things clear. Even pro players make exactly the same mistakes again and again, even multiple times on same game. Its not bad to do mistakes. Pro players understand what they or teammates done wrong and how to avoid it next time. Happens.

I see that you make a choice to play so, low influence game. Its ok and i respect that. But i think its bit relevant. Your performance is measurable. They know what champs you play, they know what role you play, they know how agressive you play, kills+assist/deaths ratio, jungle clearing speed, lasthit minion gold gaining speed. If you put it all together, you get something like a gaussian curve, there a few very good players and there are very few bad players. if you happen to be on the extremes, there will be shortage of same skill players, matchmaker need to compensate to have reasonable que times. Both challenger and bronze can have same gold league player in their team, because there where no other around to match on that time.

I might be wrong, but i see that you are playing comparable to afk players on many ways on statistics. And there are very few who play on that style. Matchmaking tries to find same skillevel players and closest they can find are either feeders or afkers. Maybe matchmaking is doing that job too well. Still, proof is needed. Preseason makes it hard, every patch resets everyone ranks and mmr. That makes teams more random then usual.

I want to askyou, on last 20 games played, how many you win, how many you lose, how many times you dodged? How many times you got afk or intentional feeders on either team and dont count bad players who play just badly?

ps. one thing what i noticed, if you clear jungle, try to autoattack and move slowly to the side where you are planning to go next, literally you will win seconds on this kind of clearspeed.

Gevlon said...

@Anon: what you are saying is "rigged". That matchmaker looks at not winrate but some arcane "performance" stats that have some correlation with winning.

Remember what Sirlin said about "skill". It doesn't matter how many minions I kill, how many assists I have. Only one thing matters: "Victory" or "Defeat".

Intentional feeder is hard do separate from plain bad player as they do the same: feed.

Anonymous said...

@Gevlon: total miss today.

First of all, AFK punishment is probably the biggest punishment in LoL, it's even worse than a temporary ban! You get the stupid timer which can't be avoided after just 2 games, and after 4 you're banned. Rigging with non-punished AFKers is a stretch, but possible. Legit player afking under current stick&stick&longerstick policy is impossible.

I used a version of your "queue dodge" strategy in season 3. Except I wasn't checking and qdodging, but simply judging in-game performance of my team, and if they are degenerate feeders, I'd simply quit and hop to next account. That was golden time in LoL, you coulnd't get permaban for quitting, only repeating 1 week ban, I had a few accounts which I hopped between and I didn't have to deal with my team's stupidity any longer than I wanted to. We certainly need more afkers in LoL to change the current stupid policy of sticking and more sticking, because the alternative to afk is intentional feeding (playing is out of question at the point where decision to afk is made, obviously), and it's definitely worse than afk would ever be.

And finally you can totally see afk games by their stats. Certainly he won't be having good ones with afk, and it's totally discernible from feeder because afker doesn't die nearly as much.

Anonymous said...

There is big correlation of amount of kills and assists and victory chance.

If we compare you to a afk player, you will have better score, you get gold, exp, minion, even assist and kills. Not many, but more then 0. So you are better then afk player from those stats.

If we compare to feeder, you will have very some comparable statistics, amount of gold, exp, minion, they are low but not 0. Same as your playstyle. Some of them are comparable but not all.

I make one example more, players who learn new champ tend to be clueless and have weaker statistics. Again, thats should be reason why matchmaking puts it in your or opponents team instead of better player. You are comparable to them from numeric perspective.

Extrem cases tend to bend any system. In your case, i think dodging is one of them. Because players enter very randomly into que and have different levels of skill, if you dodge enough, you will get favorable match. Dependednt on criteria, it might take a while, better chances on early morning, where there few people to pick into teams.

Second thing what makes me suprise. You have over 100 games, and still in silver? how? 2:1 ratio should get you up really fast, ~15 games per league level. thats 10:5, 5 loses count as -75 LP, 10 wins 250 LP, its 150 LP to waste on dodging to stay on same league level. thats alot.

I still like to hear, on last 20 games played, how many you win, how many you lose, how many times you dodged? How many times you got afk or intentional feeders on either team and dont count bad players who play just badly? feeders should have 0 kills, and 5+ deaths in short time. assist wont count.

Baelnor said...

Keep going Gevlon! This is good stuff. There is no way that the matchmaker can be completely random, therefore, it has to be manipulated. The extent and design of the manipulation is what I hope you can demonstrate, and that you find a method of manipulating the manipulation that actually makes that game fun.

I personally quit years ago(black Alistair!) because the game was so frustrating to play.

Unknown said...

I agree with you I get the same thing. There is no way I should be 5\0\4 and losing and I am not good enough. When not had 2/6/0 adv and 0/8\2 SUPPORT an nung who refused to gank and a top that mains mid and Is 0/5/0 too.And I am supposed to get good. You are right riot has to be inside trolling. No on ever talks about all the multiple accounts these smurfs have stuck in bronze and that the only way to really climb is to have multiple accounts.

Kornilus Patrick said...

I feel like you guys are riot
Employees defending a game that is already proven to be stacked against you. Cursed has thing that lets you see who you are playing against. I remember doing my b2 placement 4 times all of which contained s3 and below. While my team consisted of newly ranked players, and some on my lvl but none higher. It has also happened like that in regular ranked matches. Why should I have to face s on b how does that help me Improve this is not a game that you improve in by playing better people, and riot knows that. Again I default to the people who get stuck on b4 on smurfs but are Golds and plats. No one says anything about this. It's like the missing link hidden awa.

Kornilus Patrick said...

In my last 20 I have had an afk all but 5. 10 of those games exactly bot lane fed because they wanted another lane so 1 of them was not committed. I usually stop trying if there are 20 deaths before 15 minutes. If you are higher rank then you have escaped for the most part the pit. In low Elo once your bottom lane usually has fed and lost tower usually in under 12 minutes. Enemy team adc 20 farm at 12 minutes gets a few more triple kills and your 60 farm at 12 minutes matters for nothing as you constantly have 4 people in your lane as bot troll jungle for buffs or worse lets there opposing lancers roam as they painfully try to get retaliation tower.