Welcome to this patch's General Discussion thread for the League of Legends subforum. This thread is for discussion around League of Legends. Free feel to talk about anything LoL related here that does not already have its own thread.
On July 26 2016 04:41 General_Winter wrote: As jungle, do the tower changes mean I need to camp bot lane now?
I've been moving back and forth between mid and top to good results with Ranmus recently.
as a feminist, how does this patch affect me???
Thats right, jungle is a historically oppressed group
But seriously, I've only been playing for a year and a half. I know you guys have way better developed intuitions than I do and can theorycraft the implication of the tower change on map movments much better than I could.
On July 26 2016 04:41 General_Winter wrote: As jungle, do the tower changes mean I need to camp bot lane now?
I've been moving back and forth between mid and top to good results with Ranmus recently.
as a feminist, how does this patch affect me???
Thats right, jungle is a historically oppressed group
But seriously, I've only been playing for a year and a half. I know you guys have way better developed intuitions than I do and can theorycraft the implication of the tower change on map movments much better than I could.
how about gank lanes that get you kills who cares if a tower is easier to kill or not advantage is advantage
In theory the turret change won't kill lane swaps, because the "first blood" gold from the turret could also be used to jumpstart a carry top laner's item build/progression (Fiora, Jax, Hecarim are the first things that jump to mind) and turn them into an unstoppable snowball, split-pushing juggernaut.
... In all seriousness, of course it's going to kill laneswap meta and go back to tank island top, gank fiesta mid and bot, because ADC is every team's late game insurance policy for if they fuck up in early or mid game.
with people so concerned over giving up tower first blood I don't know if this buff will increase his playrate. He's probably strong but he wasn't that bad before and his main weakness is sorta magnified by the tower change. Even in solo q, bot lanes are more incentivized to take tower rather than leave it to continue dominating lane, so if you're playing kass there is a decent chance if your bot loses the adc will show up in mid to take down your tower before you come online.
he should def be strong if you don't get punished for no wave clear though
Kass just seems to be a constant problem with the balance team. I dont know what they can do (aside from things they definitely wont do) that keeps Kass intact without making him a competitive or soloq problem again (while also being a not-terrible choice).
I'm not crazy, here, right? I feel like I am usually the only person in my games that realizes that blue buff gives +15% AP. A few games ago, playing Sejuani, I had to practically beg our Katarina to take the 30 minute blue buff.
On July 27 2016 01:58 GrandInquisitor wrote: I'm not crazy, here, right? I feel like I am usually the only person in my games that realizes that blue buff gives +15% AP. A few games ago, playing Sejuani, I had to practically beg our Katarina to take the 30 minute blue buff.
People don't read patch notes unless they're about things they directly care about.
whatever mate it lasts like 30 seconds by the time you're back in lane it's half gone just let me use it to get a full mana bar while farming so i can make plays
yea basically what teut said. when they made that change they also adjusted the time you have the buff to be absolute dogshit if you dont take increased buff time, junglers take that, laners take the potion/assassin mastery. So it doesnt last very long.
I only give up blues as jungler once I cap CD, which is pretty fast these days
It's like a triple infernal Elder Dragon only for AP casters, though? A triple infernal Elder Dragon gives you +36% AP instead of +24% and lasts thirty seconds longer than blue buff. For an AP champ +15% AP on a 1:30 buff that's way easier to take makes blue nearly as valuable as triple infernal Elder, not to mention the 10% CDR and unlimited mana regen.
On July 27 2016 01:58 GrandInquisitor wrote: I'm not crazy, here, right? I feel like I am usually the only person in my games that realizes that blue buff gives +15% AP. A few games ago, playing Sejuani, I had to practically beg our Katarina to take the 30 minute blue buff.
funny, my last ranked game i was ahri and i was begging my sej to give me a blue buff. nope, he last hit every blue with smite the entire game. too bad he couldnt do the same for 1 infernal dragon and 2 barons that got stolen by the enemy lee
I've felt the nerf with Shen. He doesn't sustain as well. It's a small nerf but one that has made matchups where it's about skill, learn to the other champion now.
Nothing like getting connection issue in your placements on a smurf so you have to now fight through toxic low elo :<
This is giving me a chance to see though and why is it people in low gold and below seem to have no idea what to do past laning. Do you think there is a lack of educational stuff in lol for post lane? In sc2 say you know once you BO is done you will want to take more bases or tech up. In fps I don't think phases are as defined. In lol people just mill around not doing anything or just go to the lane they started in and push it till something happens. Basic idea of sticking as 4/5 and moving forward is totally alien .
It's hard to tell but I think the biggest difference between low gold and mid plat is that plats are better at playing from behind and map (gank) awareness. I didn't notice any big changes past laning phase, we just play aram until dragon or baron becomes an option. It's hard to get your team to siege side lane towers but it's like that on all levels I played on.
I don't think you can fix that with education, Most players play intuitively and don't do any research about the game. You can't force them to learn.
So. New champ now has abilities posted. Any early thoughts?
Seems like they went out of the way to make him not work in the jungle by having courage not work on large monsters. Also seems clearly not mid or adc material. Looks like they want him too but he might not be the worst support. Has a grab/hook, a late game utility ability that doesn't require gold, and the courage mechanic might bring a lot of sustain if you fight till the guy runs away, then wait and all in with your adc to get the mount and its health back.
On July 28 2016 21:23 MuddyJam wrote: Nothing like getting connection issue in your placements on a smurf so you have to now fight through toxic low elo :<
This is giving me a chance to see though and why is it people in low gold and below seem to have no idea what to do past laning. Do you think there is a lack of educational stuff in lol for post lane? In sc2 say you know once you BO is done you will want to take more bases or tech up. In fps I don't think phases are as defined. In lol people just mill around not doing anything or just go to the lane they started in and push it till something happens. Basic idea of sticking as 4/5 and moving forward is totally alien .
Its three fold
1) You spend a lot less time post lane than in lane. So people tend to learn and get better at laning a lot faster. Even when you don't have a "win lane win game" the laning phase will often take up a significant chunk of the game. So people play it a lot more.
2) There isn't as good a feedback mechanism for post-lane play. When you're in lane everything is almost always your fault, whether that is getting ganked, or not following through on one. And figuring out the solution to this is usually pretty simple since the cause and effect are close together and the actions very discrete.
But after lane its a lot harder to determine why you didn't win or why something went poorly. Should you have followed your AD or should your AD have followed you? What objectives do you go to and why? These rarely have clear answers that follow directly from cause to effect. Its hard to see why you should be sieging bottom tower when no objectives are open as opposed mid lane or top lane... or farming. The loss that can follow from that is often disconnected from the action. You lose a team fight later in the game at dragon and the lost team fight is the loss creator and not the failure to act when you had a lead.
3) There are coordination problems inherent in any game where people aren't on consistent teams and/or on voice coms. Teams can have strategies but this requires everyone on the team to be on the same page. 5 Individual strategy ideas means that none of them get executed. It often takes only one player not on the same page in order to wreck a strategy.
Because of this "follow the team" is often a very good strategy and may people use it. But "follow the team" doesn't have a coherent direction unless someone can take control and lead it. This is sometimes for reasonable reasons sometimes for not. If no one is leading then "follow the team" tends to just congregate at map points rather than going and doing.
So basically its practice, feedback, and (lack of) shot callers.
On July 27 2016 01:58 GrandInquisitor wrote: I'm not crazy, here, right? I feel like I am usually the only person in my games that realizes that blue buff gives +15% AP. A few games ago, playing Sejuani, I had to practically beg our Katarina to take the 30 minute blue buff.
i don't think it's worth having if you have any other mana/cd dependent characters on your team. the primary purpose of the buff is still those 2 things which kat needs neither of
Has there ever been a champion with lower win ratio than the new Ryze? ~38% average is pretty extreme considering the second lowest is Aatrox sitting at ~46%
The real miracle is that Ryze not only has the lowest winrate but he also has an absurdly high banrate. On the other end, Sona has an absurd winrate but still has a banrate below 10%.
On July 29 2016 21:02 Jek wrote: Has there ever been a champion with lower win ratio than the new Ryze? ~38% average is pretty extreme considering the second lowest is Aatrox sitting at ~46%
Urgot? Didn't Leblanc get close to that win rate as well for a while? And Yasuo?Also from my searching, Azir right now has a similar or even lower winrate depending on region and league.
Personally when I tried the new Ryze he seemed like a really fun champion (also great VU, his animations are fantastic), but the lack of utility that his ultimate gave him hit very hard. His ultimate is near useless lv 1, and he kind of needed that MS boost and spellvamp (although the new range on his Q means it's near impossible to harass him out of lane).
On July 29 2016 21:02 Jek wrote: Has there ever been a champion with lower win ratio than the new Ryze? ~38% average is pretty extreme considering the second lowest is Aatrox sitting at ~46%
Release Syndra I think? Maybe Kassadin at some point during the endless nerfs?
I still hold that Ryze's Q being so shitty is a big problem for his kit. Pre-Rework-to-the-rework its "utility" was that it could bypass tanks and you could simply drop it on a person's head even if they were behind minions, like an AA that didnt draw aggro, or just CS anything you wanted. Now it really is just mystic shot with no on-hit effects.
With such a shitty Q the rest of his kit really needs to do work, so hes gone through the infinite root stage, the super damage burst, the tanky bursts, etc.
Q needs an identity if they are going to keep it as his bread and butter spell. I mean, his E is an atrocity of a skill as well. Its just like, "oh, right, better press E it does...something"!
On July 29 2016 21:02 Jek wrote: Has there ever been a champion with lower win ratio than the new Ryze? ~38% average is pretty extreme considering the second lowest is Aatrox sitting at ~46%
On July 29 2016 21:02 Jek wrote: Has there ever been a champion with lower win ratio than the new Ryze? ~38% average is pretty extreme considering the second lowest is Aatrox sitting at ~46%
Azir was 34% at one point and nidalee/gragas had <38 just after their reworks.
I think the more recent release champions (Azir and Taliyah) had low 30's to mid 30's for win rate percentage, and that's mainly due to the complexity of their kits along with bugs (for Azir) and movement on the map (surfing on the wall with Taliyah's ult).
Ryze win percentage will always be lower then average due to the fact he takes a good chunk of games to be anything but absolute shit on. I like to see what type of win rates he has on people with 50 plus games on since the rework
On July 30 2016 08:40 IamPryda wrote: Ryze win percentage will always be lower then average due to the fact he takes a good chunk of games to be anything but absolute shit on. I like to see what type of win rates he has on people with 50 plus games on since the rework
Ryze was referred to as the face roll champ for so long with his initial version. I smile when I read about his complex kit now.
Actually I'd prefer the original Ryze. It could have been made to work. Maybe a bit of power out of the offense and embrace him building tanky mana mage.
Ryze win rate was almost exactly 50% in Plat+ before the rework, so from that standpoint there was no reason to rework him.
I think if they dumped his stupid ult for some form of utility (like, an MS boost???????) he'd be in a lot better shape. He actually does quite a lot of damage right now, although his button mashing has gotten harder since you need to think about whether you want the longer root or AoE dmg (actually like that they moved that part of his ult to his basic abilities). But that doesn't matter if the only way he can get to you is by flanking or running ghost.
Actually I'd prefer the original Ryze. It could have been made to work. Maybe a bit of power out of the offense and embrace him building tanky mana mage.
I too think the first Ryze (that's the S2 one right?) was best Ryze. But Riot thought the only way they could balance him was by making his Q a skill shot for outplay potential. But this Ryze is better than the one before, except for the dumb ult. His basic abilities are far more interactive now.
If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Reworked Ryze with the skillshot Q was still terror because even though he wouldn't shove you out of lane at level 1 he'd bully you super hard without pushing much with E, and later on 100-0 you thanks to his passive during a perma-root. Not exactly fine either.
The original top ryze was dumb because he was incredibly safe vs almost every pick (1v1 at least), outscaled everybody, and if he made it to tear without dying was just stupid to play against. I mean there were a few picks that could snowball off Ryze's weak level 1 when pressured with a mana crystal opening, but otherwise was a monster.
Rework with skillshot was dumb. I knew somebody in bronze who couldn't CS, had horrific map awareness and decisionmaking, but every few games on Ryze he would manage to just combo the other guy at level 3-5 simply because he practiced that combo in customs for a long time. You put pressure on him and he panics, you harass him and he panics, but severely misposition and he can combo you from 100% without letting the snare drop. Not really any counter play available.
The original Ryze wasn't bad until the anti-snowballing mechanics and toplane TP became the meta. Also the jungle not being a cakewalk to clear meant things like lvl 2 ganks and camping Ryze's lane were no longer options. That Ryze probably still wouldn't work in modern LOL, although IMO he still fits in better than both iterations of the rework.
As an aside, competitively, shouldn't "the worse" team just pick Jihn every game? Preferably with other ways to fillow up on picks like Karma, Leblanc, Ahri, etc? I almost never see the "better" team lose without getting picked a lot (almost always by Jihn) or getting the weird extended turnarounds (usually with a TP) that only curtain call can create.
New Ryze seems surprisingly good if you know what you are doing. Kind of funny how complex the iconic face roll champion has become. This created another champion that will be very hard to balance though with very different performance at low and high level of play.
I was bored and decided to make this calculator to predict the result of an ARAM game using the champions selected. Does anyone want to test it for me? I am particularly worried about the coverage of HTML5-supported browsers.
I was bored and decided to make this calculator to predict the result of an ARAM game using the champions selected. Does anyone want to test it for me? I am particularly worried about the coverage of HTML5-supported browsers.
Given that you only see the champs after the game starts, maybe it makes sense to make it an API call based on your summoner name instead of having to manually input the champs?
I was bored and decided to make this calculator to predict the result of an ARAM game using the champions selected. Does anyone want to test it for me? I am particularly worried about the coverage of HTML5-supported browsers.
Given that you only see the champs after the game starts, maybe it makes sense to make it an API call based on your summoner name instead of having to manually input the champs?
Like you have said, I can't get the list of champions in a game until the game starts. Once that happens, I can no longer dodge the ARAM... making it sort of pointless.
But I do agree with you that manually entering 10 text boxes is annoying and largely impractical.
An additional consideration is that if I do it after the game starts, I know who the 10 players are, so I can incorporate the ranking and/or mastery information of each player into my model...
If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
.
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Ryze gained a lot of power from tp meta and lane bully nerfs. Without Renekton, pre-nerfs Darius, Irelia or other ignite bruisers to keep him in check he was allowed to get to Ryze status.
I still think that kit was the most manageable of the ideas Riot has tried since.
On July 31 2016 15:19 Goumindong wrote: If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Come on GMD I expected you to try harder. This is seriously disappointing.
1. I am not going to restrict myself to only one team: there is absolutely no reason to throw away half of my data. In particular, I am not going to throw away half of my data due to rank deficiency because rank deficiency can be easily solved via regularization. I'll leave this to you as an exercise to explain why regularization solves the rank deficiency problem.
2. It's already coded as 0, 1, -1 for "not present", "blue", and "red" - this is exactly why there is rank deficiency. I'll leave this again as an exercise for you.
On July 31 2016 19:18 JonnyLaw wrote: Ryze gained a lot of power from tp meta and lane bully nerfs. Without Renekton, pre-nerfs Darius, Irelia or other ignite bruisers to keep him in check he was allowed to get to Ryze status.
I still think that kit was the most manageable of the ideas Riot has tried since.
On July 31 2016 15:19 Goumindong wrote: If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
.
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Come on GMD I expected you to try harder. This is seriously disappointing.
1. I am not going to restrict myself to only one team: there is absolutely no reason to throw away half of my data. In particular, I am not going to throw away half of my data due to rank deficiency because rank deficiency can be easily solved via regularization. I'll leave this to you as an exercise to explain why regularization solves the rank deficiency problem.
2. It's already coded as 0, 1, -1 for "not present", "blue", and "red" - this is exactly why there is rank deficiency. I'll leave this again as an exercise for you.
I agree with him that a single team search is a better tool; would be nice to see how your team stacks vs an average aram team. If you could generate a single team win rate vs median that would be awesome, actually.
^^ That was my point wrt the "single team". Its true that the full team is a better predictor, obviously. But also simply observing the end of the game and recording it is a better predictor, obviously. But neither are useful and we want information we can use, not information we can't. Also not sure why you would have to throw out half your data[half your variables yes, but not half your data]*.
WRT Rank deficiency: Its been a while but you should only be rank deficient when you've got not enough observations for your dimensionality, or you're singular. You should not be singular on the 0,1,-1 input unless i am seriously missing some obvious co-linearity. And you should have enough data points. I suppose you could just have an ill conditioned matrix but not seeing why you would have one by just by knowing the input numbers.
*You could potentially double your data because you get two observations from each game. This could causes issues but i am blanking on what they would be.
If you observe both teams half your data points are directly dependent on the other half of your data. Ie if team red wins team blue losses. If you count both teams as data point you are considering yourself to have twice as much independent data as you actually do.
He left the meta when banshees (and something else iirc) lost mana.
Banshees lost mana in patch 3.10, Ryze got his major update in 5.8. I recall him still being meta between that. Was there another mana item he built besides Archangel, RoA, FH and Banshees? I kinda remember him having no real sixth item to build. Maybe you're thinking of the FH mana nerf, although that was quite some time before.
On August 01 2016 04:29 General_Winter wrote: If you observe both teams half your data points are directly dependent on the other half of your data. Ie if team red wins team blue losses. If you count both teams as data point you are considering yourself to have twice as much independent data as you actually do.
If you double the data kindof. I felt like it should make x'e no longer zero, but wasn't sure off the top of my head. And didn't(and am not going to) check. If sufficiency wants to use the data he can figure it out.
On July 31 2016 15:19 Goumindong wrote: If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
.
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Come on GMD I expected you to try harder. This is seriously disappointing.
1. I am not going to restrict myself to only one team: there is absolutely no reason to throw away half of my data. In particular, I am not going to throw away half of my data due to rank deficiency because rank deficiency can be easily solved via regularization. I'll leave this to you as an exercise to explain why regularization solves the rank deficiency problem.
2. It's already coded as 0, 1, -1 for "not present", "blue", and "red" - this is exactly why there is rank deficiency. I'll leave this again as an exercise for you.
I agree with him that a single team search is a better tool; would be nice to see how your team stacks vs an average aram team. If you could generate a single team win rate vs median that would be awesome, actually.
Although perhaps that would be a different tool.
An average ARAM team is an interesting idea. In my mind it might be annoying to do because of the free champions changing week to week... So your average ARAM team also changes.
Perhaps a more interesting idea is to do this in the setting of ranked solo queue instead: I give you the team comp of the opposing team, you draft a team against it, then my model will evaluate your team comp....
If you were very sophisticated you could combine them. Given an enemy pick you would have an expected enemy team based on that pick and so a comp of your own to compare/create. This way you could potentially identify bad drafting options ahead of time.
I.E. I know the enemy picked Snydra first, how does my proposed team compare to the hypothetical syndra team? Do you have enough data to deal with the interaction terms to make that meaningful?
Hey, they picked an immobile champion. Let's pick champions that are good versus that. What champions are good versus that? Well, what about champions with gapclose+CC combos. Hey, that works!
On July 31 2016 15:19 Goumindong wrote: If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
.
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Come on GMD I expected you to try harder. This is seriously disappointing.
1. I am not going to restrict myself to only one team: there is absolutely no reason to throw away half of my data. In particular, I am not going to throw away half of my data due to rank deficiency because rank deficiency can be easily solved via regularization. I'll leave this to you as an exercise to explain why regularization solves the rank deficiency problem.
2. It's already coded as 0, 1, -1 for "not present", "blue", and "red" - this is exactly why there is rank deficiency. I'll leave this again as an exercise for you.
I agree with him that a single team search is a better tool; would be nice to see how your team stacks vs an average aram team. If you could generate a single team win rate vs median that would be awesome, actually.
Although perhaps that would be a different tool.
An average ARAM team is an interesting idea. In my mind it might be annoying to do because of the free champions changing week to week... So your average ARAM team also changes.
Perhaps a more interesting idea is to do this in the setting of ranked solo queue instead: I give you the team comp of the opposing team, you draft a team against it, then my model will evaluate your team comp....
They'd both be pretty cool, actually.
I read your Reddit thread, lol. "Your model predicts the winning game 66% of the time in an environment where an algorythm places player teams together who are equally likely to win. Wow that is trash." Wut, no that is pretty sweet.
On July 31 2016 15:19 Goumindong wrote: If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
.
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Come on GMD I expected you to try harder. This is seriously disappointing.
1. I am not going to restrict myself to only one team: there is absolutely no reason to throw away half of my data. In particular, I am not going to throw away half of my data due to rank deficiency because rank deficiency can be easily solved via regularization. I'll leave this to you as an exercise to explain why regularization solves the rank deficiency problem.
2. It's already coded as 0, 1, -1 for "not present", "blue", and "red" - this is exactly why there is rank deficiency. I'll leave this again as an exercise for you.
I agree with him that a single team search is a better tool; would be nice to see how your team stacks vs an average aram team. If you could generate a single team win rate vs median that would be awesome, actually.
Although perhaps that would be a different tool.
An average ARAM team is an interesting idea. In my mind it might be annoying to do because of the free champions changing week to week... So your average ARAM team also changes.
Perhaps a more interesting idea is to do this in the setting of ranked solo queue instead: I give you the team comp of the opposing team, you draft a team against it, then my model will evaluate your team comp....
They'd both be pretty cool, actually.
I read your Reddit thread, lol. "Your model predicts the winning game 66% of the time in an environment where an algorythm places player teams together who are equally likely to win. Wow that is trash." Wut, no that is pretty sweet.
If you could get on LoL games, a hit ratio like that would make you stupid rich.
On July 31 2016 15:19 Goumindong wrote: If you're thinking about dodging then it's better to just use one team. Which should also let you increase the complexity of the calculation(I.E. Include more interaction variables) and should negate your deficiency issue.
Additionally with two teams you should be able to code a single champion as 0,1,-1 for "not present", "blue", red" on a single variable. Which should also iirc fix your rank deficiency issue.
The enemy team won't matter in the long run to dodging except insomuch as your tool modifies what is played so just going to the one team is probably ideal even if it's less accuracy in the model
Edit: it's been quite a while since I've done any work in that area so I could be way off base. Also not sure if your models have input requirements with regards to a trinary structure. (Red side/blue side effect should be the same so it should work fine)
.
On July 31 2016 08:44 GrandInquisitor wrote: If I remember correctly, original Ryze was OK for a long time until various meta changes and people waking up led to the realization that he was a total terror top lane. Once that happened they had to Olaf him while they tried to figure out how to rework him. He was simultaneously a lane bully but also a hyperscaler, which was completely obnoxious.
Original ryze was on/off strong. He left the meta when banshees (and something else iirc) lost mana. He came back with tear and roa buffs. (Mainly). Which is what caused him to be reworked.
Come on GMD I expected you to try harder. This is seriously disappointing.
1. I am not going to restrict myself to only one team: there is absolutely no reason to throw away half of my data. In particular, I am not going to throw away half of my data due to rank deficiency because rank deficiency can be easily solved via regularization. I'll leave this to you as an exercise to explain why regularization solves the rank deficiency problem.
2. It's already coded as 0, 1, -1 for "not present", "blue", and "red" - this is exactly why there is rank deficiency. I'll leave this again as an exercise for you.
I agree with him that a single team search is a better tool; would be nice to see how your team stacks vs an average aram team. If you could generate a single team win rate vs median that would be awesome, actually.
Although perhaps that would be a different tool.
An average ARAM team is an interesting idea. In my mind it might be annoying to do because of the free champions changing week to week... So your average ARAM team also changes.
Perhaps a more interesting idea is to do this in the setting of ranked solo queue instead: I give you the team comp of the opposing team, you draft a team against it, then my model will evaluate your team comp....
They'd both be pretty cool, actually.
I read your Reddit thread, lol. "Your model predicts the winning game 66% of the time in an environment where an algorythm places player teams together who are equally likely to win. Wow that is trash." Wut, no that is pretty sweet.
If you could get on LoL games, a hit ratio like that would make you stupid rich.
Well maybe? The key is that there isn't a 50% win chance afte champions have been randomly selected. The champions selected is information and so there isn't any reason a person could not put down those kind of numbers. It's like saltyteemo betting. It's not hard to guess right, you often come in when one team has a, more or less, "insurmountable" lead. Bots (maybe it's bots maybe not) get to like 90% accuracy. The 50% base line is before champions are selected, not after. I would wager that the majority of ARAM players hit better than 50% prediction were they to keep track.
Now I don't know the complexity of the model, whether it accounts for team composition structures, champion synergy, etc. Those are the sorts of things the human brain is quite good at sussing out and so they could easily have an advantage over an algorithm that cannot be that complex due to mathematical limits. That doesn't mean 66% isn't good but it means we don't really know if it is.
I think a very knowledgeable player may be able to do better, but it's sort of a null point because but we can't know either way - I can't ask such a person to rate enough games (which could be thousands of games depending on the actual performance difference) to formally demonstrate the difference.
I thought ARAM uses an MMR system which pushes the majority of players at 50% win rate. If that is the case dodging games or whatever seems rather pointless because your overall win rate will not change anyway.
On August 03 2016 19:34 Redox wrote: I thought ARAM uses an MMR system which pushes the majority of players at 50% win rate. If that is the case dodging games or whatever seems rather pointless because your overall win rate will not change anyway.
In the long term kindof. The law of large numbers works to put the average towards 50%. But the difference between wins and losses is not an average. Even if you play all games at 50% as you play more games the difference between wins and losses can go to infinity or negative infinity or nothing at all.
And since, well you won't be at 50% until you hit your proper MMR and since dodging poor games increases that MMR. If you are about winning it's ideal to dodge bad games.
Granted you probably shouldn't care about winning ARAMs but I bet loads do.
Varus appears to have a pretty awful bug this patch where his smartcast Q fires prematurely (aka, before you release the key). Although it doesn't start right away, it took me about 10 minutes to trigger whatever it was and then it happened pretty consistently.
You can sort of hedge against it by not moving around as much when you're in the windup phase of the shot, but it only sort of works. I managed to get a game ending quadra kill in an ARAM with this stupid bug making every shot a miserable chore. Holy crap, was it annoying.
On August 04 2016 18:09 zer0das wrote: Varus appears to have a pretty awful bug this patch where his smartcast Q fires prematurely (aka, before you release the key). Although it doesn't start right away, it took me about 10 minutes to trigger whatever it was and then it happened pretty consistently.
You can sort of hedge against it by not moving around as much when you're in the windup phase of the shot, but it only sort of works. I managed to get a game ending quadra kill in an ARAM with this stupid bug making every shot a miserable chore. Holy crap, was it annoying.
I have felt this same thing on Xerath a few times. It might have been more lag in my case than a bug, but it made me feel like I couldn't trust smart casting at least for the charging spells like these.
On August 05 2016 05:53 MuddyJam wrote: So the one time I don't check my teams op.gg and just tab out once locking in my first pick I get premade of kalista jungle and 29% win rate vayne (29 games) they proceed to blame everyone who is not premade with them, 3v2 bot lane while enemy take 2 towers and our inhibitor mid.
On August 05 2016 08:58 IamPryda wrote: just got pulsefire ezreal in hextech surprised they even allow ultimate skins even if its the oldest one to bad I don't play ez ever
Could be strider hiryu (the sash mainly) but not actually sure what you're complaining about. Arcade skins always have a theme of "gamer person" for whatever shitty definition of that it is. I mean look at Sona. Arcade is an established skin line. Ahri gets all the skins because she is a perennially popular, and gets lots of pro play time. (Recently(?)). And of course this is one of the most modest Ahri skins
On August 06 2016 05:57 GrandInquisitor wrote: Coming up with more "series" skins is a damn smart financial move. People love playing with matching skins.
agreed, I really like how they've swapped to almost a calendar release scedule in the last 2 years with Blood Moon >> April fools > Pool Party > Project > Arcade > Halloween > Xmas
sure I'm missing a couple but the point is there, I actually anticipated Pool Party this year.
On August 06 2016 05:57 GrandInquisitor wrote: Coming up with more "series" skins is a damn smart financial move. People love playing with matching skins.
agreed, I really like how they've swapped to almost a calendar release scedule in the last 2 years with Blood Moon >> April fools > Pool Party > Project > Arcade > Halloween > Xmas
sure I'm missing a couple but the point is there, I actually anticipated Pool Party this year.
Yeah it is just more fun to show a united front where everyone is under the same theme.
But Kled looks like a jungler. LOL. AS steroid fo free w/ percent damage? big shield that comes back from autoing stuff? Strong mobility with a couple Knockbacks? Wall jumps? 2 AoE clearing moves?
Seems like a good candidate for a toxic full tank build. Shield based on Bonus HP, 2 % hp attacks and a Sivir ult that requires him to be the first one to go in.
Yeah. They went out of their way to make it hard to jungle and force him top. Not sure it will work though, I feel like his kit pushes him towards sup since he has a thresh hook and he has %health damage that keeps him relevant even if he doesn't build any damage. Likewise he has a good initiation ult for the team that doesn't require damage. Build tanky like braum
EPT
![[image loading]](https://2.bp.blogspot.com/-ZzXth_UJDxs/V6TyFZQh0RI/AAAAAAAA9XU/ma0keE8eofoUQr5RpDtcmQ1cjq3HgGxrQCLcB/s1600/2eaaf8b0ae33f1b2.jpg)
![[image loading]](https://4.bp.blogspot.com/-EEFXAWQDAu0/V6TyFb1l9PI/AAAAAAAA9XY/_uVsVQkjFvg9JqZf4BEGXjjHBRCLq74PwCLcB/s1600/7527c2231c723142.jpg)
