So, Madred's razors on Skarner? I've been treating Razors/Mercs/Philo/Sheen as my base items and moving on from there depending on the enemy composition, with an eye for attack speed when I can sneak it in. Starting Cloth Armor + Pots.
So, Madred's razors on Skarner? I've been treating Razors/Mercs/Philo/Sheen as my base items and moving on from there depending on the enemy composition, with an eye for attack speed when I can sneak it in. Starting Cloth Armor + Pots.
BRazor's are too expensive and Razor's/Wriggles feel unnecessary. Damage-wise, I rarely get more than a Wit's End and Sheen, Shurelya's + FH too strong
So, Madred's razors on Skarner? I've been treating Razors/Mercs/Philo/Sheen as my base items and moving on from there depending on the enemy composition, with an eye for attack speed when I can sneak it in. Starting Cloth Armor + Pots.
BRazor's are too expensive and Razor's/Wriggles feel unnecessary. Damage-wise, I rarely get more than a Wit's End and Sheen, Shurelya's + FH too strong
I've just been getting the Razors, and saving the Bloodrazors for the end in case I have extra gold laying about. I've found it has good symmetry when focusing attack speed anyway. I took it because I felt that Skarner's damage on the larger jungle creeps was a bit mediocre with just his Q, and it helped save some mana which I was hurting for.
Wriggles while unnecessary in some situations should never feel like a bad choice for a jungler or top or anyone who auto attacks (its even OK on ranged carries). Indeed, as a jungler i find wriggles is almost as necessary as smite just because of the strength it gives you in securing buffs.
Yeah, wriggles always helps out. While it's not ideal on all junglers, most of them can use it simply because it saves a great deal in mana-usage. If you're someone like Mao, Amumu or Skarner, who are likely to spend a fair amount of mana on a gank, wriggles allows them to go back into the jungle without severe delays, especially if you came back from the gank with enough helath to make continued jungling viable.
I used to raise my eyebrows at solo top people getting a lantern, but in the long run it saves plenty of gold on wards, in addition to the inherent stats.
Now that I think about it. Is GA severely underrated? Or rather under utilized?
GA is good if you meet the following prerequisites:
You are very sure that you're going to be the first person to be focused in a fight.
You require armor and MR.
You have restricted slots for defensive items.
Quicksilver Sash would not be more effective.
GA is bad if the enemy team just kills you last anyway and then 5v1s your soon-to-be-corpse. GA is good if it requires the enemy team to focus you, break off, and then focus you again.
@Skarner: Doesn't need wriggles at all. Past level 4, he should be able to solo any camp as long as he has 100hp left, and Sheen ends up being better deeps against dragon and baron, and of course in ganks.
Most solo-tops use all stats on it anyway. A Trynd with Wriggles, Boots, IE and PD is ready to go, and will badly outdamage a Trynd with BT+PD+ something else.
Yeah 1300 elo is what, top 25%ish of ranked? It is not the bottom.
LOL wiki says that 1250 is the starting point for top 25%
But that can't be right because i could have sworn that people started at 1275 or something and the starting ELO should be the 50/50 point (at least, ideally)
Yeah 1300 elo is what, top 25%ish of ranked? It is not the bottom.
LOL wiki says that 1250 is the starting point for top 25%
But that can't be right because i could have sworn that people started at 1275 or something and the starting ELO should be the 50/50 point (at least, ideally)
It could be there's a significant population in the 1000-1200 ELO range.
I'd love to see Riot talk a bit about Elo ratings and distribution and stuff.
i believe they said that only 10% of the community had gold rating at the end of season 1, and only 0.3% had platinum
Which one of those does Frosteey have because the only person I know with a funny frame is Frosteey and otherwise I don't know what you're talking about.
I'd love to see Riot talk a bit about Elo ratings and distribution and stuff.
i believe they said that only 10% of the community had gold rating at the end of season 1, and only 0.3% had platinum
Which one of those does Frosteey have because the only person I know with a funny frame is Frosteey and otherwise I don't know what you're talking about.
Here is the season 1 ranks Gold is top 3%, Silver is top 10%. 1249 does indeed mark the top 25% of ranked. Keep in mind that only like, half of level 30s are ranked in the first place as well.
Yeah 1300 elo is what, top 25%ish of ranked? It is not the bottom.
LOL wiki says that 1250 is the starting point for top 25%
But that can't be right because i could have sworn that people started at 1275 or something and the starting ELO should be the 50/50 point (at least, ideally)
It could be there's a significant population in the 1000-1200 ELO range.
Yes and no. Since the ELO rating system is not guaranteed to be normal (your ELO is guaranteed to be distributed approximately normal but the sum of everyone's is not*) Riot doesn't have much control over where in the distribution people end up. And since leaving/dodging automatically applies a loss even if your team wins (or no game is palyed) we expect the distribution of ELO's to be skewed to the bottom end compared to the distribution of true ELO's.
But that doesn't change the fact that the ideal place to start a player whose value is unknown is at the mean of your distribution. The reason for this is basically that since you expect most people to be in that range putting them higher makes the game not fun for everyone until they drop out.
That is, when matchmaking you want to minimize the ELO difference between teams/players because that makes the game more fun/competitive. When we place a player who hasn't played ranked yet, we want to minimize the difference but rather than minimize by choosing players with current values we minimize by giving a rating to that player that will minimize the expected difference between him and groups he would be paired with.
That point will (iirc) always be the mean of the distribution. Buuuut we can do a bit more and say that we don't care about the shape of the ELO distribution, only that it gives us relative ratings, we can simply redefine values until the mean of the distribution is the median of the distribution**. Then choose the median point.
So if 1275 is the start and P(ELO>1250)= .25 then we are doing a really bad job of putting starting players in places where its both fun for starting players and fun for the players who really are at 1275 ELO.
*Sum of two uncorrelated normal distributions tends to be normally distributed if they aren't dependent on each other. But our distributions are dependent on one another and not all the distributions will be approximately normal by the time we are adding them together... and leavers distort the normality of the distribution and so break the guarantee again
**This is not precisely true since our matchmaking and ELO assumptions make assumptions about the relative strength of players. But the leaving system breaks this assumption. If the leaving system did not break this assumption then we could probably say that the distribution was approximately normal and so median=mean. Which is to say that for the purposes of actually matching/starting people it would be best to assume they're at the median if the mean=/= the median.
Edit: Better way to say **. We can believe that player skill distribution ought to be roughly normal and if we do, then the median of the distribution that we actually see will correspond to the median of the normal distribution we believe is true even if the distribution we see isn't normal.
The elo will balance +- 1200 simply because that is the point people enter.
The 'button' days actually do infuse a lot of elo inflation, that offsets quite a bit. On a win, the losing team loses nothing instead of 12-13 elo, and the winning team gains only half. You have a net of +6-7 per person. Not that many people actually dodge a ranked game. The ones that actually dodge quite a bit are either doing it intentionally or playing an alt account and would rather not deal with bullshit.
If i take my own account I have been inflated a total of 28 points, and havent dodged a single game, or left queue. I think a vast majority of people that play ranked dont have more than 1 dodge this season.
You have outliers actually affecting the statistic. One person dodging themselves down to 0 elo has a huge statistical impact on elo inflation vs deflation. As far as actual inflation vs actual deflation when you scrub away alt accounts and scrub away the statistical outliers, we actually have moderate elo inflation.
You also have the people who play ranked for 3-10 games, get to 800 elo and quit playing ranked.
The best thing about wriggles top though is the fact that you dont have to base at any point after getting wriggles just to buy a ward. Which lets you stay top longer, you really cant measure that benifit that accurately.
One thing too is that everyone starts at 1200, yet 1250+ is considered top 25%. You might think this is unfair to people legitimately in the 1200-1300 bracket, but a change in the starting point would only slide the whole ladder, making it 'unfair' for anyone +/-100 elo from the starting point. There is no really fair way to do it, you just need to work through it.
I see lots of 1400 players that are actually just 8-2 in their placement matches. Do you think they belong there? I don't, and I'm sure if they keep playing ranked, they will lose their teammates some games. Lots of people say 'lol, I'm 1350 elo, you aren't', then you check their profile and they only just finished their placements. Most of the time they know they don't belong there and are too afraid to play more.
One thing too is that everyone starts at 1200, yet 1250+ is considered top 25%. You might think this is unfair to people legitimately in the 1200-1300 bracket, but a change in the starting point would only slide the whole ladder, making it 'unfair' for anyone +/-100 elo from the starting point. There is no really fair way to do it, you just need to work through it.
I see lots of 1400 players that are actually just 8-2 in their placement matches. Do you think they belong there? I don't, and I'm sure if they keep playing ranked, they will lose their teammates some games. Lots of people say 'lol, I'm 1350 elo, you aren't', then you check their profile and they only just finished their placements. Most of the time they know they don't belong there and are too afraid to play more.
I was so afraid.
0
SpectrumArcher of InfernoChaldea Rec RoomRegistered Userregular
One thing too is that everyone starts at 1200, yet 1250+ is considered top 25%. You might think this is unfair to people legitimately in the 1200-1300 bracket, but a change in the starting point would only slide the whole ladder, making it 'unfair' for anyone +/-100 elo from the starting point. There is no really fair way to do it, you just need to work through it.
I see lots of 1400 players that are actually just 8-2 in their placement matches. Do you think they belong there? I don't, and I'm sure if they keep playing ranked, they will lose their teammates some games. Lots of people say 'lol, I'm 1350 elo, you aren't', then you check their profile and they only just finished their placements. Most of the time they know they don't belong there and are too afraid to play more.
I was so afraid.
I'm pretty afraid, but that's mostly 'cause my skill these days playing maybe 3 games a week max is not reflective of my skill closer to the start of the season.
do people really think fiora's w is useless. I parried some 500 ap+lichbane fizz Q that thought he could be cleaver and bush juke me then killed his ass.
Parry doesn't reflect damage back at the attacker, it cancels out the physical damage and deals ~150 magical damage. It's great for laning phase and 1v1s, but in late game team fights you're very likely to end up dealing a pitiful damage to a tank or a bruiser. It'd be pretty sweet if you could reflect a crit back at the AD carry, but that's not possible due to how the skill works.
I think it'd be better if it negated any side effects of the attack, like Garen's silence.
Ive been playing more ranked and diggin myself out of >1000 elo finally with jungle/top riven.
Just wanted to ask quick what your jungle routes were. I basically start with dorans blade go wraiths then red. Quick gank/gank invade jungle gank all day pretty much. Farm up to double bts/guardian and win games.
My biggest problem with riven right now is i don't buy wriggles and therefore don't ward as much as my other junglers.
Also is it better to go half half on runes with ad and arpen or just full AD. Right now Im doing half and its been working well.
[SIGPIC][/SIGPIC]
Gamertag(SSF4/MW2)StokedAidzzzSC2 ID Stoked.655 Uploaded SC2 Replays
I used to raise my eyebrows at solo top people getting a lantern, but in the long run it saves plenty of gold on wards, in addition to the inherent stats.
Trynd solo top Wriggle's can be game definingly powerful, in my experience, depending on the matchup. It also combines well with zerkers / BC to secure buffs very quickly.
Posts
It's the LoL equivalent of not understanding how you couldn't just use the backup Mercedes-Benz to get groceries.
nerds l2p
BRazor's are too expensive and Razor's/Wriggles feel unnecessary. Damage-wise, I rarely get more than a Wit's End and Sheen, Shurelya's + FH too strong
I've just been getting the Razors, and saving the Bloodrazors for the end in case I have extra gold laying about. I've found it has good symmetry when focusing attack speed anyway. I took it because I felt that Skarner's damage on the larger jungle creeps was a bit mediocre with just his Q, and it helped save some mana which I was hurting for.
tumblr | instagram | twitter | steam
GA is good if you meet the following prerequisites:
You are very sure that you're going to be the first person to be focused in a fight.
You require armor and MR.
You have restricted slots for defensive items.
Quicksilver Sash would not be more effective.
GA is bad if the enemy team just kills you last anyway and then 5v1s your soon-to-be-corpse. GA is good if it requires the enemy team to focus you, break off, and then focus you again.
@Skarner: Doesn't need wriggles at all. Past level 4, he should be able to solo any camp as long as he has 100hp left, and Sheen ends up being better deeps against dragon and baron, and of course in ganks.
But that can't be right because i could have sworn that people started at 1275 or something and the starting ELO should be the 50/50 point (at least, ideally)
It could be there's a significant population in the 1000-1200 ELO range.
i believe they said that only 10% of the community had gold rating at the end of season 1, and only 0.3% had platinum
Which one of those does Frosteey have because the only person I know with a funny frame is Frosteey and otherwise I don't know what you're talking about.
frosteey has gold
tumblr | instagram | twitter | steam
Yes and no. Since the ELO rating system is not guaranteed to be normal (your ELO is guaranteed to be distributed approximately normal but the sum of everyone's is not*) Riot doesn't have much control over where in the distribution people end up. And since leaving/dodging automatically applies a loss even if your team wins (or no game is palyed) we expect the distribution of ELO's to be skewed to the bottom end compared to the distribution of true ELO's.
But that doesn't change the fact that the ideal place to start a player whose value is unknown is at the mean of your distribution. The reason for this is basically that since you expect most people to be in that range putting them higher makes the game not fun for everyone until they drop out.
That is, when matchmaking you want to minimize the ELO difference between teams/players because that makes the game more fun/competitive. When we place a player who hasn't played ranked yet, we want to minimize the difference but rather than minimize by choosing players with current values we minimize by giving a rating to that player that will minimize the expected difference between him and groups he would be paired with.
That point will (iirc) always be the mean of the distribution. Buuuut we can do a bit more and say that we don't care about the shape of the ELO distribution, only that it gives us relative ratings, we can simply redefine values until the mean of the distribution is the median of the distribution**. Then choose the median point.
So if 1275 is the start and P(ELO>1250)= .25 then we are doing a really bad job of putting starting players in places where its both fun for starting players and fun for the players who really are at 1275 ELO.
*Sum of two uncorrelated normal distributions tends to be normally distributed if they aren't dependent on each other. But our distributions are dependent on one another and not all the distributions will be approximately normal by the time we are adding them together... and leavers distort the normality of the distribution and so break the guarantee again
**This is not precisely true since our matchmaking and ELO assumptions make assumptions about the relative strength of players. But the leaving system breaks this assumption. If the leaving system did not break this assumption then we could probably say that the distribution was approximately normal and so median=mean. Which is to say that for the purposes of actually matching/starting people it would be best to assume they're at the median if the mean=/= the median.
Edit: Better way to say **. We can believe that player skill distribution ought to be roughly normal and if we do, then the median of the distribution that we actually see will correspond to the median of the normal distribution we believe is true even if the distribution we see isn't normal.
Man, I thought I was pretty terrible.
3ds friend code: 2981-6032-4118
there's the saying that 99% of everything is terrible
We'll see what happens when I'm finally 30. Any year now!
The elo will balance +- 1200 simply because that is the point people enter.
The 'button' days actually do infuse a lot of elo inflation, that offsets quite a bit. On a win, the losing team loses nothing instead of 12-13 elo, and the winning team gains only half. You have a net of +6-7 per person. Not that many people actually dodge a ranked game. The ones that actually dodge quite a bit are either doing it intentionally or playing an alt account and would rather not deal with bullshit.
If i take my own account I have been inflated a total of 28 points, and havent dodged a single game, or left queue. I think a vast majority of people that play ranked dont have more than 1 dodge this season.
You have outliers actually affecting the statistic. One person dodging themselves down to 0 elo has a huge statistical impact on elo inflation vs deflation. As far as actual inflation vs actual deflation when you scrub away alt accounts and scrub away the statistical outliers, we actually have moderate elo inflation.
You also have the people who play ranked for 3-10 games, get to 800 elo and quit playing ranked.
I see lots of 1400 players that are actually just 8-2 in their placement matches. Do you think they belong there? I don't, and I'm sure if they keep playing ranked, they will lose their teammates some games. Lots of people say 'lol, I'm 1350 elo, you aren't', then you check their profile and they only just finished their placements. Most of the time they know they don't belong there and are too afraid to play more.
I was so afraid.
3ds friend code: 2981-6032-4118
I think it'd be better if it negated any side effects of the attack, like Garen's silence.
Just wanted to ask quick what your jungle routes were. I basically start with dorans blade go wraiths then red. Quick gank/gank invade jungle gank all day pretty much. Farm up to double bts/guardian and win games.
My biggest problem with riven right now is i don't buy wriggles and therefore don't ward as much as my other junglers.
Also is it better to go half half on runes with ad and arpen or just full AD. Right now Im doing half and its been working well.
Gamertag(SSF4/MW2)StokedAidzzzSC2 ID Stoked.655
Uploaded SC2 Replays
Trynd solo top Wriggle's can be game definingly powerful, in my experience, depending on the matchup. It also combines well with zerkers / BC to secure buffs very quickly.