"It can hardly be a coincidence that no language on earth has ever produced the expression 'As civil as solo queue.'"
-The Long Dark Tea Time of the Soul, Douglas Adams
Friday, September 28, 2012
Thursday, September 27, 2012
Patch Day: Kha'zix
The most exciting parts of today's patch are obviously Soraka's visual rework and the arrival of Kha'zix, but there are a handful of balance changes to discuss.
Having only just been buffed barely a week ago, it's clear that the changes to Rengar made him stronger than intended. Nerfing the innate survivability and damage of his W significantly weakens Rengar's ability to ignore defenses early on and largely build assassin. The early rank cooldown nerf on Q is similar, but doesn't make him any less dangerous at full Ferocity. He'll still be a strong champion, but I'd probably avoid jungling him.
Syndra is receiving extensive buffs, and in light of Rengar's nerfs I'm concerned Riot is repeating their mistake. Rengar had the worst winrate of any champion upon his release, up until Syndra broke his record. However, with only bug fixes he eventually reached a winrate that was within the margins of acceptability. Syndra wasn't on course to do likewise, but the extent of these buffs is far greater than Rengar's. Despite currently having a winrate of 30%, it would not surprise me to see her pass 50% with these buffs and incur nerfs later.
Those are the only direct buffs and nerfs in the patch, but they represent a much more drastic approach to balancing recent champions than we've seen in recent months. It'll be interesting to see if I'm correct about Syndra's status post-buff.
Having only just been buffed barely a week ago, it's clear that the changes to Rengar made him stronger than intended. Nerfing the innate survivability and damage of his W significantly weakens Rengar's ability to ignore defenses early on and largely build assassin. The early rank cooldown nerf on Q is similar, but doesn't make him any less dangerous at full Ferocity. He'll still be a strong champion, but I'd probably avoid jungling him.
Syndra is receiving extensive buffs, and in light of Rengar's nerfs I'm concerned Riot is repeating their mistake. Rengar had the worst winrate of any champion upon his release, up until Syndra broke his record. However, with only bug fixes he eventually reached a winrate that was within the margins of acceptability. Syndra wasn't on course to do likewise, but the extent of these buffs is far greater than Rengar's. Despite currently having a winrate of 30%, it would not surprise me to see her pass 50% with these buffs and incur nerfs later.
Those are the only direct buffs and nerfs in the patch, but they represent a much more drastic approach to balancing recent champions than we've seen in recent months. It'll be interesting to see if I'm correct about Syndra's status post-buff.
Wednesday, September 26, 2012
Champion Misdesigns: Thakkrad Goldtooth
I often spend time mucking about with champion designs. Occasionally I happen upon an idea so horrifically bad that it deserves to be put into the wild. Thakkrad Goldtooth is just such an idea.
Thakkrad was born through the epiphany that a champion could have abilities which cost gold to use, which is possibly one of the worst champion design concepts ever imagined. At the same time the idea was so unique and hilarious I couldn't help but attempt to make it work.
At the outset I recognized that there were two paths to take in order to balance this mechanic. The first path was to compensate Thakkrad for his trouble by making his abilities significantly stronger than normal. As he would be expending a meaningful amount of gold on his abilities, he would normally be weaker by virtue of having fewer items. The second path was to find a way to augment Thakkrad's gold income to make up for his extra expenditures. Both paths have significant issues.
The first path has the potential to snowball very powerfully. If Thakkrad were to get a lead he could easily leverage his over-potent skills to dominate the early and mid-game. While he'd have a slump toward the late-game, he'd be exceptional in very long games simply because by the end he'd eventually have a full item build plus naturally stronger abilities. This path seemed too difficult to balance, so I avoided it.
The second path has its own pitfalls. With an enhanced gold income, Thakkrad could potentially pull a "Bank Plank" and farm incessantly, show up with a trifecta of super-items, and destroy the game. It also had some snowball potential, though not as much. As this seemed easier to balance, and more unique, I took this route.
The net result was the following monstrosity.
Thakkrad Goldtooth, the Wealthy
Thakkrad is an extremely wealthy warrior from Lokfar. He gift for acquiring riches is matched only by his ability to spend them. He fights by swinging a giant bag filled with gold.
Passive: Avarice
All of Thakkrad's gold per 10 is doubled (includes base gold per 10).
Q: Crush
Thakkrad crushes an enemy under the weight of his wealth, dealing 50/100/150/200/250 (+ 1.5 Bonus AD) physical damage, and stunning the target for 1.2/1.4/1.6/1.8/2.0 seconds. 14/13/12/11/10 second cooldown. Costs 17/19/21/23/25 gold.
W: Fast Money
Thakkrad passively gains 15/20/25/30/35 movement speed. When activated the passive bonus is doubled and Thakkrad gains 25/35/45/55/65% attack speed for 7 seconds. 12 second cooldown. Costs 20 gold.
E: Clear the Field
Thakkrad swings his bag around up to three times, dealing 20/35/50/65/80 (+ 0.3 Bonus AD) each swing and moving 30% slower than normal. At the end of his spin he hurls his bag, dealing 30/60/90/120/150 (+ 1.5 Bonus AD) damage to the first target hit and knocking them back. Thakkrad can activate this ability again to end his spin and throw the bag early. The knockback distance is based on the number of spins. 10 second cooldown. Costs 25 gold.
R: Ransom
Thakkrad gains 5/7.5/10 gold per 10. When Thakkrad dies, he can activate this ability to spend 30 + Level * 15 gold to revive immediately where he fell with full health. 180 second cooldown.
This is not a good champion design. This is, in fact, a terrible champion design. Without even taking his ultimate into account, Thakkrad can burn 400 gold per minute spamming his abilities without cooldown reduction. Despite looking like a bruiser he'd probably end up as a support, sitting in bushes waiting to crush opponents while abusing his massive gp10 advantage to get good items.
I can't overstate how bad this design is, or how fun it was to make.
Thakkrad was born through the epiphany that a champion could have abilities which cost gold to use, which is possibly one of the worst champion design concepts ever imagined. At the same time the idea was so unique and hilarious I couldn't help but attempt to make it work.
At the outset I recognized that there were two paths to take in order to balance this mechanic. The first path was to compensate Thakkrad for his trouble by making his abilities significantly stronger than normal. As he would be expending a meaningful amount of gold on his abilities, he would normally be weaker by virtue of having fewer items. The second path was to find a way to augment Thakkrad's gold income to make up for his extra expenditures. Both paths have significant issues.
The first path has the potential to snowball very powerfully. If Thakkrad were to get a lead he could easily leverage his over-potent skills to dominate the early and mid-game. While he'd have a slump toward the late-game, he'd be exceptional in very long games simply because by the end he'd eventually have a full item build plus naturally stronger abilities. This path seemed too difficult to balance, so I avoided it.
The second path has its own pitfalls. With an enhanced gold income, Thakkrad could potentially pull a "Bank Plank" and farm incessantly, show up with a trifecta of super-items, and destroy the game. It also had some snowball potential, though not as much. As this seemed easier to balance, and more unique, I took this route.
The net result was the following monstrosity.
Thakkrad Goldtooth, the Wealthy
Thakkrad is an extremely wealthy warrior from Lokfar. He gift for acquiring riches is matched only by his ability to spend them. He fights by swinging a giant bag filled with gold.
Passive: Avarice
All of Thakkrad's gold per 10 is doubled (includes base gold per 10).
Q: Crush
Thakkrad crushes an enemy under the weight of his wealth, dealing 50/100/150/200/250 (+ 1.5 Bonus AD) physical damage, and stunning the target for 1.2/1.4/1.6/1.8/2.0 seconds. 14/13/12/11/10 second cooldown. Costs 17/19/21/23/25 gold.
W: Fast Money
Thakkrad passively gains 15/20/25/30/35 movement speed. When activated the passive bonus is doubled and Thakkrad gains 25/35/45/55/65% attack speed for 7 seconds. 12 second cooldown. Costs 20 gold.
E: Clear the Field
Thakkrad swings his bag around up to three times, dealing 20/35/50/65/80 (+ 0.3 Bonus AD) each swing and moving 30% slower than normal. At the end of his spin he hurls his bag, dealing 30/60/90/120/150 (+ 1.5 Bonus AD) damage to the first target hit and knocking them back. Thakkrad can activate this ability again to end his spin and throw the bag early. The knockback distance is based on the number of spins. 10 second cooldown. Costs 25 gold.
R: Ransom
Thakkrad gains 5/7.5/10 gold per 10. When Thakkrad dies, he can activate this ability to spend 30 + Level * 15 gold to revive immediately where he fell with full health. 180 second cooldown.
This is not a good champion design. This is, in fact, a terrible champion design. Without even taking his ultimate into account, Thakkrad can burn 400 gold per minute spamming his abilities without cooldown reduction. Despite looking like a bruiser he'd probably end up as a support, sitting in bushes waiting to crush opponents while abusing his massive gp10 advantage to get good items.
I can't overstate how bad this design is, or how fun it was to make.
Tuesday, September 25, 2012
Thought: Secret Passive Escalation
Kha'zix, the upcoming champion, brings with him the latest in secret passives. If Kha'zix and Rengar are in a match together and Rengar possesses his Bonetooth Necklace, then shortly after Kha'zix hits level 16 a special buff appears on both of them indicating "The Hunt is On". Whoever first kills or assists a kill on the other receives a reward. Rengar's Bonetooth Necklace becomes the Head of Kha'zix, granting Rengar permanent full trophy stacks. Kha'zix gains a fourth evolution point. This new feature has caused quite a stir.
My own thoughts are divided. I see reasons both for and against this special interaction, and at any given time I lean one way or the other. It's a complicated issue that treads new ground, and it can be very hard to predict how these issues might affect the game in the long run.
A large number of players assert that the advantage conferred by this interaction is overpowering, but I contend that the advantage is both irrelevant to the vast majority of games and marginal even in games where it plays a role. Most games won't feature both Rengar and Kha'zix; assuming each champion appears in 25% of all games (a very high rate of play) only 6.25% of all games will have both. The majority of those remaining games will have been decided well before Kha'zix ever hits 16 and some will not even feature the Bonetooth Necklace (e.g. any game with AP Rengar), leaving only 1% of all games or fewer where the advantage even plays a role. In those few remaining games only a very small fraction will be determined by that advantage. At worst maybe 0.001% of all games will see an outcome changed by this feature.
Part of this is the lackluster nature of the advantages themselves. Rengar is practically guaranteed to have full stacks on his Bonetooth Necklace in a reasonably close game so long as he doesn't constantly run into the enemy team on his own and die without any assists; the advantage conferred by gaining the head of Kha'zix only applies when Rengar is very far behind (at which point it is unlikely to swing the game, its bonuses are good but not that good). Similarly, Kha'zix has two evolutions which are interesting but extremely situational, his W and R. Both are largely countered by wards and Oracle's, and even when they aren't they require very specific situations to make a big difference. While it's fun to have permanent stacks and that fourth evolution, they simply don't have the impact many players assume.
All of that said, there are some strong points to be made against this design decision. This secret passive is different from all others in that it confers a tangible advantage, however rarely it will actually determine a game. Players who are not aware of the special interaction will actually have a measurable disadvantage against those who are. This is unprecedented in the history of the game, and skirts the lines of what is an acceptable burden of knowledge.
The magnitude of this new secret passive is also many times greater than all its predecessors. All previous secret passives were either pure fluff or so mathematically unlikely to change an outcome that perhaps one game in the entire history of League of Legends will ever be determined by them. With hundreds of thousands of games being played daily, even small percentages can become meaningful. The "worst case" I outlined before could easily lead to hundreds of games every day where this interaction plays a significant role (though the chance of you being in that game is obviously very small).
It's a neat interaction, and it tickles my obsessive-compulsive tendencies to be able to get all four evolutions in a single game on Kha'zix. At the same time, it sets a dangerous precedent for secret passives. Were I in Riot's position I wouldn't even think of implementing further interactions for at least a half-year or more as the effects of this are sorted out.
My own thoughts are divided. I see reasons both for and against this special interaction, and at any given time I lean one way or the other. It's a complicated issue that treads new ground, and it can be very hard to predict how these issues might affect the game in the long run.
A large number of players assert that the advantage conferred by this interaction is overpowering, but I contend that the advantage is both irrelevant to the vast majority of games and marginal even in games where it plays a role. Most games won't feature both Rengar and Kha'zix; assuming each champion appears in 25% of all games (a very high rate of play) only 6.25% of all games will have both. The majority of those remaining games will have been decided well before Kha'zix ever hits 16 and some will not even feature the Bonetooth Necklace (e.g. any game with AP Rengar), leaving only 1% of all games or fewer where the advantage even plays a role. In those few remaining games only a very small fraction will be determined by that advantage. At worst maybe 0.001% of all games will see an outcome changed by this feature.
Part of this is the lackluster nature of the advantages themselves. Rengar is practically guaranteed to have full stacks on his Bonetooth Necklace in a reasonably close game so long as he doesn't constantly run into the enemy team on his own and die without any assists; the advantage conferred by gaining the head of Kha'zix only applies when Rengar is very far behind (at which point it is unlikely to swing the game, its bonuses are good but not that good). Similarly, Kha'zix has two evolutions which are interesting but extremely situational, his W and R. Both are largely countered by wards and Oracle's, and even when they aren't they require very specific situations to make a big difference. While it's fun to have permanent stacks and that fourth evolution, they simply don't have the impact many players assume.
All of that said, there are some strong points to be made against this design decision. This secret passive is different from all others in that it confers a tangible advantage, however rarely it will actually determine a game. Players who are not aware of the special interaction will actually have a measurable disadvantage against those who are. This is unprecedented in the history of the game, and skirts the lines of what is an acceptable burden of knowledge.
The magnitude of this new secret passive is also many times greater than all its predecessors. All previous secret passives were either pure fluff or so mathematically unlikely to change an outcome that perhaps one game in the entire history of League of Legends will ever be determined by them. With hundreds of thousands of games being played daily, even small percentages can become meaningful. The "worst case" I outlined before could easily lead to hundreds of games every day where this interaction plays a significant role (though the chance of you being in that game is obviously very small).
It's a neat interaction, and it tickles my obsessive-compulsive tendencies to be able to get all four evolutions in a single game on Kha'zix. At the same time, it sets a dangerous precedent for secret passives. Were I in Riot's position I wouldn't even think of implementing further interactions for at least a half-year or more as the effects of this are sorted out.
Monday, September 24, 2012
Quote: On Understanding
"I do not want the skill that passeth understanding. I want the understanding which bringeth skill."
-Helen Keller
-Helen Keller
Thursday, September 20, 2012
Quote: On Learning to Jungle
"By three methods we may learn jungling: First, by reflection, which is noblest; Second, by imitation, which is easiest; and third by experience, which is the bitterest."
-Confucius
-Confucius
Wednesday, September 19, 2012
Hotfix: World Championship
Riot is releasing some last minute hotfixes to prepare for the World Championships. Most of them are bug fixes or quality of life improvements, but there are some actual balance changes that warrant attention.
The buff to Caitlyn's Q is notable in that it increases her ability to clear waves and her damage in team fights. It's not huge, but it's significant. The change to her primary escape simply brings it in line with similar abilities, though it does constitute a nerf. Overall neither of these changes should dramatically affect Caitlyn.
Jayce's Hammer Form E contributes a lot to his laning strength, so it's no surprise that Riot targeted it for a nerf. The increase in mana cost is a conservative change, suggesting Riot believes Jayce's problem isn't damage but mana-independance. Small nerfs such as these typically test the waters; more nerfs will follow if Jayce continues to dominate top lanes.
Miss Fortune's buffs carefully target the weak areas of her kit. Strut isn't bad, but it does take forever to ramp it back up after a hit. With this change it will take a reasonable 10 seconds, down from 20, to hit full speed. Lowering the windup time on her ultimate is the bigger of the two buffs. It allows Miss Fortune to react in a timely manner to AoE initiation. Previously her ult would often be interrupted before firing a single wave, or would simply arrive too late to be meaningful.
Rengar received the greatest number of changes, though a significant number solely affect his quality of life. The damage increase on his Q is largely inconsequential, but the attack speed change is not. Because Q's attack speed buff didn't apply to his first attack he lost at least a third of the effect every time, severely dampening the buff's strength. The AP ratio on Rengar's W has been nerfed, with good reason. Being able to spam W twice every 7 seconds was extremely deadly when its ratio was 1.0. That may be the end of AP Rengar, but I think with all the other changes the more obvious bruiser style will do quite well.
As a small tangent, allowing Rengar's W to be cast while moving represents the latest design fad at Riot. Starting with Skarner Riot has favored making point blank AoEs castable while moving, including the mechanic with many new champions and applying it to old champions as well. I predict it's only a matter of time before Akali's E gets the same treatment.
Getting back on topic, Yorick's ghouls got nerfed during the early game. Players familiar with how sluggish Malzahar's voidlings were at launch will instantly recognize how big of a difference this hotfix will make. It will be fairly trivial for players to outrun Yorick's ghouls during the early phases, and at the same time it will cost more mana for Yorick to spawn the critical ghoul that slows. These changes should be a big step in bringing Yorick's laning to a sane level.
Those are the major/notable changes in the hotfix. I'm looking forward to giving Rengar a second look. He always seemed fun, but simply had a number of unfortunate bugs and quality of life issues which all appear to have been addressed at this point.
The buff to Caitlyn's Q is notable in that it increases her ability to clear waves and her damage in team fights. It's not huge, but it's significant. The change to her primary escape simply brings it in line with similar abilities, though it does constitute a nerf. Overall neither of these changes should dramatically affect Caitlyn.
Jayce's Hammer Form E contributes a lot to his laning strength, so it's no surprise that Riot targeted it for a nerf. The increase in mana cost is a conservative change, suggesting Riot believes Jayce's problem isn't damage but mana-independance. Small nerfs such as these typically test the waters; more nerfs will follow if Jayce continues to dominate top lanes.
Miss Fortune's buffs carefully target the weak areas of her kit. Strut isn't bad, but it does take forever to ramp it back up after a hit. With this change it will take a reasonable 10 seconds, down from 20, to hit full speed. Lowering the windup time on her ultimate is the bigger of the two buffs. It allows Miss Fortune to react in a timely manner to AoE initiation. Previously her ult would often be interrupted before firing a single wave, or would simply arrive too late to be meaningful.
Rengar received the greatest number of changes, though a significant number solely affect his quality of life. The damage increase on his Q is largely inconsequential, but the attack speed change is not. Because Q's attack speed buff didn't apply to his first attack he lost at least a third of the effect every time, severely dampening the buff's strength. The AP ratio on Rengar's W has been nerfed, with good reason. Being able to spam W twice every 7 seconds was extremely deadly when its ratio was 1.0. That may be the end of AP Rengar, but I think with all the other changes the more obvious bruiser style will do quite well.
As a small tangent, allowing Rengar's W to be cast while moving represents the latest design fad at Riot. Starting with Skarner Riot has favored making point blank AoEs castable while moving, including the mechanic with many new champions and applying it to old champions as well. I predict it's only a matter of time before Akali's E gets the same treatment.
Getting back on topic, Yorick's ghouls got nerfed during the early game. Players familiar with how sluggish Malzahar's voidlings were at launch will instantly recognize how big of a difference this hotfix will make. It will be fairly trivial for players to outrun Yorick's ghouls during the early phases, and at the same time it will cost more mana for Yorick to spawn the critical ghoul that slows. These changes should be a big step in bringing Yorick's laning to a sane level.
Those are the major/notable changes in the hotfix. I'm looking forward to giving Rengar a second look. He always seemed fun, but simply had a number of unfortunate bugs and quality of life issues which all appear to have been addressed at this point.
Tuesday, September 18, 2012
Thought: Syndra and Malzahar
Yesterday I talked about winrates and how they do not necessarily indicate a champion is overpowered or underpowered. While that is correct, large deviations from the average typically indicate a problem.
Today Syndra's winrate has risen to 30%, up from 25%. This is 20% below the average. To an extent one expects technically challenging champions to start out below-average and work their way upwards. However, even Rengar had a 37% winrate at worst, and that included a bugged, unreliable snare. You could argue Syndra is more technical than Rengar, but to that large an extent? Not likely.
There isn't one problem which defines Syndra, but a plethora of small issues. Aiming her E to push her spheres into targets isn't easy, and is a critical skill for asserting lane control and sieging turrets. The substantial delay between the command to throw a sphere or minion to a location and its arrival is also an issue. Her mana costs across the board are hefty, with players claiming even Chalice is insufficient to solve her mana problems. To cap it off the range on Syndra's ultimate is dramatically shorter than all of her other abilities, a mismatch which cripples her. Taken all together, Syndra is a difficult champion to play for very little reward (aside from buff grabbing hijinks).
Yet, if LolKing or Elobuff's data went back two years one could find a champion who was even worse than Syndra. Malzahar, everyone's favorite prophet of doom, had the most underwhelming skillset of any champion ever launched.
The issues affecting Malzahar were many. His Voidlings moved significantly slower and attacked whatever they pleased, which was rarely anything useful. His mana costs were higher, making him even more reliant on his Malefic Visions to maintain his mana pool. His ultimate's range was pathetic, an amazingly bad 475. To add insult to injury his ultimate's duration, and by proxy damage, were shortened by Mercury's Treads and completely countered by Cleanse. Malzahar was utterly horrible.
Malzahar was so bad that in the patches that followed he received a total of fifteen buffs, with one small nerf reducing the effect of a previous buff. It was half a year after those buffs, and a substantial shift in the meta, that he saw any further changes. Since then he's received as many buffs as he has nerfs, making it clear just how underpowered he was initially.
Syndra reminds me of Malzahar because her issues are similar in nature, though not as severe. I expect to see changes to the responsiveness of her abilities, the range of her ultimate, and her mana costs in the near future, mirroring many of Malzahar's early buffs. She's also very fun despite her issues, which was the case for Malzahar as well.
Today Syndra's winrate has risen to 30%, up from 25%. This is 20% below the average. To an extent one expects technically challenging champions to start out below-average and work their way upwards. However, even Rengar had a 37% winrate at worst, and that included a bugged, unreliable snare. You could argue Syndra is more technical than Rengar, but to that large an extent? Not likely.
There isn't one problem which defines Syndra, but a plethora of small issues. Aiming her E to push her spheres into targets isn't easy, and is a critical skill for asserting lane control and sieging turrets. The substantial delay between the command to throw a sphere or minion to a location and its arrival is also an issue. Her mana costs across the board are hefty, with players claiming even Chalice is insufficient to solve her mana problems. To cap it off the range on Syndra's ultimate is dramatically shorter than all of her other abilities, a mismatch which cripples her. Taken all together, Syndra is a difficult champion to play for very little reward (aside from buff grabbing hijinks).
Yet, if LolKing or Elobuff's data went back two years one could find a champion who was even worse than Syndra. Malzahar, everyone's favorite prophet of doom, had the most underwhelming skillset of any champion ever launched.
The issues affecting Malzahar were many. His Voidlings moved significantly slower and attacked whatever they pleased, which was rarely anything useful. His mana costs were higher, making him even more reliant on his Malefic Visions to maintain his mana pool. His ultimate's range was pathetic, an amazingly bad 475. To add insult to injury his ultimate's duration, and by proxy damage, were shortened by Mercury's Treads and completely countered by Cleanse. Malzahar was utterly horrible.
Malzahar was so bad that in the patches that followed he received a total of fifteen buffs, with one small nerf reducing the effect of a previous buff. It was half a year after those buffs, and a substantial shift in the meta, that he saw any further changes. Since then he's received as many buffs as he has nerfs, making it clear just how underpowered he was initially.
Syndra reminds me of Malzahar because her issues are similar in nature, though not as severe. I expect to see changes to the responsiveness of her abilities, the range of her ultimate, and her mana costs in the near future, mirroring many of Malzahar's early buffs. She's also very fun despite her issues, which was the case for Malzahar as well.
Monday, September 17, 2012
Thought: Winrates and Balance
It's becoming extremely popular to judge the value of a champion by their winrates (as recorded by LolKing or Elobuff). Champions with winrates below 50% are judged to be underpowered, while those above must be overpowered. This idea seems logical enough, but it has a few flaws.
The primary problem is that the League of Legends community is both stubborn and fickle. The community is quick to label flavor of the month champions as overpowered, underplayed champions as underpowered, and throws new champions into either category based on initial impressions alone. It generally takes months for the community to changes its mind, and even then the process is slow and begrudging. Only in extreme cases, such as the abusive Nunu top build pioneered by SmashGizmo, is there a sudden flip in perceptions.
As a result, champions or meta-shifts that might counter those with high winrates often go unnoticed. When Nautilus was released he was considered a subpar jungler, unable to keep up with Shyvana, Mundo, and Udyr, the titans of farming. It took three months for him to slowly shift from being underpowered and ignored to a first-ban meta-breaking ganking machine. Nautilus almost single-handedly shifted the jungle meta toward tanky, gank-heavy initiators, where it has remained ever since.
Before Nautilus everyone had simply assumed that Shyvana, Mundo, and Udyr were overpowered farmers with no counters. While LolKing and Elobuff weren't around in those days, it's very likely we would have seen all three with above-average winrates. Today, long after Nautilus received nerfs and faded from the spotlight, those three champions still sport below-average winrates. Defeating them only required a simple shift in community awareness and the meta.
This sort of obliviousness is pervasive, and means we cannot take winrates at face value. There is a correlation between winrates and a champion's strength, but that isn't the only possible cause. It is necessary to account for the community's blindspot and carefully the role the meta, typical champion picks, and playstyles play in determining a champion's strength. Winrates, like all data, are at their most useful when appraised carefully.
The primary problem is that the League of Legends community is both stubborn and fickle. The community is quick to label flavor of the month champions as overpowered, underplayed champions as underpowered, and throws new champions into either category based on initial impressions alone. It generally takes months for the community to changes its mind, and even then the process is slow and begrudging. Only in extreme cases, such as the abusive Nunu top build pioneered by SmashGizmo, is there a sudden flip in perceptions.
As a result, champions or meta-shifts that might counter those with high winrates often go unnoticed. When Nautilus was released he was considered a subpar jungler, unable to keep up with Shyvana, Mundo, and Udyr, the titans of farming. It took three months for him to slowly shift from being underpowered and ignored to a first-ban meta-breaking ganking machine. Nautilus almost single-handedly shifted the jungle meta toward tanky, gank-heavy initiators, where it has remained ever since.
Before Nautilus everyone had simply assumed that Shyvana, Mundo, and Udyr were overpowered farmers with no counters. While LolKing and Elobuff weren't around in those days, it's very likely we would have seen all three with above-average winrates. Today, long after Nautilus received nerfs and faded from the spotlight, those three champions still sport below-average winrates. Defeating them only required a simple shift in community awareness and the meta.
This sort of obliviousness is pervasive, and means we cannot take winrates at face value. There is a correlation between winrates and a champion's strength, but that isn't the only possible cause. It is necessary to account for the community's blindspot and carefully the role the meta, typical champion picks, and playstyles play in determining a champion's strength. Winrates, like all data, are at their most useful when appraised carefully.
Friday, September 14, 2012
Thought: Season 2 Rankings
The cutoffs for the final brackets for Season 2 have been revealed. Of note is the addition of a diamond ranking beyond platinum for the best of the best, and the use of top rating instead of current rating.
The latter of these changes addresses an issue that arose at the end of Season 1. Because the season kept getting extended, players who had achieved rankings had to fight with Elo decay to maintain them. Each extension caught players unaware and unprepared, resulting in lots of confusion and irritation. By making top rating the key to a ranking, players need not fret about Elo decay once they've achieved their goal.
However, this change also allows players to use friends or Elo boosting services to push their accounts into brackets they have not earned. It doesn't matter that the boosted player's Elo will dramatically crash shortly thereafter, they'll have achieved a gold/platinum/diamond top rating regardless. It should surprise no one that this possibility has many players up in arms.
I don't begrudge anyone for wanting rewards, but these sorts of shortcuts cheapen the accomplishments of others. While nothing can change the fact that it was my own skill which carried me to platinum, the context surrounding that fact can be changed. The value of my accomplishment is muddled when other players abuse the system.
How widespread a problem this will become I don't know. It won't keep me from shooting for diamond, but it gives me pause as I consider how people will view rankings in Season 3.
The latter of these changes addresses an issue that arose at the end of Season 1. Because the season kept getting extended, players who had achieved rankings had to fight with Elo decay to maintain them. Each extension caught players unaware and unprepared, resulting in lots of confusion and irritation. By making top rating the key to a ranking, players need not fret about Elo decay once they've achieved their goal.
However, this change also allows players to use friends or Elo boosting services to push their accounts into brackets they have not earned. It doesn't matter that the boosted player's Elo will dramatically crash shortly thereafter, they'll have achieved a gold/platinum/diamond top rating regardless. It should surprise no one that this possibility has many players up in arms.
I don't begrudge anyone for wanting rewards, but these sorts of shortcuts cheapen the accomplishments of others. While nothing can change the fact that it was my own skill which carried me to platinum, the context surrounding that fact can be changed. The value of my accomplishment is muddled when other players abuse the system.
How widespread a problem this will become I don't know. It won't keep me from shooting for diamond, but it gives me pause as I consider how people will view rankings in Season 3.
Thursday, September 13, 2012
Patch Day: Syndra
There are absolutely zero balance changes in today's patch, unless you count sunglasses as balance. This leaves Syndra as the most important point of interest for the patch.
Syndra's design has changed slightly over the course of the PBE, in ways that have impressed me considerably. Initially her spheres lasted longer, and their lifespan was increased by grabbing them. With CDR it was feasible to have six spheres in the field, though it required considerable setup. This presented a significant balance issue, as balancing her ultimate around this ideal would have severely crippled it in more practical use cases.
In response Riot made a few clever changes. First they increased the cooldown of Syndra's Q slightly, slowing her ability to generate spheres. Next, they made it so that her W only extended the life of spheres that were already near death. Finally, they reduced the lifespan of spheres slightly. The net result is that the maximum number of spheres is three with both 0% and 40% CDR. The elegance of the solution is the creation of a limit on the number of spheres through emergent properties rather than hard limits.
That said, I'm still mad she isn't Kassadin's daughter.
Syndra's design has changed slightly over the course of the PBE, in ways that have impressed me considerably. Initially her spheres lasted longer, and their lifespan was increased by grabbing them. With CDR it was feasible to have six spheres in the field, though it required considerable setup. This presented a significant balance issue, as balancing her ultimate around this ideal would have severely crippled it in more practical use cases.
In response Riot made a few clever changes. First they increased the cooldown of Syndra's Q slightly, slowing her ability to generate spheres. Next, they made it so that her W only extended the life of spheres that were already near death. Finally, they reduced the lifespan of spheres slightly. The net result is that the maximum number of spheres is three with both 0% and 40% CDR. The elegance of the solution is the creation of a limit on the number of spheres through emergent properties rather than hard limits.
That said, I'm still mad she isn't Kassadin's daughter.
Wednesday, September 12, 2012
Thought: Splash Art
One of the great mysteries of childhood is how a bossy, primadonna sister who never once gave a thought to the incredible impositions she placed upon her victimbrother changes with age and comes to trust and value the opinion of the younger sibling. My sister is one such person, and as a result of this change I've been asked to critique her art on numerous occasions. This doesn't make me an expert, but I like to think that I'm reasonably cultured.
Over the past few weeks the PBE has been updated with new splash arts for a few champions with outdated or problematic art, and some of them have caused a bit of a stir. In particular I'd like to focus on Lux's new splash as well as Morgana's and Miss Fortune's new splashes. I've included the full links so you can view everything at your leisure.
The reactions to Lux's new splash art have been thoroughly divided, critics seem to either love the new art or hate it. While to an extent there were similar reactions to the new Morgana and Miss Fortune splash arts, they were significantly more muted. I believe there are a number of important factors at play which have caused the polarization.
First and foremost, consider Lux's current splash art. The style comes from a very different era in Riot's artistic direction, and lacks the crispness and detail that more recent champions and skins feature. Simply comparing classic Lux to Imperial Lux makes the difference plain. As a result the current splash is, for many players, superior in large part because it doesn't look dated.
In addition, it's hard to deny that the new splash art has a number of prominent features that are very well executed. The prismatic light behind Lux is stunning, and affects the background in interesting ways. The slight translucence of her shawl is very well done, and her armor and garments are engraved with excellent detail. A great deal of skill was obviously employed in the art's creation.
Where the splash art falters, and what causes many players to stumble, is in aspects of Lux herself. Her hair and face are somewhat washed out, causing her to have a distinct "Barbie" feel. More importantly, it's hard to imagine what sort of magical diet is required to maintain that figure, and what sort of Demacian yoga she must do to be comfortable in that position. Her waist is roughly half the size of her hips or twice the width of her wrists, and her posture is awkward (particularly around the neck and her right wrist). For many players these flaws distract from the other qualities of the art and cause them to dislike it.
The Morgana and Miss Fortune splash arts are similarly very well executed, but because they have fewer distractions or flaws they failed to draw quite the same reaction as Lux. The only major flaw in the Morgana art is her surprising lack of kidneys (an issue she shares with Lux), but she tends to get a little extra slack as she's not exactly human. Miss Fortune might also have a similar problem, but because the view point is from her side it isn't as obvious and her hips aren't wide enough to accentuate the potential issue. Having only one notable, and either explicable or obtuse, flaw saves these two skins from the polarized reactions to Lux.
Part of what caused our heroines to evolve oddly thin waists, and many players to ignore them, are the same forces that sparked outrage when Arcade Sona's splash art appeared with significantly smaller breasts than her other skins. A significant portion of the playerbase consists of players who, due to hormones and/or cultural norms, tend to obsess over the unrealistic standards of beauty we expect from women and girls. While I'm sure there were some well-meaning players who simply balk at inconsistency, I would be shocked if "big breasts good, small breasts bad" wasn't the underlying motivation for most of the uproar.
Similarly, the impossible anatomical proportions of these skins are overlooked because they are what our culture expects from women, even though Photoshop is required to achieve them. Even though the splash arts display a great deal of skill, they still fall prey to the cultural demand for impractical figures.
Despite Runeterra being a fantasy, part of the appeal of such worlds is the idea of living there amongst the heroes and villains we control on the battlefield. Lux is no Riven, but when the proportions and pose of a splash art break suspension of disbelief it becomes harder to place ourselves in that world.
To be honest, I don't think the new art is beyond salvaging. Much like the original Leona splash art, only a few, relatively minor modifications should shift it from oddly uncomfortable to strong and compelling. That said I don't imagine players will care all that much a week or so after it's released, there are far more problematic splash arts (Tango Evelyn anyone?) which no one brings up save for a quick joke.
Over the past few weeks the PBE has been updated with new splash arts for a few champions with outdated or problematic art, and some of them have caused a bit of a stir. In particular I'd like to focus on Lux's new splash as well as Morgana's and Miss Fortune's new splashes. I've included the full links so you can view everything at your leisure.
The reactions to Lux's new splash art have been thoroughly divided, critics seem to either love the new art or hate it. While to an extent there were similar reactions to the new Morgana and Miss Fortune splash arts, they were significantly more muted. I believe there are a number of important factors at play which have caused the polarization.
First and foremost, consider Lux's current splash art. The style comes from a very different era in Riot's artistic direction, and lacks the crispness and detail that more recent champions and skins feature. Simply comparing classic Lux to Imperial Lux makes the difference plain. As a result the current splash is, for many players, superior in large part because it doesn't look dated.
In addition, it's hard to deny that the new splash art has a number of prominent features that are very well executed. The prismatic light behind Lux is stunning, and affects the background in interesting ways. The slight translucence of her shawl is very well done, and her armor and garments are engraved with excellent detail. A great deal of skill was obviously employed in the art's creation.
Where the splash art falters, and what causes many players to stumble, is in aspects of Lux herself. Her hair and face are somewhat washed out, causing her to have a distinct "Barbie" feel. More importantly, it's hard to imagine what sort of magical diet is required to maintain that figure, and what sort of Demacian yoga she must do to be comfortable in that position. Her waist is roughly half the size of her hips or twice the width of her wrists, and her posture is awkward (particularly around the neck and her right wrist). For many players these flaws distract from the other qualities of the art and cause them to dislike it.
The Morgana and Miss Fortune splash arts are similarly very well executed, but because they have fewer distractions or flaws they failed to draw quite the same reaction as Lux. The only major flaw in the Morgana art is her surprising lack of kidneys (an issue she shares with Lux), but she tends to get a little extra slack as she's not exactly human. Miss Fortune might also have a similar problem, but because the view point is from her side it isn't as obvious and her hips aren't wide enough to accentuate the potential issue. Having only one notable, and either explicable or obtuse, flaw saves these two skins from the polarized reactions to Lux.
Part of what caused our heroines to evolve oddly thin waists, and many players to ignore them, are the same forces that sparked outrage when Arcade Sona's splash art appeared with significantly smaller breasts than her other skins. A significant portion of the playerbase consists of players who, due to hormones and/or cultural norms, tend to obsess over the unrealistic standards of beauty we expect from women and girls. While I'm sure there were some well-meaning players who simply balk at inconsistency, I would be shocked if "big breasts good, small breasts bad" wasn't the underlying motivation for most of the uproar.
Similarly, the impossible anatomical proportions of these skins are overlooked because they are what our culture expects from women, even though Photoshop is required to achieve them. Even though the splash arts display a great deal of skill, they still fall prey to the cultural demand for impractical figures.
Despite Runeterra being a fantasy, part of the appeal of such worlds is the idea of living there amongst the heroes and villains we control on the battlefield. Lux is no Riven, but when the proportions and pose of a splash art break suspension of disbelief it becomes harder to place ourselves in that world.
To be honest, I don't think the new art is beyond salvaging. Much like the original Leona splash art, only a few, relatively minor modifications should shift it from oddly uncomfortable to strong and compelling. That said I don't imagine players will care all that much a week or so after it's released, there are far more problematic splash arts (Tango Evelyn anyone?) which no one brings up save for a quick joke.
Tuesday, September 11, 2012
Team Composition: Protect
"Protect Comp" (AKA "4 Protects 1" or "Protect the Carry Comp") is a very simple concept: Build a team that empowers and protects a strong carry. Half a year ago the composition was fairly popular, but it fell out of favor for a number of reasons.
The principle problem with the composition was its single point of failure. Every champion that wasn't the carry was almost wholly dedicated to buffing and protecting the carry. If the critical carry failed their lane, got caught out of position, or was too far afield the team was toothless. Investing so heavily in a single champion limited the team's presence, and if something went wrong it was often a catastrophe.
Changes in the meta was another big factor. The shift away from farming junglers to ganking junglers made ensuring a safe early game for the carry harder. At the same time players started using AD carries who were stronger early, especially in lane. Offensive supports also rose in popularity. In essence, the game conditions which allowed some assurance that your carry would reach the mid-game safely and with good farm disappeared.
Despite these problems it's still possible to make the composition work, but it requires a markedly different approach. Before it was normal to pick champions for little more than their ability to buff or protect the carry. Today champions like Gangplank or Warwick don't bring enough to the team to be worthwhile except in very specific cases. While every champion in the composition should bring some benefit to the carry, it's necessary that they have other compelling reasons for their presence.
Protect Comp Champion Picks:
When playing Protect Comp items like Aegis of the Legion and Zeke's Herald are absolute musts. Anything and everything that increases the potency and defense of your carry is extremely useful. The often ignored Locket of the Iron Solari becomes significantly more potent when your carry has a mix of defensive steroids on them. The summoner spell Heal also can be significant.
Beware of Olaf and Poppy, they are unique in their ability to completely bypass your team's crowd control effects. While they may not be able to outright kill your carry if your team is coordinated, the pressure they apply and the time they buy their team can be critical. Shut them down early and often.
Protect Comp is more demanding than many other compositions, but can be potent when practiced. If the best player on your team is your AD, give it a whirl.
The principle problem with the composition was its single point of failure. Every champion that wasn't the carry was almost wholly dedicated to buffing and protecting the carry. If the critical carry failed their lane, got caught out of position, or was too far afield the team was toothless. Investing so heavily in a single champion limited the team's presence, and if something went wrong it was often a catastrophe.
Changes in the meta was another big factor. The shift away from farming junglers to ganking junglers made ensuring a safe early game for the carry harder. At the same time players started using AD carries who were stronger early, especially in lane. Offensive supports also rose in popularity. In essence, the game conditions which allowed some assurance that your carry would reach the mid-game safely and with good farm disappeared.
Despite these problems it's still possible to make the composition work, but it requires a markedly different approach. Before it was normal to pick champions for little more than their ability to buff or protect the carry. Today champions like Gangplank or Warwick don't bring enough to the team to be worthwhile except in very specific cases. While every champion in the composition should bring some benefit to the carry, it's necessary that they have other compelling reasons for their presence.
Protect Comp Champion Picks:
- Top Lane:
- Lee Sin: Lee Sin is extremely versatile both offensively and defensively. Between his shield, Attack Speed debuff, and ultimate he can do a lot to protect a carry. At the same time he's can be a threat in his own right.
- Kayle: It goes without saying that three seconds of invulnerability is huge for protecting a carry. That Kayle can potentially serve as a second AD, a ranged bruiser, or even AP support helps her remain relevant against many compositions and in many situations.
- Yorick: The "gentleman's ban" is one of the best picks for a Protect Comp. His ultimate greatly increases your carry's damage through the clone, and in the worst case gives them additional time to fulfill their role. Yorick is also extremely self-sufficient and rarely loses top lane outright.
- Shen: When he isn't banned, Shen is an incredibly strong pick. The ability to protect the carry all the way from top lane, shield them in team fights, and taunt dangerous enemies makes him exceptionally strong. He's also a ninja, what more could you want?
- Nunu: This increasingly popular top lane bully is an excellent choice. Nunu brings the incredibly powerful Bloodboil buff to the table, and with all the farm top lane is a huge threat when he ults.
- AP Carry:
- Galio: Bulwark is a ridiculous 90 Armor and Magic Resistance. Combined with shields and heals the amount of survivability you can give to a carry is immense. That he also brings a speed buff, a slow, and an extremely strong ultimate is just icing on the cake.
- Orianna: Orianna's Command: Protect combines a shield with an Armor and Magic Resistance buff. While not as strong as Galio's Bulwark, it's a significant tool for protecting a carry. Orianna herself is a strong AP carry, and her other abilities can also be used to hold a team at bay.
- Morgana: Black Shield is incredibly powerful, making all crowd control effects worthless for its duration. Combined with other shields and defensive buffs your carry becomes an unstoppable force. Her ability to catch targets and control the entire enemy team are other great assets.
- Mordekaiser: An odd but honorable mention. Mordekaiser can make up for your team's singular damage source with his ultimate, and help protect your carry with an additional defensive steroid. He's not an ideal pick, but serviceable.
- AD Carry:
- Kog'maw: Even though his long range harass with his ultimate has been nerfed, his straight up auto-attack damage is still fantastic.
- Tristana: She can be hard enough to deal with without an entire team making her as indestructible as any tank. Her weak mid-game is her primary issue.
- Twitch: His ability to damage the entire enemy team at once is incredibly strong, and his Expunge is no joke at full stacks. It's also extremely hard to gank him without Oracle's or pink wards.
- Draven: With two axes spinning Draven out-damages every other carry in the game. The only issue is finding someone with sufficient mechanical skill.
- Support:
- Lulu: Without any AP Lulu can provide a massive burst of 840 health and shielding irreducible by Ignite or similar effects. Adding this massive boost to strong defensive steroids will cause a carry's survivability to skyrocket. That she has a slow and a polymorph effect makes her an extremely strong pick.
- Janna: Janna's combination shield and AD steroid tops off a toolkit that includes some of the best crowd control in the game.
- Nunu: When he isn't top, Nunu is a perfectly good support option for this team composition. He lacks the hard crowd control and shielding of many of his peers, but makes up for it with his ridiculous Bloodboil buff.
- Sona: Her auras boost both the offense and defense of your carry, and she has the single most game-changing ultimate of any support.
- Soraka: While her healing can be countered by Ignite, the sheer volume of her healing combined with her super safe laning phase and the strength of her Armor steroid makes her a decent choice.
- Jungle:
- Maokai: Maokai's ultimate is hands down one of the best in the game for protecting a carry, especially one that's going to receive multiple shields, heals, and defensive boosts. He also brings considerable crowd control and excellent jungling to the table.
- Lee Sin: When he's not dominating top lane, Lee Sin is a fantastic jungler and can still fulfill his role in protecting a carry.
- Nunu: Bloodboil is so good it gets Nunu on this list three times.
- Jarvan: While his flag doesn't boost your carry quite as much as other effects, the power of his ultimate in containing threats, his knockup, and his ability to reduce the armor of multiple targets makes him strong.
- Kayle: When top lane isn't an option, Kayle can do just as well in the jungle.
When playing Protect Comp items like Aegis of the Legion and Zeke's Herald are absolute musts. Anything and everything that increases the potency and defense of your carry is extremely useful. The often ignored Locket of the Iron Solari becomes significantly more potent when your carry has a mix of defensive steroids on them. The summoner spell Heal also can be significant.
Beware of Olaf and Poppy, they are unique in their ability to completely bypass your team's crowd control effects. While they may not be able to outright kill your carry if your team is coordinated, the pressure they apply and the time they buy their team can be critical. Shut them down early and often.
Protect Comp is more demanding than many other compositions, but can be potent when practiced. If the best player on your team is your AD, give it a whirl.
Monday, September 10, 2012
Jungling: Movement Speed
Movement Speed is incredibly valuable to any jungler. Whether you aim to perpetually farm while ignoring the ire of your teammates, to counter-jungle a foolish opponent, or to gank continuously Movement Speed makes a monumental difference
It's no coincidence that nearly every competitive match starts with eight or more players sporting Boots of Speed. That 50 Movement Speed is so valuable that it can easily determine many lanes even without the influence of a jungler. Small exchanges can become outright routs, and many normally safe positions become unsafe without that seemingly small boost. The power of that Movement Speed difference is why junglers are beginning to invest so heavily in the attribute.
Increasing Movement Speed means reaching important locations sooner, and closing the distance to vital targets faster. It's difficult to comprehend just how effective Movement Speed is until you've played without it. Try jungling Cho'gath with a Regrowth Pendant and no Movement Speed runes or masteries, then try again with Boots of Speed, and then again with every Movement Speed bonus you can get. Jungling without any Movement Speed increases is simply a handicap.
It's for this reason that Mobility Boots and Movement Speed runes and masteries have become increasingly popular. Between the two a jungler moves as much as 20% faster from place to place, an incredible advantage. Whether rushing somewhere in time to counter-gank or running in to gank, that boost easily turns would-be escapes into kills.
This is also another reason why Janna is such a coveted support, especially for bruisers without gap closers. The power of that seemingly innocuous 3% Movement Speed boost is huge, and is even more painfully felt when the other team picks her. Nothing makes an Udyr cry more than to see Janna on the enemy team.
Not every jungler can take full advantage of Movement Speed. Some derive other critical benefits from runes, masteries or boot upgrades that preclude zooming around the map at Mach 5. For those that can, such as Cho'gath or Maokai, it simply amazing. Try it once and you'll find it very difficult to go back.
It's no coincidence that nearly every competitive match starts with eight or more players sporting Boots of Speed. That 50 Movement Speed is so valuable that it can easily determine many lanes even without the influence of a jungler. Small exchanges can become outright routs, and many normally safe positions become unsafe without that seemingly small boost. The power of that Movement Speed difference is why junglers are beginning to invest so heavily in the attribute.
Increasing Movement Speed means reaching important locations sooner, and closing the distance to vital targets faster. It's difficult to comprehend just how effective Movement Speed is until you've played without it. Try jungling Cho'gath with a Regrowth Pendant and no Movement Speed runes or masteries, then try again with Boots of Speed, and then again with every Movement Speed bonus you can get. Jungling without any Movement Speed increases is simply a handicap.
It's for this reason that Mobility Boots and Movement Speed runes and masteries have become increasingly popular. Between the two a jungler moves as much as 20% faster from place to place, an incredible advantage. Whether rushing somewhere in time to counter-gank or running in to gank, that boost easily turns would-be escapes into kills.
This is also another reason why Janna is such a coveted support, especially for bruisers without gap closers. The power of that seemingly innocuous 3% Movement Speed boost is huge, and is even more painfully felt when the other team picks her. Nothing makes an Udyr cry more than to see Janna on the enemy team.
Not every jungler can take full advantage of Movement Speed. Some derive other critical benefits from runes, masteries or boot upgrades that preclude zooming around the map at Mach 5. For those that can, such as Cho'gath or Maokai, it simply amazing. Try it once and you'll find it very difficult to go back.
Friday, September 7, 2012
Mechanics: Mana and Mana Regen
Mana and Mana Regen are not particularly mystifying mechanics, but they interact in ways that spawn some interesting math.
An important concept, one that started in MMOs such as Everquest or World of Warcraft, is "Time to OOM". It's a measure, given a Rate of Mana Expenditure, of how long it will take to run out of Mana. The equation is fairly simple:
Time to OOM = Mana / (Rate of Mana Expenditure - Mana Regen)
From this equation we can tell that there are three ways to increase the Time to OOM. We can increase our Mana, increase our Mana Regen, or decrease our Rate of Mana Expenditure. We can also tell that Time to OOM is directly proportional to Mana; any increase in Mana linearly increases Time to OOM. If we graph Time to OOM in relation to changes in Mana, it looks like this:
So long as Mana Regen and Rate of Mana Expenditure are constant and their difference is positive this is how the graph will look. The picture looks very different, however, if Mana is kept constant and Mana Regen varies:
Mana Regen becomes increasingly effective the more of it you have. This is because Mana Regen has a recursive influence on Time to OOM. As Mana Regen increases so does Time to OOM, which means there's now more time spent regenerating Mana, which increases Time to OOM, which means there's more time to regenerate Mana etc. When there's a large difference between Mana Regen and Rate of Mana Expenditure this effect isn't all that noticeable, but as they converge it explodes.
These nice, crisp mathematical equations become grossly complicated once you attempt to factor in Blue buff or Chalice.
Blue buff mixes things up by tying Mana to Mana Regen. If we try graph Time to OOM versus Mana for a player with Blue, it's no longer linear.
Because increasing Mana also increases Mana Regen the graph curves gently upward. Extended far enough, or with a sufficiently low Rate of Mana Expenditure, we'd see a familiar asymptotic pattern emerge.
Chalice is even more complicated. It directly scales Mana Regen based on how little mana is remaining. To properly calculate this we have to derive a new equation for Time to OOM which accounts for the rate at which Mana Regen changes. Unfortunately this involves differential calculus, so if the following equation is utterly baffling you're probably in good company.
Time to OOM = -(Maximum Mana/Mana Regen) * (ln|Mana Regen - Rate of Mana Expenditure| - ln|2*Mana Regen - Rate of Mana Expenditure|)
I won't assail you with the intricacies of the math involved or how this equation was derived. Instead you can enjoy the wonderful graph this generates:
As before we can see that Time to OOM swings upward, but there are a couple of key differences between this graph and the previous comparison with Mana Regen. Both have identical Rates of Mana Expenditure (40 Mana per second), but this one begins its asymptotic swing much sooner. Chalice essentially halved the distance to the point of infinite Mana, making it the single most potent Mana Regen item in the game.
In practice it isn't necessary to reach the point of infinite Mana, and can even be counter-productive. Even the most indomitable of laners need to go back periodically to heal and buy items, and past the early-game returning to base frequently is a matter of course. As fun as being completely independent of Mana is, it's a better idea to obtain just enough Mana Regen to sustain you between recalls and invest the rest of your gold elsewhere.
At the end of the day managing a mana pool is going to be different for every champion. Many AP champions skirt a fine line where it isn't clear if items like Chalice are worthwhile or not. Applying the math above in practical situations is difficult due to the many extenuating circumstances in play, but if you can the difference in effectiveness, especially for AP champions, can be huge.
An important concept, one that started in MMOs such as Everquest or World of Warcraft, is "Time to OOM". It's a measure, given a Rate of Mana Expenditure, of how long it will take to run out of Mana. The equation is fairly simple:
Time to OOM = Mana / (Rate of Mana Expenditure - Mana Regen)
From this equation we can tell that there are three ways to increase the Time to OOM. We can increase our Mana, increase our Mana Regen, or decrease our Rate of Mana Expenditure. We can also tell that Time to OOM is directly proportional to Mana; any increase in Mana linearly increases Time to OOM. If we graph Time to OOM in relation to changes in Mana, it looks like this:
So long as Mana Regen and Rate of Mana Expenditure are constant and their difference is positive this is how the graph will look. The picture looks very different, however, if Mana is kept constant and Mana Regen varies:
Mana Regen becomes increasingly effective the more of it you have. This is because Mana Regen has a recursive influence on Time to OOM. As Mana Regen increases so does Time to OOM, which means there's now more time spent regenerating Mana, which increases Time to OOM, which means there's more time to regenerate Mana etc. When there's a large difference between Mana Regen and Rate of Mana Expenditure this effect isn't all that noticeable, but as they converge it explodes.
These nice, crisp mathematical equations become grossly complicated once you attempt to factor in Blue buff or Chalice.
Blue buff mixes things up by tying Mana to Mana Regen. If we try graph Time to OOM versus Mana for a player with Blue, it's no longer linear.
Because increasing Mana also increases Mana Regen the graph curves gently upward. Extended far enough, or with a sufficiently low Rate of Mana Expenditure, we'd see a familiar asymptotic pattern emerge.
Chalice is even more complicated. It directly scales Mana Regen based on how little mana is remaining. To properly calculate this we have to derive a new equation for Time to OOM which accounts for the rate at which Mana Regen changes. Unfortunately this involves differential calculus, so if the following equation is utterly baffling you're probably in good company.
Time to OOM = -(Maximum Mana/Mana Regen) * (ln|Mana Regen - Rate of Mana Expenditure| - ln|2*Mana Regen - Rate of Mana Expenditure|)
I won't assail you with the intricacies of the math involved or how this equation was derived. Instead you can enjoy the wonderful graph this generates:
As before we can see that Time to OOM swings upward, but there are a couple of key differences between this graph and the previous comparison with Mana Regen. Both have identical Rates of Mana Expenditure (40 Mana per second), but this one begins its asymptotic swing much sooner. Chalice essentially halved the distance to the point of infinite Mana, making it the single most potent Mana Regen item in the game.
In practice it isn't necessary to reach the point of infinite Mana, and can even be counter-productive. Even the most indomitable of laners need to go back periodically to heal and buy items, and past the early-game returning to base frequently is a matter of course. As fun as being completely independent of Mana is, it's a better idea to obtain just enough Mana Regen to sustain you between recalls and invest the rest of your gold elsewhere.
At the end of the day managing a mana pool is going to be different for every champion. Many AP champions skirt a fine line where it isn't clear if items like Chalice are worthwhile or not. Applying the math above in practical situations is difficult due to the many extenuating circumstances in play, but if you can the difference in effectiveness, especially for AP champions, can be huge.
Thursday, September 6, 2012
Quote: On Proverbs
"I have always been fond of the Solo Queue proverb 'Ignore all and carry like a boss; you will go far.'"
-Theodore Roosevelt
-Theodore Roosevelt
Wednesday, September 5, 2012
Quote: On Skill
"Skill is terribly revealing. Try and vary your masteries as you will,
your runes, your champion, your attitude of mind, and your soul is
revealed by your actions."
-Agatha Christie
-Agatha Christie
Tuesday, September 4, 2012
Thought: Little Things That Kill
There are various mathematical proofs and logical arguments for the non-existence of Elo Hell. These proofs are entirely accurate and largely ineffective; fighting personal anecdotes of injustice with math is rarely a fruitful pursuit. The purpose of this post isn't to engage in that futile debate, but to use math, logic, and a little common sense to help improve one's Elo.
The first, and most important, mathematical point to be made is simple: Every game is worth, on average, 22-28 Elo. If you win a game you are only 11-14 Elo higher than you were before, but you're 22-28 Elo higher than you would have been had you lost. Because every game you win is also a game you didn't lose, and vice versa, the stakes on an individual game are higher than most players realize.
In practice this means that any improvement in play, even a minor one, can lead to fairly large increases in Elo. If improving your last hitting leads to 5 additional wins out of 100 games played, then instead of ending that series at 1200 Elo you'll be 1310-1340 Elo. Add an additional win for learning to consistently keep Baron warded, two for knowing when you're overextended, and a couple more for timing buffs/objectives and you're 200 Elo higher than you would have been. The key to increasing your Elo is to improve your play.
The hard part is identifying one's flaws and correcting them. It's easy to tell when you CS poorly, but most other flaws are tiny, infrequent, and unquantifiable. Did you lose that fight at Dragon because of poor positioning or because you pulled Dragon? Did you pick the wrong champion or were you outplayed? Could you have saved top lane if you'd rushed that way instead of continuing to farm? These are the little things that kill, small decisions which are innocuous but critical to the outcome of a game.
The single easiest step you can take to help yourself is to install LoLReplay. With the replay files it saves you can observe your games objectively, separated from the moment to moment stresses of actual play. Flaws that might have escaped your notice during an intense match are much easier to see when you're relaxed and detached from the proceedings. Best of all, you can share these replays with friends and mentors and get their input. Skimming through your past games may not be the most exciting activity, but it's better than repeating your mistakes and turning them into habits.
As a final philosophical point, focusing on increasing Elo is a detrimental mindset. It makes it harder to identify mistakes, make observations, and learn. In the short term focusing on improvement might lead to losses, but in the long term you'll see substantial Elo gains. Whether you think your true Elo is 500 higher than your current Elo or don't care at all, improving is the surest way to move up the ladder.
The first, and most important, mathematical point to be made is simple: Every game is worth, on average, 22-28 Elo. If you win a game you are only 11-14 Elo higher than you were before, but you're 22-28 Elo higher than you would have been had you lost. Because every game you win is also a game you didn't lose, and vice versa, the stakes on an individual game are higher than most players realize.
In practice this means that any improvement in play, even a minor one, can lead to fairly large increases in Elo. If improving your last hitting leads to 5 additional wins out of 100 games played, then instead of ending that series at 1200 Elo you'll be 1310-1340 Elo. Add an additional win for learning to consistently keep Baron warded, two for knowing when you're overextended, and a couple more for timing buffs/objectives and you're 200 Elo higher than you would have been. The key to increasing your Elo is to improve your play.
The hard part is identifying one's flaws and correcting them. It's easy to tell when you CS poorly, but most other flaws are tiny, infrequent, and unquantifiable. Did you lose that fight at Dragon because of poor positioning or because you pulled Dragon? Did you pick the wrong champion or were you outplayed? Could you have saved top lane if you'd rushed that way instead of continuing to farm? These are the little things that kill, small decisions which are innocuous but critical to the outcome of a game.
The single easiest step you can take to help yourself is to install LoLReplay. With the replay files it saves you can observe your games objectively, separated from the moment to moment stresses of actual play. Flaws that might have escaped your notice during an intense match are much easier to see when you're relaxed and detached from the proceedings. Best of all, you can share these replays with friends and mentors and get their input. Skimming through your past games may not be the most exciting activity, but it's better than repeating your mistakes and turning them into habits.
As a final philosophical point, focusing on increasing Elo is a detrimental mindset. It makes it harder to identify mistakes, make observations, and learn. In the short term focusing on improvement might lead to losses, but in the long term you'll see substantial Elo gains. Whether you think your true Elo is 500 higher than your current Elo or don't care at all, improving is the surest way to move up the ladder.
Subscribe to:
Posts (Atom)