TL;DR in featured comments.
It seems apparent to me that League wasn’t designed with melee carry as a viable alternative to a Marksman. We have melee carries in the game, but they fill sort of a questionable role at the moment.
I’m not saying that’s how League should be, but in terms of why there is a must pick Marksmen in most games, the answer is tradition.
If you ask me how LeagueÂ shouldÂ be, that’s a much larger question that requires a big discussion. For starters, maybe it would be nice if players picked positions (top, middle, bottom, jungle) but not roles for those positions. Would players be ready for something like that? Is there benefit in having the tacit agreement that if you pick a Marksman, she’s probably going bot? Would you want to see games with two Marskmen or none at all? Does that sound exciting or destablizing?
Right, something like that. The current melee = top etc. setup has been in place for so long that teammates will call you out if you’re “doing it wrong” and they will certainly report it to us as something broken. (The pros are often more adventurous in trying what works rather than what is expected.)
But there are definitely benefits in having an assumed configuration. It’s easier to coordinate a pickup team for example because you can infer a lot of position intent just from their role choice. The question is whether those benefits outweigh some amount of fatigue for every game ending up melee-top, mage-mid, ADC-bottom.
I’m not saying players are ignorant and choose ADCs because they don’t know any better and cling to tradition. I’m saying Riot has ended up supporting ADC mandatory bottom because it’s a relatively stable configuration of the meta. There is no reason that has to be the only configuration. It’s just the one we have traditionally used for at least the last couple of years.
We have discussed whether we should rename the role of support to something else (“utility?”) just to avoid the potential confusion there. So “support” would be a position, as you describe, when you share a lane but don’t go for CS, and could conceivably be performed by a melee like Alistar or a mage like Annie or a “role-formerly-known-as-support” like Sona.
It may cause more disruption just trying to change it though.
Before we go on, let me ask you guys something honestly.Recall the last time you reached a Series. Whether it be from Bronze IV to Bronze III, or Diamond II to Diamond I, it doesn’t matter – just think back to that time.
Now tell me, honestly – which of these two feelings were you leaning more towards?
1)Â Super excitement at how important these next matches are!
2)Â Extreme frustration at the possibility of feeders/trollers/AFK’ers in your following promo matches.
As I recall, it has been Riot’s stance in the past that they have Promo Series between divisions because it makes certain games feel more important and more exciting.
But with the end of the season coming up, I feel like one of the big changes for S5 should beÂ removing the existence of Promotional Series.
Please, hear me out.
I honestly think the existence of Series is at the core for a lot of the games major issues. Trolls, feeders, AFKers, etc. would not be as harmful if Series didn’t exist, simply because they couldn’tÂ completelyÂ ruin your work by forcing you to lose that one important series match. Toxicity and raging would be lessened in a similar way – if you’re super-anxious and frustrated because you’ve proved you deserved to be a rank up, but have to winÂ a select few specific games in a SeriesÂ just to actually go up that rank, then you’re more likely to rage/be toxic to people who don’t do well in that game. Why? Because the stakes are so much higher, for no good reason. Because you’re not just losing more LP from that one feeding botlaner, you’re losing your entire series. You’re not just losing more LP from that one ‘Diamond I Smurf Warwick Mid’. You’re losing your entire series.
While much of the player base is capable of handling these situations without resorting to toxicity, if a player is going to rage, it’s liable to be in that moment; when all of their hard work is crushed by one other allied player, because of the way Riot’s ‘Series’ System functions.
Another benefit of removing series is that it’d help destroy the concept of ‘ELO hell’, whether it exists or not, by making transition between divisions much smoother. Currently, a player could effectively win almost every single one of his games, but still be stuck in Bronze V because he had 1 or 2 AFKers on his team in each series. This situation is highly hypothetical and unlikely, but itÂ isn’t unlikely that players are often screwed out of ranking up because of a troll/feeder/afker in their Promos.Â In fact, I’d say this probably happens a lot.
Last but not least, lag. Lag is obviously a major factor in winning or losing in a game like this, especially lately with all of the apparent DDOS attacks and major server lag spikes. The Loss Prevented system is helpful but we all know how unreliable it can be. Lose a game due to lag, lose full LP – win the next game, have your LP cut in half because of loss prevention? Those sort of issues. The biggest problem is that Loss Prevention doesn’t actually apply to series. Correct me if I’m wrong, but even if everyone in your game lags out and the servers are being DDOS attacked and Riot turns on Loss Forgiven, you’re STILL counted as having a ‘loss’ for that game in your Promo Series.
Basically, Promo Series increase the excitement of the game on a small scale, by making certain games more important. I believe Riot has specifically said this is their reason for wanting series in the game.
But that increase in ‘excitement’ comes at the cost of;
- Increased stress & anxiety
- Increased punishment from trolling teammates
- Increased punishment from AFKing teammates
- Increased punishment from feeding teammates
- Increased toxicity
- Increased rage
- Increased belief in & rage over the concept of ‘ELO hell’And finally,Â
Increased factor of ‘random chance’ in determining your placement, rather than skill.
How is it worth it?
EDIT:Â An excellent compromise has been suggested. Keeping Promo Series only between ranks, but not between divisions.Â
So, for example, there’d still be the increased stress & excitement of going from Platinum to Diamond, or even just Bronze to Silver – but you wouldn’t have to go through that stress for every division, between Bronze V, IV, III, etc.
Thanks for the feedback. A few of the reasons you called out for removing promos are issues that should be addressed across all ranked games and not just in promos. For example, whether you’re in promos or not, having somebody rage or AFK on your team isn’t a great experience. Also, lagging out in the middle of a ranked game whether you’re in promos or not totally sucks.Â
We’ve been working with the player behavior team to create a more robust system to address behavior related issues in ranked. For example, we’re testing a feature on PBE called Ranked Restrictions that blocks players with a consistently negative behavior from playing ranked games. We will have more details to share on this in the coming weeks.
I’m not able to speak to lag issues in a super educated way, but I do know there is aÂ team at work to address those issues. When we see connection issues, we enable compensation mode which forgives lost promos and lost LP, but still counts wins. This isn’t a perfect solution, but it does help shield players from a situation outside of their control.
One of the reasons you called out I disagree with. That is making ranked a more smooth experience. There is a bit more to series than just making the matches feel important and exciting. Division promotions reinforce the value of achieving competitive milestones. If we removed promotional series it would be easier to climb. In ranked easier doesn’t mean it’s strictly better though. Ranked play is about accepting the challenge of being measured by your skill. This is really what separates it from normals. Every change to make the system easier undermines what the accomplishments mean. What separates you from the players in lower tiers is not only the LP gains, but the series you fought through where you proved yourself and came out on top.Â
If you consider the example you gave where a player is at the top of Bronze V and wins consistently, then fails to win his promos repeatedly, should he really qualify for Bronze IV? Probably not. If he’s going 1-1 assuming equal gains and losses he’s at 100 LP still. If he’s going 1-2 he’s lower than 100 LP, which is why LP can be lower after failing a series (especially in the case of going 0-2). There are a also number of safeties in place to ensure if you’re more than qualified to move on to the next tier you can either skip two tiers after completing your promos, or skip them entirely. If you’re in promos and not skipping it’s because the system isn’t confident enough in your skill to move you to the next tier yet.
Ranked solo isn’t for everybody, this is why it’s an opt in experience. Some players want to know how they measure up to the competition as an individual. It exists to serve this purpose. I challenge your assumption that solo queue position doesn’t reflect skill because it’s a team game though. Players consistently climb or fall to an accurate position on the ladder even after resets or when playing on alts.Â
Not familiar with Cassiopeia’s rework?Â Here’s the changelist.
Yes it should function in the improved way you describe here. If it’s not as forgiving as it needs to be, we can make it more so or less so as necessary. But yes, this problem was in the scope of these changes ;D
Great thinking here, you bring up a lot of interesting points. The ones I’m most interested in talking about are 1) 180 degree shift in play pattern , 2) that her damage is no longer reliant on her skill shots, and 3) that the update was just tacked on to bugfixes/reliability changes.
1) We tried to pay very careful attention to this one in our brainstorming. The reason the passive grantsÂ LOTSÂ more stacks when hitting champions than minions/monsters is to preserve a lot of her old play pattern. That is, in order for lane bully Cassiopeia to be successful, she needs to land her abilities on her lane opponent carefully and frequently, and in order for scaling Cassiopeia to be successful, she needs to land her abilities on her lane opponent carefully and frequently. In terms of actual play pattern, not much has actually changed. The outputs (or, what her opponent notices) has changed a bit– Cassiopeia is less of a threat early in the game but more of a threat late game.
2) While her damage (number wise) will mostly come from her Twin Fang, she is entirely dependent on consistently landing Noxious Blast for both poison applicationÂ andÂ the movement speed to stay in range of Twin Fang. Miasma also helps her here, in much the same way.
So, where her damage is coming from in particular and at which points of the game her damage will be highÂ arechanging, I think we’re preserving much of the traditional Cassiopeia play pattern.
3) So, I can definitely see how this update comes across as a misguided attempt at making a character ‘better’ for no good reason. I assure you, this is not the case! There were problems on Cassiopeia’s kit that, while not on the level of some other champions in the game, still needed to be addressed. Shurima and her texture update provided a great opportunity to do all of these things. I wont go deeply into the problems here, but they generally involve non-interactive lane bullying patterns, game-stalling, and problem with unsatisfying or under-appreciated DoTs (though, not by all people).
I hope all of this makes sense/seems fair. I also don’t seek to invalidate the concerns you expressed, as they are all certainly valid. I just wanted to provide some context on the changes and hopefully explain that there was a reason for everything. It’s true that all of this comes at costs, and as you explained, it may have left “her feeling like not-Cassiopeia for some people.” This is probably true for some people, and I’m sorry for thatÂ . However, in the end, I think Cassiopeia will still feel a lot like Cassiopeia in many of the ways that people love, express less problems than she has historically, and have a stronger identity overall. Time will certainly tell, and I promise to be around to make right anything that goes wrong.
I could definitely see how some of the changes come across like this, but I’d have to disagree on a few points. Primarily, I think she hasÂ moreÂ of a kite pattern now than before, with her kiting tools having more power in their movementspeet related aspects and with her ult being more reliable.
I the end, it’s true, a lot of the math and comparison points for this update were ADCs, but I’m very confident that there are enough defining and distinct features in Cassiopeia’s gameplay (i.e, the way her kit comes together) to make her feel significantly different than ADCs.
As for mana, maintaining the mana to cast spells in lane should be very doable with some practice on priming minions with poison/basic attacks and last hitting them with E. If this turns out simply not to be true, then we’ll have to retune things accordingly.
I think we landed on a coherent kit that flows well, especially when you invest time into mastering it, so I encourage you to play around with it a bit. I really do hope that you enjoy it, as it definitely makes me sad to disappoint Cassiopeia playersÂ
So this is a little unclear, but the way these stack is as follows:
Aspect of the Serpent AP boost.
Archmage on top of Aspect of the Serpent AP boost.
Deathcap on top of Archmage and Aspect of the Serpent AP boosts.
So, the way the numbers break down, when both Deathcap and her passive are both giving 30% amped AP, Deathcap will actually be giving more, numerically, because it stacks on top of her passive.
Does that make sense?
It’s not a farming passive D=. Well, it does include farming but…
The minion/monster farming part of the passive functions almost entirely as a fallback pattern. Though, it does come across as being the primary stacking mechanism. I found this very difficult to solve through messaging, but I assure you, if you crunch the numbers or just play enough games on her you’ll quickly realize that farming minions/monsters in hopes of stacking your passive is MUCH less efficient than hitting enemy champions with your abilities.
The minion/monster part of it is just so that when you can’t find any champions to fight, you have something to do to work towards your end goal. If this wasn’t the case, enemy champions could just play keep away all game and you’d get reallllly frustrated. That said, this is still an effective way of slowing down Cassiopeia, but at least she won’t be full-stopped.
Cassiopeia gains a stack of Aspect of the SerpentÂ
for each second that an enemy champion is poisoned,Â
for each poisoned unit that she kills and when sheÂ
hits an enemy champion with Twin Fang,Â
up to a max. of 400 stacks.
- 75 STACKS: 5% increased ability power.
- 150 STACKS: 10% increased ability power.
- 225 STACKS: 15% increased ability power.
- 300 STACKS: 20% increased ability power.
- 350 STACKS: 25% increased ability power.
- 400 STACKS: 30% increased ability power.
At start of the game, and for each milestone she will also getÂ
an evolution point. These points can be spent on 7Â
different upgrades. She will eventually get everyÂ
upgrade, but the order is completely player dependent.
(Sort of like how Kha’Zix evolves)
- Twin Fang now heals for 6 / 8 / 10 / 12 / 16 (+ 10% AP).
- 25% bonus cooldown reduction.
- 15 (+ 2% AP) bonus flat movement speed.
- +15 HP/lvl (+ 25% AP)
- 150 bonus range to Noxious Blast
- 3 second lower cooldown to Miasma
Latent Venom – 5% of the damage from her basic abilities will be
added as Latent Venom, which will then burst 4 seconds after
the last damage taken from these abilities, refreshing with every new
instance of damage. (Imagine a refreshing Zed ult)
I have to say, this is super exciting. However, if the passive we landed on had complexity costs, boy does this one haha. It would be fun to try something more like this, but it would require implementing all sorts of new systems to work with evolving a trait from a pool of 7 and all sorts of stuff like that, which the game currently doesn’t support. Maybe this could be a cool project some time in the future, but seems unlikely for now.
That said, I like the way you think!
We actually had a version like you describe in testing, and the first version of Cassiopeia to hit PBE did this smooth scaling for the AP bonus. However, neither tested well, as the incremental increases were simply unappreciable, and other bonuses are intended to come into effect at specific points in the game and not until then. The stack mechanism and numbers that we landed on solved all this pretty well.
It’s true, things might be overtuned. However, itemizing one large many item (seraphs or RoA, say) allows you to over come this pretty handily because of how valuable they make the mana restore. As you grow your mana pool, you grow the rate at which you recover it, which should mean that both Cassiopeia has more ‘ammo’ and takes less time to ‘recharge’ as the game goes on.Â
But it’s true, we’ll have to see if the numbers are just a little off.
Yeah, this seems to be how everyone is reacting. I see that I replied to one particularly positive one, and can see how that drew some suspicion. Though, I believe the rest of my posts have largely been in response to dissenting opinions. Maybe I’m wrong here. If that’s the case, I’m really sorry, that wasn’t my intention =[
She has not been adjusted on alternate maps. We could take an awkward stab at the numbers and hope they are close, but we simply don’t have enough data on how she would perform on those maps to make an accurate guess. The levers are in place to adjust her accordingly (namely, the number for her breakpoints changing based on the map), so we should have an easy time adjusting her for the different maps as we get more information.
I totally agree with you on clarity. There should be some particles in to more accurately represent this, like a poison explosion around her when it happens and snakes spiraling around her for each tier she has hit.
A lot of players have mentioned that “indication of improvement” and Ranked as a measurement of skill might be key drivers of player growth, but I haven’t found that to be a compelling argument for a few reasons.
One, when we look at Normal Summoner’s Rift versus Ranked Summoner’s Rift, we can see that most players prefer to play Normal Summoner’s Rift (and that the queue has done really well over the course of time).
It may surprise some, but Ranked is not the the mode the majority of players play! Although I agree that some players would play Dominion more if there was progression and a measurement of skill in something like Ranked, to say that it would drive general player growth is a stretch because most players prefer not being faced with high-intensity, high-stakes games every time they play League.
In fact, depending on how you slice the data, it’s possible that removing MMR in Dominion/Normal Summoner’s Rift actuallyÂ increasedÂ the engagement of these modes because you’re removing the constant “stakes” in each match.Â
Regarding Twisted Treeline Ranked, in retrospect, it’s hard to say whether it was the right choice. As a games studio, I think it’s OK sometimes to admit mistakes, and try to learn from them. Game design and game development is really hard, and it’s pretty difficult toÂ predictÂ exactly how players will receive a new feature, system or queue. In the old days, it seemedÂ obviousÂ that we should open a Ranked version of Twisted Treeline, but did it actually damage the overall growth of Twisted? Did players get frustrated with the queue times and lopsided matches, and quit Twisted altogether? These are key questions that help us learn from the choices we make.
I intend on reading the full post, but am curious about this point. How do you explain the growth of ARAM when it was a Custom Game? This was before the blog post, before being able to queue for it, before any support at all.
This is another interesting data point. If there was a large number of Normal Draft Dominion players, you might be able to argue that most of these players would be interested in Ranked Dominion and that would be the “starting” playerbase of Ranked Dominion plus some additional group of incoming players.Â
However, Normal Draft Dominion is completely unplayed and has never had any engagement. In fact, after reviewing the data, we should probably close Normal Draft Dominion completely.
I didn’t see any times when Dominion’s population drastically rose. However, the official release of Howling Abyss did reduce the numbers on Dominion. This could suggest that players that enjoy “alternative” maps that have different pacing prefer Howling Abyss to Dominion, and that these maps fill the same need for these players. I understand that some players might think of Dominion as a competitive map, but it might not be the majority opinion. For example, I bet there are some players that think Howling Abyss is a competitive map and would prefer to play Ranked Howling Abyss too–but, it’d be a far stretch to say the majority of players think that way.
Interestingly, Team Builder hasn’t pulled much from the alternative maps like Dominion and Twisted Treeline. Team Builder has mainly split the Normal Blind Pick population, which is something we’re very sensitive to. We have some key design choices we want to make in Team Builder to make the queues as fast as Blind Pick, and want to see how these changes play out before discussing the replacement of Blind Pick. For example, we may run some experiments where we do have players only select Position and see if the experiences are just as positive, but queue times are much lower.
Yes, we’re working on a few short-term things to combat Bots in Dominion. Drevarius is a designer that has been working on that problem. As you might know, we usually don’t discuss anti-hacking or anti-botting measures because it typically is an arms race.
Introducing Ranked Dominion might translate to an increase in the population that plays Normals, but the question is what the increase might be? Secondly, would the increase in Normal Dominion players be enough to compensate and overcome the pull that Ranked Dominion would have on current Normal Dominion players?Â
As an aside, realistically, turning on new queues isn’t as simple as flipping switches. Rioters might have trivialized this in the past, but as someone directly involved in Team Builder, new queues require continuous monitoring and maintenance by a team.
Hey all, with the launch of Master Tier Diamond players may notice some changes to their LP gains and losses. Here’s what is going on in these cases:
* The top 200 challenger spots were protected by a concept called ‘clamping’ before Master Tier was introduced. This ensured the top 200 dudes on the server had both the highest LP and MMR. This greatly slowed gains and losses, making small gains appear normal. Removing clamping means gains (both up and down) are more fluid now. Since in some cases clamping was actually protecting players from falling down too quickly, these players will see some larger losses and smaller gains until they’re back where the system expects them to be.
* In other tiers (bronze – plat) league standing is a very accurate mapping of player skill, but there is a buffer to slow players LP losses temporarily if they start to go on a losing trend. Since clamping was removed, but challenger and master must contain the very best players, the loss shielding is thinner in diamond than other tiers. In other words, gains and losses can feel more swingy.
Gains should become more stable over time. We’ll be monitoring how LP changes stabilize over the next few days and evaluating if we need to make adjustments or not based on the results.
Want to check Patch 4.16’s notes? Here’s a link:
If you have any questions, feel free to ask me atÂ @NoL_ChefoÂ or e-mail me at email@example.com.