Boards›Dev Corner

Early Snowball 8.13

RiotNot Surrendering·6/14/2018, 8:01:56 PM·1 votes·30,225 views

Hello,

Since it missed Quick Gameplay Thoughts, we are looking at reducing early snowball. Target is 8.13 and the changes are intended to be relatively low disruption (not particularly noticeable, but have good impact).

We have seen early snowball increase slowly, but steadily over this season. As I pointed out in a reddit post earlier this week, this could be due to any number of factors that our data analysts are picking apart and will hopefully share when we have stable results.

Some of the tactics we're looking at:

Re-introducing the bounty on 2 kills and potentially a small increase to bounties 3 and above.
Early death timer reductions from levels 1-7, while keeping an eye on what this does to teleport summoner.
Outer turret global gold reduction 100 >>> 50.

What we're not looking at for this pass:

Towers getting more tanky, especially in laning phase.

I know this is a common complaint, but we have reasons to believe that elongating laning phases is not the best course of action right now. This is partially because having elongated laning phases means more chance for lanes to snowball out of control without meaningful chances for other lanes to intervene as well as just reducing the amount of game variance (through interaction with other champions). I personally think the average length of laning phases is about right at the moment, but the standard deviations (the length of the longest laning phases or the shortness of the shortest laning phases) are too large. This can result in some unsatisfying laning phases where the first turret goes down as early as 7 minutes.

It also has the tendency to make games stall out even longer when already won, so I would want us to have a more robust (and flexible with game state) solution. Examples here could be: killing towers doesn't grant as much gold (which is frontloaded snowball), but instead grants a deferred or distributed bonus of some sort. Tower fall times could also be regulated more strictly (ie. very hard to drop before a certain time and easy to drop after a certain time).

75 Comments

lol you wild wyd6/14/2018, 8:07:14 PM42 votes

Tower changes should be looked at in general in my opinion. The reduction of health of base towers was not needed so was the change to Baron damage. I think Baron iteration before the damage reduction was literally PERFECT, dangerous and rewarding at the same time. Fast games are not League of Legend's identity, being able to execute multiple strategies is (like scaling comps). I hope this increase in snowball was not an attempt to cater to the casuals as they say. It feels amazing when you can win in multiple ways like stomping or a LONG intense strategic game of League where meta tactics come into play.

Smithy2Q6/14/2018, 8:12:09 PM23 votes

We have seen early snowball increase slowly, but steadily over this season. As I pointed out in a reddit post earlier this week, this could be due to any number of factors that our data analysts are picking apart

Maybe it's the damage being too high? I'm betting it's the damage being too high.

jadelink6/14/2018, 8:06:31 PM13 votes

I'd actually like to see kill gold reduced to 250. Snowballing from kills has lead to so many dull one sided games. Games so often are effectively over before we have even a single teamfight now, and it feels dull, even when I'm the one doing the stomping.

TheMinionLeader6/14/2018, 8:16:18 PM12 votes

The damage is too high in this game. No one is having fun. Why does true damage exist? It was one thing when it was only on Vayne and Yi. They were/are hyper scaling carries that can be absolutely rekt in the early game. Now when you put it on someone like Camille its just like 0.0

Venedictos6/14/2018, 9:32:08 PM11 votes

Why are you not entertaining the question of what kind of snowball is preferable?

Say towers get beefier and take forever to destroy. Sure. I start killing my enemy laner again and again and again and every time and they are worth less every time I kill them. But because the tower is still there, my snowball remains contained inside my lane and its up to my opponent to keep it contained in it.

Time and again we've seen that no matter what lead you acquire in the laning phase, it isn't worth much if you can't translate that into a snowball that affects the entire map and eventually leads your team to victory.

So in that case, is it better that I am able to take down the tower much faster right now, shove all the way to T2 tower and then have plenty of time to disrupt the enemy jungler, ward their jungle, provide my team with plenty of useful information without much risk, gank another lane and appear in a skirmish in the map way earlier than the enemy laner?

And anyway, yeah sure. Laners can't really intervene because they might be busy in their lanes too. But isn't that what junglers are for?

I feel that the reasoning behind not changing the towers is way too shallow.

On the side note: Yes, beefier towers do make the game last longer. And there's a lot of people that have complained that games are too fast. There's also a lot of people that enjoy longer games. There's a lot of people that enjoy the late-game fantasy of their mains and they rarely ever get to that point. I don't think 40-45 minute games would be bad or tiring. And if there's too many situations where a game is decided 10-15 minutes earlier than when it actually ends, you've got plenty of levers to pull - Elder Dragon and Baron. Baron can be a raid boss and that could be fine, so long as the rewards it provides are meant for teams that are extremely ahead and want to close out a game or teams that make extremely smart calls.

nuletic6/14/2018, 8:53:24 PM10 votes

I really wish that leads were cemented into turrets. The game could be about territory and strategy, just like football. Playing against and around turrets should be much more meaningful. The lead should come FROM THE MAP not just from gold, items, autopushing minions etc. That's basically just "oh hey, I'm stronger than you and you're useless haHa xD. Come "outplay" me while several levels behind and down in gold!" It just leads to stomps. It should be about map control, map pressure, turrets and so on. Our team is ahead because we have 2 more turrets and it's not just the gold that matters, it's that the turrets themselves are HUGE objectives and difficult to take. The lead becomes external to raw stats and the game becomes richer adn more strategic... more about experience, more about understanding, more about knowledge.

Shut down gold is artificial and randomizing and so are autopushing minions. They are unearned leads/comeback mechanics imo. Just make turrets much stronger and much more gold. That way, you won't need so many of these snowball and anti snowball mechanics to begin with. It will feel more earned. Right now, almost no objective feels that rewarding to take in terms of difficulty. Baron is still really easy to take. I'd add at least 2000 HP more to it. It's good that the damage was reduced though.

Also, I think the ADC buffs on PBE might be a bit much combined with this. Twitch more or less switched to BoRK first anyway. Vayne is scary enough late and might also be gravitating towards BoRK first. The other changes seem OK. These 2 might be overbuffed if the snowballing is reduced though and the meta hasn't settled down enough to notice how strong or weak they are. I might be wrong on this though. My impression is that Twitch will become the strongest ADC in the game and the meta will revolve around pairing him with a great defensive ardent support.

One that I don't see on PBE though is Corki. He got severely nerfed (both abilities and AD growth) and his build got hurt a lot too. His core build was TF-Sorcs-RFC-Shiv-IE. He's essentially down 600g and even his late game isn't performing well because he has no self-peel, no CC etc. in an assassin meta. Tanks are too weak overall and he doesn't get a frontline to allow him to be useful outside of when he has package.

EarthyMadness6/14/2018, 8:12:06 PM7 votes

What are the thoughts regarding the change made to the Inhibitor and Nexus turrets regarding HP regen limits (100%/67%/33% if it falls below those points)? A similar system could be adapted for outer and inner turrets, giving a much lower regen rate and dividing them into quadrants or fifths instead.

Still allows for more concerted efforts to nuke them down if intervention is necessary from jungle or other lanes, without it being an "all is lost because I took a bad trade and died" scenario like now. If someone gets a kill and immediately runs turret down to 5%, having a chance for it to regen a little is a whole lot better than the current "I'll just dive and auto the turret twice before it aggros me because they can't fight me without the turret" situation that happens.

This wouldn't really extend laning phase too much, and should damper early snowball by making it just a little harder to get the turret that last little bit, possibly turning a losing lane around if you're able to strategically use the turret to your advantage.

Overall, thoughts?

ANARMYOFOVER1006/15/2018, 6:40:32 AM7 votes

Something that may help is reverting the longer ward cool downs. If we are killed once or twice by a successful gank, our tower is gone. And then our laners get to roam and snowball the game really hard. Then we cant ward to make smart plays or keep from being hard punished from a random bush. Decreased ward timers would help us not get killed so easily from pushing right from under tower.

ModKnightsKemplar6/14/2018, 9:28:33 PM6 votes

Have you considered "First Tower" and "First Blood" rewards? I can understand them making sense when they were introduced, but now they seem more like a "win more" mechanic.

You already get a gold advantage if you get a clean kill, and if you don't, I'm not convinced you should be rewarded for that. Same with tower trading, but I understand why that one matters for pro play.

Shuriman Curator6/14/2018, 9:34:13 PM6 votes

Have you guys not considered reducing damage an option to decrease snowball? Maybe turn it down a few notches?

wait

what the fuck am i thinking! lmfao

https://i.imgur.com/0yUdQ3e.jpg[/img]

Dog of Pavlov6/14/2018, 8:55:47 PM6 votes

PLEASE INCREASE TURRET HP AND DEFENSES THIS IS WHAT EVERYONE WANTS [sg-soraka]

DovahKing6/15/2018, 2:24:08 AM5 votes

Since you guys are aware that laning phase has shortened, and prefers to keep it that way, do you have plans to compensate Nasus for that? Nasus really favors an extended laning phase in order to stack and hit his mid game power spike. Particularly the time after level 9 and before enemy can send most of their team top to bulldoze your lane, that was the prime stacking time for Nasus. But now that time period is basically non-existent, and when enemies are all grouped up around the map, you don't have the time to sit in a side lane and leisurely stack up, since you at least need to hard shove the lane to generate pressure, and that results in suboptimal stacking.

Kyunsei6/14/2018, 8:13:55 PM5 votes

Those. ARE. GREAT.

I mean it, that's really good news, keep it up please !

iTaLenTZ6/15/2018, 1:40:36 AM4 votes

Everyone has been shouting for buffs to towers for 3 years now and if you are the one who keeps nerfing and nerfing and nerfing them then maybe you are the problem we should be start looking at. If we want better towers fucking give it to use already instead of patronizing us on why it would be bad blablabla. All I have seen is this game going more and more downhill for 3 seasons now so maybe it is time for you to accept that your ways aren't working.

Lunarahi6/15/2018, 2:25:34 PM4 votes

I actually rather agree about keeping turrets rather squishy. What I don't like is how little damage they do. It feels really bad getting dove and it not really doing much because 3-4 turret shots only takes out half an assassin's health bar. Considering most assassins have an aggro reset or other way to quickly go in and out of a dive, and tanks can sit under turret indefinitely, it feels like turrets only mean something to adcs and mages (who can usually stand outside their range anyways).

Increasing turret damage would also help reduce snowball drastically, as a team with a lead can pretty much ignore the other team's base right now. It also makes it so a team that devotes resources and plays around keeping a turret alive can actually be rewarded. Something I would also consider is adding armor pen to the inner turrets in some form, so that tanks can't just sit under turret for free in those mid game teamfights.

TsunamiSS6/15/2018, 10:27:10 PM4 votes

At no point in your post were you even close to discussing damage being too high, as has been mentioned ad nauseam on Reddit and these boards. I award you no points, and may God have mercy on your soul.

Bootleg Gay6/14/2018, 8:34:42 PM4 votes

Would you consider adding back the damage reduction to bot lane towers too? Because there are two, sometimes a third player down there, they can run down a tower in a single push, all the while you're either dead, (Which smaller death timers might help with.) or you are forced to back off.

I know it was originally to make it more contestable for first tower, but too often is bot tower the first to go, or top lane if your bot lane get ahead and decide to wander off. Mid Lane could easily be more contestable if made so, and it's one of the lanes that can snowball the hardest because of all the assassins.

Plus, there are twice as many champions there to deal damage to it over the course of a few waves, which the damage reduction could slow down for the first few minutes, if only for those moments.

Ataraxas6/14/2018, 11:01:56 PM4 votes

Also I warn riot against having the 10 player's power levels all over the place. The solutions you guys are coming up with to reduce snowballing makes it so there can be a huge gap between players' power levels within the game. (Your solution is increasing bounty gold which only one player gains.) This will only lead to even more snowball.

butt grabber6/15/2018, 12:27:14 AM4 votes

I would like to know why your company likes to fuck everything up, then try to fix it again. the meta right now is a big issue to why i have quit. i will re install the game when it gets fixed

Daddy Ants6/14/2018, 10:50:27 PM3 votes

I miss the times when the old client said

"Average Game Length: 30 - 45 minutes" and games actually lasted that long.

{quoted}

Have you considered "First Tower" and "First Blood" rewards? I can understand them making sense when they were introduced, but now they seem more like a "win more" mechanic.

You already get a gold advantage if you get a clean kill, and if you don't, I'm not convinced you should be rewarded for that. Same with tower trading, but I understand why that one matters for pro play.

First Tower gold contributes a ton to early snowball.

Personally I'd be fine with it being removed.

Spideraxe6/14/2018, 8:10:44 PM3 votes

How do you feel about snowball XP?

yándere6/15/2018, 3:09:30 AM3 votes

I know this is a common complaint, but we have reasons to believe that elongating laning phases is not the best course of action right now. This is partially because having elongated laning phases means more chance for lanes to snowball out of control without meaningful chances for other lanes to intervene as well as just reducing the amount of game variance (through interaction with other champions).

I would like to know what the reasons are for thinking that longer laning phase means more chance for lanes to snowball, if you're able to respond. I did see your reddit post and also responded to it, it was very interesting.

I personally think the average length of laning phases is about right at the moment

I'd also like to know your reasons for thinking this. I can't remember a time in the past five years where laning phases were this short, so to me it feels kinda wrong, but if it's better for the health of the game it doesn't matter to me.

Zullar6/14/2018, 8:57:19 PM3 votes

I blame patch 3.9! https://boards.na.leagueoflegends.com/en/c/gameplay-balance/YZ1ewJIq-i-blame-patch-39

Any thoughts on reducing gold gain on repetitive kills on the same target? I think it's a problem an enemy can get super fed by repeatedly killing his opposing laner over and over and over. If the enemy top laner gets 4 kills on his enemy laner should he get moderately fed? Sure. Should he get extremely fed to the point where there's not much you can do? I don't think so.

I suggest completely overhauling the bounty & gold decay on deaths system. -Slight bounty even after 1 kill (maybe 350 gold) -Further reduce gold value after 1-2 deaths. (Maybe on 2nd death a champion gives 225 gold, and 3rd death gives 150gold for example) -Decay of gold reduction over time and/or gold gained. (Prevent proxy Singed. If you died 3 times a while ago but then manage to survive for 10 min and farm 150 cs then you are worth more money again). -Hit Tab to see bounty value on all enemy champs so you know who's juicy and worth killing.

The benefits would be 1: Be able to control snowballing to moderate level 2: Reduced frustration. Currently if you are having a bad game and are 0/3 it makes sense for the enemy jungler to kick you while you're down and gank you again. It's low risk (gank is likely to succeed, and likely to win 2v2 if countergank occurs) and high reward (since you almost give full gold and they can get a tower afterwards). Right now the game encourages ganking your winning lanes and ignoring your losing lanes. Is this really what you want?

Sona Ping6/14/2018, 8:18:26 PM3 votes

we are looking at reducing early snowball. Re-introducing the bounty on 2 kills and potentially a small increase to bounties 3 and above.

So you're looking into adding even more gold into games as way to reduce snowballing? When you are looking to stop an avalanche, it doesn't strike one as a good idea to add more snow on the opposite side. Like the avalanche can take that as well, for but one example.

AirKingNeo6/15/2018, 12:05:21 AM3 votes

Re-introducing the bounty on 2 kills and potentially a small increase to bounties 3 and above.

DO NOT, DO NOT, DO NOT! This at least allows for solo player snowballing and keeps pro play more interesting.

Do things that prevent global snowballing, like turret global gold.

Also, increase Baron's damage. It's not risky enough. Half reverts the nerfs.

Gunnerkreig6/14/2018, 9:37:06 PM3 votes

It's a tough issue to crack. There are two main areas I think you guys need to watch... First is the impact early kills have. Second is the rate/way towers are falling.

Honestly, kill gold I think is in a decent place. A small bounty at 2 kills might be fine, but I'm not sure it will do that much. I'd be worried about the first teamfight, where the enemy team's 3/0 carry is trying to shut down your team's [last hope] 2/0 carry. The enemy team kills your team's only strong person and gets a shutdown for it? Rough. But hey - shutting down my team's last hope is the right play for them, so maybe they deserve that bounty.

In regards to TOWERS... I think there are some problems there, and that's the real thing that is causing the snowballiness. There are a number of factors that are working together now to make games end faster and more decisively. (FIRST) Last season (?) you introduced the rift herald as a way to break through the tough outer/inner towers early, which were stalling out games, especially against high waveclear comps. Fair, and we all love the big guy. (SECOND) But a few seasons earlier (again not sure of my timeline, but doesn't really matter), you changed Baron as a way to give teams lategame - and especially inhibitor turret - pushing power. (THIRD) Right now, laners are barely punished at all for leaving lane/roaming. As long as you shove your wave and don't miss a cannon, you're fine.

What does all of this mean, when you put it all together? Well, with laners (esp. mid) not being punished for roaming, it means more early-game grouping. Not inherently bad! But it means that it's not uncommon for a 4 man to wind up tower diving bot at 7 minutes, because towers don't deal much damage, and oh look, then the tower is gone. Meanwhile, the mid who didn't roam got to do what? Push two non-cannon waves? Probably two lanes lost in one decision, if the roaming midlaner got any kills.

Going back to "more early game grouping" - this is also how you take rift heralds! Early grouping and forcing a tower is pretty common now, so the natural next step for the team up a tower is [take mid tower, unless it's down or you can't siege, then] Rift. Rift was originally introduced to provide an incentive for early game skirmishes, but also to help deal with outer turrets. Early Grouping is already happening though, to take towers and get kills, so in the current environment Rift almost as often takes out an inner turret. It's VERY common nowadays to see, at 20 minutes when Baron spawns, a team that's behind have at least one lane down to the inhibitor turret. And then it's just a matter of taking baron and pushing THAT tower down. We all know what a team's chances are when they're down towers, kills, and an inhib - next to nothing.

All of the problems I've stated are just things that are slightly overtuned - first baron, rift herald, towers being weak, and most importantly roaming being more valuable than farming. But they all encourage the same outcome - faster games. Up to Riot to decide which lever they want to use, if they want to slow it back down.

TomiMan76/14/2018, 9:04:33 PM3 votes

What we're not looking at for this pass: Towers getting more tanky, especially in laning phase.

Not for this pass but in general you should have look at this a ~~**LOOOOOOOOOONG **~~time ago. [zombie-brand-facepalm] [zombie-brand-facepalm]

ModCaptainMårvelous6/14/2018, 10:29:26 PM2 votes

As a quick question: Why aren't tower changes being looked at? Do you not think towers are the issue? Do you have a better plan? Do you consider tower changes the "nuclear" option?

Astutely6/14/2018, 8:28:39 PM2 votes

This is the right direction. We're all struggling to adjust to the meta. One bad play can be exacerbated heavily.

Ataraxas6/14/2018, 10:53:29 PM2 votes

Perhaps try increasing passive gold gain? This will reduce snowballing all while giving late game scalers more opportunity to reach their fantasy. Currently they have none given how quickly games end. I say this because I know Riot is against longer games and this seems like a good compromise between short game time and giving late game scalers an opportunity.

Ataraxas6/14/2018, 11:03:35 PM2 votes

Perhaps try removing first tower gold? And maybe even first blood? After all that disadvantage is the beginning of all snowballs.

Headmans Pride6/14/2018, 11:09:46 PM2 votes

Hey! Have you guys considered making it harder for a team to break through inhibs and inhib turrets? I think making it harder to breach this part of the base (many many ways to do this), will result in teams being able to stall out a game for longer but keep very similar snowballing. Maybe teams need the ability to stall rather than only break open turrets.

Cosnirak6/15/2018, 12:07:59 AM2 votes

If you want to reduce early snowballing you need to give players worthwhile defensive choices for the early game. Like more rune choices that help you not die in the first few levels. Or the ability to be fairly safe under tower (which would require towers to actually hurt champions significantly beyond level 3). Perhaps reverse the buff ignite got. Maybe reintroduce viable early game items that give defenses for classes besides tanks (like the stats on old abyssal scepter and/or mask, but without the unhealthy aura issues).

Xyltin6/15/2018, 7:43:17 AM2 votes

0/90/210/300/360 (+30g) up from 0/0/200/250/300 (+25g). Make the bounty lvl go down to lvl 0 when at lvl 1-3, to lvl 1 when at lvl 4-6 and to lvl 2 if 7+.

Death timer reduction is likely ok if it is just ~1 second. The TP problem should solve itself if TP gets reworked.

The gold reduction is HUGE (-250g). I agree that this could solve a part of the problem, but why not make turrets finally decent again?

Instead of giving turrets scaling armor/MR over time, let them start at full Armor/MR and remove the scaling. That way early turrets are more durable but they are still easy to cut down later. It would make it harder to just take a turret at 8 minutes but not significantly at 12+ minutes.

Turrets need an overall overhaul. Slightly more dmg, more protection in the first ~5 minutes againts dives and more durability early on (base turrets need more durability overall).

The long version:

ALL: Turrets now have 70 Armor/MR up from 55-70.
Mid lane (outer/inner/inhib): turrets now have 80 Armor/MR up from 55-70. Why that? Because too often you start to grp mid lane and run down the turrets for the gold. Or after a won teamfight you rush mid and taket 1-2 turrets. They are closer together and faster to take because getting minions there needs less time and they are closer to the teamfight most likely. This should reward people trying to splitpush more often and take the side lane turrets.
Nexus/Inhib/inner: +10 dmg. They lack power right now.
Nexus: HP up to 3000 from 2700. They are way too squishy
Nexus: Gold up to 100 global from 50 (more rewarding and helps finishing the game)
Nexus/Inhib: Now have markers for the 33% and 67% HP values (for the HP reg)

Outer turrets (all 3 not just top and mid):

40/25/10/0% dmg reduction till minute 0/5/6/7 changed from 50% till minute 5. This makes turrets harder to get down too early. Reduced in power very early due to the increased Armor/MR
+10% dmg reduction (50/35/20/10% combined with the upper) agianst ranged champs if an ally is close to it. This is to make turrets last longer when the enemy can defend them. Mostly to reduce the poke ranged champs can easily get in early on while melees can't take that risk at all most of the time.

Now this would reward lane swaps again because all outer turrets get that, which we want to prevent:

Outer Turrets gain +20/10/5/0 % bonus dmg against champs at minutes 0/5/6/7 per enemy champ extra over allied champs. So at a 4v2 turret dive at minte 5 the turret would do +20% dmg (2 times 10%). Or in a 4v1 top lane turret dive at minute 4 +60% dmg. Due to it scaling with numbers and and falling off pretty early already it doesn't really prevent turret divesv like a 2v2 or a 3v2 or a 2v1. The +10% dmg at minute 5 then are something to consider but nothing too major if the dive is executed well, but also a larger risk.

This would prevent the 4v0 lanes we saw with lane swaps most likely because they can't just shove that guy out from under his turret so early that he can't get to lvl 2 or 3 and at that point his jungler can start to help out. So I hope this would prevent 4v0 lane swaps but enable possible 2v1 lane swaps.

Tr8torian6/15/2018, 11:41:20 AM2 votes

Hey Riot, Nerf Singed?

DW Diana6/19/2018, 10:47:01 AM1 votes

How do you feel the short game times are effecting champions that need scaling to have kill pressure? Are you think of giving that kill pressure back?

Doctor Liber6/20/2018, 8:27:48 AM1 votes

After looking through https://www.leagueofgraphs.com/rankings/win-stats/all it seems obvious how impactful certain kind of objectives are:

Herald has 71% winrate;
Nashor has 82% winrate;
First Inhibitor has 92%.

I think , as most of all do, that damage is obviously too high, but since it seems not to be the point that you are going to touch, i guess that moving five minutes forward the timer of both Herald of Nashor would made the most impactful change to get slightly longer games.

When herald is taken too early by the team with a lead, who's behind in the opposite team has hardly ever any chance to contest cause he didn't farm enough to get the first item finished and give an huge advantage to the team who took it. Same for Nashor, 20 minutes are not enough for who's behind to get anything and get back in the game, they simply can't contest and lose the game anyway. In this way you should have more time to farm and get your items to contest such important objectives that determines Winners and losers of the game.