Boards›Gameplay

Are games really snowbally? Let's see the stats

PopcornBunni··50 votes·2,780 views

https://www.leagueofgraphs.com/rankings/win-stats/all

First blood gives a 60% winrate and a 40% loss rate. So just being the first team to have a death means you lose 10% of your odds to win. Okay technically a 20% loss in odds, but a 10% loss in probability to win.

Taking the first dragon gives you a 68% winrate. So if you get pushed in early botlane, tough luck, that dragon loss means a game loss 18% more often.

Taking Rift Herald gives a similar 70% winrate, mostly due to the fact that the team which is already winning has a much easier time taking Rift Herald in the first place. Not so much a cause of snowballing as a symptom.

First Tower gives a 72% winrate. So once a tower is down, that lane gets behind and can not catch up on farm because they now have to ward twice as much to keep on top of the minion wave safely. Meanwhile a first tower means that advantage starts spreading to other lanes, getting them ahead while the behind team tries to catch up.

Meaning if you die once in bot lane, your tower will fall, and you lose 22% to your odds of winning. Playing blackjack with no basic strategy or card counting gives you better odds of winning (42% vs 38%).

So yes, it turns out the games are in fact ridiculously snowbally. These factors have always contributed to a rise in winrates, but in the past, first blood was something like 54% to win, and first tower 60%.

There is no such thing as a comeback in 2018.

Oh and for fun, turns out warding doesn't matter anymore* because a single mistake in laning makes a bigger difference than the impact an unwarded river does now.

https://www.leagueofgraphs.com/rankings/warding

42 Comments

Eedat42 votes

Until you realize these rates havent changed really ever.

Look at those winrates from say, patch 5.22

First Blood: 60.62% WR

First Tower: 68.86% WR

First Dragon: 70.44% WR

First Baron: 79.78% WR

First Inhib: 91.31%

In contrast if you look at patch 7.24 the numbers are practically the same

First Blood: 59.81% WR

First Tower: 72.27% WR

First Dragon: 68.74% WR

First Baron: 80.04% WR

First Inhib: 91.15%

So the numbers are practically identical actually

Danjeng10 votes

You dare use EVIDENCE to refute the general consensus and will of the boards that games are too snowbally? HERESY! Seize him!

Z3Sleeper6 votes

This goes against our 0-tolerance policy to anti-circlejerking.

Troll for Trump2 votes

Don't be a dick

Xonra2 votes

"Evidence" doesn't mean crap out of context. When Games are on average shorter, because the snowball happens significantly faster than the "Evidence" trying to refute OPs claims.

deadlychuck10 votes

What matters is time scale though. If it takes 40 minutes to take the first inhib, as opposed to 20.

The concept people have when it comes to snowballing, isn't just about the stats of objectives to winning.

junglerboy162 votes

He's also pulling stats from barely 2 seasons ago, which was still arguably after many of the overall changes that pushed snowballing in the game.

He should be comparing to stuff like season 3 or 4 if he wants to find what people are saying they miss.

Spank the Fox5 votes

{quoted}

He's also pulling stats from barely 2 seasons ago, which was still arguably after many of the overall changes that pushed snowballing in the game.

He should be comparing to stuff like season 3 or 4 if he wants to find what people are saying they miss.

You’re shitting me right? While patch 5.22 was stupid, it does capture the essence of Season 5, which most Board members will say mid S5 was the most fun/balanced tme in League history.

4262877 votes

[{quoted}](name=Spank the Fox,realm=OCE,application-id=3ErqAdtq,discussion-id=yqMTbXpQ,comment-id=0002000100000000,timestamp=2017-12-31T06:04:05.917+0000)

You’re shitting me right? While patch 5.22 was stupid, it does capture the essence of Season 5, which most Board members will say mid S5 was the most fun/balanced tme in League history.

What? Most board members say that? Where is your evidence of this? Stahp.

NineTailedMystic2 votes

Season 5 was da best though. Tanks were the strongest but every class had their place.

Anags5 votes

Season 5 had all kinds of stupid changes in it. If anything the game has been in a constant state of flux ever since that time.

The changes since then have only been more and more dramatic it seems.

Seasons 3 and 4 were much more tame imo.

deadlychuck1 votes

Naw, the game was already too deep into the burst AD meta at that point. People think season 5 was good, because so many people who still play are the cancer that made shit like zed and yasuo not get completely reworked, while the mages get fucked by the design team.

Eedat2 votes

I'm not cherry picking stats. Thats the earliest patch Leagueofgraphs provides an infograph for

Eedat7 votes

I mean it even breaks down average game times at the bottom. It looks like games average slightly longer now than they did then

Darkeus1 votes

Actually they are 7 minutes shorter on average. Someone on Reddit did the math and research. It is on front page so easy to find.

deadlychuck1 votes

I have a hard time believing that, when every late game strategy has been completely gutted and neglected by riot.

They can't make the game not shit, if they're being forced to design the game like it's a mobile game rather than a PC title. Which is what Tencent makes them do, because they see it as a market competitor to mobile mobas.

Lit Corn4 votes

the stats are pretty much the same , but what happens in the game is actually different , the way snowball is set up , how fast tower falls and so on , there is no such thing as a tower falling at 5 min back in s5 , and tier 2 had shields , tier 3 had lasers , so the way you could siege was really different and when enenmy comp had lot of waveclear and played well they could definetly stall , but right now its basically a rofl stomp and you get one shoted if you try anything , also , vision control and overall strategy is far less important.

PaffWasTaken2 votes

Two seasons have similar stats considering win rates with taken objectives

The first one happens to have a medium length of 40 minutes, with the shortest ones being utter crushes at 20 minutes

The second has a medium game length of about 18-24 minutes, with some games ending under the 15 minutes mark

Which one seems like the most snowbally one to you?

Eedat5 votes

This is just straight up wrong. It actually outlines game lengths at the bottom of both. Games were actually shorter on 5.22

Xonra1 votes

Guess Riot was lying, given that THEIR stats say games are shorter now, just not by a ton. So just throwing that out there.

Eedat1 votes

Except Riot doesnt release their stats. Source.

AMYS GRAVE8 votes

5.22 was when they added A LOT of snowballing mechanics. To name some : Rift Herald, upped the team gold on kills, removed turret shields, upped the gold on turret takedowns, upped gold values on farm (both lane and moreso from jungle), lowered the gold on scaling ADC builds, raising the gold on AP items (which often help scale defenses like shields and heals), increased the cost of defensive items, added keystones many of which were fairly hard snowballing at the time of release (hi release warlords), marksmen update which removed many of ADC early game weaknesses, ...

Like, we've been complaining about game snowballing ever since this patch (give or take a patch or two) so saying that is the benchmark for dynamic gameplay that should be the standard candle for snowballing compared to "back in the day" isn't exactly fair. I dare say most people complaining about snowballing are comparing to patches earlier than 5.22.

Sire Hippington3 votes

well, 5.22 was one of the most snowbally patches ever, so the fact that we have about the same snowballiness than in one of the worst patches in lols history is not good at all... I'd like to see some numbers from S4 or befor on that...

Oxyrotten7 votes

You are forgetting:

Heralds eye. Auto-pushing minions. Dragon buff. Dragon buff multiplicative. New masteries stacking off kda. Super minion stats. Tower stats. Tower scaling. Movespeed increases. Flat stat buffs (new runes).

There's a few more but I forget. Intentional snowball mechanics implemented by Riot to shorten game times by putting more focus on early game.

PopcornBunni9 votes

{quoted}

You are forgetting:

Heralds eye. Auto-pushing minions. Dragon buff. Dragon buff multiplicative. New masteries stacking off kda. Super minion stats. Tower stats. Tower scaling. Movespeed increases. Flat stat buffs (new runes).

There's a few more but I forget. Intentional snowball mechanics implemented by Riot to shorten game times by putting more focus on early game.

Of course. The mastery that seems to be flying under the radar is Triumph. It gives 25 gold per take down, regardless of the enemy's bounty, eliminating in large part the component of the game that rewards killing high-bounty targets.

Even the stacking defensive masteries are snowball oriented. Grasp rewards characters that can trade well by making their trading even better, and overgrowth rewards characters that can lane bully and counterjungling that sets that role behind.

Conditioning is the only defensive rune that actually makes a difference if you durdle to get it active, while the hunter line and legend line make winning characters win more.

No Fearless to protect against burst damage No Seigemaster to protect from dives and make tower farming possible No Legendary Guardian to improve defenses durring ganks and teamfights No Toughskin, making characters that are bad at trading even worse at it Bandit locked to bad support items preventing its use in hard matchups Windspeaker equivalent lost its resistance boosting effect Aftershock vs Courage of the Clossus only works better blocking extremely high burst damage (1000+) and is much worse at cc comboing to trade in-lane (a 5% health shield blocks damage entirely, 20% bonus resistances don't, and do nothing against true damage)

The end result is only a handful of the tankiest of tanks that are overtuned can do the "play safe and farm up" route while everything else is throwing damage at each other 1- and 2-shotting and causing games to be decided by the 6 minute mark.

Oxyrotten2 votes

{quoted}

Of course. The mastery that seems to be flying under the radar is Triumph. It gives 25 gold per take down, regardless of the enemy's bounty, eliminating in large part the component of the game that rewards killing high-bounty targets.

Even the stacking defensive masteries are snowball oriented. Grasp rewards characters that can trade well by making their trading even better, and overgrowth rewards characters that can lane bully and counterjungling that sets that role behind.

Conditioning is the only defensive rune that actually makes a difference if you durdle to get it active, while the hunter line and legend line make winning characters win more.

No Fearless to protect against burst damage No Seigemaster to protect from dives and make tower farming possible No Legendary Guardian to improve defenses durring ganks and teamfights No Toughskin, making characters that are bad at trading even worse at it Bandit locked to bad support items preventing its use in hard matchups Windspeaker equivalent lost its resistance boosting effect Aftershock vs Courage of the Clossus only works better blocking extremely high burst damage (1000+) and is much worse at cc comboing to trade in-lane (a 5% health shield blocks damage entirely, 20% bonus resistances don't, and do nothing against true damage)

The end result is only a handful of the tankiest of tanks that are overtuned can do the "play safe and farm up" route while everything else is throwing damage at each other 1- and 2-shotting and causing games to be decided by the 6 minute mark.

To be fair I think grasp is aimed a bit more at late game, but it definately has snowball potential as percentage health grows with items and stacks hp from trading.

But the issue is that snowballing is aimed at shortening game times, it does this in compound by the mechanics that focus on early game as well.

I was thinking more along the lines of:

stacking attackspeed, stacking lifesteal, stacking tenacity, stacking movespeed, stacking on first hit on hit damage (WITH NO CAP), healing rewards for burst or getting kills in team fights (I mean is it necessary to make it even harder to trade even in a team fight?) pred could even be considered factoring in the intention is to increase early game gank potential, stacking eyeball (adaptive damage), stacking heal from abilities, stacking cdr, manaflow to eliminate mana dependencies (which mostly effect laning and are an integral part of champion balance), stacking ult cdr, increased cdr stacking damage, movespeed stacking damage, demolish, % resist spike (conditioning is debatable), stacking hp from creeps (debatable but still snowballing), klepto... blink (debatable but aimed at making ganks more efficient even without flash), debt... overall a bad investment until you realize it's aiding early game item completion and staying in lane longer stacking damage to creep based on type plus a farming aid,

I mean literally one fifth of all the new masteries are some sort of lifesteal or in combat healing effect? Sounds a lot to me like riot wants someone to win team fights by cleaning up for multiple kills. And what happens when someone wins a team fight by way of clean up penta? It gets pretty hard to win the next team fight against them...

RankOneUganda5 votes

the 68% first drag winrate isn't really so much about the drag itself. its more of a result of your bot lane losing or underperforming compared to enemy. sounds about right.

I would like to know the wr of the more fed adc by 15 minutes.

xxxDogmanxxx5 votes

{quoted}

https://www.leagueofgraphs.com/rankings/win-stats/all

Meaning if you die once in bot lane, your tower will fall, and you lose 22% to your odds of winning.

Unsubstantiated statement. You made the argument that statistically speaking, First Tower leads to a 72% win rate, which is valid enough, but did not give anything at all to prove dying once in bot lane means First Tower.

Another unsubstantiated statement:

There is no such thing as a comeback in 2018.

There are unacceptable leaps in logic in this post.

Jobriq4 votes

Triumph is definitely overtuned. 12% missing health on takedowns is huge. Presence of mind is niche on some mana heavy champs (kassadin/Swain) and overheal is practically useless. A shield when you're at full health, but it decays when out of combat. I can see it being decent if you have a Soraka/Sona support (or maybe on Sona since her W heals herself) but compared to Triumph, which is always good, there's little reason to consider two very niche only kinda useful options

Jo0o3 votes

Them's be some strong number-words.

I Do Not Recall3 votes

https://imgur.com/a/brmeo#OjzlPAd

Playing macro games for late game in season 8 LUL

DunkinNoobs2 votes

OP needs a big lesson on correlation and causation.

Zach65082 votes

If you have a problem with the opponents taking first dragon, follow these 2 easy steps:

  1. WARD!!!! Not enough people do it.

  2. Camp Bot Lane. My brother doesn't do it and I lose quite a bit because of it.

And just don't die. It's easier than you would think. Use COMMON SENSE. It helps a lot. Also, having a good support also helps (Bard is amazing for whatever reason)

Jew Stomper1 votes

Yeah, why tf would you go toplane if you can gank bot and disrupt 2 opponents AND might be able to get dragon.

ghosthunter5232 votes

I won a game where the enemy Kayn was full build at 26 mins in so I'd have to disagree

Saianna1 votes

There is no such thing as a comeback in 2018.

But my fathers friends uncle gets a comeback liek veery gaaeam!11!11111 (typos on purpose)

PhantomPrankster1 votes

despite all this people still think dying doesnt have a large role in how they do

Broken Bandle1 votes

I feel like this is all information that Riot has access to, and if this was not their intention that they would change it. It would seem that they are attempting to follow closer to other games with shorter timers which while I don't like it and feel isn't in the spirit of league might be their new business model. It seems like its "working as intended".

PixelPicksTW1 votes

It's almost like gasps violently teams that focus on getting objectives win the game. Now I too agree that this META is way too snowbally, but the winrates are not a measure of this, win times are.

RocketRelm1 votes

I think the reason the "win rates with x statistic" matter is because this isn't a heavy focus on getting objectives. First blood on tower is often (but not always) "we got a single kill bot lane", and jungle was there to shove the tower because they did the gank.

It's not the elaborate "we actively tried in any way to coordinate and push the right things at the right times" like "teams that focus on getting objectives" makes it sound like.

Wolfeur1 votes

Does first blood means you get better chance of winning? I see correlation, but no causation. Maybe you get first blood because you're better, which explains why you have better odds at winning.

PaladinNO1 votes

You sure got a point. As a Support main, I feel much less inclined to leave my ADC now, for any period of time. And Sightstone has gone from a 1st-2nd item to 4th-5th item - if ever.

And I initially feared I would get "if there had actually been placed wards..." a lot, but too rarely happens. Again, if ever. So at least there is something positive: The majory recognizes the need for the duo lanes to stay together.

...Either that, or all just fears dying so much, and they are so stressed out they don't find the time to type in anything.

More from Gameplay

3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay764 comments
7,167
votes
3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay156 comments
3,510
votes
UPVOTE FOR URF
Gameplay606 comments
2,440
votes
Yes plz.
Gameplay456 comments
1,721
votes
3 Refunds per season, up vote if you want 3 refunds per season.
Gameplay114 comments
1,716
votes
When is ANYONE at riot going to address the awful new player experience?
Gameplay634 comments
1,383
votes

More by PopcornBunni

Weekly reminder that Taliyah STILL --
Gameplay83 comments
201
votes
Replace "none" ban with "pass" to another teammate
Gameplay174 comments
144
votes
Somehow 10 bans doesn't seem like nearly enough anymore
Gameplay146 comments
139
votes
Games are too snowbally and Towers are made of candy floss
Gameplay46 comments
134
votes
"Win a PVP game with a player level 15 or lower"...
Gameplay71 comments
131
votes
Imagine being allowed to be busted for 2 years straight
Gameplay155 comments
114
votes