Red Post Collection: Practice Tool live, Valentine's Day 2017, Marksmen Thoughts, Ask Riot, DDR in LCS, & more!

Posted on at 12:49 AM by Moobeat
Tonight's red post collection includes the Practice tool live in all Riot regions, Valentine's 2017 community collaborations, Meddler with thoughts on Marksmen, the latest ASK RIOT, context on Kill XP changes currently on PBE, info on Riot's new Deterministic Disaster Recovery Tool seen in LCS, & more!
Continue reading for more information!

Table of Contents

Practice Tool live! 

After a server by server roll out, the new PRACTICE TOOL is now available on all Riot regions!


You can access this new tool through the new client - go to play, hit change game type, select "training", then launch PRACTICE TOOL! You can enter with up to one bot and have a variety of testing optionsdummies to hit, and more! Check it out in action with this montage:

"At its core, Practice Tool is a place where you can train your mechanics. It’s the training gym where you can go to brush up on your Lv1 last hitting, that early jungle gank route or your Riven mechanics. A simple, straight forward, single-player environment that lets you test out things like checking wall flash points or if you can solo dragon with a certain item build, without wasting hours in Custom Games. As League gains more and more similarities to a sport, it makes sense that there be a place where you can go to shoot hoops! 
Read More:"

Feel the love with our 2017 Valentine’s postcards 

Love is in the air! Share some with all your friends with these adorable Valentine's day cards made in collaboration with artist Zuqling!
"Whether or not you have a duo partner queued up to help you celebrate Valentine’s Day this year, it’s always a great time to reflect on the many people who help us feel the love all year long. 
We’ve collaborated with community artist Zuqling (whose work we have a huge crush on) to create some Valentine’s postcards for you to share with the special people – or pets! – in your life. Enjoy!"

Blitzcrank's Fleshling Compatibility Service

Blitzcrank is offering his services with his new Fleshling Compatibility service quiz!

Check this link to take the quiz find out who your match is!
"Find out which champ is your soulmate with help from the ultimate matchmaker: Blitzcrank."
Here's the full background art from ffsade!

Some follow up marksmen thoughts 

Meddler hit the boards with another round of thoughts concerning the current state of marksmen, things that changed since the last discussion post, and changes looking forward:
"Hi folks, 
Figured it'd be good to touch base on our thoughts on Marksmen again after talking about them at the start of the year. 
Things that have changed since then: 
  • Jungler influence is down due to nerfs to XP and other systems. We increased jungler influence unintentionally in the pre-season and have been reducing it to around previous levels since. 
  • Courage of the Colossus is significantly weaker, reining back champions that utilized it especially well in particular. Combination of that and jungle stuff above's restored agency to a range of positions/champs, marksmen included. 
  • Lethality's been buffed since it wasn't serving the needs of a range of physical damage dealers from multiple classes. That's given caster and early game marksmen itemization to better serve their needs and we're seeing a lot of success from them (and other lethality users) now as a result (too much so in some cases). 
  • We've made some changes to individual marksmen where appropriate (Draven, Kalista, Lucian, Trist etc).  
Current thoughts: 
  • Utility marksmen (e.g. Ashe) seem to be in a reasonable spot. Caster/early game marksmen (e.g. MF) are performing strongly too, too well in some cases. Crit using marksmen (late game AA focused) however may need further item work (not straight buffs, but adjustments to who they're strong/weak against when). 
  • Armor pen needs some followup changes after the Lethality buffs. Some of that's Black Cleaver being too broadly appealing, leading to some excessive strength sometimes when combined with Lethality. We'll be shifting BC to a higher health/lower AD profile as a result. Longer term (mid season potentially) we'll also likely take a larger look at the different armor pen types. As a stop gap before that we're testing Last Whisper (the component of Lord Dominick's and Mortal Reminder) having the same % pen as those upgrades, but less AD as a safety valve against mid game armor stacking. plus a reduction in the Lethality from Ghostblade and Masteries. 
  • We'll also probably be nerfing some individual marksmen in the near future. Corki and Jhin are the most likely candidates. 
  • We'll be making changes to Fervor and Warlord's. Warlord's being an effective default choice leads to a situation where you either take the thing that lets you not lose lane or the thing that lets you be appropriately effective later in the game IF you didn't get crushed too much in lane in the first place. We want Warlord's to be more niche as a result. 
  • We're not seeing other classes edge marksmen out of bot lane. So far Ziggs is the only mage getting meaningful play as a farming bot laner and, while he looks a bit strong, we're still happy with his presence there. Long term we'll likely look at whether there are a couple of other mages we could get into bot lane in a similar way (tower taking power's probably some of the answer there)"
When asked about the Jhin nerfs mentioned in the post, Meddler noted:
What are your plans for Jhin currently? 
Is there any future plans for Lucian? I still feel he's very weak and isn't being as strong a bully in-lane as he should be.
For Jhin we've been testing a nerf to the ult slow (shorter duration or decaying slow) to reduce his ability to nuke down squishies with just his ult in some circumstances. Also considering reducing damage to minions from his traps to reduce his wave clear. Those aren't locked changes yet though. 
For Lucian we'll want to see the impact of Warlord's and Fervor changes before assessing him again, since he's likely to be one of the more affected champs by those."
Meddler continued:
Speaking of Lucian, how do you feel about him in regards to strategic identity? He doesn't really seem to have a distinct/unique niche as a Marksman aside from "duelist," which other Marksmen can arguably do better (i.e. Quinn, Vayne). Could be a reason why he's struggling.
I'd agree he's too generalist. Early game bully who's short ish ranged but mobile is part of the identity he should have I think so we've got a decent start at least. Would be nice to see him bring something unique to a team as well though."
As for thoughts on Kindred, Meddler noted:
Hi Meddler, 
Since the topic's on Marksmen, are there any changes in line for in the future?
We'll probably make it so the first mark spawns on a Scuttle crab consistently, rather than sometimes on Raptors, given how hard that camp is now for them to deal with. Was a problem case someone called out on Boards the other day that looks like an example of bad RNG (low room for adaptability, little cross team fairness, feeling of 'get screwed'. 
Beyond that not at present. Some interest in looking at their scaling/power curve once we've got the mid-season stuff figured out, nothing concrete nailed down though."

When asked about support oriented changes mentioned earlier in the season, Meddler noted:
Meddler, I know this isn't the place for this but: What's the deal with supports after your experiment? More mages are being played in the support role recently, and I'm wondering if you guys are still working on support and making it have greater broad appeal. 
Thanks for your time.
We'll have a number of changes coming out gradually over time. Next up are some XP changes in 7.4, should have details of those posted shortly (today or early next week at a guess)."
Meddler also commented on the Warlord's Bloodlust changes currently on the PBE, adding:
"Warlord's could be more engaging, interactive, cooler. We'd like to look into options for it at some point, though can't make any definite promises there. 
In the short term though we believe the game's just straight better without Warlord's being a common choice in lane. That much sustain from game start results in some trends that lead to less interesting game states (poke doesn't stick so burst or farm strats get favored, some champs go warlords to avoid losing lane at the expense of mid game+ effectiveness against other positions, comparative impact of lifesteal items gets devalued etc)."

Ask Riot: I Need More Keys 

Here's the latest Ask Riot this time answering questions on the drop rate of keys, different teams working on balance of  the game, and the 10 bans in professional play.
Welcome to Ask Riot! 
This week it’s key fragments, League balancing, and an update on the 10-ban system. 
What exactly do I have to do to win key fragments? I haven’t received one in weeks.

When we originally sat down to build Hextech we played with a few loot systems in other games to get an idea of what was out there. We noticed a pattern of hard monetization gates — in a lot of cases, the game would give you a chest, but ultimately you had to pay to open it. As a player this felt pretty crappy, so early on we knew we wanted players to be able to earn Hextech content without having to pay. 
On the flipside, giving away millions of skins each month can quickly limit our ability to keep investing in better servers, balancing the game, and developing other features that help the game for all players (regardless of spend). From that we came to key fragments as a way to balance giving out free content at a pace that made sense for all players and felt rewarding whether they played a little, or a lot. 
Now that the system has been in player’s hands for the better part of a year, one issue we’ve seen is that no matter how many fragments and chests players get, they say “I’m not getting enough.” Some of that can probably be solved with a clearer feedback system, but some of it might be unavoidable. Who doesn’t want more free stuff?
The basic rules for key fragment drops are:  
  1. Key frags are only dropped from wins.
  2. Key fragment drop rates work off of a per player cyclical reset timer, similar to chests. Higher drop rates are slightly front-loaded and gradually slow based on the total number of key fragments you’ve received in a cycle until the reset.
  3. It’s possible to earn 4 keys (12 fragments) in a 4 week period. This matches the number of chests players can get.
  4. There is a bad-luck protection rule that puts a maximum number of wins between key frag drops.
  5. Players get a small drop chance bonus if they were in a premade. 
While this ruleset is functional, there is definitely room for improvement. We are looking at some possibilities for where we can take this system next, but aren’t ready to share all the details yet. 
As always we appreciate all the feedback you’ve given around this and hope you’ve all gotten a ton of great loot over the last year. 
SOCRATES, Memer (Meta Game Systems) 
Sometimes it feels like Riot has no idea how to balance League. Why don’t you just hire Challenger/Pro players to help judge the impact of changes?

Getting advice from players who are extremely good at playing League of Legends is really valuable when it comes to working on the game. We’ve got a number of Masters, Challenger or ex-Pro players at Riot who are constantly involved in the balance and testing process as a result. However, there are a lot of other skills that are also really important in addition to being good at LoL (in many cases more important), so our game designers are spread across a range of different divisions. 
We’ll take a look at a team-by-team breakdown below for the teams that work on balance-impacting content. Before that, though, it’d be good to do a quick runthrough of some of the skills we look for. This isn’t a completely comprehensive list (I’m sure I’m forgetting a bunch of things), but hopefully captures the core concepts: 
Analytical ability and problem-solving 
  • Being able to assess a situation or element, break down what’s created it, propose different approaches that could be taken and their likely effects, etc.
Creativity and thematic ability 
  • Understanding what themes resonate with players and fit the LoL IP. Finding new ways to put a spin on broad archetypes. Recognizing the right matches for thematics and particular gameplay mechanics that are intuitive rather than forced 
Useability/accessibility and feel 
  • Making player interactions with the game (using an ability, navigating the shop, natural 
Vision building 
  • Ability to get others bought on and excited about an approach 
  • Helping others grow, learn new skills 
Written and verbal communication 
  • Concise, able to generate excitement, clear, right method for the circumstances and so on
  • Ability to communicate well with other disciplines (this one is particularly important; e.g. forging strong relationships between art and game design). 
  • Managing others (performance, etc.) professionally 
Technical skills 
  • Scripting abilities, statistical analysis, writing code, etc. 
Organizational skills and reliability 
  • Being able to juggle multiple projects, deliver work on time, solve resourcing conflicts, etc.
So, looking at the individual teams then: 
  • The team responsible for quick changes to the game. Tends to be focused primarily on game health and balance
  • Values understanding of the current and past states of the game very highly
  • In ranked: Gold through to Diamond, average high Plat/low Diamond 
  • Systems focuses primarily on long-term work, working on the foundational rules that LoL runs by. Stuff like map layout, how runes/masteries work, how XP/gold are generated, etc.
  • Systems works in more conceptual/abstract spaces on average than the other teams. Design principles and theoretical analysis are especially important here
  • Bronze up to Gold
  • A team whose primarily responsibility is to test upcoming changes throughout the day and offer feedback and analysis on them
  • This is where ability to play at a really high level is particularly important in order to validate in-game how things will play out in really high level play.
  • Mid-Diamond upwards, with some former Pro players (LCS). Some players who move in and out of Challenger occasionally 
  • Champion team makes new champions.
  • Thematic understanding, feel/useability tend to be particularly important here, along with game health understanding (strengths/weaknesses of kits in particular).
  • Pretty even spread from Silver up to Diamond 
  • Updates old champions, whether full VGUs, like Warwick recently, or class updates like the assassins
  • Similar set of skills to Champion, though with a heavier focus on game health
  • Silver up to Diamond, clustering in Plat 
Hope that’s been informative and offers some useful context on the range of different skills we look for and how the different teams compare. People who are awesome at everything are certainly ideal, realistically though there’s always some degree of tradeoff of different skills for any given role we’re looking to fill. 
MEDDLER, Lead Gameplay Designer, (League) 
When is the 10-ban system coming to regular games (not pro)?

We’re still a couple patches away from detailing the specifics of the 10-ban system outside the pro scene. Increasing the number of bans has a different set of goals for an organized team of players versus individual players on a team. For now, here’s a look inside a few of the design thoughts around a 10-ban system for all draft queues.

Agency – Giving every player a ban should give everyone more control over their game experience. While hard lane counters aren’t too common in League, this means you can grab Tryndamere and not worry about Teemo ruining your day. 
Fairness – Your order in champ select is essentially a dice roll. We want to make sure that no matter where you sit in the champ select order, there’s less potential of feeling kinda disadvantaged. For instance, being second pick on blue side is arguably the weakest position in the current system – you don’t get a ban, and you’re fourth pick overall, so you aren’t especially likely to get a power pick or a counter pick. 
Time – We’re not aiming directly for a time reduction in champ select, but we’re always looking for new ways to shave seconds off anything from getting into game. Adding four more bans to the system in the same format could stack up to two extra minutes between clicking ready and hitting the loading screen, but that’s not the direction we want to go. We could counteract this by shortening the ban window, but stacking even more time pressure when your proposed ban gets sniped by the enemy team feels bad. 
We know changes like this have far-reaching implications so we’re proceeding cautiously – we’ll have more to share in the coming weeks. 
GORTOK, Designer (Get in Game Team)
Have a question? Head over to Ask Riot and sign into your League account. Check out the Dos and Don’ts, then ask away. 
We promise to read every question, but we can’t guarantee they’ll all get answers. Some questions may already be answered elsewhere, and some won’t be right for Ask Riot. This isn’t the best place to announce new features, for example, and we might skip conversations on issues we’ve talked about in depth before (though we can clarify individual points). 
We are listening, though, so keep asking. We’ll make sure your questions are heard by the Rioters working on the stuff you’re curious about."

Kill XP Changes on PBE 

There are currently several changes to XP awarded for champion kills testing on the 7.4 PBE!

Here's Riot Colin with more information and context on the changes:
"Howdy everybody, 
Riot Colin here to give you guys some information/context about some exp changes that are currently live on PBE. We are seeing certain characters (usually supports or junglers) fall significantly behind their team in XP, even when they are playing well. They are still providing value to their team in team-fights and through utility, but these actions are frequently not rewarding enough to pull them back into the game. (Their kill XP is frequently divided across many people.) We want to increase the amount of XP they receive through rewards for kills they facilitate. We also are going to be keeping an eye on future changes for other sources of EXP that these players help earn, like Baron, Dragon and Turret kills. 
The changes listed below address this problem by adding additional XP on champion takedowns specifically for players who significantly lag behind their team. We also are slightly tuning up the “underdog” bonus on kill XP - the bonus from scoring a takedown on champions who are higher level than you. Essentially, if you outplay someone who is more powerful than you are, you should receive some amount of increased reward for your action. 
Teamwork Bonus: Bonus XP has been added on champion takedown (kill or assist) for players who are behind their team in level. The amount of bonus XP scales with how far behind your team you are.  
  • The bonus starts kicking in when you are just over one full level behind.
  • By two full levels’ difference, you get an additional ~25% of your share of the base kill experience. The bonus reaches its max value when you are 4 levels behind your team’s average, giving an additional 60% of your share of the base kill experience.
  • We put a limit of 60% on the maximum size of the bonus to prevent extreme slingshotting 
Underdog Bonus: XP bonus for takedowns (kill or assist) on champions higher level than you has been increased. Expect around 10% extra XP if you are 4 levels behind your opponent."

Phroxz0n on Fizz changes for 7.5

While Fizz's 7.4 PBE cycle changes may have been reverted, phroxz0n jumped on reddit to note the changes will return next cycle and shared an updated tentative change list:
"Urchin Strike (Q)
  • AP Ratio: 0.35 >>> 0.55

Seastone Trident (W)
  • Bleed Duration: 4 sec >>> 3 sec
  • Mark Duration: 4 sec >>> 6 sec
(Trident cries and bubbles come out of them while DoT is on, otherwise it doesn't cry).

Fish (R)

Fizz' ult is currently too load-bearing, meaning we need to cheat the ult to make his kit functional - leading to a lot of player dissatisfaction when playing against. 
By removing some of this dependency and adding some versatility back, we can have a more sustainable character."
He also posted a few thoughts on Fizz and Melee Design for those looking for more discussion on Fizz, philosophy on character design, and more. 

Normal Draft Returns to EUNE 

The normal draft queue is returning to EUNE! Here's Riot Husker with more information, noting they be evalutating match quality and queue times to see if the mode is healthy enough to maintain on EUNE!
"Hi all, 
TL;DR - We are re-enabling Normal Draft on EUNE between the hours of 12 P.M. CET (Noon) and 12 A.M. CET (Midnight) from Friday, February 10th. Based on match quality and queue times in the coming weeks we will evaluate if the queue is healthy enough to maintain. 
Why did we turn Normal Draft off? 
At the start of preseason we enabled Flex alongside Solo/Duo queue. In a few regions including EUNE where queue health did not appear sustainable, we disabled Normal Draft. Based on the information we had at the time, which showed that having all three queues active (Normal Draft, Ranked Flex and Ranked Solo/Duo) would lead to a deteriorated experience for players in queue times and matchmaking. The trade was that players could no longer play with position select in a non-ranked queues, which we heard loud and clear some of you weren’t happy about. 
Why are we turning Normal Draft back on? 
We’ve been monitoring boards and other communications platforms very closely here in Europe. We’ve seen and heard your frustration around the decision to turn Normal Draft off. 
Some of the points that were raised include: 
  • Blind Pick is not a good substitution for Normal Draft
  • Some players only like to play Normals. Removing Normal Draft restricts the way in which these players can enjoy League of Legends.
  • Normal Draft often serves as a less competitive environment to ranked play, where you can try new champions and builds etc
  • EUNE players should have the same queues available to them as EUW players 
A few things have changed since we launched Ranked Flex and Solo/Duo. We’ve seen participation in these queues settle into a healthy place. We’ve also greatly improved the viability of queues with position select through autofill and a few other features, which means there does not need to be as many players in queue overall to maintain acceptable matchmaking times and match fairness. 
The plan going forward 
With the reintroduction of Normal Draft, we’ll be actively monitoring queue health on both Normal Draft and Ranked Flex. This means that we’ll be looking at the amount of players playing in each queue and what way they decide to queue up (solo, duo, trio etc) to understand any changes in queue times and the quality of matchmaking. Be sure to keep an eye on the queue health update in our patch notes. 
If we see a degraded play experience, we’ll look at introducing some fixes to alleviate this, but there is the possibility that we may have to remove Normal Draft or adjust settings in Ranked Flex. 
We’ll be sure to communicate here on boards, ahead of time, any changes that we’re looking to implement which will affect the queues on EUNE."

Deterministic Disaster Recovery Tool (aka Chronobreak) during LCS

During the FlyQuest and Cloud 9 game in NA LCS Week W4, Riot utilized the Deterministic Disaster Recover Tool to successfully recreate the game and avoid the bug after FlyQuest was offered and accepted a remake.

The game was remade then resumed at 10:20 game time before the buggy Miss Fortune ultimate was used.  This is the first time Riot has publicly referenced or used this tool.

Over on reddit, RiotBrentmeister dove in to comment:
"Hey guys, one of the engineers that worked on this feature here. This feature was a long time in development. Obviously relieved to see it work when needed! 
I'm happy to answer any questions people have about the feature and we're also planning on doing some technical deep dives on the Riot tech blog and throuFgh other posts. 
Major props to the esports broadcast team for working so closely with us to make this successful. As well as many of the other teams at Riot that pitched in time and effort to make this tech possible. The continued effort and work at Riot to improve our codebase has made tech like this even feasible."

As for what the tool actually does from a technical stand point, RiotBrentmeister shared:
What does "Chronobreak" actually do? How did you 'remake' the game with the exact same instance to the point before the bug happens?
We use an aspect of programming called determinism. Shortly put, it means that if you give the same inputs to a program you should get the same outputs. So then it just becomes a matter of capturing 100% of the inputs (incoming network traffic, server configuration, initial game setup, game data, etc...) and replaying them back with the same timing information. 
Lots of old games (Age of Empires II for example) used determinism to build their multiplayer. It's fallen out of favor with modern multiplayer titles so we had to "backport" this feature into the existing League of Legends."
RiotBrentmeister  continued:
What was it that finally made it possible to develop this tool? Did creating replays + practice mode help facilitate this in any way?
In order for this tool to work we needed to make sure that we capture 100% of inputs for the game and when we play them back the server executes in the EXACT same way every time. 
Replays are client-side which is a bit different because the client just visualizes what it gets from the replay file. With DDR we have to get 100% of the server state recorded just right. 
To do this we rewrote the in-game clocks, random number generators and more to do this. The biggest factor that helped this was the tireless efforts of all teams at Riot to make the code more sustainable and predictable. 
Things like the missile refactor: 

Automated Testing: 
and other projects we never wrote articles about all helped us greatly."
He continued:
"It's roughly ~10MB for one game. We keep the data for about a week and throw it out when we patch the tournament realms. If there is a bug in one of the games we'll back it up in our bug database for fixing."

As for the use of this in competitive over remakes and the time it has been in dev, RiotBrentmeister noted
How long was this in development and or will this replace remake for competitive?
We hope that this can replace remake for competitive. There are classes of bugs that this cannot fix however. For instance, if the bug was something like "Aatrox is doing double the damage he should with his passive." DDRing perfectly would just leave that bug in place. 
Nothing like that would ever happen though right?/s Seriously though we did do a review of all instances before starting feature work and found DDR can recover from nearly all issues we've actually seen in competitive play. 
Development has roughly been ~8 months. You can see a more detailed explanation here:
As for how "Deterministic Diaster Recovery Tool" got it's name, they shared:
"Let me tell you an interesting dev tale. We once had a tool called "Nashor's Tooth", who's primary goal was to get players back through the login queue (remember that guy?) as quickly as possible after the platform went down. (The whole "joke" was that Nashor's Tooth gave CDR) 
Anyways, there was problem with it. However, instead of calling up the platform team who worked on the tool, the gameplay team responsible for the item got called to come into work and fix it. After much initial confusion we did eventually get the right team called in. People weren't happy with the confusion, mainly because it delayed fixing the player pain. 
So lesson learned, try to pick un-ambiguous names. 
That being said there are certainly sexier names than Deterministic Disaster Recovery Tool. Mostly people just liked the acronym... I was overruled :("

Riot  Engineering -  Under the Hood of the League Client's Hextech UI

The Riot Games Engineering blog has put out a new article by Jules Glegg that takes a tech savvy look under the hood of the new League Client's Hextech Ui!

Here's a brief introduction from the start of the article:
"Hi! I’m Jules Glegg, a Software Architect working on the League client update. We’ve been fielding a lot of questions about how the updated client’s Javascript/HTML/CSS works, and there’s a huge amount to cover there. I’m here to give a very broad overview of the most interesting parts of our UI architecture - we won’t get to cover every bolt and weld in this post alone, but we should get a good feel for the general shape of things."


Quick Hits - Internal Playtesters, Impact Challenge winners, & more

[Quick hits is our own collection within a red post collection, often including easy to digest stories, specialized information, and/or repeat info you may have missed in other posts!]

  • Several of Riot's internal playtesters dog piled into this reddit thread to answer the question - "Internal play testers, what's the funniest experience you've had in game?"

  • ricklessabandon tweeted the previous Kindred mark PBE changes will not be going out in 7.4 but other changes will return in the 7.5 cycle.
"heads up, this won't be shipping in 7.4 — going to spend a little more time on it and ship something in 7.5 (should be off pbe now)"

  • RiotAxes commented on reddit that they would like to change Bond of Stone but won't be in 7.4:
"We are exploring Bond of Stone changes, yes. No specific timeline except that they won't be in 7.4 for sure. Agree that it could use some love, though."

  • The 2017 EU LCS Spring Finals are headed to Hamburg, Germany! More information [here].


To round out this red post collection, here are a few reminders on current promotions or limited time events!
  • Valentine's Day 2017 content has been released, including new Heartseeker Lucian and Quinn skins, new summoner icons, bundles and more! Legacy content is available through February 15th.

No comments

Post a Comment