Doom Bots of Doom Q&A
Originally Posted by Riot (View Original Source)
Would you consider removing the level restriction on Doom Bots?
I was trying to play with an NA friend earlier, but couldn't because his EUW account was only level 2. Really unfortunate and super sucky.
I get why you would have a restriction, but featured game modes can be a fun, non-serious way for us to play with friends from other regions. Especially if we have a group of 5 and it's against AI.
We generally add these restrictions because game modes are alternate ways of exploring League of Legends in the context of League of Legends. We don't want a new player's first experience in League of Legends to be in an alternate game mode, where the rules don't make sense if you aren't familiar with League of Legends.
That said, we should empower players to make those decisions, and your situation is likely not an uncommon one. We will likely revisit our stance on queue restrictions for future modes.
Why did you make ip gain so low? I had to stop playing after the first day because the income made is like worse than that of beginner bots for level 30s when you consider all the losing that you will be sure to be doing. There is no motivation to keep playing after beating it once, only punishment.
I mean what kind of reward for finally triumphing over 5 bombs doom bots in a long grueling match is 60 ip? You can say it's about the journey not the reward, but when someone chooses to give you a reward so insultingly low, it diminishes the whole experience, makes you lose any sense of accomplishment and certainly does not make you want to undertake it ever again.
Our goal is to make fun experiences for people that are rewarding in their own right (yes it is about the journey). Our hope is that replayability will come from players being excited to try new champions and challenge themselves to beat the bots in interesting ways.
If we do decide to focus on rewards for participating in game modes it is unlikely that they will be IP based, we could certainly come up with some systems to celebrate game mode related achievements. That said, we could have done a better job of tuning the IP as compared to other COOP vs AI queues, we'll try to be more mindful of that going forward.
Maybe give a reward other than IP?
It sounds crazy at first because that could be seen as interfering with the financial structure, (and indirectly you)
but i was thinking, maybe give a special ward skin or something, similar to the icon for a certain number of games, or something special you have to do in-game to unlock it, like triple-kill a group of doombots.
My ideas aren't necessarily the best but the idea might be worth some thought, especially since it is about the journey and new, fresh experiences.
We would definitely like to do more rewards, and have experimented with it for events like Showdown. More experiments are coming :)
I don't think that they need to be particularly rewarding. But when the IP gain is SO abysmal, you are actually giving something up in order to play.
That's - you know - the opposite of a reward.
You are effectively PAYING IP to play the game mode, because if you didn't play it, you'd be earning more IP no matter what you did.
If the game mode at least gave the same IP as playing vs Intermediate Bots, it wouldn't feel like a total kick-in-the-nuts, but would still be less IP-favorable than playing normal games.
We agree the economics of it were a fail on our part. In the end we hope the mode was fun enough to be worth some of your time. Going forward we will make sure IP rewards are in line with what you guys would expect.
Soooo this had to be asked... but why no Bot? Seing as he is kind of Staple Mascot of Bots his absence made me kind of disapointed and sad :(
Haha yeah that was a bit of a missed opportunity. We did a bit of early exploration with Nunu Bot (including a hypothetical version where he would periodically freeze the river) but decided against it because a lot of his kit didn't lead to interesting gameplay patterns with the skill paradigm we had generally adopted in favor of skill shots and/or telegraphed sidegrades.
Level 5 Lux Bot shoots 20 laser beams when she casts Final Spark. The Light Binding would chain up to 5 times - we originally had it chaining indefinitely, but it was kind of too ridiculous. =P
Why does the Doom Bots mode have to go away?
Featured Game Modes are designed from the ground up as short-term engagement experiences. This short-term window affords us the creative wiggle room to try new things and see what works, listen to your feedback, then pull back and iterate before re-releasing. Trying to build a long-term sustainable game mode would actually constrain us from doing things like Doom Bots, URF, etc. Featured Game Modes have also been shown to taper off in popularity sharply after a short period of time. Riot Brackhar actually made a great post on this topic late last year titled, "Behind the Scenes on Featured Gameplay Modes", explaining many similar points with some cool graphs.
We're keen to re-release and bring back popular modes. When we do, they should be better experiences than the original version, with everyone's help. ^_^o
You only made AP champs as Doom bots, right? Why not make ADC champs with extra doomy items, like for example give them a ranged version build of Ravenous Hydra, and combo that with Ruinaan's Hurricane... Are items so much more difficult to modify than champions?
Suggestion: since they are doom bots, you should have given them vision of the jungle. In the games I played, it seemed that just moving between lanes a lot confused the bots (even 5 bomb bots) to the point that pushing to the nexus was easy and then the challenge was only to push the last two towers down.
Items aren't difficult to modify, but they generally lack the identity that champion abilities have. Part of what made Doom Bots entertaining was seeing the champions that you were familiar with doing things that were still distinctly reminiscent of that champion, even if what they were doing was ridiculously over-the-top. Giant Tibbers is a good example of this - it needs no explanation, and you would never fail to make the connection between it and Annie.
For this first iteration, we concentrated more on giving the bots unique identities via ability modifications. Creating new items would definitely add power and difficulty, but a lot of the uniqueness that made the bots immediately identifiable would be lost.
That's not to say that this kind of thing won't happen in the future. You guys have proven that you want something even harder, and Doom Bot-specific items is definitely a way to achieve this. =)
Why did you decide to go from lvl 1 - lvl 2 - lvl 5?
Because counting is hard. =(
It's because we wanted to more accurately portray the jump between difficulty levels to better set expectations. The gap between 1 bomb and 2 bombs was objectively smaller than the gap between 2 bombs and 5 bombs. Also 5 seemed like a pretty good number.
Hi again, :-), another question. Will we ever be able to see a game mode where we'll be able to have the Doom Abilities. Even though we would need more champions with Doom Abilities to make that more fun, I would love to be able to use Doom Tibbers >:-)
The Doom Bot's skills were designed specifically with bots in mind, and that's key to why they work at all. The bots do not optimise their behaviour in any way to take real advantage of the skills. This constant 'predictable' logic is what creates the clear counterplay patterns for Doomed abilites (Lux's laser pattern. Fiddle's ghost ulti's). This would not be the case if controlled by humans in their current shape.
There's still the seed of an idea there though, where players get crazy / warped abilities, but that's a different game mode (and would need to be balanced accordingly).
what's the crazyest thing you guys thought of when making these insane kits?
We had a version of Malphite Bot who would basically use a version of Unstoppable Force with a global cast range as a substitute for Teleport.
...His cross-map ganks were INSANE.
Can you like release a test video of that? i would LOVE to see that!
It was pretty cool, but also broken in many places. Like he could use it to just bail from a team fight he was losing.
One time Doom Malphite was 1v4'ing us, even killed a few but was starting to lose... so when things looked bad, he just used his ult to peace-out back to his base.... as far as Doom Malphite was concerned, he was "ganking his own fountain".. :/
Were the ones included the only champions you worked on? Or were there some scrapped ideas for bots that didn't make the cut?
We had a couple bots that didn't make the cut. There was a Malphite Bot who used a global version of Unstoppable Force as a Teleport substitute, a Sivir Bot who had an enormous Boomerang Blade and perma-old-school-Ricochet, and a couple of other exploratory bots that ended being leveraged for Bonus Doom abilities (e.g. Warwick, whose Blood Scent is now a bonus doom - the original Warwick Bot version had permanent Blood Scent on the bot team and IT SUCKED).
Other than having buffed/mutated abilities. What else specifically was done to the coding of the bots? It seems they are a more intelligent than even intermediate bots. Can we expect "hard" bots in future?
We wanted to keep AI changes to a minimum, for sanity's sake, but I ended up tweaking some of their sieging behaviors slightly for this mode.
A lot of the "new" behavior I noticed players commenting on were part of the pretty significant overhaul of our regular bots a few months ago (it included warding behaviors, flashing, skill shot dodging, etc).
For Doom Bots, we made sure they were taking full advantage of those new toys.
As the community and I discuss here,
I feel that Doom Bots of Doom (DBoD) fails as a Featured Game Mode (FGM). While 'fail' probably ended up being too strong of a word, I still believe in the premise. DBoD doesn't deliver the experience that past FGMs delivered on (PvP, replayability, challenge) , and isn't in line with Brackhar's description of FGMs (that is to say, DBoD doesn't deliver on the "classic League of Legends formula". With those two things in mind, my conclusion was that DBoD should have been released as a permanent addition to the Co-op versus AI game choices, and not as an FGM. Thus, my questions to the team are as follows:
Is Brackhar's quote from 8 months ago still the driving vision of Featured Game Modes? If not what is? If yes, do you agree with my assessment of DBoD and how it meets (or fails to meet) expectations?
Any chance you can/will release the statistics relating to frequency of play for DBoD relative to other FGMs?
Do you agree with other player's assessments (not mine), that players often found themselves fighting around the bots instead of against them in 5v5 teamfights?
What is the biggest lesson learned from making this mode (i.e. what would you do differently in hindsight)? What aspect of the mode are you most proud of?
Despite my disappointment in this endeavor, I'm still very grateful for all of the hard work you guys do. Thanks Rito, and looking forward to the next one even more (since I suspect it won't be against bots and I'll actually play it more than 3 times) :D
FGM's aim to let you guys discover new and unique ways to engage with your favourite champions. Whether that's encouraging interesting play patterns, new meta styles, or skill mastery, we're certainly not bound to any particular ruleset that might preclude everyone having a heap of fun. If Doom Bots were able to stir feelings of triumph in the midst of seemingly overwhelming odds, and make players feel like a badass on their champion of choice.. I'd call that a positive. ^_^o If this can be fulfilled without PvP, then I'd challenge that as a necessity of FGM's.
As with all game modes, it's still too early for us to know the full impact of Doom Bots, until the mode has run its course and we've had time to analyse it. Current anecdotal evidence though would suggest that many player's faces were melted and they laughed while it happened. :D There was much laughter and melting.
We might do another dev blog as a post-mortem on the Doom Bots of Doom, but only if it turns out there's super cool stuff worth it for you guys to check out.
hi, i really enjoy this mode (almost 50 games played!) and have some questions (which im sure may have been asked already so dont worry about it.
- why so heavily focused on ap champions?
- why couldnt we spectate friends playing (i get why, but did you think maybe you could have changed it?)
a nice thing about this game mode is that it heavily relies on your team to actually use teamwork and adjusting between offence and defence (except when you get matched with the always split pushing people, which is a good sign for an easy loss).
i look forward to future game modes, thanks.
- TL;DR: Skillshots rock, "bullet hell" gameplay sounded cool enough to experiment with, Auto Attacks were not as sexy.
- We definitely missed an opportunity with spectating doom bot matches. We weren't sure of how compelling of an experience it would be, so it stayed out of mind. Once shipped, it became clear that watching your friends get pounded was a big part of the fun.
Should we revisit this mode, we'll certainly want to make spectating games possible.
Do you have any plans to create a rotating cycle of game modes that switch out each week? Eg. Champion Free Week but with game modes instead.
This is something we've been considering. If there's a healthy way to do this, we'd like to give it a try!
What gave you the idea to make doom bots? Especially the idea to give doom bots random abilities and the factor, that they get stronger each time they die?
Where did Doom Bots come from? From a collaboration between our AI and the Featured Game Mode teams :) Once we shipped our updated AI, we wanted to explore a collaboration and see what was possible in a PvE space for League.
For the record, Doom Bots get one (2 bomb) or two (5 bomb) random Bonus Doom (randomized every time they spawn). Killing them never grants them more power; as designers, we don't want to put you in a position where not killing bots was a valid strategy.
I had a question: Where did the inspiration for the bush-doom abilities come from? Was it just there for epic, laughable fun, or was it to encourage vision warding? :3
May the supports be happy... ;-;
Having come off of implementing warding in bots, I was probably the one pushing most for some vision and brush related nightmares.
It's an aspect of the game that many players under appreciate and I wanted to create clear incentives for players to care about vision.
Also, epic, laughable fun. Because Garen.
A large part of the draw to most contemporary videogames is the feeling of empowerment that the player experiences in taking on the persona of a character that can do far more than the person behind the screen can. In Doom Bots, it seems there was an attempt to take that power away from the player. I'd love to hear about the thought process that went into creating this mode in a way that doesn't frustrate players while still retaining the difficulty that seemed to be a core goal of this mode.
We were actually very conscious about NOT giving any extra power to players against Doom Bots. As a challenge, the goal is to overcome with only the tools you would normally have at your disposal. Any extra XP, gold, power, etc, would only cheapen players' sense of accomplishment when you eventually did beat the bots. It would just feel like charity. :/
I really hope you guys read the thread I linked to earlier which explains, in part, why the challenge of overcoming doom bots is not the most engaging, enjoyable type of challenge it could be on a deep game design level. Taking power away from the player leads to the cheesy tactics described by some players in that thread (aka split push to win, do anything but fight the bots). Spinning this off as 'well lots of players seem to be enjoying the mode' is a cop-out and a fallacy. The game has approximately 3.7 gazillion players, no one doubts that lots of games were played. Competitive players probably enjoyed the novelty of it, no doubt, but novelty for novelty's sake does not good game design make.
We are well aware that bots have a weakness to split push and that players in the know could capitalize on this to secure a win. That is why we took care to design 5 Bombs to be quite difficult but doable without cheese tactics. What is enjoyable and engaging is learning the patterns of the bots and overcoming them. The fallacy is believing that the only way to win is split push with your best most OP champ.
Another thing we do with all game modes is collect and crunch a ton of data around how players are engaging with the mode. This lets us compare the success of modes so that we can say with certainty whether players are enjoying the game and coming back or just trying it out and responding with a, "meh." So do not fear, it may be that this mode didn't scratch an itch for everyone, but we're confident it hasn't fallen flat either.
Hi! I have a few questions:
- Doombots could clearly have been much harder, or much easier. How did you choose the difficulty you did, and what metrics did you use to make sure you were approaching your target?
- At what point did you get the idea to have the three difficulty levels and bonus Dooms?
- What Dooms did you reject for being too doomy?
- What changes, if any, did you make to the AI for Doombots, and what were you trying achieve with these?
- When was it decided to alter the lighting and sound for SR?
Hi! I have answers:
We focused on making these bots scarier, not necessarily more difficult. As for how we settled on the difficulties: playtests, playtests, playtests!
We knew we wanted Doom Bots to a fun experience for a large segment of our players, spanning a very disparate skill range; one-size-fits-all was not going to cut it. We ended up with 3 simply because it provided enough granularity without risking high queue times.
I don't think we shied away from anything too doomy per say, rather we scrapped ideas that didn't create new cool opportunities
See other red posts
We decided pretty early that normal SR was too bright for what this mode was turning into, really glad we managed to get our gloom on!
Ghostcrawler Big Talk Continuation
Originally Posted by Riot (View RedTracker Source)
Ghostcrawler, you use the two terms "slap fight" and "binary" as if both were bad ways for fights to play out. But in my understanding any fight which has no possibility of a "binary" outcome has no tension in it, which is exactly what makes it a "slap fight". In other words there is no other way to stage a fight, no way to avoid the "slap fight"-"binary" dichotomy.
If you have some ideal in mind of a fight which is neither binary nor slappy please enlighten me, because I don't think it is possible.
I definitely think it's possible. Your definition seems to be that an encounter will always have a victor, but even if you accept that's true, you can have encounters whose outcome is not predictable from the onset.
By "binary" I mean a situation where a melee champ is trying to fight a ranged champ. The melee champ largely can't do any damage unless he closes with the ranged champ. So if the ranged champ can effectively kite, she's going to win the encounter. This means the tendency is to buff melee damage to really high levels. The logic works like this: if the ranged character is doing 100 damage all the time, and the melee character is doing 0 damage while being kited, then the melee character has to do 200 damage when in range in order to be even. That's the binary aspect -- if I effectively kite, you lose. If you ever close, you win. It's not that extreme in reality because some melee do have ranged attacks and gap closers with limited windows and duels occur on maps with structure and other players and so on.
By "slap fight" I mean a situation where two melee champs are wailing on each other. Maybe one pops a steroid or something, but largely it comes down to damage vs damage mitigation and less about the specific choices each player is making. I agree that how those encounters interact with laning and with other champions can affect the outcome in interesting ways. Several of you have also taken me to task for over simplifying melee duels as two dudes beating on each other with axes until one falls over. I acknowledge that's an exaggeration. It's just something I'm using as a shorthand for driving melee diversity and enhancing the distinctions between fighters and tanks or bruisers and divers or whatever categorizations we can come up with. (I further acknowledge several of you are completely happy with melee as they exist today, minus their interaction with ranged perhaps, and that feedback is also good for us to hear.)
Obviously, either of the two extremes are problematic, and we've seen them both in the past -- junglers shouldn't just farm all game with no interaction with laners but at the same time they shouldn't be nonstop ganking machines.
Either extreme is okay, so long as either doesn't become the only way to play. When there is only one way to play, that hurts strategic diversity, which risks making the game boring over enough time. Strategic options are what make game 500 feel different from game 499. Too often, jungling is kill creatures, get buffs and gold, go try and gank, repeat. That's fine as an option but it would be nice if you started some games thinking "I'm really going to focus on ganking this fight," and others (even with the same champion!) you thought "I'm going full carry this time." Swapping strategies in the middle of a game depending on what else is going on is even better.
BTW, I think it's a misconception that people don't want to see Mundo or Shyvana in an LCS game. I think what people don't want to see is Mundo and Shyvana in EVERY game. I think diversity in toplane is what people want to see in LCS, and less island-like, snowbally gameplay. I think the answer to this is probably in map design less than champion design.
Maybe it's time to bring back league of bruisers since at least there the bruiser champs had a chance against ranged characters.
I get that a lot of you are saying you prefer "bruiser vs. bruiser" to "ranged dominates melee," but we don't think either encounter is a success. We'd rather fix the underlying problems than oscillate between the two. Fixing the underlying problems requires more than just buffing or nerfing stats or mana costs. We have to look at itemization. We have to make sure champions have strengths and weaknesses. We have to make sure all bruisers aren't interchangeable with regard to how they interact with other champs and the map. This is why the "melee problem" hasn't been solved yet, which I think was the initial question that launched this (pretty interesting TBH) discussion.
I feel like a lot of complaints about champion strength relative to others comes from the fact that most players in ranked refuse to experiment outside of the meta, and immediately get mad at and start giving heat to people who do pick these different champions
It's a problem for sure. It wasn't that long ago that Leblanc was the highest banned champ despite having lower winrates that many other champions. It could be that she was just so annoying to play against that it was worth the ban, but we suspect it was more that players felt like they would get grief for not banning her, since that was the conventional wisdom.
Most of the solutions here lie in the social space, which is really Lyte's purview, not mine, but part of the reason Team Builder has a leader is so you can establish "Is this a get our points and get out game, or can I try a crazy new strategy that isn't likely to work?" The best answer of all is to play with your actual friends, because they'll tolerate just about anything.
Usually this leads to "nerf x champion" instead of making other champions more viable or attractive. By nerfing a champion, you destroy their play and everything that makes them appealing.
I'd challenge this logic a little bit. If Lucian is dominant and you just buff Vayne, you get the same result than if you nerfed Lucian; you're just trading one for another. Now, if the two have different strengths and weaknesses such that the question doesn't just come down to who kills the enemy team the fastest, then the decision that gets more interesting. Using your example, if we just give MF a gap closer and more CC then it feels like an easier to solve problem of who has the best damage + gap closer + CC. If you have to choose over A+ gap closer with B+ CC or the reverse, then gameplay starts to have a bigger effect than just who has the best stats.
I agree champion diversity is important and as I said far, far above, not yet acceptable.
On the topic of Melees being "Balls of stats" vs melees that Riot considers healthy such as Renekton:
Olaf is pretty clearly an example of "Ball of stats". What does Jax count as in your view? What about melees like Wukong? Riven?
I don't think I have the bandwidth to give my two cents on every melee in the game, but let me call out Jax as a melee who doesn't have the ball of stats problem IMO. He's cooldown limited, which gives him some powerful moments but also opportunities for counterplay. Another melee can make decisions about when to engage or what abilities to use based on the state of Jax's cooldowns. Jax can make similar decisions. Once you're making decisions, the duel is much more interesting. He's the most healthy of the ones you mentioned, again IMO. We don't think the Jax solution works for every champion though or that just leads to all melee feeling the same.
So what Statikk said in the Q&A about the state of Garen..It's understandable that you guys want him to be an easier Champion for people to pick up and use, as he is a beginner Champion, but what's the actual difficulty in raising his Skill Cap? There are quite a few "Beginner Champions" that have low skill floors, but really high potential and skill caps i.e. Ryze, Warwick, Master Yi.
I'm not a fan of having "beginner champions." It does a disservice to players who really like that champion and feel like they can't have nice things because of players who might not even stick with the game. I'd say rather that there are aspects of Garen that are really popular (spin to win from brush) and we'd want to make sure to keep that in whatever work we did to him. I would have no problem raising his skill cap if we can keep those familiar aspects.
WoW was more balanced in TBC than it's ever been since. What he wanted to do was constantly change the game, the talents, and the champions to keep things fresh and attract new customers. Which led to the game being unbalanced and long-time customers getting frustrated and quitting with the changes.
BC was more balanced in a sense because there were fewer classes. Many specs were just not viable in either PvE or PvP and often both. It's much easier to balance a game with fewer classes. We could have a dozen champs in League, but many of us would probably get bored pretty quickly. What the class team did for WoW from LK and beyond was to make sure specs like Ret and Balance were something you could play beyond the leveling experience. It took some time and is not unlike trying to get all of the champs in League more competitive.
Changing any game constantly to attract new customers isn't really a sound business model IMO. True new players don't appreciate the changes anyway because they have nothing to compare it to. It is often the veterans who beg for game changes because they have played the most and get burned out if things feel stale. We change League for the veterans, but we still don't make arbitrary changes just to keep things fresh. We try to make strategic changes.
Lyte on Punishment Systems
Originally Posted by Riot (View RedTracker Source)
@Lyte what do you think about the fact that intentionally feeding and trolling in ranked results in chat restrictions?
Chat restrictions wont stop them from trolling/feeding.
Can we have an update on your plans to stop that type of ****?
Intentional Feeding is going to be handled by a different system that was built just to detect intentional feeding.
Right now, sometimes Intentional Feeding gets hit with a Chat Restriction because they also were verbally abusive, but the overlap makes it confusing. We'll make it more clear in the future why you got a specific ban, and what behavior it was for.
Still seen worse [toxic persons] that have yet to be punished.
Systems are tuned to minimize false positives. This means that most of the players that are punished probably deserve it; however, there are players that deserve punishments but aren't punished yet.
As we improve the accuracy of systems, we can increase the thresholds so that the systems start aggressively searching for wider nets of players, but, we always want to ensure low false positive rates so that neutral or positive players are never affected by these systems.
How about that full chat log and not just selectively quoting?
That was the player's full chat log that game. We don't post other players chat because it's an invasion of their privacy.
Hey Lyte, I don't know that this is the proper place to ask this, I just noticed you were here and figured I would ask.
What do you think about the impact that having a chat restricted player on your team has, in regards to lowering your chances of victory? I realize that they would be even lower if that person was cussing your team out, but I was wondering if anyone ever thought of preventing chat restricted players from playing ranked during the duration of their restriction.
Again, sorry if this is a poor choice of location to ask you this.
Actually, we find that many players never chat at all and only use smart pings and perform pretty well. For a large number of players, chat restrictions actually increases their win rates, because being verbally abusive to teammates tends to make the team's performance spiral negatively.
We have considered restricting players from playing Ranked if they currently have a penalty, RiotSocrates should have more information about this in the future.
So Lyte... your goal right now is not to get systems in place that will ensure people get punished without a person ever looking at the file in stead of just trying to improve the efficiency of the system?
No, we're testing these systems today and manually reviewing them. When the new Tribunal launches, we can use these lessons with players reviewing them instead.