League of Legends XXV: Who is your Summoner, and what does he do?

[Eldariel]

Personally, I think at the end of the day any character who revolves around being a worse version of Blitzcrank's pull is going to be a bit underwhelming.

Well, it's not worse; you can pull through enemy teams and minions to initiate on the key targets and you can flash in to do it on people currently cowering and you get a longer accompanying disable too but still, yeah, Blitz's Q isn't strictly worse and that's ult we're talking about.

[Dogmantra]

Well, it's not worse;

quicksilversashface.jpg

[dgnslyr]

I feel a bit late to the party, but what's the whole "gluttonface.jpg" joke and its variants about? It's never fun being the only not in on the joke...

[Math_Mage]

Ahh, and Qtpie gets caught and Dig pretty much auto-loses the fight. Good job, TSM.

edit: Game is very back and forth, but I want to say that if Kass had gotten all of the kills Oddone got, as well as all the blues, this game wouldn't even be going on anymore.

Yeah, I was watching that and punching the side of my head and saying "No, Phreak, get it right, the #1 most important thing in this game is Kassadin trying to contribute without kills or blue buff." Hell, TSM delayed through half their first Baron buff just waiting for blue to respawn so Kassadin could poke.

Contrast game 2 where Regi gets early kills on Zilean and proceeds to get RoA-Deathcap when Zilean only has RoA. Easy win.

[Penguinizer]

It's related to Bloodline Champions so I don't know if Dog wants to talk about it.

Anyways. It's about this.

Spoiler

Show

[Dogmantra]

I feel a bit late to the party, but what's the whole "gluttonface.jpg" joke and its variants about? It's never fun being the only not in on the joke...

so basically I was playing BLC and wanted to use the phrase "you got trollfaced" but I thought "oh, glutton looks like a troll" and it was HILARIOUS

I figure if something's funny once, it's funny a hundred times, right?

[toasty]

Yeah, I was watching that and punching the side of my head and saying "No, Phreak, get it right, the #1 most important thing in this game is Kassadin trying to contribute without kills or blue buff." Hell, TSM delayed through half their first Baron buff just waiting for blue to respawn so Kassadin could poke.

Contrast game 2 where Regi gets early kills on Zilean and proceeds to get RoA-Deathcap when Zilean only has RoA. Easy win.

I mention it on Reddit and get downvoted. Sigh...

Edit: I like how EPIK's team right now is basically "**** it. We're gonna play our best champions."

[ex cathedra]

It was a very specific line up to counter Dignitas favored champs, with Kennen filling the role the ad carry usually plays during tower sieges. While we might see more change ups in comps like that, I doubt it.

Psh, implying that they picked that team as a comprehensive counter? Outrageous. I think Dyrus just wanted to play his Karthus/Soraka metagame with Nhat.

[Joran]

Woo, Maokai, Riven, Heimerdinger, LeBlanc all seeing play!

Err... Why are there two top? Vlad + Taric vs. Ezreal + Sona.

Yorick and Karthus battling bottom.

[Eldariel]

Woo, Maokai, Riven, Heimerdinger, LeBlanc all seeing play!

Err... Why are there two top? Vlad + Taric vs. Ezreal + Sona.

Yorick and Karthus battling bottom.

Ezreal + Sona is a ridiculously strong lane. They put it on top to deny the weakest laner on EG (solo Vlad). EG had to respond by either doing a full lane switch or moving Taric to top; they chose the latter since it costs less time.

EG's problem right now is that while Dan is a good Heimer, he isn't a good AP Player which means while he does good plays with Heimer, his base understanding of AP laning and positioning is inferior to other players in this game and costing stuff like that first kill (which was 100% avoidable; Salce wouldn't have died there - Heimer was too far and too greedy). And West's jungle mechanics are mediocre at best too.

[toasty]

EG was trying to figure out how to play their "best" heroes while ignoring the meta. 100% solo queue mentality, basically.

I think they need to simply practice more. if they actually did that, then they'd probably actually get first place some time.

GG to TSM, I'm very happy that they won. Only dropped 2 games the entire event.

[ArcanistSupreme]

Probably wouldn't be too difficult.. they already have triggers/decision making processes that say 'bottom lane is alone at half health, go gank him/push his tower', it shouldn't be hard to add 'or take Dragon' to that. If the bots start semi-intelligently pursuing other map objectives, tho, I'd want to see the bot advantage settings turned down- it'd be near impossible to have a serious fight over Dragon/Baron (at level/build points where these take a while and are significant threats on their own) when the bots respawn fast enough to just swamp you in wave tactics. You win the first fight, try to take the monster, and then even if you get it you get cleaned up by the freshly respawned enemy.. it's like fighting a full team of no-cooldown Revives.

See, the respawn timers are totally borked and should be adjusted anyway. Players that know how to abuse the AI just get stupidly fed and ones that don't have a ridiculously hard time until they figure out the AI. So just adjust the respawn timers to reasonable levels and throw in some objective-based if/then statements. Call it "Hard Mode" or something, even though it'd still be easy.

Really? I would swear I've seen bots disappear and come back to lane with a buff, and I don't *think* they stole it from one of my teammates.

From what I've noticed, bots seem to take the buff camps if they ambush someone doing one of them and the critter is heavily damaged. I got ganked at red once, and I'm pretty sure I remember the Annie taking the lizard elder after unloading everything on me.

What he said. That or a teammate, who probably didn't need the buff anyway, decides to dive the bots 1v4.

EDIT: Also, go TSM! I'm glad that they've turned things around and came together. They should probably dial things back for a little bit now so they don't burn out. 2 5-hour Energy drinks a day probably isn't a healthy long-term diet.

[Makensha]

So just adjust the respawn timers to reasonable levels and throw in some objective-based if/then statements. Call it "Hard Mode" or something, even though it'd still be easy.

The respawn timer adjustment just means that the bots will have even less time in lane.

In order for a Hard Mode to really be hard, the bots HAVE to cheat, simply because they are otherwise outplayed, even if they actually go for map objectives (People will just win fights at objectives).

My Suggestions
Fair Stuff
-Make bots attempt to dodge enemy skill shots.
-Make them replace their first item slot at the end of their build.
-Fix the level 1 tower dive "cheat"
-Fix pathfinding so they don't take unnecessary tower hits.
-Don't make them AA you when you attack them under tower during early stages.
-Have them replace a defensive item with thornmails if 3 or more secretly labeled characters are labeled "auto-attackers" (Obviously can be abused by AP Trist and what not, but can one really have perfection?).
-Jungling.
-Teach Ashe how to kite.
-Have MF stop using ult when she stops hitting champions
-Switch out Cho/Shen for Rammus or Galio against AD/AP heavy teams.

Cheats
-Dominion-like passive exp
-Give Taric/Soraka a built-in blue buff, and Ashe/MF a built-in red buff.
-Give bots a free ward every 2 minutes, which they place while moving from place to place.
-Quadra Damage! Well, maybe like 5%...
-Reduce Summoner Spell cooldowns

[ArcanistSupreme]

The respawn timer adjustment just means that the bots will have even less time in lane.

In order for a Hard Mode to really be hard, the bots HAVE to cheat, simply because they are otherwise outplayed, even if they actually go for map objectives (People will just win fights at objectives).

The bots will be outplayed every time the players realize that they act the same way every time. The suggested changes are just to make the game more interesting/reasonable for both experienced/inexperienced players. That being said...

My Suggestions
Fair Stuff
-Make bots attempt to dodge enemy skill shots.
-Make them replace their first item slot at the end of their build.
-Fix the level 1 tower dive "cheat"
-Fix pathfinding so they don't take unnecessary tower hits.
-Don't make them AA you when you attack them under tower during early stages.
-Have them replace a defensive item with thornmails if 3 or more secretly labeled characters are labeled "auto-attackers" (Obviously can be abused by AP Trist and what not, but can one really have perfection?).
-Jungling.
-Teach Ashe how to kite.
-Have MF stop using ult when she stops hitting champions
-Switch out Cho/Shen for Rammus or Galio against AD/AP heavy teams.

Cheats
-Dominion-like passive exp
-Give Taric/Soraka a built-in blue buff, and Ashe/MF a built-in red buff.
-Give bots a free ward every 2 minutes, which they place while moving from place to place.
-Quadra Damage! Well, maybe like 5%...
-Reduce Summoner Spell cooldowns

...these are all really good ideas. The increased exp gain would allow the removal of the respawn hacks, and the Thornmail thing could be further refined to "If 2+ enemies have >150 Ad, Then buy Thornmail." Heck, giving the bots dynamic builds would improve them hugely. Then maybe their late-game health could be toned down to reasonable levels so that Shen can't escape 1v5 after everyone has blown their ults on him.

[Douglas]

the Thornmail thing could be further refined to "If 2+ enemies have >150 Ad, Then buy Thornmail."

If you're going to go that route, take all autoattack stats into account - AD, attack speed, crit chance, crit damage, everything that affects total damage per second that would be subject to Thornmail's damage return. Also, I'd base it on total autoattack dps of the whole team, possibly with ranged champions weighted more because they'll have an easier time applying their autoattack dps.

[Eldariel]

My Suggestions
Fair Stuff
-Make bots attempt to dodge enemy skill shots.
-Make them replace their first item slot at the end of their build.
-Fix the level 1 tower dive "cheat"
-Fix pathfinding so they don't take unnecessary tower hits.
-Don't make them AA you when you attack them under tower during early stages.
-Have them replace a defensive item with thornmails if 3 or more secretly labeled characters are labeled "auto-attackers" (Obviously can be abused by AP Trist and what not, but can one really have perfection?).
-Jungling.
-Teach Ashe how to kite.
-Have MF stop using ult when she stops hitting champions
-Switch out Cho/Shen for Rammus or Galio against AD/AP heavy teams.

Few I'd add:
- Teach them to last hit. Supereasy for a machine that keeps track of all the minion HPs and knows to a decimal his autoattack damage.
- Switch to maximum range laning if outnumbered or outpowered. In general, implement both aggressive and passive laning. One harasses whenever in range and pulls back, trying to harass and push. Other just last hits and sits back avoiding any aggression thrown its way. Maybe a third one for "deny" where it harasses but doesn't push.
- Have them use more optimal setup; jungler, solo top, solo mid, duo bot.
- React to ganks; in general, react to outside stimuli as soon as they're aware (and use Wards or cheat with maphack).
- Have triggers for group-ups for strong pushes and object takes.
- Register if opponent is missing; switch to less aggressive posture.
- When in fights, calculate who's going to win at the given pace. Too often bots try to run when they can't run, and when they could just win the trade of blows.
- Have bots adjust their defensive items based on the magic/physical damage on the enemy team (doesn't need to be based on champs; they can just read the bonus AD, ASpd, ArPen & Crit vs. MPen, AP split on the enemy team - this kinds of calculations are supereasy for machine to do on the fly and should be utilized in the AI for optimizing defensive purchases)
- For crying out loud, give them better starting items! They know how to use Pots so there's no excuse to have Ryzebot start with friggin' Doran's Ring out of all things, for instance. Boots + 3 for all or specialize further.
- Better summoners. This takes no effort. They can already use Flash-like abilities (though poorly and the enemy detection when using those needs to be enhanced so they react to opponents other than the one they're currently fighting and try to run to a direction that makes sense), and everyone should have Flash, Ghost or Cleanse at least.
- Program brushes into the AI. Currently AI might randomly lose targeting if opponent disappears into brush. They also never check brushes and they never use brushes other than accidentally themselves. For such a huge part of the game, it'd be supereasy to program them to use 'em more efficiently (since all that involves is program their existence into the AI; "if move past brush, check if enemy there" and "if attacking a weak target, check brush if losing line of sight").
- For crying out loud, teach them to use Elixirs (as "OH **** WE BEHIND ELIXIR AND PUSH!"-buttons and when finishing builds) and Oracle's ("IF ENEMY TEAM HAS STEALTH HEROES, X GETS ORACLE'S AND GROUP UP TRIGGERS ARE LOOSENED, AGGRESSIVE LANING TRIGGERS HARSHER").
- Yeah, teach them to sell items after finishing builds and upgrading until reaching optimal build for the game (even just programming 3-4 complete builds for each would be a massive improvement, far beyond your average LoL player in fact, and would increase the challenge going into the game with a subpar comp tremendously).
- Teach all of them how to move in combat, in general. Melees to stutterstep, supports, mages & carries to stick back, tanks initiate; and all switch targets to the perceived easiest target (but never abandoning a sure kill).
- Reacting, in addition to skillshots, to AOE damage abilities. Currently they happily run into Anivia's Glacial Storm or Singed's Poison Cloud (and then turn around) which leads to hilarity. Also, they notably ignore Heimer turrets instead of just killing them. In other words, just program them to register the existence of non-minion non-champion objects on the map.
- Calls for help trigger teleports from any available champions.
- For crying out loud, make them not cross the river early without a near-guaranteed kill, or at least program them not to run into enemy jungle and then try to run back through a live enemy turret without friendly minions nearby. In general, make them register turrets as "BIG THREAT, AVOID IF FRIENDLY MINION NOT NEARBY!" early on.


It would take about 10000 words of code to make them a threat to about 50% of LoL's userbase, especially if still allowed to cheat. With cheats, they can obviously be tailored to be problematic for everyone.

[Makensha]

With cheats, they can obviously be tailored to be problematic for everyone.

The cheats aren't there to make them problematic, they are there to cover the bots' inadequacies. Like blue buff Taric to cover up that Taric hates his mana pool and attempts to rid himself of it as soon as possible. Exp so that they won't fall so far behind in levels. Extra damage because they suck at fighting. That kind of stuff.

[Eldariel]

The cheats aren't there to make them problematic, they are there to cover the bots' inadequacies. Like blue buff Taric to cover up that Taric hates his mana pool and attempts to rid himself of it as soon as possible. Exp so that they won't fall so far behind in levels. Extra damage because they suck at fighting. That kind of stuff.

Aye. But if desired, more could be done to bring them in line with stronger players. Of course, the problem is making them too strong which forces abusing the AI to win, which is obviously not desirable.

[Arbitrarity]

I'm not sure, but I bet they don't have a very developed framework for creating bots. Examples of things I'd want to make as developer if I was going to implement their AI:
getTargets(self, class=champion) gets targets based on visibility and distance (reasonable distance, i.e. 1500)
Predict(self, skill, target, intersect, exact): Calculates skillshot vector based on current target speed, skillshot projectile speed and delay, and current target vector. While it's a pretty naive algorithm, this should about as well as most players do, though it doesn't deal with blinks or jukes well (which is fine). I doubt bots are going to be able to predict jukes or blinks anyways; not enough information. Add an optional (exact) parameter for non-line skillshots (and Fish/axe). This also doesn't do well isofar as efficient use of AOE's goes, so call it mostly when targetting preference is determined.
chooseTarget(self, [targets], priority): Calculates preferred target among a group of enemies. Should be based on target health and damage output by default, but can be passed an optional function to calculate priority in other ways (i.e. tanks might have a more damage-oriented targetting). Should also call the next function to avoid some silly baiting that you see among bots now.
isReachable(self, skill=self.autoattack, target): Calculates movement speeds and available gap closers to see if the character is able to use the specified skill on the target. Slightly naive about movement speed boosting skills and blinks, so this might need some refining functions to become useful. The big test is really behavior though.
isKillable(target): Checks if it's possible for the bot to kill the target in some reasonable period of time using all skills and autoattacks. I'd say if the bot can't reach and kill the target in under 20-30 seconds, that would be non-viable.
Damage(character, skill, target). Calculates exact damage on target given their stats. Should have a few calculations calling it, like total damage and possibly threat assessment. Other functions might include calculating target health after damage, damage as a percentage of maximum or current health, and so on.

I'd probably want a championAI object, containing fields for states and so forth. It should be subclassed down to various AI "types" and those should describe general behaviors. The specifics can be calculated based on the individual champion abilities, or overriden, as the case may be. For example, you might have tanks with a general CC(chooseTarget(getTargets(enemyChampions)), expectedDamageOutput), which would prioritize champions based on damage output and ability of the tank to reach them, and attempt to apply CC.

A problem this might have is that bots could go all over the place in teamfights. For that reason, having a Team object that encompasses the bots would be a pretty good idea, as it could dictate more general strategy, and call bot functions as appropriate. By having bots filling fields (like Tank) in the Team's roles, the Team object could coordinate bot strategy.

The most important part of this, however, is developing the initial scripting language, where you get useful behaviors encompassed in simple functions. Additionally, if you make those part of an object that bots inherit from, bot classes can override (virtualize, if you've done C++) those functions to tweak individual bot behavior, which lets you do complicated customization for specific champions. I'd hope you develop robust functions in the first place, but it's quite possible that, say, Zilean might need a very specific isReachable, given the nature of his movespeed boost/slow, and his Rewind. I doubt (distance-self.gapCloser()-self.msBoostTime()*self.msBoostSpeed)/(self.ms - target.ms) would quite cover Rewind. And look at how many functions I need to make that work. At least I can see how I'd calculate those....

Dammit now this sounds like a really interesting project. Too bad it's so low priority.

[Math_Mage]

Waitwaitwaitwait. I thought the Rioters vs. Bots match had hard AI? Because that's the impression I get from other players, but I've never actually gone to look at the match they broadcast.

[toasty]

Waitwaitwaitwait. I thought the Rioters vs. Bots match had hard AI? Because that's the impression I get from other players, but I've never actually gone to look at the match they broadcast.

THe Rioters lost, and we were given the impression that the AI was "good" but as soon as it was released, it became apparent that the Bots were terrible and easy to abuse.

[Arbitrarity]

I thought the point of that game was to demonstrate that there are a lot of Rioters who aren't very good at the game?

[toasty]

I thought the point of that game was to demonstrate that there are a lot of Rioters who aren't very good at the game?

They lost to bots while trying. No offense, but I suspect that they . . . weren't trying very hard.

In other news I have an interesting observation about the matchmakers and Solo Queue mentality:

Very often I will go on a winning streak. This winning streak will usually be a result of me playing solo top or jungle and doing very well. What will then happen is I end up 4th or 5th pick on my team. In these matches I am forced to play support or a role I am not good at: mage, AD carry, support. What then happens is that someone on my team is terrible and I am not playing a role I am good at and despite trying and maybe even doing average, I fail.

For instance, today I played 2 games: 1 as jungle WW (so actually got a role I am good at, but I didn't pick a very good ganking jungler and I was aware of this. I felt that I was good at WW tho, and I wanted to test Zach's theory that he was still a good jungle) and then 1 as AP gragas middle.

First game: we had an akali that said "I am top" even tho we had a nasus. They both picked at the same time and I am sure our nasus assumed Akali would simply go middle. But she refused. She then started boots 3 with neither of her passives: not a good sign. We ended up sending Ashe middle and Nasus bottom. I did my best to provide pressure while fast tracking to six. In the end, me and Nasus got pretty fed (I felt the pain of having no items as the jungler, since I haven't played anything but lanewick in forever), but our Akali died constantly after a lane change (she blamed Ashe for not switching, I honeslty don't entirely blame her, but I still feel that Ashe was a "nicer" person so to speak). Ashe middle did okay until tryndmere just started to wail on her and beat her with superior lane sustain. Ashe didn't buy any lifesteal the entire game. You can't gank Trynd post 6 middle so QQ. We lost.

Second Game: Oh my god our teamcomp. A player I suspect was Korean picked Yi after our first pick picked amumu and our second pick picked Gangplank. I was facepalming. I ended up playing Gragas against Ryze. I lose a lot of health giving amumu a strong leash because he started Regrowth pendant. I do okay against ryze despite this. In fact, in a game where all lanes do "okay" I would have been in a very good position, despite not getting a single blue from amumu and ryze getting 1-2 from Lee Sin. HOWEVER: Yi cannot stop Trynd from pushing. Pretty sure this was, again, because Yi build no lifesteal in lane and couldn't push as fast as Trynd. Trynd got 2 towers. Amumu is level 2 when Lee Sin is level 4. Why? Because Lee Sin does what any good lee sin would do: punish our amumu for first picking amumu. I dunno what bottom did but it wasn't enough. I really felt I did good: I held my lane against some crazy pushes and got 2 kills off ryze (1 with a gank). I also outfarmed him despite my ****ty start with basically half health and no pots (I start dorans ring on gragas. Dat Passive).

Anyways, I think my complaint about the solo queue MM is this:
I understand that it should be a "No brainer" that the 1st pick is the best player and is carrying his team (I mean... yeah) but it seems to me the power of first pick in solo queue is that you get to play whoever you want and whatever role you want. This means you get to play the role you are best at. If I could garuntee that I played top solo every game I played, I'm sure my winrate would skyrocket. But I can't, because in ranked, people don't let you do that. *sigh*

What's even more frustrating is this: our akali player, game 1? 100% winrate with Nid after 6 games. Lost both games with akali she played. Methinks she is one of those people who play Akali "because she is OP" not because they know how to play Akali.

[endoperez]

The bots are challenging for starting players, which is good enough. Since good players can beat bots even if they're made more clever, there's no reason at all to ímprove bots.

A level 5 team can easily lose aginst bots, unless at least one of them knows how to abuse the bot AI to get super-fed, and even that might not be enough. The same is true for a level 10 team. My smurf is currently level 12. I last lost a bot game at level 11, but it might happen again.

Why does this happen?

The bots have better stats (level 30 masteries + runes, perhaps even over 30 mastery points), get more gold than poor farmers, and do build defensive items. If at least two of the 5 human players are new (as they usually are) and not that good, that means they're starting at a disadvantage compared to bots, and since they can't get early kills and aren't good at farming yet, they're likely not getting the items they'd need to stay relevant against bots. Since they're new, they might end up dying a lot, and won't have the map awareness to know when to defend towers. So towers are lost, the game takes lower, the gold difference between teams grows bigger, and the bots usually win after that.

And that's assuming everyone knows what they're supposed to be doing and don't build attack speed as triple boots as Kassadin.

[Dogmantra]

So I've been playing a lot of Janna recently (she's scarily good at carrying ranked games at my current ELO, lots of ways of saving team-mates who screw up or just get caught) and I had a thought last night which was supposed to be a silly thought but it turned out quite interesting:

chalice of gold/5

Basically, the less gold you have, the higher your gold income (to a point, obviously there is no cap on gold so if it were to work like chalice, I suppose you'd instantly get effectively infinity gold) but what I'm talking about is an item that gives you maybe three extra gold/5 when you're below a low amount of gold, two when below a higher number and one beyond that or something. It's just that with the slow gold income that supports tend to get (even though I want to I have a hard time taking farm in bot lane when I'm with a carry - 's partly why I think deliberately trying to not get farm is silly, if the carry could have got it, they almost certainly would have) I always feel bad when I port because I know my team expects me to buy wards and I expect me to actually get any items at all. This item would make me feel better about saying "oh well I had to go back but I didn't have enough for any proper items, guess I'll load up on wards" because normally I think if I didn't have enough money for a proper item that I should save my money so I can get that proper item sooner.

[Math_Mage]

Okay, yeah, I went ahead and watched the bot match, and now I have no idea where people get the idea that Riot has some secret hard mode they haven't released yet. If those were hard mode bots, I'm Dyrus.

The bots are challenging for starting players, which is good enough. Since good players can beat bots even if they're made more clever, there's no reason at all to ímprove bots.

A level 5 team can easily lose aginst bots, unless at least one of them knows how to abuse the bot AI to get super-fed, and even that might not be enough. The same is true for a level 10 team. My smurf is currently level 12. I last lost a bot game at level 11, but it might happen again.

Why does this happen?

The bots have better stats (level 30 masteries + runes, perhaps even over 30 mastery points), get more gold than poor farmers, and do build defensive items. If at least two of the 5 human players are new (as they usually are) and not that good, that means they're starting at a disadvantage compared to bots, and since they can't get early kills and aren't good at farming yet, they're likely not getting the items they'd need to stay relevant against bots. Since they're new, they might end up dying a lot, and won't have the map awareness to know when to defend towers. So towers are lost, the game takes lower, the gold difference between teams grows bigger, and the bots usually win after that.

And that's assuming everyone knows what they're supposed to be doing and don't build attack speed as triple boots as Kassadin.

Nobody's suggesting getting rid of either Easy or Intermediate bots.

So I've been playing a lot of Janna recently (she's scarily good at carrying ranked games at my current ELO, lots of ways of saving team-mates who screw up or just get caught) and I had a thought last night which was supposed to be a silly thought but it turned out quite interesting:

chalice of gold/5

Basically, the less gold you have, the higher your gold income (to a point, obviously there is no cap on gold so if it were to work like chalice, I suppose you'd instantly get effectively infinity gold) but what I'm talking about is an item that gives you maybe three extra gold/5 when you're below a low amount of gold, two when below a higher number and one beyond that or something. It's just that with the slow gold income that supports tend to get (even though I want to I have a hard time taking farm in bot lane when I'm with a carry - 's partly why I think deliberately trying to not get farm is silly, if the carry could have got it, they almost certainly would have) I always feel bad when I port because I know my team expects me to buy wards and I expect me to actually get any items at all. This item would make me feel better about saying "oh well I had to go back but I didn't have enough for any proper items, guess I'll load up on wards" because normally I think if I didn't have enough money for a proper item that I should save my money so I can get that proper item sooner.

-The times when I'm bothered by supports that try to CS, it's usually when my IRL buddy plays Maokai and starts spamming Saplings on incoming caster minions. Also there's that issue where you start to autoattack for a last hit but it dies to someone else's autoattack, and then you end up attacking a different minion and missing the last hit on that one as well.
-Chalice of g/5 would seem to incentivize bad gameplay by making supports want to back more than they should. But at least it's one support item that wouldn't get abused by non-supports.

EDIT: Is it just me, or is Tryndamere roflstomping post-patch? And that's just Lanedamere, who's at least counterable with Nasus/Teemo and a few others (my last game saw one go down to Irelia top and I dominated the previous one with Singed, but I don't know if that's expected); I bet his jungle clear got a lot safer, and he's still a really strong ganker.

[Dogmantra]

-The times when I'm bothered by supports that try to CS, it's usually when my IRL buddy plays Maokai and starts spamming Saplings on incoming caster minions. Also there's that issue where you start to autoattack for a last hit but it dies to someone else's autoattack, and then you end up attacking a different minion and missing the last hit on that one as well.
-Chalice of g/5 would seem to incentivize bad gameplay by making supports want to back more than they should. But at least it's one support item that wouldn't get abused by non-supports.

On the first point, I think that's not supports trying to farm, that's just bad gameplay. Spamming saplings in lane is going to push and get you ganked (and run you out of mana quickly).

On the second point... ehhhh, I don't see a problem with it. They can go back a lot if they can get away with it, but if they back at a stupid time then obviously the opposing lane will be 2v1 and they can kill whoever they're laning with. It does sort of incentivise bad gameplay by making nofarm more viable though.

[ArcanistSupreme]

What would be the best way to call Riot's attention to how simple it would be to improve the bots? Could we create a petition on the LoL forums? I feel like having bots demonstrate how important basic skills are would drastically improve the average skill level of players and the value of playing bot games.

I wonder what what happen if someone did all of the code and then sent it in to the summoner showcase.

[Djinn_in_Tonic]

What would be the best way to call Riot's attention to how simple it would be to improve the bots? Could we create a petition on the LoL forums? I feel like having bots demonstrate how important basic skills are would drastically improve the average skill level of players and the value of playing bot games.

Honestly? It takes a long time to build an AI, especially one with as many variables as you guys want to see. Especially so if the AI isn't meant to be perfect...a perfect AI is much easier to build than one that it just *good.* If they focused on the AI improvement you want, we'd see a marked decrease in all other content, including champions and patches (or an incredibly long time before any AI improvement could ship out).

I don't know about you, but I don't value AI that highly at my level of play. It works for beginners, and I can always solo queue or ranked queue if I want a challenge.

I wonder what what happen if someone did all of the code and then sent it in to the summoner showcase.

What would happen is that that person would be hired on the spot, and I'd be serious impressed at someone who managed, by themselves, to build a complex, champion-specific AI for a game with 90+ champions while still holding down a different job and/or not going mad.

[Terazul]

What would be the best way to call Riot's attention to how simple it would be to improve the bots? Could we create a petition on the LoL forums? I feel like having bots demonstrate how important basic skills are would drastically improve the average skill level of players and the value of playing bot games.

I'd honestly rather not, and feel quite the opposite. The new (and old) tutorial serves its function pretty adequately in demonstrating the basic skills; "Don't tower dive." "Last hitting is important." "The brush hides you from sight.". A couple of those take lots of practice to get good at, especially across different champions, but I really am starting to dislike the dependence on bot games.

It's already been a pain in the you-know-what trying to get my newer friends to the game to unclench their teeth from the teat that is bot games, because they were initially too scared to go into pvp, and despite being well into their 20s in levels (from bot games), some still are. I mean, it was fairly useful early on so that I could play with my lowbies without wrecking their ELO, but now it's kinda... ugh. I feel like having people NOT play bot games all the way to 30 would drastically improve the average skill level of players, because they'd learn to do deal with things they can't telegraph, place wards, and fight more than a handful of champions that always go to roughly the same lanes and always buy the same items. Maybe if they played more regular games I wouldn't end up with Sivirs who think they can just run past enemies and they'll ignore them because they're already chasing somebody, or other zany things.

I mean, it's kind of the point of the game anyway, and I made it to where I am just fine without bots. Most of the suggested changes to make them "harder" could just be avoided by making people play games against actual people in the first place.