Boards›[ARCHIVED] Gameplay+

The Cost of Failure: Does League's Design Discourage Champion Interaction?

ModThe Djinn·10/4/2018, 6:35:56 PM·20 votes·27,458 views

With Worlds upon is, it's worth taking a look at the comparatively slow and reserved gameplay that we often see in the pro scene. It's definitely a factor of the superior ability of high-skill players to see opportunities and make fewer bad decisions, but it leads me to wonder something else:

##Does League's core gameplay loop discourage interaction at the highest levels of gameplay?

Let's take a look at a few key ideas:

  • Long-term vs. short-term gains/losses.
  • The multiplicative cost of failure.
  • Loss aversion.
  • Design of Aggressor / Defender Vulnerability.

#Long-Term vs. Short-Term Gains/Losses League's objectives can be summarized as either long-term (gold, xp, tower advantage, dragon buffs) or short-term (vision, red/blue buff, Baron, Herald). Where I think this becomes an issue is that the gameplay interaction moments are mainly tied to short-term interactions (scuttle fights, baron/dragon fights, etc.), while the potential losses accrued in lost XP, gold, map presence, and towers is a long-term loss that can put you much further behind. The exceptions are Dragon buff and Baron buff but, while fairly huge, most teams would rather stall and attempt a last-minute attack on the taking team (or drive them away and delay the objective) than actually attempt to fight for the objective in the first place.

My suspicion is that the potential long-term loss outweighs the perceived benefit of the short-term gain. You can always take that tower later, or stall that dragon fight for 2 more levels, or deal with the enemy team having a Dragon buff, or fight against them with Baron, and doing so is almost always preferable to picking a potentially losing fight and losing both...and the cost of losing that fight and losing both is magnitudes greater than the cost of losing one or the other.

The result? Cautious gameplay that strongly encourages avoiding risks, despite those risks making for some of the most exciting gameplay in League.

#The Multiplicative Cost of Failure Losing a fight sucks. Losing an objective sucks. What's worse, however, is knowing how much these mistakes can affect you far into the game. Because of the long-term gains associated with victories and the long-term losses associated with failures, each failure sets you up for future failure through mechanics much harder to fight against than some other genres. In a game like Overwatch you may be behind some ultimate charge, but you're not fighting from a position of dealing less damage for the rest of the game -- you just lose momentum and positioning. In League, a loss puts you at a level and gold disadvantage going forward -- something that only victories (now harder to earn) or a long game can turn around. Additionally, as more of the objectives are designed for long-term advantage, you can't even make as many risky pre-objective plays as you could in a game like Heroes of the Storm, where a risky, XP-losing play that ends up netting you a 5v4 advantage going into a major objective can be completely viable. Since League's objectives are happy to wait for longer and offer less immediate reward (for the most part), this strategy is far less functional.

This magnifies the existing point -- it's not worth risking a compounding failure for a chance at reversal until it already feels like a last-ditch effort. The incentive to attempt risky plays earlier isn't made clear, and so that leads teams to play safely while behind in most cases.

#Loss Aversion Hand-in-hand with the previous point is loss aversion. Taking risks and losing feels bad, and sometimes your team will even yell at you. It feels a lot worse to get stomped than it feels to win a fight, so human psychology plays a part in telling us "don't take that risk -- you can't win," and in making our team feel that way even if we see a good opportunity to go in. Combined with the multiplicative cost of failure and the likelihood of winning long-term gains instead of short term games (short term gains FEELING stronger even if they aren't) and you have an environment where winning an engagement often doesn't FEEL worth it even if it was. If you can't get an objective or map control off of the victory, it rings hollow even if it's really a crucial turnaround.

#Design of Aggressor / Defender Vulnerability. This is an odd one, but I've been wondering lately if League could benefit from some tuning here.

In many cases (Dragon, Baron, Invades, Tower-Attacking, lane-dueling) the aggressor has a natural disadvantage -- they're hurt by the monster and in a specific, known area, they're walking into unknown territory, they're taking minion aggro, they're taking tower aggro. This is discouraging to aggressive plays, especially from a team or player who is behind. The further you are behind, the more aggression is punished, unless you're willing to wait and hope for that one perfect moment.

Yet if you're not willing to be aggressive, the cost is high -- a team that holds back can end up fighting into multiple Dragon buffs, or into Baron, which put them at either a game-long disadvantage or on a timer that effectively says "better defend for X minutes unless we mess up." Yet -- as mentioned above -- the risk of dying into that and giving the team more time with Baron that you CAN'T defend is typically seen as the better option unless the situation is truly desperate.

In short, the team that makes the first play is typically at a disadvantage unless that play succeeds, and a team that is behind is often penalized more for not having opportunities to attack.

#CLOSING THOUGHTS I'll begin by saying that I'm not sure this IS a problem or, if it is, how big a problem it is. It seems, however, that League's heavy and multiplicative penalties to failure, along with its tendency to favor long-term advantages over short-term advantages, may lead to a game where teams are hesitant to take actions unless/until comfortably ahead or in an ideal position, and feel like they're struggling uphill to attempt to regain any lost advantage or momentum.

I can't help but wonder if switching to more immediately impactful and appreciable objective design (such as Herald), combined with more frequent objectives and less penalty for losing teamfights, might be a net benefit to both the pro scene, casual scene, and player frustration in general. Nexus Blitz, for all the problems you could probably find in it, seemed to have a fun, frantic feel that encouraged big plays and engagement, and had victories and losses that felt more immediately impactful while not feeling as detrimental if you failed to secure them. Is there something in that feeling that could perhaps benefit Summoner's Rift and improve the gameplay feel of League? Maybe.

Thoughts?

23 Comments

ModUlanopo10/7/2018, 5:38:48 PM6 votes

I think this ultimately comes down to the question of what League is "about". Is it a strategic game where the location and timing of a fight are important or is it a fighting game with strategic elements? My feeling is that League used to be the former and has gradually morphed into the latter.

Fisherman Fizz10/15/2018, 11:39:08 PM3 votes

For the aggressor/defender thing, I think it's pretty important that the aggressor is put at a disadvantage. The extent to which they're put at a disadvantage is a lot more complicated but interesting to think about.

To start, I'm going to operate off the assumption that snowballing/comeback mechanics are inherently necessary for league to be a good/fun game. The length of SR games pretty much dictates that. Without snowballing, you'd regularly have the first 30+ minutes of a game not matter, because a team that plays better for that time won't get any meaningful advantage out of it, and instead only the final minutes of the game would matter. Conversely, without any sort of comeback mechanics, only the first few minutes of a game would matter, and anything after that would be a "trapped game," where the winning team has to spend a lot of time taking objective after objective while the losing team is helpless to do anything but watch themselves lose for 20-30 minutes.

Now, having this kind of aggressor disadvantage on objectives serves as a really important comeback mechanic and acts as a lever that can help control how strong snowballing is in the game as a whole. This is because the pressure is on the ahead team to be taking these objectives almost always. Once a team has an advantage, it's on them to leverage their lead to take more objectives so that they can continue to extend their lead and close out a game. Any time spent not doing this is beneficial to the behind team.

For the behind team, there's no reason for them to actually make any aggression. It's in their best interest to focus on minimizing losses, stalling, and punishing mistakes. Aggression made by the behind team will almost always have to be in reaction to a mistake the ahead team made, because if they try to be aggressive on even terms they just lose. So if the ahead team isn't using their lead to take more and more advantages, the game stalls out and benefits the behind team. If a team is up 23k gold to 20k gold and does nothing with that lead, over time that 3k gold advantage will mean less and less. If it stalls out and that team is up 53k gold to 50k, the % difference in strength between the two teams is much more even.

So then, the reason why it's important for the aggressor to be at a disadvantage is because it gives a losing team some sort of advantage they can leverage to try and come back into the game. Without that disadvantage, they'd have fewer chances to try and take a fight and come back into the game. For example, lets say Baron does 0 damage. There will still be a few small disadvantages a team has when taking it (restricted positioning because they have to be near the pit, time, mana and cooldowns required to actually kill it), so the losing team has no chance of actually contesting it and they're almost guaranteed to just give it up and fall more behind.

On the other hand though, too much of a disadvantage can also suck. The higher the disadvantage, the more of a lead a team has to have before they can safely take something, and the higher the risk and uncertainty there will be whenever making the decision to take something. The first of those effects means that games are much harder to close out and will stall out way more often. The second is that risk aversion will play a bigger part in making games passive, which would get worse the higher in MMR you go and then drastically get worse still in pro games.

So it's important to have that disadvantage, but it's mostly a matter of finding a sweet spot where said disadvantage isn't too much or too little. There's also the problem of tackling the disparity between low MMR games, high MMR games, and pro games where risk aversion will be more/less important. Generally, the more skilled a player is, the more risk averse they will be. That's because, with 0 risk involved, a better player will win against a worst player every time. The higher the risk, the higher the chance that a worse player can get lucky and win despite the skill difference. In pro games, you see much more risk aversion because they have so few games to play, each with a huge impact. Worlds finals is one best of 5 series, so losing one game because of a risk is huge. In solo Q, it doesn't matter because players can play 100s of games, so the law of large numbers will prevent risk from really mattering that much.

Lastly, I dont think any of this is disagreeing with you btw, especially since you mentioned tuning specifically. I just thought it was an interesting point and wanted to expand on it a little :3

Galiö10/7/2018, 5:07:35 PM2 votes

I think the largest factors of this issue is the current state of the game.

The game had been trying to make the Game focus more on individual play over team play which leads to a more snowball-y game.

What I mean by this is Solo Q players wanted to have more agency on determining the outcome of the game or be able to 'solo carry' games; viewers of competative wanted to see this style of play more in pro.

Pro players however, do not spam 100s of games to reach the top of the worlds ladder while as in Solo Q if you lose a game you do not lose much.

Pro players are more likely to be risk adverse especially when the game is understood more at that level, and even individual levels of skill/weaknesses are understood more.

Example: If you know your bot is weaker or your overall early game as a group is weak, you choose a hard scaling bot laner and forgo an interactive laning phase.

If your mid is weak choose a champion with a safety net that can clear creep waves well and assist another player to win the game to mitigate interaction and/or draw resources into your lane.

Players will want to be risk adverse due to losing a lane will require the team to pool more resources into another lane: more jungle presence, harder shoves penalizing your own CS, more cross map play.

In metas where it is more about the team, players can take some more risks because they game isnt actually decided well into late game. Leading in a more interesting (sometimes) early but huge risk aversion late game until the big BARON POWER PLAY.

In metas where its more about the individual the game can be decided in the mid game and sometimes even early. Leading to not wnqting to take risk at any point in the game.

For Solo Q players tend to not be as risk adverse, although this does happen sometimes, due to not having as much faith in other players and more 'fun' in 'going ham.

While as pros will never want to have huge risks because their career/pay/glory is on the line in a slow team style meta or a fast snowball meta.

I dont think itll ever matter how many objectives or incentives there are in the game to change this other than a mode like nexus blitz that FORCES players to react.

chipndip110/7/2018, 9:38:46 PM2 votes

Here's the thing: High level play in any game will ALWAYS boil down to "Most gains possible through the least risk taken". Pro play will never be about making the riskiest, shounen protag play possible and managing to barely make it through via the sheer will power you gain through friendship and the power of justice. Calculated risks are king in pro play for ANY game. League is no different. In Street Fighter, you land "hit confirms" off your fast, safe attacks before doing the big hits, sacrificing total damage for a reliable, safe way to do the damage. In League, you farm and fish with poking abilities or wait for your jungler to force an interaction.

PandaNator4310/8/2018, 4:27:38 PM2 votes

It sounds like most of what you talk about in Long-Term vs. Short-Term Gains/Losses and The Multiplicative Cost of Failure are granular level explanations of snowball.

By that definition, it seems Riot themselves agrees with you, as reworking snowball (its strength and where it comes from) is a major focus for preseason.

I think you bring up a lot of good design challenges.

In a game like Overwatch you may be behind some ultimate charge, but you're not fighting from a position of dealing less damage for the rest of the game -- you just lose momentum and positioning. In League, a loss puts you at a level and gold disadvantage going forward -- something that only victories (now harder to earn) or a long game can turn around.

This is something I both love and hate about league. Its fun to get Big and destroy ppl, but also sucks to play from behind.

**It would be interesting to see if kills/objectives no longer granted bonus gold, but instead grant increased passive gold generation. **So a kill gives less of an immediate power increase, but potentially more over time. I could see it allowing the behind team more of an opportunity to catch up (desirable), but also make the winning team more passive to wait out the game and let their passive gold gen advantage snowball (not desirable).

What do you think?

ModKnightsKemplar10/8/2018, 6:34:32 PM2 votes

I'm sort of with Ulan on this, but to put a finer point on it, I don't see any of the problems you mention as being problematic.

The other wonderful thing about League is that a game only lasts 30-45 minutes (typically). Yeah, you sometimes get into a bad loop where it feels like you can't do anything, but that's no different from any other strategy game. This is true in classics, like the board game Risk, all the way up to a 4x strategy game like Civilization. League benefits from what is a relatively quick turnaround window, compared to many other strategy games. When I look at it from that perspective, League starts looking a lot better.

The temptation to compare to Overwatch is tantalizing, since they are both popular eSports, but I actually think that's apples and oranges, because Overwatch, at its core, is not a strategy game. It's a fast paced shooter. The goals are entirely different.

On top of that, the bounty system is in place to counter loss aversion and multiplicative costs of failure. The onus should be on the winning team to snowball as best they can. If they get lazy or sloppy, the defending team can quickly turn it around because of the way bounties work. We've seen comebacks at worlds, and I see it all the time in my games. I'm a sub-par laner, especially considering my rank, but I excel at macro compared to my League peers. As a result, a lot of my wins are comeback wins. From my perspective, reports of snowball are exaggerated, particularly by those with poor macro. For them, if they ever manage to get behind, they'll likely lose. But that's a problem with their play, not League.

The reality is, there's a risk associated with being ahead because of bounties. Wouldn't it be interesting if they added bounties to things like towers, or baron? They're already moving this direction a bit by associating bounties with cs leads... and I think that's a good thing in the long run, since kills aren't the only thing that gives a team an advantage. If that looks good, and they want to push it further, they could start giving gold bonuses for breaking the enemy's drag or baron streak, as well. There are levers they have yet to touch that can deepen the strategy component of the game, and I think that's what a large portion of the playerbase likes about League. I expect to see more of this in the future.

ModUlanopo10/7/2018, 2:03:33 AM1 votes

Approved

Linna Excel11/10/2018, 1:54:24 AM1 votes

Something I'm pretty sure I talked about before, or meant to, was that League lacks a comeback mechanic like real sports.

What happens in basketball when one team scores? The other team gets the ball. What happens in football when one team scores? The other team gets the ball. What happens in soccer when one team scores? The other team gets the ball. What happens in hockey when one team scores? The other team gets the puck. What happens in baseball when... okay baseball is a little wonky in that the onus is on the defense to shut down the other side's offense but eventually the team gets the ball.

League's problem is there isn't any ball that the team who lost the objective can get control over. In all other real sports, what happens in one point of the game doesn't directly affect anything going forward but league with it's RPG elements of gold and levels kind of screws over that concept. This is why I think that overwatch has an advantage over league in terms of game design: it doesn't have those punishing RPG mechanics holding the other team back. I think that the easiest way to fix this is that when one team gets an objective of any kind, the other team gets gold/xp instead of the team taking the objective. This would be a way of giving the other team the ball.

So what does this have to do with interaction? Well to be honest if you know that you'd get a chance to come back into the game if you screwed things up and lost an objective, it'd be easier to interact with the other team because it's not a total loss.

Take the most severe example of the punishing nature of interactions in league. Imagine being in a lane where you are hard countered and if you die once, the lane is lost. Maybe even the game depending on the champ you fed and the tempo of the meta. You absolutely can't interact in that situation because there's such a heavy cost for a failed interaction. One screwing literally throwing the game for your team is a huge burden. That match up is both unfair to the player on the weak end and a toxic game design because even the smallest mistake in laning can cost your team a game.

Then you've got that 10-20 minutes after laning in which the rest of the game is just a formality.

With 3 people flaming you.

Why would you put yourself through that?

HaIlMonitor10/7/2018, 8:34:06 AM1 votes

I basically would say the problem lies in the scaling of champions. A thing pro teams do is just simply pick the high scaling, low risk champs which lead to that. The only way I could really see this changing would be to make scaling more linear.

What I mean by more linear is either make champs about scale the same which would require a whole game rework. I don't think that is possible with items.

The other option is to make scaling champs not so damn strong in early game. I am not sure how others feel, but there is zero reason for Jax/Irelia/(mostly just fighters and Juggs) to be lane bullies and late game monsters. Now I am not 100% sure this would fix lol and the snowball but it would put the game on a timer. Which could be healthier for everyone.

Pika Fox10/8/2018, 6:39:26 PM1 votes

It doesnt. NA pro scene has a tendency to always play reserved. Everywhere else, however, you see constant bloodbaths and risks being taken for the rewards. Its pretty much just the western teams that play safe.

The Bad Touch10/10/2018, 10:02:18 PM1 votes

Long short short.

Defense wins championships.

1 800 MID LANE10/14/2018, 11:05:49 AM1 votes

I only want to touch on one of your points here: Multiplicative cost of failure.

Losing objectives and key fights suck, but something that Riot and we as players need to take into consideration is low ELO play. If these situations weren't very snowball-y in high tiers and pro play, Bronze and Silver games would take FOREVER. Like it or not, Bronze and Silver tiers takes up slightly less than 2/3 of the entire playerbase, so this is definitely something that needs to be taken into consideration. There's already a noticeable discrepancy in game time between high and low ELO, and while I'd like to see my games last longer and pros be more willing to take risks in their games, I also understand why the game has to be this way in order for low ELO game times to be as healthy as possible.

ZaFishbone10/26/2018, 12:44:00 PM1 votes

TLDR, but to answer the question in the title: Yes, yes it does!