|
![[image loading]](http://i.imgur.com/a3myj.png)
(Part 1; Part 2; Part 3)
A lot of viewing has to do with the visual stuff, but we cover that last post. This is more about the experience of viewing, suspension, and expectation, the experience.
![[image loading]](http://i.imgur.com/Eli7M.png)
The viewer experience hinges on two basic things: we need a time to be in suspense and a time to cheer. Death match FPSs are hard to watch for a lot of reasons, but one of the issues is that there is no time to be in suspense, no set up. It's just a bunch of people dying really fast. It's very exciting to play, since you're always on the edge of your seat. But you never have that experience of watching a player carefully maneuver and prepare, all the while biting your nails hoping he'll succeed or fail. The anticipated event in a deathmatch is a kill, and since the setup for an anticipated event is usually the tiny time it takes to aim, we don't have a chance to anticipate.
Capture the flag is a step up, allowing us to see the team progress towards their goal, and hope they reach it. Because the flag is a point of interest, it centers the action around there so it's easier to spectate. Still a long way from ideal, though.
Now let's imagine a golf game where the player only gets one swing. They either make the whole or they don't. We wait in suspense as they take 20 min. to set everything up just right. Then they hit the ball, and it's over. For either satyr we cheer. This does not make for good spectator sport.
League of Legends (and other MOBAs I’m less familiar with) gives you one overarching anticipated event, killing the opponent's nexus, with dozens of small anticipated events leading to that one. Where in suspense as a gank is set up, and we cheer when a escape or they die. We grow in suspense as turret starts losing health each time the players push and the other play rushes to defend. We anticipate the battle over the turf dying of living. We are also anticipating seeing a champion snowball and become extremely strong with the power to crush the enemy champions. League of Legends layers the suspense, and creates many opportunities of varying importance for us to cheer.
Starcraft II has us anticipating battles, but those anticipations change with each set of strategies we see. Sometimes the anticipated event is a unit drop into the opposing players base to snipe economy or tech, and sometime it's an army versus army clash. The goal of the FMB, or fewer minerals per base movement, is to increase how many anticipated events we have, by focusing the game more on small skirmishes. In this way, the viewer experiences moving more toward League of Legends and away from the single hit game golf we imagined earlier. This is a very good thing.
Creating many anticipated events, layered anticipated events, creates a compelling esport. As viewers, the build and release of tension is essential to our experience, and a whole lot of fun.
But we have a more meta issue we need to explore which builds on that, and an issue that often leaves me feeling unsatisfied as a spectator when watching Starcraft II: viewer expectations. We are expecting those chances to anticipate, that’s the excitement, that’s a big part of the reason we’re here to watch. Game design that doesn't deliver this consistently sabotages its viewership. It is an anticlimax, a disappointment for the viewer. You turn to your favorite TV show to get the feel it gives you every time, and so fans turn to a sport to get caught up in the struggle. The better a game delivers consistent results, the easier it is to hook an audience.
Viewer Expectations
I have never seen a build order loss in pro League of Legends. Every time I have seen one won, it was won through skill, teamwork, and sometimes really ballsy, do-or-die risks. I have never seen the equivalent of IdrA just barely not scouting a marauder hellion all in, and just barely losing because he yat just barely under prepared.
![[image loading]](http://i.imgur.com/D7B7V.png)
This was the last game of a best of 3, and it was crushing to watch. Whether it was the skill of the Terran player, luck, or IdrA needing just a 7th attempt at scouting (no seriously, he tried so hard and was being careful), it was still disappointing to watch. Hoping to see a dozen anticipated events and only having one without struggle is the definition of an anticlimax. That’s not the kind of experience that keeps me tuning in again and again (though I do, for other reasons).
I have seen the players I'm rooting for get outplayed, and I can't help but cheer for how good the opponent is. Because the game was exciting. Because I got to see their skills matched. Because they struggled. I can't see build order loss as anything other than weakness Starcraft II's gameplay design as a spectator sport.
![[image loading]](http://i.imgur.com/3XLAZ.png)
But what does the possibility of build order loss offer to the viewing experience? Surely it does something good.
1) The underdog has a chance to out-strategize instead of out-play their opponent. This adds some excitement to matches that still feel like a lost cause.
2) Even the greatest champions can be laid low. Let's say we had a Bonjwa, there is a chance that someone could topple him in every match.
While 1) sounds good on paper, the experience of watching that marauder hellion all-in end the game without out a struggle disappointed.
On the other hand, 2) is supposed to add tension to every game. In a best-of-seven the more-skilled Bonjwa is almost certainly going to take the set, but if he lost in the round of eight in a tournament, in a best-of-three, we would all feel cheated. Yet, that's the only place 2) really enhances the experience, best-of-sevens, which are rare in SC2. Also, the Starcraft II scene is harder to follow because we have so many people taking the championship and then disappear, there is less continuous story. Build order losses are one contributing factor. Overall, it's bad news.
For the health of an esports scene and the pleasure of viewing, designing a game to always deliver struggle and anticipated events is critical.
But how?
It is said that Starcraft II is a game of inches at the highest level. But I would argue otherwise: it's a game of inches to get an edge, but a game of miles to lose it. One miss-click with a ball of Mutas, one unattended moment in EMP lands on your Templar, one second not watching the mini map and Banelings nuke mineral line. It's a game of inches to win, but a game of miles to lose.
For good or ill, League of Legends and other MOBAs aren’t that way. If you misjudge a teleport, you don't lose everything you spent the game investing into, you only give the opponent a small gold advantage. In order to win, the opponents have to repeatedly best you, it's a built-in version of the effect that the best-of-seven set has. By having a game of small advantages and small losses, the viewer always knows they're going to have a chance to see the struggle. Every kill matters because of the snowball effect, but players still have the opportunity to come back if they consistently outplay the opponent. Starcraft II doesn't offer that opportunity. It makes individual engagements more exciting, but, for the average match, I find that a League of Legends match keeps me excited for longer, keeps me rooting for my team to the end. No “this is over, he’ll just have to GG once he sees how far behind he is”, there is a chance for the team to come back.
League of Legends strikes a great balance between too small a reward and too big. Emulating it's style, which allows for comebacks at any time (but only if the losing team repeatedly outplays the opponent), is a great choice for an esport.
It's also why we like macro games in SC2, players have a chance to lose a fight and come back and struggle again and again.
We have an understanding of the basics of a viewing experience, now for putting it together.
Pace of the Game
Pace is the speed, conflict, and value of the things happening on screen. In an esport, casting can have a profound effect on the pace we experience. Most Starcraft II games have an extensive lul in the action at the start of every match, sometimes I'm just bored and it feels so slow. I go to another tab in my browser and try not to forget that there is a match happening in the meantime. However, I haven't even noticed this watching the NASL the past few days when Mr. Bitter and Rotterdam are casting. They fill the time with stories from trips, hilarious squabbling, and other entertainment, and that lul I always felt magically vanished.
So what does this mean when designing a game's pace? Should we have those down times? Before we get to that, let's examine the pace of existing esports a bit more closely.
Starcraft II and League of Legends, the two esports I know best, have wildly different paces. Starcraft II can be intense from about the 1 minute mark onward, or can take up to 12 min. when players macro hard. This inconsistency makes it hard to set our expectations, which isn’t inherently a bad thing. Some might argue that it's exciting because you never know when something's going to happen.
Yet, truth is, if someone's not canon rushing or dropping an early pool or proxy barracks, we know it's going to be a while before things get exciting. The tension only remains until we see the players play their first cards. Once we know they're going for a long macro game, or even a two base timing, we know the next 4 to 6 min. are going to be slow. While excellent casting can cover this time, it's just that, a flaw that needs to be covered. Excellent casting can express itself in other ways besides trying to burn time for a game that takes its time getting exciting. Pregame warmups, post game analysis, and burning time while the players set up the match, are just a few. Designing your game to be exciting from start means it won’t need to lean so hard on casters to engage a larger audience.
On the other hand, having wall-to-wall action from start to finish doesn’t tell a good story, every story writer knows this. Humans acclimate to the level of stimuli, and so having a diversity of intensity enriches each part of the experience. There do need to be luls in the action. Champion kills in League of Legends are much more exciting than a kill happening in the Halo game, partly would be because we get to see cool flashy abilities, but also because we’ve had a chance to anticipate the kill and come down off the excitement of the previous attack.
![[image loading]](http://i.imgur.com/KThca.png)
So what about League of Legends? League of Legends is more consistent in its pacing at the pro level. There's almost always some level 1 shenanigans, a chance to start that ever so important snowball roll to victory. Often it's followed by some downtime while the players farm their lanes, or gets exciting again with an early jungler gank. Most of the early and midgame proceeds that way, either passive farming, setting up to kill enemy champs, or pushing objectives.
Passive farming does give us a chance to breathe, but I don't think it's an inherent advantage. On the other hand, Elimination of all downtime is a bad thing, so you need to take the role of the storyteller and do what all the best stories do: balance intensity and calmness. A series of time-sensitive objectives with lul in between can help with pacing if your current design is just not delivering the dynamic.
Another interesting note for pacing for League of Legends is that, unlike Starcraft II where people claim you "never know what is going to happen" when you clearly can see that a player is expanding and playing passive again, at any moment in the game, League of Legends players can move to slay the money objective, a Dragon. In Starcraft II the players must choose a series of committed actions that indicate whether excitement is coming now or in a while, but here all they need to do is choose to walk away from their land to the objective. With constant sources of excitement from fighting over the Lane, to the opportunity to go for the Dragon and start large team fights, League of Legends always has a sense that something big might go down. It's a permanent layer of tension engaging the viewer, but not one that overwhelms the luls we need to appreciate it.
So to recap the objectives for designing the pace of an esport: 1) create excitement from start, 2) balance downtime, 3) design a mechanic that creates constant tension, by always having exciting action available to the player no matter what they do.
The last issue to discuss with pace is the end. In Starcraft, there is no "long tail" end; players usually gg swiftly once clearly defeated, and often end right after a climactic battle. The only main issue is when players stay in the game trying to eke out a win once they've received a staggering blow with no real hope of coming back. I'm not talking about situations where a player has a lot of upgrades on their Marines and really good micro can earn back their win, I'm talking about situations where the opponent has no real hope of defeating the enemy, but the enemy can't quite kill them yet.
This is an issue also occasionally seen in League of Legends, but is mitigated by the fact that, in most cases, the team does have a chance of coming back. It's going to take outplaying them, but still a chance.
Neither of these games suffer much from a long tail, but something to keep in mind when designing. The winner needs to be able to kill their opponent once they have no chance, or the loser needs to have a legitimate chance of coming back. Often, players surrendering can help solve this, but for the fun of gameplay it's also good to build in some measures to prevent the longtail.
![[image loading]](http://i.imgur.com/ObO0o.png)
Points of Interest
Points of interest, such as objectives, our ways of creating anticipated events. They focus the action so the camera can view it all, and they tell the player when to get excited when the enemy starts approaching that objective. Letting your viewers know when to get excited is so key to a spectator sport. In football (American), we know to get excited when they approach the goal line. In Starcraft II, we know to get excited when the armies approach. In LoL, we know that something important has happened when a tower falls or champion it being chased. Starcraft II, we know something important has happened when Colossus falls or siege tank line is sniped by mutas. It's visual language says these things are important, and they are. Create points of interest to create excitement.
So what makes a good point of interest?
They have to be valuable to the outcome. It's often best if there is a small amount of preparation required to take the objective. In League of Legends, a player must extend across the map to push a tower, or the team must collect around the dragon and kill it. While this is more of a visual thing, having a visual way to communicate what happens when achieving the point of interest can help set expectations. For instance, we know that when the Colossus dies the giant laser beams are no longer going to be burning up units. However, in LoL, we have no clue that killing the Dragon will give the team money. He should be protecting a horde of chests of gold spilling out, then players can see what you're fighting the Dragon for. Barron should have in addition to the chests of money some glowy magic crystals, or floiting magic orbs, or at least his buff circling him, though those buffs do a poor job of communicating what they do. When the players slay him, they get his magic and cash. Show what the objective means.
Conclusion
The rest of the viewer experience has to do with intricate balancing and timing for anticipated events. That will just take testing and experimentation, but you know you're having success when viewers know when to get excited, have a chance anticipate it, and this building and fulfillment of excitement happens often throughout the match.
If you want more awesome information on pacing (targeted more towards narrative driven games rather than an esport) watch:
http://penny-arcade.com/patv/episode/episode-07-pacing
 
|
It's a game of inches to win, but a game of miles to lose.
This is something that's very true of another strategy game: Chess. It is often said that the winner of a game of Chess is the player who makes the next-to-last mistake. I actually see this as a strength, because it means players need to exert their full energies in all phases of the game. A good midgame can easily be undone by a single serious misstep in the endgame. So long as the path to walk is sufficiently broad to allow the occasional small inaccuracy, this game feature promotes the pursuit of perfection, which in turn allows a serious competitive scene with an extremely high skill ceiling. If the path is too narrow (e.g., if there were only one correct move in any position), then the game becomes too volatile, and is neither fun to watch nor play. While it is certainly frustrating to lose a game of either SC2 or chess on one bad decision after what feels like an hour of good decisions, I don't think either game is too volatile.
|
I was thinking about this for a while and then i saw this fantastic blog. Part 3 hasn't been published but for single-player experiences in game design flow is a central element. It's the balance between player skill and game difficulty. When times passes by really fast your probably under the spell. Can the experience felt by the player be felt by the spectator?
As HotS or other e-sport titles are being developed are there any theoretical framework like flow that can be applied or are game design an organic, feedback driven and creative process?
My second point based off your graph although games can vary in length, action, skill and strategy there should be a general rise in action. As this link explains plot-structure there should be time for rising action, hopefully a climax and falling action where the loser is not procrastinating the gg for too long. If you apply the authors thoughts in the link there should be room for an introduction of how both players are going to play the game out. If the rising action is too short there won't be a climax and this is especially true in starcraft matches where for example a 4gate vs blue flame hellion rush is boring to watch.
A third point comes from traditional sports where most fans can appreciate the skill the players has when they are doing a slam dunk or an amazing goal. So therefore I think it's important that all units should have one or several micro trick than can enhance their performance in the battlefield.
|
Fantastic responses guys, I'm excited to continue the discussion to these deeper depths.
On April 26 2012 09:26 whatthefat wrote:This is something that's very true of another strategy game: Chess. It is often said that the winner of a game of Chess is the player who makes the next-to-last mistake. I actually see this as a strength, because it means players need to exert their full energies in all phases of the game. A good midgame can easily be undone by a single serious misstep in the endgame. So long as the path to walk is sufficiently broad to allow the occasional small inaccuracy, this game feature promotes the pursuit of perfection, which in turn allows a serious competitive scene with an extremely high skill ceiling.
This does make chess an intense and engaging game play, and you'll notice that it's crowd of spectators are rather small. But maybe that's due to other factors, and not the "miles to lose" factor we're talking about, so let's that observation side. Instead, let's talk about more fundamental of game design.
One of the most basic elements of any competition is determining the point of victory. In the finding community, victory is not determined by a single match, it's determined by many. Why is this? So the better skilled player has the opportunity to repeatedly beat the less skilled player, reducing the effects lucky hits on the match. A baseball game doesn't end after two innings, because the opponents mismatch each other several times. A football game doesn't end after a drive by each team. In all the most popular spectator sports, players must match their strengths and skills many times, again, with the intent of allowing the better skilled team to shine.
The American football, a poorly held defensive line does not result in a defensive team losing all their points, results in the other team getting a few -- or could result in the other team getting a few, if the defensive team continues to underperform. This element of repeated struggle provides more opportunities for spectators to anticipate and more chances for the better player to shine through, both which are very good for a spectator sport.
Chess does not have a mechanical element, so whoever does the better thinking wins. This is not true with highly watched sports. With real-time elements with a non-grid board, the world of the game becomes massively more complex, so many of the things that work for chess and its highly restricted choices do not translate well. But even that doesn't matter so much, were creating a spectator sport, and seeing that repeated struggle is essential.
when I sit down to watch a football, a rare thing these days, I don't ever want to see a defensive team make a few bad plays and lose the whole game. The worst that can happen on the play is that the opposing ballplayer gets free from the defense and makes a touchdown. That is so little compared to what his wrist in almost every Starcraft 2 engagement. again, it's as if one misstep in the defense of team lose all the points they work so hard to score the entire game. This would be terribly disappointing game to watch. Even when there's a runaway game, and losing team is losing by a lot, it's because they're being repeatedly outclassed over and over. I don't know how else to describe this, except to point out all the popular sports that are watch: tennis, golf, soccer, football, volleyball, basketball, baseball etc. All of these are won and lost a little at a time, and it makes for more consistent results and better viewing experience.
However, you say it raises the skill ceiling. I don't know how, you'll need to describe that in more detail, because all I see right now is that the loser has a higher probability of winning. In order to win consistently a top player must play more perfectly. This is different than a skill ceiling, this is volatility. A skill ceiling is how good a human can get at a game, and what I hear you saying is that a skill ceiling is how much a player stands to lose over every mistake. The thing is, if those same mistakes lost you again my inches instead of miles, the better player would still come out on top, losing my inches and losing by miles is still losing. The problem is that those mistakes are so easy to capitalize on, to make the person doing the mistake lose everything they've worked for. The more a mistake cost the player, the larger contributing factor is the victory, edging out the other factors for victory and making them insignificant. This reduces the breadth of skills and mechanics that contribute to victory. A high skill ceiling requires this breath.
So, now having established that losing by a lot losing my little is still losing, and losing by a little reduces the volatility in the matchup, what makes a good skill ceiling? There are many elements: the breadth of skills a player can master, the breath of knowledge a player can learn about the game, the amount the mechanics of the game allowing player to influence the outcome (more on this in part three), and the amount of differing mechanics. None of these offer volatility, and all raise the skill ceiling. They also make room for more play styles, where the "lose by a mile" mechanic suppresses the expression of various play styles by focusing the game around those mistakes ( since they are the largest single contributor and founder to victory).
As for the pursuit of perfection, having many of these skills ceiling raising elements, as well as having money and prestige on the line will encourage the pursuit of perfection sufficiently, as it does in modern sports.
Notice, I'm not making the gameplay more narrow, all I'm saying is that the defeat in one of the struggles is a point for the opponent the same way that victory is a point for you (that is, a bad defensive play in a football game loses you seven points the same way an excellent offense of wakings you seven.) In fact, the volatility of the "losing by a mile" mechanic doesn't create more diversity it's just a result of the current design which is diverse. Diversity is achieved by allowing many things influence victory, and the more "lose by a mile" elements there are in a game, the more the game centers around those elements to the exclusion of the other parts of the game, crushing diversity.
But, some volatility is exciting! I think that's where you're coming from, right?
I think it's about having in the right places. For instance, each battle in league of legends is very volatile, capable of swinging wildly one way or another. Therefore, almost every battle in league of legends exciting. It's the fact that victory is determined by many of these battles along with a dozen other elements that allows the more skilled team to come out ahead. It's all about resolution, volatility in the small repeated elements with the overall victory of the game allowing the more skilled team to come out ahead.
And finally, to determine whether or not anything is too volatile you have to set objectives for the thing. I think it objectives in designing the sports is to consistently deliver on the expectations of the viewer, that is, they get to see what they tended to see. More people tune in to see struggles, and therefore you once your game to focus around that. I often find feel I have to sit through many okay games to get the thrill in SC2. The goal is to deliver on thrill more consistently. If that is the goal, I feel SC2 is too volatile. Games can be designed to be more consistently exciting. Though traditional sports are not perfection, the point the way to good, tried and true elements, and the losing by a little bit method is constant.
On April 26 2012 09:46 archonOOid wrote:I was thinking about this for a while and then i saw this fantastic blog. Part 3 hasn't been published but for single-player experiences in game design flow is a central element. It's the balance between player skill and game difficulty. When times passes by really fast your probably under the spell. Can the experience felt by the player be felt by the spectator? As HotS or other e-sport titles are being developed are there any theoretical framework like flow that can be applied or are game design an organic, feedback driven and creative process? My second point based off your graph although games can vary in length, action, skill and strategy there should be a general rise in action. As this link explains plot-structure there should be time for rising action, hopefully a climax and falling action where the loser is not procrastinating the gg for too long. If you apply the authors thoughts in the link there should be room for an introduction of how both players are going to play the game out. If the rising action is too short there won't be a climax and this is especially true in starcraft matches where for example a 4gate vs blue flame hellion rush is boring to watch. A third point comes from traditional sports where most fans can appreciate the skill the players has when they are doing a slam dunk or an amazing goal. So therefore I think it's important that all units should have one or several micro trick than can enhance their performance in the battlefield. Part 1 and 3 I totally agree with, well said.
If you haven't watched the video at the end of the post, you need to. The graph that article demonstrates is very inaccurate, placing the climax at the middle of the story in creating a constant tension increase, where oscillating tension with an overall increase creates a better experience.
If you guys have counter points I'd love to hear them. I know I don't understand it all (or even close!), so I love to debate things out to gain greater understanding. As iron sharpens iron... ^_^
|
Man, this is well written....O.O
Keep em coming, can't wait to see what you have to say for part 3.
|
As soon as you said you had never seen a "build order win" in League of Legends I almost stopped reading to be honest. Games are won and lost all the time at the picking phase, meaning before the game has even really begun (depending on where you draw the "game has started" line). Games are won and lost because of sneak Barons, games are won because of missed smites, literal milliseconds of not pressing 1 button can win or lose a game, just like flying a flock of mutas over marines can. One bad team-fight, much like 1 bad battle in sc2, can basically seal the game even when you're ahead.
Not to mention that they don't need to "repeatedly best you" when the game an extremely large amount of snowballing built-in; just like in sc2, you don't have to repeatedly best them in many scenarios. You get ahead and should win, unless the other guy fucks up. Just like in sc2. On top of that, there are HUGE down-times in league of legends, absolutely MASSIVE. There are some games where there isn't a FB or anything exciting for an upwards of 12-15 minutes, and it's just boring to watch.
For many newer players (for either game) this is true, but to any experienced player it's quite easy to tell when someone is going towards a dragon/baron and when they aren't well in advanced. How they're controlling their lane (pushing, freezing, etc.), the movements in the other lanes, whether they have an advantage or not, etc. can give it away minutes before they actually go towards an objective. Just like an experienced player can tell you "player A is doing build X and player B is doing build Y so there won't be any tension until time T" in Starcraft. Most of what you've said I don't agree with, and I want to say it's because you don't grasp how LoL works as well as you grasp how SC2 works. To someone who DOESN'T know how SC2 works there is tons more of "What's going to happen next?!" moments, whereas your average diamond+ league players have a much better sense of when the action is really going to start, and it works the exact same way in League as well.
|
Vancouver14381 Posts
On April 26 2012 16:04 Skwid1g wrote:
As soon as you said you had never seen a "build order win" in League of Legends I almost stopped reading to be honest. Games are won and lost all the time at the picking phase, meaning before the game has even really begun (depending on where you draw the "game has started" line). Games are won and lost because of sneak Barons, games are won because of missed smites, literal milliseconds of not pressing 1 button can win or lose a game, just like flying a flock of mutas over marines can. One bad team-fight, much like 1 bad battle in sc2, can basically seal the game even when you're ahead.
Not to mention that they don't need to "repeatedly best you" when the game an extremely large amount of snowballing built-in; just like in sc2, you don't have to repeatedly best them in many scenarios. You get ahead and should win, unless the other guy fucks up. Just like in sc2. On top of that, there are HUGE down-times in league of legends, absolutely MASSIVE. There are some games where there isn't a FB or anything exciting for an upwards of 12-15 minutes, and it's just boring to watch.
For many newer players (for either game) this is true, but to any experienced player it's quite easy to tell when someone is going towards a dragon/baron and when they aren't well in advanced. How they're controlling their lane (pushing, freezing, etc.), the movements in the other lanes, whether they have an advantage or not, etc. can give it away minutes before they actually go towards an objective. Just like an experienced player can tell you "player A is doing build X and player B is doing build Y so there won't be any tension until time T" in Starcraft. Most of what you've said I don't agree with, and I want to say it's because you don't grasp how LoL works as well as you grasp how SC2 works. To someone who DOESN'T know how SC2 works there is tons more of "What's going to happen next?!" moments, whereas your average diamond+ league players have a much better sense of when the action is really going to start, and it works the exact same way in League as well.
Countering something pretty much rules out a build order win because the information is readily available during bans/picks so I don't think I've seen a build order win either. It is the responsibility of the captain to pick appropriately to respond to any counter picks (unless it was last pick). In this way, it's poor decision making rather than a build order win. There are certainly hard matchups but the players have the information to try and change things up so they're not as disadvantaged. If this doesn't happen, it is most likely because 1) they are too proud to admit they can't handle the enemy or 2) they're not comfortable with a lane change to shake up the enemy.
Also, the difference between 1v1 and a team game is that the viewer can reason out what the 1v1 player is thinking depending what they see on screen while it's not always possible to tell what a 5 man unit can do. It is possible to fill the lull in team games with action elsewhere (e.g. very little action in bot lane so you move camera to mid or top). This isn't possible in a 1v1 RTS because there are only 2 noteworthy viewpoints.
|
I enjoyed part 1 and agreed with almost all of it. In part 2, I think there are a lot more questionable arguments.
You talk about the ideal scenario being a building of suspense through a series of engagements (rising action) and then finally a big engagement (climax) that essentially ends the game with one guy GGing out soon afterwards. This may be the ideal, but I don't think any fair competition can really enforce this structure in a meaningful way. It is a structure that rarely happens in traditions sports, RTSs, or even MOBAs.
Many sports are anti-climactic after the first quarter/period/inning as one team takes a huge early lead, but still has an hour or more to play out. Most fans tune out at that point, and that's a problem. The only way to keep an audience interested after a noticeable lead has been obtained is through comebacks. Comebacks come in 3 varieties:
1) Luck based.
2) Devastating mistakes.
3) Systematically win small engagements to get back into the game.
#1 is generally bad for skill-based competitions (although it has shown some popularity in the televised Poker boom not too many years ago). I think we can agree that this is bad for serious competition without getting too deep into the subject.
#2 is not ideal, but still has its place in this world. It's still skill based, but it puts a premium on extreme precision and never making mistakes. Starcraft 2's terrible-terrible damage tends to follow the devastating mistake model and it does bring the benefit of the comeback and the upset to the game without making the game luck based. Extreme precision and mistake-free play are skills that can be practiced and improved and you can get relatively consistent winners. However, when those skills are at a premium, it discounts other skills that are important to the fan base... multitasking, strategy, and creativity. IE, the ability to control one big blob of units perfectly is more important in SC2 than controlling three groups of units well at the same time. Precision > multitask.
#3 is great, but rarely happens in any sport/game. Why would a team/player be so much better as to obtain a noticeable advantage, but then be so much worse as to allow the other team/player to systematically get back into the game? True comebacks are rare for a reason. Without the physical (tiring) nature of sports, it makes comebacks based on the systematic small wins even more difficult. I'd even argue that physical sports rely on emotions (get pumped) much more than e-sports (try to be emotionless) and big advantages can often be negated due to the emotional swings. E-Sports doesn't have those physical/emotional factors. The chances of #3 happening in a well constructed e-sport is very small.
You list MOBAs as fitting your ideal structure, but I highly disagree. From what I've seen in League of Legends at a pro level, comebacks almost never happen based on #3. Every once in a great while you'll see a comeback based on #2 where the clear winning team botches a team fight and suddenly has three towers and a nexus knocked down before they can respawn. That's it. After 10 minutes of a LoL game, I can tell you who will win with about 80% certainty and I'm no expert LoL player. The climax of most LoL games passes by nearly unnoticed as one team gets a small advantage early in the game that snowballs. The next 10-40 minutes are mostly a waste.
When you come down to it, an E-Sport can rely on Luck (bad) or Devastating Mistakes (questionable) to keep an audience interested after someone takes a lead. It cannot rely on systematic small wins because these will happen way too infrequently. So what should an e-sport strive to achieve? If you're trying to sell a tournament of games, you cannot have a longperiod of falling action unless the game has a devastating mistake mechanic built in. If you do have those long periods of falling action without the legitimate possibility of a comeback, you've just killed your viewer base for your next game. If the choice is between:
1) trying to build suspense over a long period of time at the risk of creating gigantic falling action periods
2) ending a game quickly once an advantage is taken at the risk of jumping straight to the climax
Any event that relies on a series of games should always err on the side of #2. When you have a series of games (Bo3+), each individual game acts as one or more peaks and valleys on your ideal chart. You don't need to encompass the entire chart in one game. The "climax" of game one can just be the first rising action in the series.
Instead, the competition needs to rely upon the series of games to create the individual peaks and valleys that would be desired out of a game. It also needs to rely on intelligent casting to help. You give the example of an Idra game where he barely can't quite scout and loses to a "build order loss". The truth was that he lost due to poor scouting and poor adjustment to his inability to scout... if you can't get a good scout, then you need to play safe. He didn't. I don't know specifically what Idra game you're talking about, but that scenario has been a fatal flaw in Idra since I've first seen his games. An intelligent caster who truly understands the game should be building the moment of tension in the scout or in the lack of adjustment by Idra. Properly done by the casters, those things can be the climax of the game. But the caster has to understand that Idra is putting himself in an unwinnable situation. He needs to focus in on that production tab ("where is that baneling nest? He needs to make that baneling nest now!"), zoom in on the round of eggs making drones, or even zoom in on the empty spot where a spine crawler should be placed, etc. It's not the glamour play, but it is the climax of the game. It's like the block that springs the running back free for an easy jog into the end zone. It seems anti-climactic, but the professional casting crew in football will show 5 instant-replays of the amazing block from various angles. When a non-glamorous climax is put into a series, it becomes the rising action that builds the story for the next game. We just have to hope that it doesn't happen in the deciding game, but sometimes the fans get unlucky. That's just a part of sports.
The other issue I have is that if you somehow were able to magically force a game's structure to fit that chart of rising action, you will create games that all look almost identical. Without evil. there can be no good. Likewise, without a "build order loss", there can be no awesome games. Every game starts to blend into each other and none stand out. The whole thing just becomes bland. Starcraft 2 does a very nice job of giving us a variety of ways to win. We see the cheeses, the build order losses, the solid timing pushes, the solid macro games where maxed armies clash 1+ times, the multitask games, and the big mistake games. Fans tend to gravitate towards the solid macro games and the multitask games as the "great" ones, but if every single game was a solid macro or multitask game, they'd all start to look bland. Games need the variety of structure to keep the plot from getting stale. Just like the structure of a story, you need the valleys in order to appreciate the peaks.
|
Ah, what a great discussion, guys. Kk, let's dive in.
Skwid1g:
As soon as you said you had never seen a "build order win" in League of Legends I almost stopped reading to be honest. Games are won and lost all the time at the picking phase, meaning before the game has even really begun (depending on where you draw the "game has started" line). Games are won and lost because of sneak Barons, games are won because of missed smites, literal milliseconds of not pressing 1 button can win or lose a game, just like flying a flock of mutas over marines can. One bad team-fight, much like 1 bad battle in sc2, can basically seal the game even when you're ahead. As JBright said, both teams have perfect scouting information on the selections of their opponent. If they fail to build a decent composition versus them (assuming Riot doesn't have something way overpowered) than it is a failure on the team's part and not the result of the game design. While we do try to look at making it fun for "Bronzies", you can't design a game to make up for their mistakes. So yes, a win that has champion selection as one of the primary contributing factors to victory is possible, but in every match I've spectated it's based on the teamwork and clever use of the champions, not simply selecting the champions themselves.
This is in stark contrast to the ZvT described in the OP, where the only thing the T did was chose Marauder Hellion and deny scouting, and a little micro, and the game was over.
Remember that the closer in mechanical skill and teamwork the two teams are, the more non-execution based decisions seem to effect the gameplay. Clone a team give them identical environments, and give the clones a slightly better team composition and theoretically the clones are going to win every time because the teams are perfectly matched elsewhere. This is a case of "the straw that broke the camels back", though, not "Everything hinged on this one decision", because without the excellent mechanical execution and teamwork champion selection wouldn't matter. But that's an entirely theoretical situation, no team is going to be perfectly matched against another in every way.
As for baron sneaks, they can win the game, and are a pretty large straw to break the camels back, but it's still smaller than losing a single army or muta/marine clump Starcraft 2 (before the players are capable of remaxing). Again, it's a straw that breaks the camels back, but another decision somewhere else could have countered it. A good team fight could turn an over-confident baron team into a dead team. This is the heart of what I talk about in part 3 of having many elements contribute to victory.
As for a single team fight turning losers to winners, this is a volatile element that counters the excitement-repression of snowballing. By raising the cost of the death as the game goes onward, losers still have a chance to come back. I don't know if LoL has an optimal balance of snowballing/volatility in the end game, I kind of suspect it doesn't. Not many traditional sports have a snowballing effect. I think a little snowballing, when countered with the appropriate amount of skill-dependent volatility in the late game can help keep games interesting, but balancing the amount of each is an area that needs to be explored and refined. Here, prototyping will be the key.
On top of that, there are HUGE down-times in league of legends, absolutely MASSIVE. There are some games where there isn't a FB or anything exciting for an upwards of 12-15 minutes, and it's just boring to watch.
Are you talking about professional matches? I know there is a huge amount of downtime and amateur play, but the professional matches I've spectated were always full of ganks and attempted dives. It was quite entertaining during that 12-15 minute period you're refering to. Or do you only see team battles as non-down time?
As for the possibility of a greater knowledge of Starcraft II making games more often driving, yes that's a good point. On the other hand, if you intend for a sport to get embedded in the culture, and needs to be entertaining to the millions of armchair quarterbacks who will know the game enough to not expect a 6 minute attack in a two-base macro with double ups on both sides. These are your most consistent and passion fans. You don't want to design a game that becomes significantly less interested as your audience becomes educated or you risk losing the hardcore fanbase. I don't think either LoL or SCII are very bad this way, but it is something to keep in mind when designing game. Good point.
@RenSC2
Good post.
Quick thing, I think MOBAs are closer to the ideal structure, but still have a long ways to go. If I said otherwise earlier, I misspoke.
I want to propose a alternative to your three well thought out options: escalating stakes. In the escalating stakes model, early-game victories offer smaller rewards than later victories. The victor is almost always decided in the end game due to actions taken then having a large effect on the outcome of the game.
The early game needs to be visually interesting with conflicts, and it does need to contribute to victory. It needs to have a "micro snowball" effect, where an advantage in the early game helps the team perform better in the late game, but the average advantage gained in a well-matched game is small enough that it can be overcome. (This is a very demanding and specific balance.)
This does many things:
1) Games will have a certain minimum length for the victor is determined. This must be carefully decided length, one that allows for best of 3s that the viewer has enough time to actually watch. In a team sport you don't have as many teams as in SC2, so each set can take longer, but keeping the viewer investment low to watch will help the growth of esport.
2) It counters the extreme power of snowballing. You need a bigger snow ball to have an impact on the end, where everything is escallated. I agree that many LoL matches are exciting until someone snowballs if the teams aren't very close in skill. This creates a dramatic curve that flatlines. We want to create a dramatic curve that escalates, and escalating stakes are one of them.
3) LoL does have escalating stakes, and it helps the game, but I think it can be vastly improved. The escalating stakes are the increasing respawn time is, making each step cost more and more. This creates volitility than can occasionally counter the effects of it's dominant snowball mechanic. However, in an ideal escalating stakes model, the players engage in frequent struggles in the end to determine the victor, not fewer. A well designed escalating stakes model will look fundamentally different.
One possible escalating stakes model is where players earn money by killing their opponents early game. In the midgame more objectives are placed on the map, focusing the action. These objectives offer sligthly more reward than the previous engagements. Then, as the midgame moves toward the end game, a giant monster for each nexus is released. They storm toward the opposing player's nexus, meet in the middle, and try to force eachother back. Each bit of damage done to the monster pushes it back toward its nexus. A player choses whether to focus his damage on enemy champs or the enemy monster. Players respawn every 10ish seconds a little outside the battle. Also, items purchased early game have 1/3ish the effect on stats that LoL's currently does.
In this model, we have the rise and fall of excitement around champion kills, replaced by the higher excitement of team battles over objectives. This also rises and falls as objectives are taken and respawn. Then, the tug-of-war beasts serve as our climax, where players ferociously fight. Spectators can see the progress of the tug-of-war beasts and anticipate and fear, can see the deaths, and thus can clearly see who's winning. Since early game items have less of an effect on the default stats, more emphasis is placed on mechanics and the proper placement of damage, rathar than the snowball of the early game.
In this model, every game has a climax, and the players win by repeatedly besting the other team in the fight around the tug-of-war beasts. Sure, good early game decisions can give a team the item edge, but that edge can be overcome. In this way, the early game is exciting, the mid game is exciting, and the end is exciting. A team can play to "win" the game in any of the phases, focusing their composition and decisions online and dominate a particular phase. Each phase has a different feel, keeping things interesting.
I don't think this is perfect, I would change further from the MOBA formula when applying all the concepts discussed, that idea serves more as an example. Feel free to pick out the design holes, that's what makes this fun and helps us all understand design better.
-
I'm not convinced that build order fails are essential to avoid blandness. I suspect that a great diversity of options and strategies, along with a game that naturally encourages rise and release of tension, can continue to be interesting. I don't find that seeing a player lose by a mile enriches my SCII viewing experience -- if every game was nailbittingly close with major stakes, I'd watch more. Not that every game can be, you can't design to only allow that, that's bad sports design. But designs can encourage that.
While it's true that having one awesome game amongst more medeocre games make the game stand out, it reduces the excitement of the sport as a whole to glorify that game. I want to design a game that makes everything as exciting and dynamic as possible while still being a great sport for the players. I don't see how single-struggle games such as a build order loss ever achieve this. I think allowing for a diversity of strategies can keep things interesting and make up where forcing a certain minimum of struggles/engagements might not.
Diversity comes from many sources, and so long as you put much of it into the game, sacrificing the one-struggle-game type to have more consistently delivering games it worth it in my book.
Maybe I'm wrong, and the proposed game type will suck the suspense out of it. Or maybe I'm right, and it will keep more players turned into the full game, enjoying the ride. Someone can make an incredible game design based off incredible game theories, and until it's tested we just don't know if he will come out as expected.
Again, please point out flaws you see in that escalating stakes model, I'd rather see them now than latter.
Great stuff guys.
|
Escalating stakes are very interesting. I didn't think of them in my initial assessment; however, one thing to understand about escalating stakes is that they are based in the "devastating mistake" model. They just keep the mistakes less devastating early on and then increase the devastating power of each mistake later on. Overall, I do agree that the general model of escalating stakes is better than a pure devastating mistake model.
There are still issues and other observations though.
The first thing is to mention that to some degree escalating stakes are already included in typical RTS games. Escalating army sizes and the effect of upgrades on armies as opposed to bases/workers means that mistakes made late in the game are much more costly than early game mistakes. For example a terran doing a reactor-hellion expand against zerg can sacrifice all the hellions (his entire army) for little or no gain and still be in an okay position because the zerg has to travel across the entire map with a small army and then run into a wall + workers + reinforcements. Whereas, sacrificing an entire army late game for little or no gain is typically game-ending.
Another thing I want to mention is that the mechanic of the escalating stakes can be very jarring depending on how they're done. In your MOBA example, the game goes along as normal and then all of a sudden the entire focus of the game completely changes with the escalating stake mechanic including a giant monster. It seems like a very jarring change to the way the game is played. When does the SC2/LoL mid game begin? Late game? It's a very natural flow and you won't necessarily be able to pinpoint a spot and I think that's a good thing. A game should generally flow well from one stage to the next where each stage's beginning/end should be essentially invisible otherwise it can create negative effects especially for a spectator.
As a spectator with a jarring change to increase the stakes, all of the battles early in the game seem almost meaningless. You know that the big boss thing is going to come out and the game won't end until that point. Does the beginning really matter at all? Sure, the beginning might actually matter in building a small lead, but the perception will never quite match. It's like a TV series where the main character is in danger of dying, but there's really no suspense because you know he has to live so that the series can continue on. Likewise, if there's no real chance of a death early in a game, much of the early game loses all of its appeal. At best, someone gains a small advantage, the majority of the drama is removed. Having jarring changes highlight those points where the viewer should start caring, but conversely highlight the parts where the viewer shouldn't care at all. Smooth stake increases do a much better job of preventing this effect. Even with a smoother curve of escalating stakes, if the mechanics of the game don't allow Idra to lose to a marauder/hellion push, then we can still tune out until the 15-20 minute mark when the real battles begin. I don't think that's good for the game.
Likewise, as players, if you know that some big event will occur in every game, then you're going to build your composition to be at max potential at that point. Why build a composition made for taking the early advantage if you're guaranteed to go to the late game and another composition will be better for the late game? Again, you're limiting interesting plays to conform to one overarching style. If a marauder/hellion push cannot kill your opponent and instead can only give you a slight lead, but still put you behind significantly in tech/economy if it doesn't work, then you will never see it. The whole early-mid game can become extremely bland extremely quickly as everybody recognizes that trying to work up small advantages are a waste when they should really be building for that one big moment. Again, there is much less of this effect if the stakes are gradually and naturally upped as nobody knows exactly what point they should build for and it won't be exactly the same as the opponent.
Finally, the increasing stakes become a huge balancing act. If the stakes start too low, all the beginning stages become meaningless. If the stakes start too high, then you risk build-order losses and other ugly losses too frequently. If the stakes increase too slowly, you can have all the problems of a game without increasing stakes. If the stakes increase too rapidly, you make everyone play for the late game and turn that late game into a single battle too frequently. If you get the balance right (along with about a million other things), then it would be a positive addition to your game and you could create a game that'd blow away everything out there.
For other things, you are correct that build order fails are not essential to avoid blandness. They are an easy way to increase variety in a game, but far from the only way and not the best way. But again, if someone can't die early after making a bad mistake, then the early game becomes a formality, not an exciting event. Even with the risk of early death, most SC2 early games are formalities at this point with casters struggling to fill time because the players have mostly figured out all the early all-ins. To make the beginning of the game interesting without major risk of death is tough. My best solution is to include 1) non-base objectives that are positioned to force small army clashes. 2) Early game units should require heavy micro.
As a little side tangent, the first "combat" unit from each race should be some sort of scouting unit (preferably with heavy micro requirements in-combat), not a primary combat unit like the marine or zealot (zergling isn't quite a primary combat unit, but close). Then a build order fail is truly a failure of scouting or an immaculate job of scout denying by the opponent and people would more readily accept it. Either way, it's much better than a true build order fail where there was no possible way to scout what was coming.
And one more thing that I forgot in the initial post. A game can have comeback mechanics built in to help defeat the snowball effect. Like, for LoL, a champion kill could be worth more gold if that champion has a good K:D ratio. They already have that for shutting down kill streaks, but I think it should probably be a little more general rather than a one-time thing. To keep this section short, clearly designed (make sense) and well integrated (not overpowered) comeback mechanics can be positive additions to games.
|
The first thing is to mention that to some degree escalating stakes are already included in typical RTS games. Escalating army sizes and the effect of upgrades on armies as opposed to bases/workers means that mistakes made late in the game are much more costly than early game mistakes. For example a terran doing a reactor-hellion expand against zerg can sacrifice all the hellions (his entire army) for little or no gain and still be in an okay position because the zerg has to travel across the entire map with a small army and then run into a wall + workers + reinforcements. Whereas, sacrificing an entire army late game for little or no gain is typically game-ending.
Hm, interesting. I'd actually argue that the escalating mistakes in SC2 is not really escalating, but rather really drunk. A terran on 5 mining bases with a field of barracks can lose an entire army to banelings and still just keep cranking the Marines and marauders, or a Terran can lose just two too many hellions in the early game, lose map control, and die to an opportunistic baneling all-in.
Truth be told, it's really hard to talk about these things because of how many different actors contribute a game. Let's try to use the principle of charity when interpreting each other's posts.
Basically, when I say it's stakes are "drunk" I mean how devastating the mistake is is all over the place. Not quite scouting the back corner of the third can mean you miss the proxy and lose the game in one instance, and sacking your whole army (poorly engaging to that it is not cost efficient) can be no big deal. It's extremely dependent on how the game has progressed. So I don't even think SC2 is a good example of an escalating states model, since in order for to be escalating must have some consistency.
A want to emphasize, I do not think League of Legends is a perfect game, far from it. I just think it does things very different from SC2, some good, others bad. This makes a great point of contrast and is the reason I keep referring to it, not because I think it's perfect. That said, what League of Legends has is a homogenizing factor, that is an element that consistently limits the devastation of mistakes in the game. Two things that create this factor: 1) the cost of the mistake is losing XP and giving the opponent Gold. This is a relatively consistent loss, where SC2's early game losses can be anywhere from snow-ball inducing to game-ending. The reason it's not game-ending is because 2) Turrets. These prevent deponent from achieving the victory condition immediately, insuring multiple engagements. They are a buffer. To make a metaphor, if achieving victory is the equivalent of climbing a cliff, in Starcraft II there is no mechanic preventing you from strapping on a jet pack and find the top, if properly executed. In LoL, both teams must climb the cliff, a handhold at a time. This buffer falls away as the game progresses, and champions can challenge turrets. In this way, the stakes of each action increases because you can swiftly lose your buffer.
That said, I suspect there are better factors for homogenizing the effective mistakes to a more crowd-pleasing intensity curved. And of course, once you factor in player skill, then you have the extreme snowball effect which counteracts this. Can you imagine LoL without the effect of turrets, though? One death in an early lane and you could lose your nexus.
Another thing I want to mention is that the mechanic of the escalating stakes can be very jarring depending on how they're done. In your MOBA example, the game goes along as normal and then all of a sudden the entire focus of the game completely changes with the escalating stake mechanic including a giant monster. It seems like a very jarring change to the way the game is played. When does the SC2/LoL mid game begin? Late game? It's a very natural flow and you won't necessarily be able to pinpoint a spot and I think that's a good thing. A game should generally flow well from one stage to the next where each stage's beginning/end should be essentially invisible otherwise it can create negative effects especially for a spectator. This is an interesting consideration. While the giant monster idea is not refined and could definitely stand some, there are some statements you make that I don't understand where they come from.
I find the transition from early game lane farming to the midgame, teambattle and team-centric play jarring in LoL. Almost every decision you make once you enter the main game of LoL is made using a completely different set of factors. The scope, movement, item purchases, power usage, and even farming timming and dynamics (lane pushing) are extremely different. And yet, I don't find it to be negative. In fact it makes playing each game more rich because I'm not experiencing the same dynamic the entire time. It also raises the skills ceiling, needing to know when to transition and how to play both phases. I also think it's better for the viewer, for a lot of the same reasons mentioned: a diversity of things to spectate, an expected series of distinct dynamics to watch. In my experience, the experience of one allows me to appreciate the experience of the other, the contrasts accentuate each other.
So to understand what you're saying, I need you to 1) define what a smooth transition is, and 2) define why abrupt changes negatively impact the viewer.
As a spectator with a jarring change to increase the stakes, all of the battles early in the game seem almost meaningless. You know that the big boss thing is going to come out and the game won't end until that point. Does the beginning really matter at all? Sure, the beginning might actually matter in building a small lead, but the perception will never quite match. It's like a TV series where the main character is in danger of dying, but there's really no suspense because you know he has to live so that the series can continue on. I find this a really interesting analogy, especially since this is a problem I can plague and otherwise excellent story.
Likewise, if there's no real chance of a death early in a game, much of the early game loses all of its appeal. At best, someone gains a small advantage, the majority of the drama is removed. Having jarring changes highlight those points where the viewer should start caring, but conversely highlight the parts where the viewer shouldn't care at all. Smooth stake increases do a much better job of preventing this effect. Even with a smoother curve of escalating stakes, if the mechanics of the game don't allow Idra to lose to a marauder/hellion push, then we can still tune out until the 15-20 minute mark when the real battles begin. I don't think that's good for the game.
okay, this statement makes an assumption I think we need to examine: the only draw of a game is to see who is the victor and thus only actions that near-directly lead to victory to be interesting.
I usually don't watch the first 5 min. of Starcraft II match, and I do a LoL. I believe we've already established that the excess downtime at the start of Starcraft II match is not something we want in the way in designing esport, so move on. Interesting part is that, even though no one can lose in the first 5 min of a LoL game, I still watch it. We could blame it on the extreme snowball in fact, that these first kills can play such a huge game who ends up winning, and we be right. But I only think partly, I think there's more going on.
Conflict is interesting, and at the start of most Starcraft II matches the most exciting conflict we get is probe harass. This does not hold my attention. On the other hand, look at a LoL early invade, where the teen groups up and marches into the enemy territory for level I engagement. We have time to build suspense anticipating what might happen, we see positioning, setting up a trap, teamwork, risk, visually pleasing spells, and a victor for that battle. This engagement often doesn't ensure the winners take the game, but it's interesting to watch.
Humans have an incredible capacity to become drawn to the suspense of something that is completely irrelevant. Even if this battle doesn't determine who gets the money at the end, I propose that we can still become psychologically invested, just like we can fictional characters battling it out in a movie. The build up of suspense, the visually pleasing and exciting action, and what we can take away from it to apply to our own game (or, alternatively, our armchair quarterbacking fantasy) still creates incentive to value an early engagement.
Early SC2 in most games doesn't offer any of that. While I agree that the game, and thus the victory and success of our favorite players, hinging on the battle makes the battle more exciting, remember the dramatic arc curve. We are trying to hit maximum excitement at the start, were trying to start with something exciting and build toward the climax (with periodic releases of tension).
In this way, early battles having less of an impact on the outcome of the game plays perfectly into the desired tension arc. (I don't mean we should make them meaningless, just not super-snowball power.)
That said, my Monster MOBA idea may not create the smooth increase in stakes, a lot of it would depend on the details. I agree, that is essential.
Likewise, as players, if you know that some big event will occur in every game, then you're going to build your composition to be at max potential at that point. Why build a composition made for taking the early advantage if you're guaranteed to go to the late game and another composition will be better for the late game? Again, you're limiting interesting plays to conform to one overarching style. If you a marauder/hellion push cannot kill your opponent and instead can only give you a slight lead, but still put you behind significantly in tech/economy if it doesn't work, then you will never see it. The whole early-mid game can become extremely bland extremely quickly as everybody recognizes that trying to work up small advantages are a waste when they should really be building for that one big moment. Small reward to high risk is a really basic competitive gameplay balancing issue that isn't specific to escalating stakes. Yes, the dev team would absolutely need to refine those ratios to ensure a diverse gameplay experience, but that's inherent in any model of competitive play.
As for building your composition to be at the max potential at the same point every game, I tend to think that the meta-game will deal with this once it develops past infancy. Once everyone gets good at maximizing that element of the game, the winners will be determined by who can also maximize their effectiveness during the other parts of the game. What once would be "the most important thing in the game is to be ready when X happens" would become "yes, X is so extremely important to victory, but with everyone so good at it, what sets the top players apart is those who can also to Y." I can practically hear Artosis is saying something like this,  .
We don't want the lower levels to only focus on X, though, true. But let's go back to the core goal of escalating stakes: ensure that any one part doesn't play too larger a factor in determining the victor. In a game like LoL the issue was the strength of snowball, in SC2 with much more sporadic risk levels, the issue was individual strategies. Since we were trying to improve upon the league of legends model, what you describe is the ditch on the other side of the road, overemphasizing the end I said in the beginning.
Which is an extremely important thing to avoid, it's just committing the same sin in a different spot. That said, let progress from here assuming that the goal is the deemphasis of any single factor.
...Which might bring up some issues with our intensity arc, which is weighted toward the end, and at first glance it seems like I'm saying now that the intensity should be equal all across the game. The truth is, I think something should play a larger but balanced role. In order to serve the dramatic curve, those things tend to me to be at the end of the experience.
And speaking of, imaging issues with diversity if you have regimented elements such as a monster appearing from the nexus at 25 min. every game. I just remembered an element for an old Starcraft II custom map: across the map were oil rigs that players needed to defend. When 5 oil rigs died, didn't matter which side, new abilities were unlocked for each side. The players who destroyed more oil rigs had an advantage in money, but to counter the beginning emphasizing effects of snowball, both teams received the new abilities. This is our 25 min. monster, except instead of time determining it, gameplay elements did. This meant that well-defending teams could stay in the first phase of the game for a long time, and when they were ready could allow themselves to losing oil rig to move into the second phase, where the optimized their builds for (where the other team might be optimizing for the third phase). They could manipulate the progression of the game to suit their strategies. I think this is a much better dynamic then I time regimented one, raising the skill ceiling and allowing for more diversity in the game.
I have to thank you for bringing these points out, because I haven't given these issues proper thought.
1) non-base objectives that are positioned to force small army clashes. 2) Early game units should require heavy micro.
Exactly what I was thinking. In the game I'm designing, players are always motivated to do 3 things: kill the creature in the middle of the map for bonus resources and experience, steal the enemy's resources for bonus resources, and defend their own resources. Each of these objectives oppose each other, in order to do one you must steal from the other. This creates interesting conflict in the players, and also interesting battles between the teams. And just like you say, the units players spawn with are heavy micro, making the battles a joy to watch for more reasons than just wanting to see who wins the prize money.
On a side note, there are half a dozen starting units teams can choose from and three players on each team. This allows for a ton of different team compositions to face a ton of different team compositions, times the town of micro opportunities, times the many different ways that team can approach those three objectives. Times the degrees of success in executing those strategies and mechanics. Do that math, and you know just how many different ways a game can open.
As a little side tangent, the first "combat" unit from each race should be some sort of scouting unit (preferably with heavy micro requirements in-combat), not a primary combat unit like the marine or zealot (zergling isn't quite a primary combat unit, but close). Then a build order fail is truly a failure of scouting or an immaculate job of scout denying by the opponent and people would more readily accept it. Either way, it's much better than a true build order fail where there was no possible way to scout what was coming. I like the way you think! I agree.
Okay, to boil down the discussion points:
1) I think dramatic shifts in games can be good. I use LoL as an example, and discussed the effects of contrast. I ask you to explain why you to define what a smooth transition is, and why dramatic shifts are bad.
2) I say that a lot of the issues you point out are inherent to any competitive game, and not unique to the escalating stakes model. But based on your concerns, I remember some sweet things that add diversity to a game (by allowing players to affect when the phases change), and remember not to drive into the other ditch.
3) We talk a bit of unit and game design, and I like the way you think.
I wanted to mention a bit more about creating conflicting objectives. This is something that was pointed out to me by the Extra Credis crew, the people making a great show that I keep linking at the end of posts. They talked about how grabbing a mushroom in Mario, the original, his great game design. Your long-term goal is to win the game, your short-term goal is to survive. Grabbing the mushroom helps you in your long-term goal, but because it races forward into unknown dangers and you have to chase it, it places your short-term goal of survival in conflict with the long-term goal. This is a great, basic example, and something to think about. I think pitting your goals against each other as diversity the the gameplay and creates interesting choices. I'm really curious what other ways we can do that in a competitive multiplayer game. Mainly I see to do it is through objectives that increase the risk of getting attack or even make you weaker, kind of like trying to take Dragon or Barron in LoL.
But that's still pretty basic, Survive Now vs Snowball Later. I guess in a way, with the multitask intensive way SCII is, every task is pitted against every other task as a demand for your attention. Great for the skill ceiling and can certainly be impressive, but it doesn't put viewers in suspense the same way. They're not sharing that conflict of interest with the player except for rooting for their APM to be fast enough.
What other pitted objectives could we have, I wonder...
|
|
|
|
|
|
EPT![[image loading]](http://i.imgur.com/3DjkS.png)
