We like it when our sports are interesting--if the losing side realistically has zero chance of a comeback, why bother playing it out? Why bother watching?
Fortunately, most games with some sort of uncapped skill test--including all games built on a stable yomi foundation--always both players the chance to make a winning move. And by extension, any number of winning moves they made on you, you could hypothetically make back. If they scored 30 points more in the first half, it's equally possible for you to score 30 more points in the second. If they won 47 points in a row in tennis, it's just as possible for you to turn around and win the next 48.
So why have comeback mechanisms at all?
The most common issue is that many competitive video game genres are built around snowball mechanisms, where success is rewarded by the winner getting stronger. So, they probably win even more, and get even stronger, until the game is super boring.
The midgame of professional SC2 can be really dramatic, but the ending that follows is always an anticlimactic "gg". (Save the rare base race)
League of Legends is chock full of exponential snowball mechanisms, and to remain a viable eSport it bends over backwards to include as many anti-snowball counterbalances as possible. Over the last 10 years, League has incorporated:
Wrong.
Snowball Mechanisms Implicit To Fighting Games
"Wait a second!" you cry. "Fighting games are based on yomi! If I take away 90% of my opponent's health in Street Fighter 2, it's still exactly as easy for me to turn around and take 90% his health back! This isn't a MOBA, his character hasn't gotten stronger! We're repeating the same decisions, with the same options!"
It turns out there are a handful of snowball mechanisms built into all fighting games pretty deeply. They are somewhat subtle relationships, but quite powerful. I'm going to look at a few of the biggest ones in detail:
Inequality of Ties
Fine, let's say I took away 90% of your character's health in SF. You say it's just as reasonable for you to do it back in reverse? Not so fast, I've got some tricks up my sleeve.
Any outcome where we trade hits and both take damage is a win for me. If I can force even a couple trades, that's gg. If my character has an easy way of doing this, the game might be as good as over.
Or, what about chip damage? Normally safe blocking tactics might be super risky in this situation, where small slivers of chip damage could end the game.
There's a lot of ways Ryu could win this--more than DeeJay.
When I was doing my 90%, I didn't suffer these disadvantages. I could throw out all the trade-able and chip-able options I wanted, without worrying that they might cost me the game. Now you are trying to do the same 90% back with significantly fewer options at your disposal, against an opponent who is fully aware of that and prepared to take full advantage.
You might say that Smash doesn't KO based on heath, so neither of these apply as much. But the fundamental idea applies exactly the same: in Smash, if I already did 120% damage to you, and you are trying to make a 120% comeback? You are now having to play around all the moves I could throw out that KO at 120%. I didn't have to do that while I was making my 120% on you first! Your comeback is harder than what I did to you.
This is a factor that many fighting games actively try to design around--Smash has rapidly regenerating shields, and SFV does not allow chip damage to KO. Still, the overall imbalance of "tie" outcomes remains.
Yomi Handicap of a "Forced Hand"
When you know this is your opponent's last chance--when you know there are some moves he HAS to defend against--you have him on the ropes.
Take baseball.
Baseball is "linear", right? The pitcher doesn't level up every time he gets a strike. There is no reason in the rules of baseball that any pitch should be more difficult than any other pitch.
But the data shows a different story altogether:
It's hard to bluff when you are one pitch away from losing it all.
If there are already strikes in the count, the batting average goes down. Way down. And the opposite for balls.
The more you understand yomi, the more intuitive this is. Yomi is all about evaluating the weights of options, and proximity to victory/defeat is one of the primary factors of those weights. It's not just psychology, but mathematically-founded game theory.
Impetus to Approach
All fighting games I know of use a timer. If the timer is up, the person "in the lead" just wins.
You know where this is going. (You've all played a Sonic before, or watched Hungrybox.) It means the winner has the luxury of assuming whatever stage position they'd prefer, and the burden of approaching them rests on the loser.
"Only a 5% lead!"
How much this matters depends wildly on the matchup, as well as the stage when we're talking Smash.
It's worth noting that this snowball mechanism is largely an implication of the external tiebreaker rules the community has imposed on the game. I'm not saying we shouldn't do this, but the snowball effect it adds is very much on us. (It's possible to design a tiebreaker system that doesn't do this or encourage stalling, but it would have to be built into the game.)
Other Snowball Factors
Many fighting games also have some form of super meter, which rewards you for landing hits. Smash doesn't really have this, except slightly for Little Mac and Cloud.
Smash does experience this pretty heavily with items, though! Item spawning largely removes the impetus to approach advantage, and replaces it with "whoever currently exerts more stage control is most likely to control the items." You see this most often in "Pokeball only!" games--the massive stage control that some Pokemon grant often let that player grab even more Pokeballs.
Our stage-picking set structure also favors whoever won the earliest matches, but this is a different topic entirely. (And frankly, not really a problem.)
And then there's player elimination in teams, which is just a super-snowbally complete mess. But that's, also, a whole 'nother topic.
Conclusion
The same principles of yomi that say a comeback is always possible also show that a comeback is always harder than whatever you are coming back from. This is the natural state of all yomi-based fighting games, in a vacuum.
Just like League, fighting games need comeback mechanisms to counter-balance their implicit biases towards whoever gets the early lead, so that the first hits are not disproportionately more important than all the action that follows. Fortunately, fighting games don't have it nearly as bad as genres like RTS or MOBAs, and can reach a more stable equilibrium with less effort.
We can hope that rage is tweaked from 15% of all knockback to 20-25% of knockback growth, but we should be thankful for the impact it's had on match excitement, moderating game length, and helping out heavies while keeping Diddy and Sheik in check. Let's hope that Smash 5 continues having eSports-friendly systems that provide us with years of tense matches.
This message paid for by Friends of Tsu Super PAC.
Fortunately, most games with some sort of uncapped skill test--including all games built on a stable yomi foundation--always both players the chance to make a winning move. And by extension, any number of winning moves they made on you, you could hypothetically make back. If they scored 30 points more in the first half, it's equally possible for you to score 30 more points in the second. If they won 47 points in a row in tennis, it's just as possible for you to turn around and win the next 48.
So why have comeback mechanisms at all?
The most common issue is that many competitive video game genres are built around snowball mechanisms, where success is rewarded by the winner getting stronger. So, they probably win even more, and get even stronger, until the game is super boring.
The midgame of professional SC2 can be really dramatic, but the ending that follows is always an anticlimactic "gg". (Save the rare base race)
League of Legends is chock full of exponential snowball mechanisms, and to remain a viable eSport it bends over backwards to include as many anti-snowball counterbalances as possible. Over the last 10 years, League has incorporated:
- Flat passive gold gain
- Inventory max
- Level max
- Exponential exp requirements to level
- "Catch up" exp when behind
- Reduced gold for kills on a repeated target
- Increased gold for kills on a target who has not died
- Reduced exp for kills on a lower level enemy
- Increased exp for kills on a higher level enemy
- Generous assist gold+exp tends to favor the losers, who are more likely to get kills as a group than solo
- Rubber-banded exp for jungle monsters based on your level (vs. average)
- Respawn timers scale with relative level
- Respawn locations relative to objective order
- Relative cost of returning to heal/buy from current objective
- Relative cost of maintaining vision/safety relative to objective order
- Inherent opportunities for passive play (existence of towers, opponents forced to leave to buy at some point)
- Towers have a hefty early armor bonus, to limit extreme early success
- Towers have an "anti-backdoor" armor bonus, to discourage base attacks that are difficult to respond to
- Multiple primary map objectives (often exclusive--"if you Baron we can at least Dragon")
- Defensive itemization structure that favors the loser--can buy items that mitigate specific gained advantage
- ...and more!
Wrong.
Snowball Mechanisms Implicit To Fighting Games
"Wait a second!" you cry. "Fighting games are based on yomi! If I take away 90% of my opponent's health in Street Fighter 2, it's still exactly as easy for me to turn around and take 90% his health back! This isn't a MOBA, his character hasn't gotten stronger! We're repeating the same decisions, with the same options!"
It turns out there are a handful of snowball mechanisms built into all fighting games pretty deeply. They are somewhat subtle relationships, but quite powerful. I'm going to look at a few of the biggest ones in detail:
- Inequality of Ties (trades, chip damage)
- Yomi Handicap of a "Forced Hand"
- Impetus to Approach
Inequality of Ties
Fine, let's say I took away 90% of your character's health in SF. You say it's just as reasonable for you to do it back in reverse? Not so fast, I've got some tricks up my sleeve.
Any outcome where we trade hits and both take damage is a win for me. If I can force even a couple trades, that's gg. If my character has an easy way of doing this, the game might be as good as over.
Or, what about chip damage? Normally safe blocking tactics might be super risky in this situation, where small slivers of chip damage could end the game.
There's a lot of ways Ryu could win this--more than DeeJay.
When I was doing my 90%, I didn't suffer these disadvantages. I could throw out all the trade-able and chip-able options I wanted, without worrying that they might cost me the game. Now you are trying to do the same 90% back with significantly fewer options at your disposal, against an opponent who is fully aware of that and prepared to take full advantage.
You might say that Smash doesn't KO based on heath, so neither of these apply as much. But the fundamental idea applies exactly the same: in Smash, if I already did 120% damage to you, and you are trying to make a 120% comeback? You are now having to play around all the moves I could throw out that KO at 120%. I didn't have to do that while I was making my 120% on you first! Your comeback is harder than what I did to you.
This is a factor that many fighting games actively try to design around--Smash has rapidly regenerating shields, and SFV does not allow chip damage to KO. Still, the overall imbalance of "tie" outcomes remains.
Yomi Handicap of a "Forced Hand"
When you know this is your opponent's last chance--when you know there are some moves he HAS to defend against--you have him on the ropes.
Take baseball.
Baseball is "linear", right? The pitcher doesn't level up every time he gets a strike. There is no reason in the rules of baseball that any pitch should be more difficult than any other pitch.
But the data shows a different story altogether:
It's hard to bluff when you are one pitch away from losing it all.
If there are already strikes in the count, the batting average goes down. Way down. And the opposite for balls.
The more you understand yomi, the more intuitive this is. Yomi is all about evaluating the weights of options, and proximity to victory/defeat is one of the primary factors of those weights. It's not just psychology, but mathematically-founded game theory.
Impetus to Approach
All fighting games I know of use a timer. If the timer is up, the person "in the lead" just wins.
You know where this is going. (You've all played a Sonic before, or watched Hungrybox.) It means the winner has the luxury of assuming whatever stage position they'd prefer, and the burden of approaching them rests on the loser.
"Only a 5% lead!"
How much this matters depends wildly on the matchup, as well as the stage when we're talking Smash.
It's worth noting that this snowball mechanism is largely an implication of the external tiebreaker rules the community has imposed on the game. I'm not saying we shouldn't do this, but the snowball effect it adds is very much on us. (It's possible to design a tiebreaker system that doesn't do this or encourage stalling, but it would have to be built into the game.)
Other Snowball Factors
Many fighting games also have some form of super meter, which rewards you for landing hits. Smash doesn't really have this, except slightly for Little Mac and Cloud.
Smash does experience this pretty heavily with items, though! Item spawning largely removes the impetus to approach advantage, and replaces it with "whoever currently exerts more stage control is most likely to control the items." You see this most often in "Pokeball only!" games--the massive stage control that some Pokemon grant often let that player grab even more Pokeballs.
Our stage-picking set structure also favors whoever won the earliest matches, but this is a different topic entirely. (And frankly, not really a problem.)
And then there's player elimination in teams, which is just a super-snowbally complete mess. But that's, also, a whole 'nother topic.
Conclusion
The same principles of yomi that say a comeback is always possible also show that a comeback is always harder than whatever you are coming back from. This is the natural state of all yomi-based fighting games, in a vacuum.
Just like League, fighting games need comeback mechanisms to counter-balance their implicit biases towards whoever gets the early lead, so that the first hits are not disproportionately more important than all the action that follows. Fortunately, fighting games don't have it nearly as bad as genres like RTS or MOBAs, and can reach a more stable equilibrium with less effort.
We can hope that rage is tweaked from 15% of all knockback to 20-25% of knockback growth, but we should be thankful for the impact it's had on match excitement, moderating game length, and helping out heavies while keeping Diddy and Sheik in check. Let's hope that Smash 5 continues having eSports-friendly systems that provide us with years of tense matches.
This message paid for by Friends of Tsu Super PAC.