Timeouts, Tiebreakers, and The Lead

[Sixfortyfive]

The ideal winner of any timeout should be the player who spent the most time closest to the center of the stage, possibly weighted towards the end of the match.

We Divekick now.

 

[Raijinken]

The ideal winner of any timeout should be the player who spent the most time closest to the center of the stage, possibly weighted towards the end of the match.

Alas, it is not to be.

My brother argues very hard that Stage Control should be the deciding factor. And as much as I agree, it's presently too subjective. You'd basically need stopwatches (or some pretty nice software) to watch each replay and track character position relative to a defined point on the stage, and then increment score at a relative rate to proximity.

Would be cool, but can't be done. Even with the software, anything short of it running during the match via an external capture device would make it take up the time of another match to process.

Also, is Little Mac really winning if he's center-stage against a Villager he can't safely go near?

 

[Touchebag]

The ideal winner of any timeout should be the player who spent the most time closest to the center of the stage, possibly weighted towards the end of the match.

Alas, it is not to be.

This sounds good in theory (if we ignore the obvious practical issues) but is this really a fair way to judge control of the game? I would assume that is what you really want to measure. Defensive characters with projectiles will want to stay at one end of the stage and just force their opponent to approach them (e.g. Fox) while taking potshots at them. Their opponent will undoubtedly be closer to the center of the stage but are they really winning if they can't even get close to Fox?

 

[ぱみゅ]

For some time I used to believe perentage lead wasn't too important until passing a certain point.
As in, a 20% deficit is not too important so they're virtually even, but a 50% deficit is kind of an uphill battle.
But where is the line drawn? How much is the breaking point of being even and not?
That's the reason I dropped the idea

/brainstorming

 

[Zonderion]

For some time I used to believe perentage lead wasn't too important until passing a certain point.
As in, a 20% deficit is not too important so they're virtually even, but a 50% deficit is kind of an uphill battle.
But where is the line drawn? How much is the breaking point of being even and not?
That's the reason I dropped the idea

/brainstorming

I think this is a good idea. If opponents are within a certain % range of each other, it is considered even. 25% would be a good indication. Anything above that means the player is at a disadvantage and vice versa.

 

[Touchebag]

I think this is a good idea. If opponents are within a certain % range of each other, it is considered even. 25% would be a good indication. Anything above that means the player is at a disadvantage and vice versa.

We still have the problem with weight as discussed earlier. Is a Bowser at 50% really losing to a Jigglypuff at 20%? Bowser is farther from KO percent.

 

[Zonderion]

We still have the problem with weight as discussed earlier. Is a Bowser at 50% really losing to a Jigglypuff at 20%? Bowser is farther from KO percent.

I don't think we can worry about which character is closer at KO percent, because that becomes a very fast moving target that depends on the characters, as well as their opponents.

Ryu can kill Jiggly with breaking her shield with his strong F-tilt. Technically in that match up Jiggly will always be closer to kill percent because she can die at 0 and beyond.

The percentages, IMO, aren't to determine who is closer to dieing, but who has more control of the match.

If the Bowser can control the match, then he can prevent Jiggly, or whomever, from tacking on more than 30% (or insert % here). If he can't, then Jiggly is the one controlling the match.

 

[SpaghettiWeegee]

El threadbumperino (oops!)

If the Bowser can control the match, then he can prevent Jiggly, or whomever, from tacking on more than 30% (or insert % here). If he can't, then Jiggly is the one controlling the match.

Against Jigglypuff, sure, but Smash 4 is a game wherein many characters can rack up an easy 30% (especially from low opponent starting percent) off of one hit confirm. Is a player actually playing better than their opponent in that situation, or is that just a quirk of the matchup? If you put a Shiek against a Ganondorf and the match ends with Ganondorf at a 20% deficit, that actually indicates significant control against the Shiek coming from the G-dawg precicely because Ganondorf doesn't have the tools to combo as effectively (and thus rack up as much damage as quickly) as Shiek (theoretically this is balanced by kill power and weight-- theoretically).

If we were gonna get really silly we'd just let the winner be decided by Twitch chat/live audience. Incentive to bring the hype.

 

[FallenHero]

What if the percent difference was less than 30% at last stock and the timer runs out they play a 1-stock 3 min. match? Some people say 20%, but I don't think getting a 20% lead is really all that difficult to do. Maybe 25% would be a good compromise.

 

[SpaghettiWeegee]

As I mentioned in this thread, traditional fighting game balance with respect to timeouts can't apply to Smash-- Smash is a different game. The percent system wants desperately for an alternative, that much I think is certain.

One of the major things that makes the percent system in Smash different is that it means different things for different characters. Since it's a record of damage taken, not health remaining, it's difficult to say when two characters have roughly even health in Smash. I would say it's analogous to how different characters have different hit point values in a MOBA, for example-- in Smash, higher weight is similar to higher max HP, since presumably heavier characters can survive for longer. We need to take this into account somehow when determining tiebreakers.

I detailed a fuller solution in the thread I linked, although here I would like to say that factoring in character weight when considering who was in the lead at the time of timeout is akin to seeing which character is closest to their kill percent (or the point in the game where they're effectively dead if not actually dead) rather than just comparing two static (and in the case of Smash, almost meaningless when taken out of context) numbers. Additionally, it'd be pretty easy to implement in tournament: just have a chart of character weights on-hand and do a quick calculation. Sure, it's not as quick as the system we have in place now, but it's much more sensible for a game system like Smash.

EDIT: I'd also like to offer that trying to figure out who has "momentum" at the moment of time-out, regardless of how much it affects a player's actual ability to win games can't factor into determining the winner of tiebreakers. Momentum is nebulous and non-measurable; additionally, it's constantly changing throughout the course of a match-- momentum can be broken in a matter of frames by a player choosing the right option. The winner in an overtime should be the player considered in the best position to be likely to win, not simply whoever looked like they were going to whenever the clock hit zero.

Possible edge cases include when the clock runs out with one of the players in freefall after clearly being gimped-- sure, it's not numeric data, but if the clock just had a few more seconds on it, the winner of the game would be obvious. Maybe edge-case rules could be constructed specifically for this sort of thing (sorry, Little Mac).

 

[teluoborg]

My brother argues very hard that Stage Control should be the deciding factor. And as much as I agree, it's presently too subjective. You'd basically need stopwatches (or some pretty nice software) to watch each replay and track character position relative to a defined point on the stage, and then increment score at a relative rate to proximity.

Would be cool, but can't be done. Even with the software, anything short of it running during the match via an external capture device would make it take up the time of another match to process.

Also, is Little Mac really winning if he's center-stage against a Villager he can't safely go near?

It can be done and in the simplest way :
If the match goes to time then the player who is the closest to the center of the main platform is the winner.

Explanation : positioning matters more than damage dealing in Smash, we're not playing Street Fighters and it doesn't matter if you reach 999% as long as you don't touch the blastzones. Having clean numbers to compare at the end of the match is nice and all, but as it has been said multiple times in this thread percentage doesn't reflect accurately the momentum of a match.
Moreover when the timer is getting close to zero the players tend to change completely their behaviour, play very defensively, sometimes go so far offstage that it would lead to a certain suicide if the timer didn't stop the character's fall.

On the other hand, having the victory decided by positioning forces the players to get aggressive, create momentum and take control of the stage. It simplifies the winning condition from "send your opponent into the blastzone" to "send the opponent as far as possible", giving players the possibility to win even if they have 100% more than their opponent, just like in normal matches.


Conclusion : Smash doesn't have health bars, so let's stop pretending victory can be decided by percentage alone.

 

[Raijinken]

It can be done and in the simplest way :
If the match goes to time then the player who is the closest to the center of the main platform is the winner.

Explanation : positioning matters more than damage dealing in Smash, we're not playing Street Fighters and it doesn't matter if you reach 999% as long as you don't touch the blastzones. Having clean numbers to compare at the end of the match is nice and all, but as it has been said multiple times in this thread percentage doesn't reflect accurately the momentum of a match.
Moreover when the timer is getting close to zero the players tend to change completely their behaviour, play very defensively, sometimes go so far offstage that it would lead to a certain suicide if the timer didn't stop the character's fall.

On the other hand, having the victory decided by positioning forces the players to get aggressive, create momentum and take control of the stage. It simplifies the winning condition from "send your opponent into the blastzone" to "send the opponent as far as possible", giving players the possibility to win even if they have 100% more than their opponent, just like in normal matches.


Conclusion : Smash doesn't have health bars, so let's stop pretending victory can be decided by percentage alone.

I like the idea a lot in theory.

But what if my opponent and I are both practically at the middle, and neither is clearly closer than the other. What about character size? It's a really good determining factor, but enforcement becomes exceptionally hard (with the sole benefit that it is rarely needed to begin with).

 

[M15t3R E]

I agree it's inadequate to decide the winner based on percentage alone. Some have suggested a calculation that takes into account damage done and damage taken but even that is skewed against the player that got the kill at an earlier percent.
One thing's for sure: sudden death is terrible.

 

[FallenHero]

But what if my opponent and I are both practically at the middle, and neither is clearly closer than the other.

1-stock 3 min match.

What about character size?

Maybe if the majority of the character's hurtbox is at the middle of the stage?

 

[Raijinken]

1-stock 3 min match.



Maybe if the majority of the character's hurtbox is at the middle of the stage?

What if I time it so that I'm intangible at match end? Teleport/Warp/Vanish?

This is obviously just worst-case nitpicking, and I think a subjective TO ruling would suffice in such a rare (and probably contrived) case.

 

[FallenHero]

What if I time it so that I'm intangible at match end? Teleport/Warp/Vanish?

This is obviously just worst-case nitpicking, and I think a subjective TO ruling would suffice in such a rare (and probably contrived) case.

I don't really know what to do then, but technically the character's hurtbox isn't there when they do something that makes them intangible, so TECHNICALLY they aren't there

 

[Raijinken]

I don't really know what to do then, but technically the character's hurtbox isn't there when they do something that makes them intangible, so TECHNICALLY they aren't there

Works for me!

Still a pain to enforce that rare case, but I think it's unlikely enough to begin with that it'd still be worth using over Percent.

 

[Zonderion]

Those last few seconds will be spent dodge rolling towards the center.

 

[FallenHero]

Those last few seconds will be spent dodge rolling towards the center.

Maybe, but all it could take is one f tilt or jab to get the opponent away from the center at the very end, so some might not want to risk getting hit away from the center at the point when they are tangible and cannot roll again.

 

[Raijinken]

Those last few seconds will be spent dodge rolling towards the center.

Better aim well or you'll move past it, or get punished.

 

[teluoborg]

I like the idea a lot in theory.

But what if my opponent and I are both practically at the middle, and neither is clearly closer than the other. What about character size? It's a really good determining factor, but enforcement becomes exceptionally hard (with the sole benefit that it is rarely needed to begin with).

If you and your opponent are practically at the same place maybe you could, I don't know, kick him away ? Of course it needs to be tested but realistically I don't think having 2 people battle for the same spot will result in both character being really close without hitting each other unless they agree to.
And if you use a move that makes you invisible/teleport/incapable of telling your position while perfectly knowing the rules then it's your fault. It's like asking "what if Link has the percentage lead but at the end of the timer he damages himself with his bomb and loses the lead ?". Link went full ****** and he lost. Don't go full ******.

 

[Raijinken]

If you and your opponent are practically at the same place maybe you could, I don't know, kick him away ? Of course it needs to be tested but realistically I don't think having 2 people battle for the same spot will result in both character being really close without hitting each other unless they agree to.
And if you use a move that makes you invisible/teleport/incapable of telling your position while perfectly knowing the rules then it's your fault. It's like asking "what if Link has the percentage lead but at the end of the timer he damages himself with his bomb and loses the lead ?". Link went full ****** and he lost. Don't go full ******.

Again, I think the rare cases are too unlikely to concern with. I just think it's interesting to ponder things like Wii Fit's up-smash timing.

Also, we are assuming we're talking center of the bottom-most platform right? And I guess a reference point could be declared for each part of Delfino or Skyloft or whatever other stages are run.

 

[MrGame&Rock]

Maybe an easier to track rule could be "the person who was most recently offstage or a certan distance above the stage when the game goes to Time loses"

 

[FallenHero]

Maybe an easier to track rule could be "the person who was most recently offstage or a certan distance above the stage when the game goes to Time loses"

Who is going to keep track of that?

 

[Touchebag]

I still think the best way to deal with time outs is to diminish the gain from doing so. As I previously mentioned in this thread I believe that the current percentage based system could be kept but that one timed out match of a set is essentially meaningless on it;s own. It simply counts as a double loss. Only after the whole set a tiebreaker is run if they are still tied.

I.e. you could win the first game, tie the second and win the third in a Bo3 and still win the set without having to play a tiebreaker.

 

[ぱみゅ]

I still think the best way to deal with time outs is to diminish the gain from doing so. As I previously mentioned in this thread I believe that the current percentage based system could be kept but that one timed out match of a set is essentially meaningless on it;s own. It simply counts as a double loss. Only after the whole set a tiebreaker is run if they are still tied.

I.e. you could win the first game, tie the second and win the third in a Bo3 and still win the set without having to play a tiebreaker.

That's actually really interesting. Not sure if worth testing, considering Timeouts aren't THAT common.

 

[Touchebag]

That's actually really interesting. Not sure if worth testing, considering Timeouts aren't THAT common.

Right. And they'll be even less common if you don't really gain anything from a single timeout. Timeouts will always be controversial and/or boring to watch. The more incentive there is to finish a match within the time limit the better.

 

[Azazel]

What are we trying to accomplish with tournaments?
tournaments are a way to determine skill in this game

Hypothetically lets think of what a no time limit tournament tests.

skill is determined by your ability to win in this game
to win in this game you need to remove your opponents stocks before they remove yours
Skill is determined by your stock taking and holding ability.

timing out has nothing to do with winning ie skill

but because of tournament logistics, we add timeso skill is now removing stocks before time, or having stock/percent lead when times up.

Tournaments need time and to accurately measure skill
we don't use sudden death because it adds time and randomness which disproportionately benefits less skilled players and therefore compressing the skill gap. this is what we want to avoid at all costs, anything that skews our measurement of skill.

for example, Online tournaments don't count to your rank because lag is randomness which disproportionately benefits less skilled players, compresses the skill gap, and skews our measurement of skill

unless you can provide a better alternative to percent lead time out, it is going to stay this way.

What is the ideal alternative to Time?
the ideal alternative needs to

• above all accurately measure skill
  
• save time

Ideas of extra tie breaker matches is non-conducive to our goals of time management.
And there is little evidence here showing that timed matches add enough randomness that undermines the Skill factor to warrant tiebreaker matches

prove that time outs are random or disproportionately benefits less skilled players vs more skilled players and skews our measurement of skill to warrant tiebreaker matches

I'd like to see an example where time was significantly benefiting a less an objectively less skilled player vs an objectively more skilled player and how time out was random.

I don't see how time is random at all. it is something that is displayed and ticks down at a regular pace. it's something consistent that you can strategize around.

If your problem is how timeouts are seen as by the public and the game as a spectator sport, crys against boring campy timeouts have more weight.

the ideal alternative needs to

• above all accurately measure skill
  
• save time
• discourage campy timeouts that spectators don't like

One idea that I saw here that does all of those was factoring in percent AND stage control (divekick) or like momentum/right of way like in fencing
but how do we objectively calculate stage control or right of way (momentum)

How'd I'd imagine an objective stage control is like this.
In the event of a timeout, percents are recorded
depending on the position of the player percent is added to the player.
then we compare the final percents

in traditional fighters timeouts go to the player with more health IE the opponent was closer to death
adding percent based on stage position makes some sense as in, you were closer to the blast zone IE death

Spoiler

lets say for example each vertical and horizontal tick mark adds 15%

If someone attempts the timeout strat, the goal is to simply not get hit and run away.
what this does is add an objective that forces the players to approach IE discouraging boring campy timeouts

Time gives the person with the lead an even greater lead as time goes on.
stage control potentially gives the person who is down a comeback factor but hey mo more boring campy timouts right?(spectators may or may not like this, this can feel like randomness although it really isn't, but considering spectators get salty over Rage. . .).

time approaches, center stage becomes an objective
a player is down 40%
the person who is down gets a clutch smash attack while in the center
because of how the lead was calculated the down player effectively made up the deficit, 15% smashattack+ 30% "positional advantage"

this idea has a slew of issues, but it sounds cool in theory. something more simple would be more ideal

so far I think this was the simplest solution as long TO's don't confuse people, although this idea does not address all the gripes/salt with timeout.

I still think the best way to deal with time outs is to diminish the gain from doing so. As I previously mentioned in this thread I believe that the current percentage based system could be kept but that one timed out match of a set is essentially meaningless on it;s own. It simply counts as a double loss. Only after the whole set a tiebreaker is run if they are still tied.

I.e. you could win the first game, tie the second and win the third in a Bo3 and still win the set without having to play a tiebreaker.