Timer Clause

[tofucocoa]

Timer Clause is in effect what I'll be calling the option of timer stalling as it is on-cart. For those of you who don't actually own the game (for shame!), there's a timer implemented in every multiplayer battle now. For regular ranked matches and local battles, battles are limited to one hour, while special rated battles and official tournament matches are limited to only 15 minutes. When this timer runs out, the player with the most Pokemon remaining or HP wins the match.

So, even though your entire team might be in the red with a Scizor threatening a sweep, you can still win if you've taken enough time off the timer beforehand. This is balanced by the fact that you still need to stall those turns, so that Scizor might still run you through if you're losing a Pokemon every turn. Conserving Pokemon to win on timer adds another aspect to the game, albeit not an incredibly common one on-cart.

So, why would we even consider Timer Clause? To put it simply, it minimizes luck. I know you guys are currently complaining about Prankster Swagger, but if you think that's bad, try battling a Minimize team on the GBU (or Battle Spot, but GBU has a better acronym) metagames. These teams can't do enough damage fast enough to put themselves in a position to win on timer, so people that know about the game's tiebreakers simply timer stall these teams out. If they don't know to do so, Minimize+Toxic will quickly bring teams to their knees. Timer stalling on-cart is still painful to do, but at least they get the win and hopefully makes the person with Minimize think twice about their team composition, especially if this happens more than once. It's not hard to imagine how Swagger 'teams' in Singles would be influenced by Timer Clause either; as long as you can knock out at least one Pokemon and if they take their sweet time trying to confuse you, you're actually guaranteed to win against them. I recall someone showing me a 'Funbro' set that used Recycle Leppa and Block, which would obviously be subject to timer stalling as well, so there's another plus for you.

Speaking from personal experience here as a VGC player, matches can go to time even when we're not trying to stall. I had one match this past regional that went to time and most definitely made a difference, even though we didn't stall for time. I can only speak from experience here since I haven't seen or heard of too many matches being decided this way (probably due in part of how they restricted battle videos this gen), but considering VGC matches have considerably less time on their timer, it's going to be more common there, so it stands to reason we should at least have a way to practice for that scenario since PS! is a great testing ground for such things before actually going out and breeding and training full teams for actual events.

The cons to Timer Clause (that I see) boil down to being difficult to implement into PS! and how timer stalling there might encourage people to timer stall prematurely on cart, making matches less enjoyable. While 'premature' timer stalling on cart probably won't be used widely apart from people who have a perpetual stick up their ass, implementation of such a clause into a simulator would be tricky. Most of the problem here is the fact that the attack animations aren't nearly as long on the sim as they are on-cart, and PS!'s battle timer, whose only purpose is to hurry along someone's move, might conflict with Timer Clause. PS!'s battle timer might be a boon to keeping Timer Clause balanced though; it keeps people from just waiting one turn after they KO one Pokemon until the timer runs dry, much like the turn timer on-cart (which is, by the way, shorter for VGC as well at 45 seconds).

So, disregarding how hard it may or may not be to code because I have no idea how computer magic works, the real problem becomes battle animations. I propose a button that can pop up when someone's ahead in the match after a certain number of turns that will cause a timer to start ticking away and when it hits zero, the tiebreakers will be applied and someone will win on the spot. The time given and number of turns that triggers the button will obviously have to be balanced, and will more than likely be a long an arduous process, and of course I'm open to suggestions on this.

Overall, it's something that needs to be looked into, especially for vgc. Without the option to timer stall, strategies like Swagger spam and Minimize make laddering a lot less consistent and less skill-based. It's a mechanic on-cart that influences matches, so I don't see why we shouldn't be concerned with at least trying to emulate it on our official simulator, unless I missed the memo and we're not actually simulating cart mechanics.

Also, I've been told this has been discussed before but user: Pwnemon has not supplied me with a link and I haven't been able to find it. Regardless, I think it's something we should look into, if only for VGC matches.

 

[Pocket]

I think it would be cool to have a global timer where all games end in 1 hour. We added Team Preview to simulate the on-cart play experience. I don't see why we cant implement an "1 hour per battle" rule for our matches if timer is always activated on every multiplayer game on cart.

Despite the differences in PS! animation versus cart animations (PS! animation is much shorter), I also agree that we should try to incorporate as many VGC elements to the Showdown!'s VGC games, including a 45 sec-per-turn timer (after 45 secs pass, the top move is selected by default iirc; lucariojr can confirm) and a 15 min-per-battle timer, where the W/L is determined by the team's total # of Pokemon or HP (which one is it, lucariojr)? This would offer the most realistic VGC experience on simulator. I think PO already has VGC's specific timer rules implemented for their VGC formats.

Of course, if any of these worsens the overall play experience on showdown (ie makes PS! to lag terribly), then don't implement it.

 

[Katakiri]

One concern I have with the 45 second turn timer is that PS! isn't nearly as responsive at times of high traffic as the carts are. I'm not familiar with how PS! registers user inputs vs server inputs but a scenario like a player selecting a move during the last seconds of the turn timer only to have it not register before the server itself chooses a move for him/her could cause some controversy. Putting a delay between the time that user inputs are greyed-out and the time it takes for the server to decide a move could help to mitigate that but, again, I'm not familiar with the processes involved so it may not even be an issue in the first place.

 

[Hugendugen]

I was gonna post a thread about this, but since one was already made, I'll post a few comments here that arose as a result of the Endless Battle Clause. Aldaron and I had a lengthy discussion about the in-game timer and the Endless Battle Clause and agreed that there are 4 distinct questions:

1) without a timer, how should we deal with endless battles?
2) should we implement a timer?
3) if yes, how should we implement a timer?
4) with a timer, should we ban "endless battles" as a win condition?

1) has already been dealt with through the Endless Battle Clause (banning Leppa Berry + Recycle/Harvest and Leppa Berry + Shadow Tag + Trick). This was always intended to be a temporary solution while we worked out how to implement the in-game timer. Without the timer, this strategy was simply used to troll the opponent, for no real competitive benefit.

2) is almost definitely a yes. It's a restriction in the games, and since we strive to adhere to what is possible in-game, we probably shouldn't be allowing battles that last for more than 1 hour in XY metagames. Even so, it's an important discussion to have, since it will introduce an entirely new win condition that we've never used before.

3) is far trickier to answer. I'll recap how V4 explained to me that the in-game timer works (I understand that VGC games use a different timer)

Every wifi game has the timer on automatically in XY

Match Time Limit: 60 minutes
Turn Time Limit: 99 seconds
If the Turn Time Limit is exceeded, the move in the first slot is used for each Pokemon that has not selected a move.
If the Match Time Limit is exceeded, tiebreaker mechanics are used to determine the outcome of the match.

1) The player with the most Pokemon fainted loses.
2) The player with the highest Current Party HP to Total Party HP ratio wins.
3) The player with the most individual Hit Points wins.

If both players are equal in all three conditions, the match is a draw.

Now the issue is that battle animations in-game are a lot slower than they are on PS. This means that you could perform several more moves in the space of an hour on the simulator. The question arises whether it is the amount of time the battle takes or the number of moves possible that should be restricted. There are a few different possible ways of tackling this:

• A strict hour limit.
• A limit based on the maximum number of turns possible in an hour, using the move with the shortest animation times.
• Calculating the animation time of every move and keeping track of this during the battle to limit based on how long the game would have taken in game. (This would be the most accurate, but also by far the most difficult to research and implement)

If anyone wants to research the last two, I will personally kiss them on the lips.

Note that we probably wont be adopting the in-game turn timer outside of VGC, since it's far more abusable than the one currently used on PS.

4) once the previous question has been answered, we need to determine whether or not we want to allow the strategies brought up in the endless battle discussion as a method of winning through timer mechanics. This could be done by drawing out the timer using one of the many strategies that force endless battles and simply ensuring that you are ahead according to tiebreaker mechanics when the timer runs out. This is a new win condition in Gen 6 and we'll need to debate whether it is broken and/or within the spirit of the game. I obviously feel pretty strongly that it's not within the spirit of the game, but the decision will be up to the OU council and the individual tier leaders. Ideally it would be nice if we could all agree on the same decision, but there is a precedent for the official metagames using different clauses (for example Ubers removing the Evasion Clause).

I'm especially curious what the tier leaders have to say (mainly with regards to questions 2 and 3), however anyone with access is welcome to post.

Tags for good measure: Aldaron, Haunter, McMeghan, Nachos, M Dragon, kokoloko, Molk, Raseri, blarajan, Joim / Pocket (still have no idea), Tobes / TheMantyke, chaos (as unofficial randbats tier leaders)

Edit: since there seems to have been some confusion, I'm asking for all XY metagames, not just VGC. VGC should definitely use the VGC timer, though we can also discuss what the best way to implement that is. And to reiterate, it's not possible to play a wifi game without using the timer.

 

[Pocket]

Why can't we implement in-game turn timer? It seems much better to have the top move selected if the turn timer runs out than losing the game at the spot. I admit that 99 seconds may be shorter than what we are used to, but if this is how it works in reality, why not simulate it? I am disregarding the difference in animation time, since there is really very little we can do about that (other than slowing down PS! animation to match the in-game one, which would be an awful request). You speak of potential abuse, but where does the abuse come into play?

 

[Hugendugen]

Since it was brought up on IRC, if there are enough people willing to help, I might start a project to research:

• A limit based on the maximum number of turns possible in an hour, using the move with the shortest animation times.
• Calculating the animation time of every move and keeping track of this during the battle to limit based on how long the game would have taken in game. (This would be the most accurate, but also by far the most difficult to research and implement)

If you're interested in helping, shoot me a pm so that I can gauge the numbers. I don't have much time to research it myself, but I don't mind running and organizing a team of covert operatives (researchers) to get it done. It could very well put you over the edge to get that researcher badge ;)

Also the offer still stands about that kiss on the lips.

 

[TheMantyke]

oh neat I get tagged in pr stuff now

So I have very little to add to this conversation on the topic of endless battle clause, but I do want to address concern #3 and lucariojr's concern.

Calculating the animation time of every move and keeping track of this during the battle to limit based on how long the game would have taken in game. (This would be the most accurate, but also by far the most difficult to research and implement)

I think that if we're going to make a timer, we go all out, take this option, and have it fully reflect game mechanics. This unfortunately would mean a boatload of research into not only how long each individual attack animation takes, but also how long things like sending Pokemon out at the start of a match, afterturn effects, shiny animations, switching animations, HP Drain animations, mega evolution, abilities activating, and time spent communicating between DSes before and after moves takes. It's a pretty large project, but wins from time outs are not uncommon at higher levels of play. I do think this is something worth implementing. This community also has a tendency to swallow difficult pills when it comes to tedious research projects with pixel incentives on the line so orchestrating a large scale research project for this sounds like a possibility. Implementation or at least the plan on how to implement it sounds fairly straightforward. If we can find times for how long each action takes, we can subtract the time the animations would take from the timer once the turn's been made. I understand that this is likely nowhere that easy for the PS! team to implement, but it seems like the most straightforward implementation plan.

I would advise against any sort of half way implementation to game mechanics, at least for the sake of VGC play where the endless battle clause is of no concern. During early 2012, I remember our PO server had a time limit for VGC matches that was intended to make battles more accurate to the event experience. This didn't really reflect truthfully on how the actual matches worked since battle animations and such we're accounted for in the time lost, so it ended up just being some feature players weren't really influenced by at all.

So essentially, I think our options should be either being completely accurate to the cartridges or to leave things as is. Out of those two, I'd prefer we take the first option. I don't think we should implement a timer that's semi-accurate since semi-accurate would still be fully inaccurate.

also sorry mingot but if we do this your minimize chansey vgc 2013 team is kaput ;x;

 

[Arcticblast]

Hugendugen

• A limit based on the maximum number of turns possible in an hour, using the move with the shortest animation times.

If you use an offensive move on your partner there will he a very short animation basically consisting of "smack!" and a small impact effect. I can't guarantee this working for every move, but if it's a short animation time we're after than Evolite Chansey Hidden Power should do it.

 

[kokoloko]

hugen tagged me so i'll oblige and go on record saying i don't care if we do any of 1. use an optional timer as we do now, 2. use a PO-style timer, or 3. use an in-game-like timer.

people will bitch and adapt no matter what so who cares.

 

[Layell]

I'm going to offer a stance on this, since I was one of the people poking a few holes in this concept in irc. To pull this off correctly and completely, which ought to be the only way we do this we would need to:

-Calculate the length of all 600+ move animations
-Be reasonably sure that different pokemon don't take different lengths of time for using the same move (
-Calculate KO animation length
-Calculate entrance animation length
-Implement stat boost time limit
-Implement Weather display time waste, and residual weather damage time spent
-Implement amount of time to restore HP via rate of HP gain
-Checking switching time
-Mega Evolution and other transformation time check
-Calculate the rate of HP loss per damage percent
-Check if speed of turns grows quicker over the course of many turns (from personal experience this seems to be true but we'd need to check)

Smogon has done a number of huge number of admirable research projects over the years that have improved the sims we used. But this would not be one of them. The average player does not want to wait 10 seconds for leftovers and sandstorm text to scroll by, then for Aegislash to go from shield to sword stance, then attack. Players play sims because we have made them quick efficient and easy to use.

In terms of accuracy for VGC I question this stance, any VGC player worth their salt uses the sims to test, builds in-game, then tests some more. By necessity they have to, they use the sims to test out teams quickly and efficiently, and make changes on the fly. We may actually do a disservice to them by making matches take longer.

Requiring a timer in any case may hurt some players who don't have the best connection, or even some of the fantastic 100+ tournament matches we sometimes have.

 

[TheMantyke]

Smogon has done a number of huge number of admirable research projects over the years that have improved the sims we used. But this would not be one of them. The average player does not want to wait 10 seconds for leftovers and sandstorm text to scroll by, then for Aegislash to go from shield to sword stance, then attack. Players play sims because we have made them quick efficient and easy to use.

In terms of accuracy for VGC I question this stance, any VGC player worth their salt uses the sims to test, builds in-game, then tests some more. By necessity they have to, they use the sims to test out teams quickly and efficiently, and make changes on the fly. We may actually do a disservice to them by making matches take longer.

I'm not for making the simulator take as long as it would on a cartridge if I didn't make that clear. The way I pictured this being implemented would be having the simulator execute everything as it does normally, then at the end of the turn when players are allowed to select their next move, an amount of time corresponding to how long the actions of the turn would take on a cartridge is deducted from the timer.

Example:
The turn starts with 10:30 remaining on the timer
Ferrothorn and Rotom-W are on the field
Rotom-W switches out to Garchomp
Garchomp takes Stealth Rock Damage
Ferrothorn uses Leech Seed on Garchomp
Afterturn effects occur: Garchomp loses health, Ferrothorn restores some of his.
Though Showdown takes maybe 10 seconds to do all this, a minute is deducted from the clock as that is how long it would take on a cartridge, leaving it at 9:30 and counting as players select their next move.

 

[Layell]

As I discussed on irc a system that removes time as accurate to cartridge, without letting the timer play out removes the important part of players thinking about their next move before they have to select it. If I have to wait for that brave bird animation I am thinking about my next move before the clock starts ticking.

Removing time to match how long it would take on cartridge will also confuse the hell out of new players.

I just can't see any situation where adding an auto timer improves the simulator experience, we already have a relatively idealized system with percentages, and sleep clause. We should prioritize matters of convenience with the sims.

 

[Joim]

I wouldn't like PS turns to take longer. It'd be detrimental for everyone and serves no clear purpose. It would only help to train and prepare for wi-fi / NDS tournaments, and that's about it. I'm fairly certain that we can adapt the in-game timer to PS without the need to make animations or turns slower. We already have much more time than the wi-fi timer gives, so who cares if we give slightly more?

What we could do is calculate how many seconds should VGC and GBU turns last on the sim to be as accurate as possible (same way maybe we should get rid of cancel move in those formats).

We should enforce wi-fi like timers on all games. It should be able to be deactivated only on challenges so friends can play without it. This would enhance competitive play (think faster, think better), would limit ghosting and, well, the sim would be more accurate and I'm all for accurateness. I'd still ban endless battle strategies, because abusing a strategy trying to make your opponent forfeit or lose time is totally uncompetitive and has absolutely no merit.

 

[Pocket]

Nobody is proposing to make the animations slower. However, I do agree with Joim and TheMantyke in making our timer as accurate as possible. Layell does bring up a good point - the ideal "calculated" timer option may take away the time that players would use to plan out their moves in wifi play. All the hard research that goes into perfecting this timer may go to waste. The work:reward may most likely not add up.

The middle option "maximum turn per game" seems like a good alternative to try, imo. It would be cool if we can make it more fluid than a static cap on turns, though. Say if the animation time for a standard move (ie 3 secs) has a value of 1. Then longer animations (ie Mega Evolution) would have a value of 2. Burn / poison / sleep / freeze would all count towards animation time as well. Super long animations such as Brave Bird would probably have a value of 2.5. Every switch and faints on both sides need to be accounted, too, as well as the summoning of weather, etc. Shorter animations (failed attacks) would have a value of 0.5 or something. Once the cumulative value hits 1200 (3600 secs/3 secs) then the game is over. So say at the beginning of the game, 2 Pokemon enter the battlefield, one Pokemon mega-evolves and attack while the other switches out to another Pokemon, the total amount of animation time value would be 7 for that turn (2 from both sides summoning their Pokemon, 2 from mega evolving, 1 from attacking, 1 from switching out, 1 from switching in). The maximum amount of turns left would then be displayed as 171 turns (1200/7), which would fluctuate as the game progresses.

Moderate amount of research needs to go into this to group animations into short, normal, long, super-long, etc animations, but I think the effort would be more fruitful than the "calculated" timer option, mainly because we're counting the maximum game duration in terms of turns rather than time (so you avoid the problem that Layell outlined). Even though we're not counting game duration in the same manner as wifi / vgc, it would offer the closest "in-game" experience imo.

If we cannot do the middle option "maximum turn per game" for whatever reason (too much strain on the simulator, etc) then we should simply set a strict time limit. We can simply go with 1 hour like wifi, or take into account the shorter animations and cut the time limit to say 30-40 mins. Not sure how we would calibrate this exact time limit, though.

PS: Joim and I are co-leaders of Doubles :>
PS 2: Keep the Endless Battle Clause intact. It's a dumb strategy that doesn't try to win games.

 

[Ash Borer]

Currently the move time limit (180 seconds on the ladder, 300 seconds in tournaments) on Showdown is not the same as the move time limit on cart, yet no one cares. Why try and accurately depict the game time limit on Showdown? There's no reason to waste time trying to record the time it takes for every animation on cart.

the actual principle of this is whats important. A game time limit would be pressuring to something that is being.. discussed as a potential suspect in XY OU and that is Baton Pass. Baton Pass works by defensively boosting, for free essentially as they can take the hits then pass off to fresh teammates. By necessity this strat loses to the game timer if it takes too long to sweep with their Espeon meaning that theyre not free to sub protect and boost for 100 turns inconsequentially.

This is important as clearly the competitive aspect of this game done by Game Freak is made with their own timers, and such in mind. Why are we not following?