Proposal Enforce Battle Timer

Status
Not open for further replies.
Karxrida

Karxrida

Eventide (art by @kzhjp)
is a Community Contributor Alumnus
With the semi-recent removal of Sleep Clause Mod in favor of a Sleep ban and a move further towards cart accuracy, I think we need to discuss the battle timer (or lack thereof) on the simulator.

Starting from I believe X and Y, on-cart PVP has employed a hard 60 minute (20 in Sword and Shield lmao) timer that forces a winner based on which player has the most remaining Pokémon and remaining HP upon expiration. Our refusal to enforce this and instead utilize an optional turn timer is a glaring inconsistency in our desire to be cart accurate, especially when it removes a valid win condition.

Even though we cannot perfectly recreate the timer with total accuracy due to differences between sim and cart in regards to move animations and overall game speed, opting to instead do nothing falls victim to Nirvana fallacy. This refusal to act creates issues since games can and have gone on forever or end up excessively long, and it's odd for a competitive community to forgo such a measure that is otherwise commonplace. If I participate in a Yu-Gi-Oh! local I have 45 minutes for each set. If I play Mega Man Battle Network I have 15 turns to win each game before Damage Judge. Etcetera.

Even if we did something like reduce the timer to 45/50 minutes to account for sim/cart differences, SV is fast enough that there is still enough time for most games to be completed without issue. And the ones that do run up against that time are probably stall v stall grindfests that really really need to end in a reasonable timeframe.

While this proposed change is largely for current gen, I do think ORAS and SM could easily implement it without much hassle. I distinctly remember doing on-cart PVP during ORAS and never hitting the end of the timer, and also remember that one time a tournament match went so long (1000 turns or something) that this issues came up here in Policy Review lol. (No I don't know where that thread is.) However, this is ultimately up to those communities to decide.

SS is fucked, though. Probably shouldn't bother trying to enforce it there.
 
One note in particular with this proposal, if it is made to be implemented:

There are two additional timers on cartridge, your Total Personal Time which is hard locked to 30 minutes (unless I'm mistaken), as well as your Move Time. Players are locked into picking their moves within 60 seconds on cartridge, with the first move slot being selected automatically if that timer runs out (ignoring choice locks or a move running out of PP).

There's a few questions that come up when talking about these timers, I do personally believe that running off Total Personal Time is a better indicator for Timeouts than total match time, since one of the main issues are MOVE ANIMATIONS.

I think these being inconsistent among both players frankly just means we should be completely ignoring the Total Battle Timer. Total Personal Time will always be shorter if we discount animation times.

Obviously Pokemon Showdown has move animations that are skippable, and this plays a lot into at what point should your timer start playing.

There's a few options here that are possible, with the first being my favorite:

1. Neither player can select a move until both players have finished watching move Animations, both Total Personal Times + Move Time begin simulteanously

This makes their timers run down at an equal pace, with no regard of someone skipping the animation. Skipping the animation gives you more time to think of the game state, but also gives the opposing player to have the chance of watching the move animation, without you losing out on time because of it.

2. Total Personal Time + Move Time begins as soon as both players have stopped seeing move animations
This makes it so you can pick a move as soon as you skip, this does give you an advantage over someone that never skips, by making it so no time passes at all until they have finished seeing the animation. This negatively penalizes a player that wants to see animations.


3. Total Personal Time + Move Time begins as soon as player has stopped seeing move animations

If a player skips the animation, their battle timer begins, this option makes it so you optimally should never skip a move animation, since it means lost time. This negatively penalizes a player that wants to skip animations.

4. Entirely remove skipping animations and calculate the animation time of every move in the game, removing their value from the Total Game Time

lol


Now here's the thing, all of these are a huge fucking headache to code (probably), if you wanted to instate true cartridge accuracy, you'd have to advocate for one of these options, since they're all "more accurate" than running off total time. Option 4 is technically the only fully cartridge accurate solution.

I personally don't think achieving cartridge accuracy for the timer is worthwhile at all, it doesn't affect the reproducibility of battles in a meaningful way compared to something like Sleep Mod.

Wanting to be fully cartridge accurate essentially means removing the QoL of skipping move animations, or doing huge gymnastics around figuring out how they should affect the timer, which ultimately devolves into a subjective mess.

The timer is fine as is, I think ignoring cartridge accuracy when it comes to the Timer is entirely ok. If we are to change the timer, I would honestly want to do it fully, by not implementing a single 60 minute timer, but the 2 others as well to not half ass it. However this would be a monumental change to the current action time code, and would not be something we have the capacity to decide unless Showdown Programmers are willing to implement it.
 
Karxrida said:
Even if we did something like reduce the timer to 45/50 minutes to account for sim/cart differences, SV is fast enough that there is still enough time for most games to be completed without issue. And the ones that do run up against that time are probably stall v stall grindfests that really really need to end in a reasonable timeframe.
Click to expand...

Reasonably long games can easily hit 45/50 minutes, assuming both players take an average of 30 seconds to make a move over 90/100 turns that would result in the timer ending the battle. 100 and especially 90 turn games aren't the majority but they happen commonly enough for timer to be an issue for games that don't involve stall mirrors.
 
There is basically no reason to enforce this lol the current timer works perfectly fine and cases where games take a long time are outliers and frankly there is nothing wrong with two players duking it out to the best of their abilities with high-longevity teams for a few or several hundred turns
 
Some of the most memorable games I've ever played/watched have easily been over an hour. The thought of removing that possibility is just terrible. Smogon mons is extremely different from cart mons, and these endeavors to move closer and closer to replicate cart in all ways are eventually going to take us further away from what playing mons on Smogon really is. If all we wanted to do was play the game according to all rules, boundaries, and limitations that the cart forces for us, then we would just play cart. It's not very difficult these days to get perfect IVs EVs etc. to do so.
 
I do think one thing not being considered is how using a match timer can make scheduling for tournaments much easier. Having a clear upper limit in how long a set can take (+15mins for them to show up on time) means not having to carve out an unknown amount of time from our important busy schedules.

Also, from my experience playing vgc/bss cart formats for a decade, time outs as a win condition can be very competitive. It very rarely comes up. Nobody plays with it as the intent from turn1, but when the games naturally flow that way and players realise the match timer is getting low, there's a really interesting shift in priorities as the wincon changes.

I really admire the pursuit of making every game play out to its fullest extent, I think I consider this a pure view of the game, but playing with timers isn't less or more competitive, they're just different and even then only in those relatively rare cases. Or I guess, you effectively choose the rough frequency with how long the timer is set to be.

Anyway, the important part of what I wanted to say was that respecting each player's valuable time in scheduling and play is really the biggest benefit to a match timer
 
All the timer decisions TCPI has made since their first tournaments (and in-game timers) have been based on the need to finish a long Swiss tournament in a single day. Therefore, they didn't choose the best option, just the possible one.

We are not running a one-day Swiss tournament, so why would we choose the suboptimal but necessary option instead of the best one?
 
The battle timer is a ruleset and does not affect in-game turns. It's not a strict mod to the game. For example, you wouldn't know whether someone timed out by their Your Time timer running out first or by Smogon's ruleset by watching an in-game SM replay. Just because we can't enforce our timer rules on modern cartridges (because GF already hard-coded their own rules into the game) does not make it illegitimate for us to enforce them on PS!. Smogon should strive for cart accuracy when it comes to game mechanics; the turn-by-turn interactions. The point of making a competitive format is that you can set the rules and Smogon has always meant for its competitive format to be played first and foremost on simulators.
 
One thing learned from timer discussions I've seen in another community (smash) is that lower timers actually *increase* the amount of games that end up going to time, which overall means the games will take longer on average not shorter. This is because players are incentivized to play for time instead of for a decisive win.

A simple example can be a player having a defensive core that cannot survive long term due to pp problems but is currently up 3-2 so they take as much time as possible and extend the game such that they are able to survive with an "advantage" long enough until the timer runs out, even though they were in a losing position overall and would have otherwise lost or forfeited.
 
Given there's pretty close to zero support for this - on top of the fact that we'd ultimately need to use up a metric ton of volunteer time to remake every move animation to align with how long the animations are on cartridge, if the goal of this is cart accuracy - drastic changes to how the timer works will not be considered at this point in time.
 
Status
Not open for further replies.
Back
Top