Thoughts on a solution to "infinite (time wise)" battles

[Scofield]

This has never been an issue for me before, but recently it's come to my attention that it is entirely possible to have a battle go on for an "infinite" amount of time...at least until the server inevitably crashes. This can occur through a few scenarios. It happens when the remaining pokemon on the field on either side are incapable of damaging the pokemon on the opposite site of the field.

For example, if one side has

Skarmory@leftovers
BraveBird
Whirlwind
Spikes
Roost

and

Gliscor@toxic orb
Toxic
Earthquake
Facade
Taunt

and the other side has

Gliscor@toxic orb
Fling
Earthquake
Acrobat
Taunt

and

Bronzong@leftovers
Gyro Ball
Payback
Earthquake
Stealth Rock

In this scenario even with full hazards on both sides of the field, it is impossible to wear down either side via only switching, since both mons on either side are both immune to spikes, toxic spikes, and resist stealth rocks (well not skarmory) with lefties to negate the damage. Eventually both sides will stop trying to waste pp, knowing they wouldn't harm the other side and just start switching to waste the other sides pp. Unfortunately once the other side catches on to this....they can just start switching also so as to avoid wasting pp...creating a standstill.

This scenario might not be perfect, but it's just to illustrate that it is very possible, I've had two battles of 200+ turns today alone with nothing but switches. For the sake of argument, if I miss something just justify it, ie side 1 probably doesn't have SR up on its side so Skarmory doesn't get worn down by switching.

Possible solutions:

I think there needs to be one, as funny as it might seem this scenario is very feasible..and common.

1. After a certain number of turns (200?) the game automatically stops and calls a tie. Unfortunately I've heard people have had legit 200 + turn battles. This number can be upped to 400+ or even higher, unfortunately that still means you have to spend forever just sitting there and switching..


2. After a certain amount of turns, a button becomes available to create a tie, both sides have to press the button for the tie to occur. I don't know how viable this is to implement.

3. Flip a coin. Contact a moderator to flip a coin, winner gets the win, loser leaves or gets kicked by the mods. It would be nice if this were able to be automated. Unfortunately, it's not a very fair solution.

That's all I can think of, any other suggestions?

 

[undisputed]

solution 1 makes the most sense. its probably the best i can think of

solution 2 won't make people want to stop. theyll just play to see who can hold out longer

solution 3 seems like it would create problems with potential bias issues, etc.

 

[shrang]

I really don't think this should be that much of a problem. People's patience have limits. If it gets down to something like that, then the winner should be decided by whoever is more patient and can stand the monotonous struggle for longer. There will be a time when one of them will yield due to reasons of 1) "I cbf anymore" or 2) they have other stuff they need to do and Pokemon can no longer occupy any more time.

 

[Oglemi]

Well, if it were possible to implement a "tie" button where both players had to push it, would it also be possible to add a "coin flip" button that only the moderators would have access to, but would show up on both players' screens?

I'm thinking the coin flip app a la Poketch from DPP? Would it be possible to add that feature into the server, and if an agreement is reached that it should come down to a mod coin flip, a mod would then ask the players who was heads or tails, then the mod could press the button and the app would appear on both players' screens, whoever called heads would get the win.

Might be too complicated, as I have zero to no experience with programming, but it's something to think about.

If that's not possible, I would like to see the "tie" button feature implemented though, or again, if the mods were called, they could simply call the match a tie.

The amount of turns suggestion seems tedious, and you know then some people will simply stall out the turns as they would the timer, simply to piss people off, even if it were possible for somebody to win.

 

[cim]

Can someone provide logs of this actually happening in a non-staged battle?

 

[Umbreon Dan]

chess has a rule that (loosely) allows either player to claim a draw if both players repeat their moves three times. more info

i think the fairest and best solution would be to implement a rule like this. after three repetitions, it becomes clear that the game is going nowhere, and it should end with a tie.

solution #1 is awful because it prevents players from have legitimate lengthy battles.

 

[Atticus]

I really don't think this should be that much of a problem. People's patience have limits. If it gets down to something like that, then the winner should be decided by whoever is more patient and can stand the monotonous struggle for longer. There will be a time when one of them will yield due to reasons of 1) "I cbf anymore" or 2) they have other stuff they need to do and Pokemon can no longer occupy any more time.

so you're saying in a potential tournament situation that the winner should be decided by who is more "patient"? anyway i think this type of thing is way too situational to warrant a large concern and should just be handled on a case by case basis. if it's a tournament just treat it as a tie and redo the match, if it's on the ladder than i guess theres nothing really you can do other than getting a mod to come in and flip a coin, at least until theres an option on PO that gives players the option to tie.

 

[Firestorm]

Solution #1 + #2 works best. That way they know it's going to happen so they'll just both press the button. Otherwise, what undisputed said may happen where one player just tries to force the other player into quitting.

#1 + #2 > #1 > #2 > #3

That of course requires someone to actually program it. Something Super can hopefully keep in mind for PokeLab? I don't see it happening in PO.

 

[shrang]

chess has a rule that (loosely) allows either player to claim a draw if both players repeat their moves three times. more info

i think the fairest and best solution would be to implement a rule like this. after three repetitions, it becomes clear that the game is going nowhere, and it should end with a tie.

solution #1 is awful because it prevents players from have legitimate lengthy battles.

IIRC there was another one where if you get up to 50 moves (Both added together, so 25 turns) without having anything significant happen, it can also be considered a draw (Obviously you can have something less for a Pokemon game).

@ Atticus: I was talking about your typical ladder match, but for tournaments, it will be better to actually implement something like a tie or coin flip.

 

[Hipmonlee]

A lot of legitimate PP wars can go for an extraordinary number of turns. PP wasting Rapid Spin by itself takes 64 turns..

I dont see a need for requiring people to play 200 turns before allowing them to offer a draw. Even if people are accepting draws just to avoid tedious but potentially decisive pp wars who cares?

It also means if you have to leave mid battle on the ladder, you can get a fairer result than being forced to forfeit.

(I believe this option already exists on NB).

Have a nice day.

 

[cim]

i think the fairest and best solution would be to implement a rule like this. after three repetitions, it becomes clear that the game is going nowhere, and it should end with a tie.

Jirachi used Calm Mind!
Skarmory used Spikes!

Jirachi used Calm Mind!
Skarmory used Spikes!

Jirachi used Calm Mind!
Skarmory used Spikes!

Under that proposal, the battle would end in an automatic draw. This forces either Jirachi to get less Calm Minds or Skarmory to get less Spikes, all to prevent a rather rare battle condition from making a battle take longer.

 

[capefeather]

What if we implemented #2 except a mod could force a tie too? So if someone's being a jerk then a mod could be called in... I'm not sure I'd want to place that responsibility onto the mods, but it seems rare enough that it could be OK.

 

[Umbreon Dan]

Under that proposal, the battle would end in an automatic draw. This forces either Jirachi to get less Calm Minds or Skarmory to get less Spikes, all to prevent a rather rare battle condition from making a battle take longer.

good point, but if we took the chess approach more literally, it would be "repeat the exact position three times". to draw, you would need the same boost levels, the same field effects, the same hp and pp (and whatever else).

IIRC there was another one where if you get up to 50 moves (Both added together, so 25 turns) without having anything significant happen, it can also be considered a draw (Obviously you can have something less for a Pokemon game).

this is true, and i think it's also a good idea for pokemon. something simple like "neither player attacking for five turns draws" could solve the scenario posted in the op, but i'd like a single, unified draw rule that catches all infinite stall scenarios, so :/

edit: and same sleep timer, funnyguy!

 

[cim]

good point, but if we took the chess approach more literally, it would be "repeat the exact position three times". to draw, you would need the same boost levels, the same field effects, the same hp and pp (and whatever else).

One Pokemon is put to Sleep for four turns. Opponent predicts that they attempt to wait out the timer and sends in own Slept pokemon to wake back up.

Both Pokemon sit "Fast Asleep" for 3 turns. Draw?

 

[Kevin Garrett]

I am completely against solution #1 unless it was a ridiculously large number because I have had countless games go over 250 turns with PP stall that were guaranteed wins for the person who conserved the most during the battle.

I wouldn't mind solution #2 because that gives people an out in the situations where you can switch forever.

 

[Eo Ut Mortus]

Can someone provide logs of this actually happening in a non-staged battle?

Don't have the log because the computer on which it occurred is dead, but a good number of people saw ToF and I have a 400+ turn (1+ hour) UU battle (in fourth gen) which could've went longer had I been patient enough to not try and end it with a risky move.

Threefold repetition only works in chess because of the near-infinite amount of scenarios that are possible to create on a chess board, and the fact that some of them are no longer able to be recreated after certain moves.

I think the fairest way to solve this would be to let moderators end battles as ties. The problem with solution 2 is that there are likely going to be stubborn people who refuse to accept the draw even if it's obvious a battle is going nowhere, and the problem with solution 1 is that there are some legitimate 200+ turn battles, as stated, although this is probably the best automated solution in my opinion.

 

[DJXO9]

Maybe something like "If no damage is incurred for over 10 turns, the battle ends in a draw." That would cover most situations I think, and would be nigh-impossible to take advantage of between weather, entry hazards, and every attack doing at least some damage, unless the other Pokemon has an immunity obviously.

 

[locopoke]

This is extremely situational and barely ever happens. I don't think anything "needs to be done", if it happens then players should just play it out until one person gets impatient and leaves or has to go.

 

[Firestorm]

Has there ever been a battle over 600 turns?

All there needs to be is a number at which the players know the game will end in a tie. Let's say even 1000. That will give them the incentive to hit that tie button instead and avoid undisputed's scenario.

 

[eric the espeon]

One Pokemon is put to Sleep for four turns. Opponent predicts that they attempt to wait out the timer and sends in own Slept pokemon to wake back up.

Both Pokemon sit "Fast Asleep" for 3 turns. Draw?

Change in sleep counter. Hidden values, but that's irrelevant since we know they are changing.

Still the "Chesslike" solution needs some refining to work. The trick will be getting a version which blocks all unending games, but does nothing to "normal" play... hard when you consider that same situation can occur repeatedly in Pokemon via double switches when players try to gain a prediction advantage, something which does not exist in Chess since all the pieces are on the board as it were. A tie button would be cool and certainly help, but requires both players to agree (one could be determined not to), or required mod intervention (yay arguments about whether a situation is really unending or not), so is an imperfect solution.

Perhaps we could make a clause which goes along the lines of "If both players are able to make a series of moves which will guarantee that they will not lose but neither can force a win the game ends in a tie.", and have it mod enforceable for ladder matches when it comes up for now (asked a couple of the PO mods, they think it should be scriptable currently, though if not a new event could be requested). Does lead to more mod work, but at least with this form of clause it cannot interfere with any normal (defined as non-infinite loop) play.

Edit: To clarify, I think both #1 and #3 are poor options. #1 has an impact on normal play, which is not required to fix this unusual situation. #3 is simply luck based, which is evidently less than ideal given that ties exist in PO and this situation is the very essence of stalemate.

 

[Dubulous]

Look, I'm not much of a Pokemon player, but I do pay rather close attention to this section of the site. Keeping that in mind, I'm finding it rather difficult to understand why this topic was made. It seems counterproductive.

First of all, the OP states that this scenario is common. Again, I do not play very much Pokemon, but I've played my fair share of battles. This scenario has never once happened to me, which leads me to believe that this situation isn't all that common at all. There is further evidence of this, as well. Earthworm is on the record saying that this situation has never happened to him. Philip7086 has said that this has happened to him exactly once. In fact, rereading the OP, Scofield admits that this has never been a problem for him. How then is this a common scenario?

Second of all, the timing of this thread is rather odd to me. Right below this thread, in fact, is a thread about removing an arbitrary win condition in tournament scenarios. This thread proposes exactly the opposite with the option #3. This option in particular is quite illogical.

I'm not a fan of option #1, either. While imposing a turn limit does not create a new win condition, it does create a draw condition that is does not exist on the cartridge games. Does that really matter...I'm not entirely sure. In general, though, the policy has been to stick to the cartridge as closely as possible. This is why #2 is the most appealing of the options presented; in a link battle, both players can hit run at the same time during a match and the match will end in a draw.

Do we really need something like this, though? For ladder matches, I don't think this is much of a problem. Tournament matches are a different matter, though. It seems fair to me, that, if this situation were to occur in a tournament, that the Tournament Directors would be able to declare that the battle was a tie and that a rematch should occur. This is not very much different from the way Tournament Directors make rulings should a dispute arise. Furthermore, this removes the massive programming overhaul of installing a tie button into the interface or implementing a "chesslike draw condition," which is by itself a massive pain to even conceptualize.

 

[Alaka]

I don't think the issue of staying true to the cartridges has any bearing here, it could just be a rule that both players have to hit run after x turns or whatever. Our win conditions don't need to have anything to do withe the carts, just mechanics. For instance the VGC institutes a completely different win condition for matches that are taking too long, and it's obviosly played on carts.

 

[Scofield]

In fact, rereading the OP, Scofield admits that this has never been a problem for him.

Maybe you should reread the op more carefully, I have said that it has been a problem for me. Happened to me twice in one night and at least once in the last month.

 

[eric the espeon]

While we do not *have to* strictly adhere to the games beyond following mechanics, it would be a mistake to start imposing arbitrary or unnecessarily complex rules. Bans and limits should be clear cut and have direct justification, they should avoid affecting the metagame in ways which are not required to improve it. Why would we limit it to 200 turns, not 199 or 150? It's the similar to argument used to dismiss "lets bring Lv. 86 Mewtwo into OU", both draw an arbitrary line and there is no need for that. Whether we like it or not we are at the very least metagame designers, if not true game designers. Modifying certain parts of the game we play should not be taken lightly and should be done with great care, even if we stay true to mechanics with all options. The win (or "ending") condition is perhaps the most important, in my opinion. Any new ending condition should be either absolutely necessary (as an end is here, leaving it as "whoever can be bothered to wait longest" is not acceptable for a competitive game. The same goes for Castform in Acid Rain.) or entirely impossible to force (as Evasion, OHKO, Item, Self-KO, and my proposed Sleep Clause along with Pokemon/Item/Ability/Move bans are) in order to leave the true win condition (KO all opposing Pokemon, tie if all yours get KOed at the same moment) intact and retain the "spirit" as well as the "letter" of Pokemon as a game.

 

[Dubulous]

This has never been an issue for me before, but recently it's come to my attention that it is entirely possible to have a battle go on for an "infinite" amount of time...at least until the server inevitably crashes. This can occur through a few scenarios. It happens when the remaining pokemon on the field on either side are incapable of damaging the pokemon on the opposite site of the field.

...

Possible solutions:

I think there needs to be one, as funny as it might seem this scenario is very feasible..and common.

...

Emphasis mine.