Other Using move time for prediction

[coffeeaddict]

I kinda want to put this out there to highlight a more obscure aspect of prediction. I've seen no previous guides ever mention of this, but I believe it can hold some degree of value to lower and intermediate level players. There has been many discussions over the creation of the perfect team or the secrets to teambuilding, synergizing cores etc. but what people fail to realise is that teambuilding only accounts for 50% of a team's overall success, another major contributor to a player's skill is their ability to predict and take calculated risks that can shift a game in their favour. Although the level of prediction required for a particular team varies between their style of play (stall teams generally utilize less prediction), it is undoubtedly a heavy asset if used correctly.

I want to introduce a facet of prediction that is rarely used but is present in many games, that is, move-time prediction. To put this simply, it merely involves taking into consideration the length of time an opponent spends on his move, then basing your decision upon that consideration.

For example, let's take a theoretical scenario where Player A's Mega-Manectric switches into Player B's Assault-Vest Azumarill after a KO where both pokes are relatively low (50%). The most logical move for Manectric is a Thunderbolt/Volt switch, as Azumarill's Aqua Jet will fail to OHKO manectric, while T-bolt/Volt Switch is a guaranteed OHKO. However, Player A notices that the opposing team has a Garchomp and decides to take a high risk - high reward prediction that Player B will switch in Garchomp to tank the fairly obvious Thunderbolt/Volt switch and opts instead for HP Ice.

Normally a player that selects Thunderbolt/Volt-switch will make their move in less than 5 seconds as they have already predetermined what they wanted to do, the moment they switched in Manectric. However, Player B notices that it has been 15 seconds and Player A has still not made their move, and decides to cancel his Garchomp switch-in and opt to keep Azumarill in and waterfall Manectric for an easy KO.

In lower/mid-level play (< 1400), this behaviour is fairly commonplace and relatively easy to predict and counter, players usually give-away their intentions unknowingly by taking a slightly longer than average duration to make their move. However, if it has been 30 seconds and the player has still not made their move, it is better to go with your initial decision as it is very likely they are going for a double prediction and using that wait period to confuse you (or they went for a toilet break).

All of this is of course, highly speculative and could potentially fall under theorycrafting, but through my many games, I've noticed that move-time tends to play a major factor in predictive decisions, especially among the lower to mid bracket of players. Of course, instinct is also important, it is essential that you make these assumptions after you have determined the average wait time of your opponent (and their risk propensity), as well as the context of the situation (i.e in late-game scenarios, many opponents are far less likely to take risks than if it were in the early to mid game).

 

[Tressed]

In a perfect world, this makes complete sense. In real life, lag happens.

There's also external stuff to consider. Your opponent is taking a long time to pick a move. Maybe they're debating staying in or not. Maybe they remembered they need to get the laundry going. Maybe they got hungry and are in the kitchen debating if they want an orange or cookies. All you see is them not making a decision, which isn't enough to base your moves on given the number of real life variables that could have an effect.

 

[heritage]

Really when they take more than a minute to make their decision it just makes me want to stab them through my computer screen cause I know they're stalling

 

[Vinc2612]

Lag doesn't matter in this scenario. In this example, coffeeadict would immediatly switch to Garchomp, and if the opponents wait more than 10 seconds, he would cancel and stay instead.
If the opponent picked a move but it looks like he didn't because of lag, the cancel won't work, so it would be the same result as if there was no lag.
If the opponent didn't pick a move yet, cancel will work and the second case can happen.

Otherwise I agree with BlackLight, there are too many possibilities to justify the opponent being slow.

What you don't consider though is to use YOUR time to make your opponent hesitate. In a situation where you know that your opponent could take a risk that would give him a great advantage, you can timestall. I've noticed most people tend to use the safest option when you give them the time needed to be afraid of such a risk.

 

[BuddhaSmite]

This is hard to quantify, because you never know if your opponent is really thinking hard or just browsing the internet between moves, or doing a damage calc.

I admit that I sometimes try to play mind games like this, though. If I'm going to make an obvious play, I'll wait a few extra seconds to mess with my opponent.

Bottom line, though. Good players will decide their best move and stick to it. Sort of like poker: you play your hand, not your opponents.

 

[SoulRed12]

Skilled players will do risk/reward calculations in seconds and make their moves relatively quickly regardless of whether they ultimately select the "obvious" or "non-obvious" play. They will consider "kinds" of plays in the same way, so neither will take them longer to decide upon.

Even going with your logic, taking long wouldn't signify any particular move, it would just indicate the player is considering alternatives to the obvious play. They could easily decide to go with the obvious play anyway. So while playing quickly might indicate making the obvious play, playing slowly just means the obvious play is no longer more likely. In other words, you might be able to conclude something from a quick play, but you can't really conclude anything about a slow play.

 

[Stathakis]

this is a gut-instinct sort of thing and pretty unreliable so I'd generally rather base my plays on the risk vs reward of each play and going for predictions based on more reliable things to consider, like how my opponent has been playing up until that point, what the risk/reward are on their options, how I've conditioned them to think about certain options I've made, etc.

though it's always nice when I go to answer my door, come back, and my opponent overpredicts giving me an easy win so maybe I should encourage this type of discussion.

 

[Lickalatapus]

It is a variable, yes, but honestly this talk whittled down to a 50/50 based on waiting time? Not worthy of discussing IMO

Plus this is something you can only experience on PS

 

[Entei The Swift]

I agree wih the OP. I find it very easy to predict players that take longer to make their move and will intentionally take longer to fool them into over predicting. I personally can't stand the "change your mind" button as abusing it numbs your instinctual senses and makes your wifi experience sub par. All of this is just opinion however.

 

[Wheezer]

I also agree with the OP. Although once you get up to mid ladder (1650ish) players tend to have a faster thought process which can cause problems. Although this is more of a gut instinct thing like Stathakis said, in about 7 cases out of 10 it works. My strategy with this is to not predict anything right away, wait 3 seconds, then predict them to predict me. If they take more than about 5 seconds I will make the best play using Risk vs Reward. However, this strategy is not set in stone, and often I will adapt it or completely scrap it depending on the play style of my opponent.

 

[Sevifar]

In a traditional tournament setting (Both players physically present) when you know there's no distractions or anything, then I think this has some merit, but only when the person takes an uncharacteristic amount of time to make a move. And even then, it's easy for a player to dummy you with a bit of time stalling. It can actually work against you just as it works for you.

Players of the calibre to be at the top level (Of any game, not just pkmn) are usually above such tricks anyway :D. The basic idea makes sense but in practice not really...

 

[KoalaNoob]

This works sometimes, but there are too many variables to consider. Maybe the opponent is doing multiple battles. Maybe your opponent is busy doing something else. Heck, he or she might even be taking a poo. Overall, not really that worth it IMO.

 

[Dee Train]

I will occasionally think like this, but i find that there are many IRL factors that realy could just delay the process...and Lag as well :)

 

[ginganinja]

However, Player B notices that it has been 15 seconds and Player A has still not made their move, and decides to cancel his Garchomp switch-in and opt to keep Azumarill in and waterfall Manectric for an easy KO.

And then PS logs your previous move anyway because the cancel button is fucking terrible.

 

[Red King]

And then PS logs your previous move anyway because the cancel button is fucking terrible.

This. Seriously though, so many things can go wrong it's not even worthwhile considering this tactic. It's more likely to screw u regardless than simply relying on your brain. Ya know, the thing that lets you predict, w/o a gimic? Yeah, that one.

 

[Sunodasuto]

This is one of the reasons why I dislike the Cancel button, along with the fact you can't cancel your command on the cartridge.

 

[Vinc2612]

The button works fine. You just tried to cancel after both players picked their move when lag made you think your opponent didn't

 

[KoalaNoob]

The button works fine. You just tried to cancel after both players picked their move when lag made you think your opponent didn't

Sometimes I click cancel, the timer says my opponent has 120 seconds left, and then five seconds later, it goes without me being able to select another move. :/

 

[River Martini]

So the moral of the story is to think carefully before choosing your move? Then you do not have this trouble.

 

[Wheezer]

This strategy isn't meant to work all the time, every game, with everyone. Once you get into the flow of the battle and understand your opponent a bit more you can get a feeling of what will work and what won't. This strategy works most of the time if you are in the right situation, but it's not something that you should be relying on 100% of the time. There is a lot going into this strategy, but if you use it at the right times, it should be successful.

 

[escarlata]

I might be going a bit into psychology philosophies here, but when making predictions, the odds of making a correct guess is either 0%, 50% or 100%. While using move time sometimes does help guessing moves correctly against the less strategic mind, but strictly speaking, "prediction" is just a glorified "guess". So long as you don't know your opponent well enough, it is hard to understand their thinking speed and habits to make a good judgement. Even if you do, there is the likelihood that they may change their thinking habit to match yours. Assuming a similar intellectual level, the odds of making an accurate prediction never deviates from 50%. Having said that, any method of "using <factor> to ease prediction>" is what I consider an "illusional" increase in odds of making accurate choices. Anyone is capable of outpredicting the other (since 50% is fair), but what really differentiates a good player and an average player is that the average player plays on the basis "what will he do if I make a correct prediction", while the good player thinks "what should I do if I make a wrong prediction".

I apologise if this derails the topic a bit, but I felt I should voice this out, and this thread is possibly the most appropriate place to say it.

 

[Karxrida]

Like BlackLight and others have said, there are way to many variables that could contribute to your opponent taking their time to chose a move, one of which can be they are trying to psyche you out when they already know what their move choice is. I wouldn't recommend relying on this even if you're facing your opponent in the same room IRL specifically because of the possibility that they're doing it on purpose.

 

[fail_at_battling]

People keep talking about how long it takes someone to make a move. I agree that this is unreliable due to factors such as lag and mind games, but I feel like there is some merit predicting on how short it takes someone to make a move. This is especially true for pokemon that could run multiple sets; if someone almost immediately makes a switch to something your pokemon could easily counter under certain circumstances, they probably have a trick up their sleeve to eliminate a threat. So even if you have no idea what set the opponent is running, it's probably a good time to switch out.

 

[Celticpride]

First, as others have said lag is an issue. Second, when I play online sometimes I am either checking my email, looking up stats of less common things, running damage calcs, fixing a mistake in the teambuilder (sometimes I forget EVs or run the wrong move on something), or sometimes I have to step away for a few seconds. There's a lot of reasons for not responding within seconds.

 

[Kabuchops]

I completely agree with coffeeaddict as its pretty easy to tell the difference when someones lagging/stalling or actually contemplating if they should switch in/out or attack. As he said the opponent generally takes 5 seconds to make a definite decision (especially low ladder) and more than that if they are considering in a risk factor. However there is a thing such as overprediction so you might want to take that into account.