Skill, or glorified guesswork?

[SevveN]

Lately, I've been seeing a few sigs with something about "92% of battlers think prediction is a skill. If your one of the 8% that think it's glorified guesswork, put this in your sig." So I just wanted more people's opinions.


Honestly, I believe it just depends what way you look at it. If you think the opponent will switch, and foresee that and play accordingly, it can quite literally be win or lose.


There's really only two cases in this argument. The first argument is that it's a skill, as being one step ahead of your opponent will always mean in an advantage. However, if your opponent out-predicts you, you could find yourself in a world of pain. The other case, is that your just guessing, based on your instincts, one would say, what the opponent will do, and hoping your right.


But see, I believe theres more to this. Prediction doesn't come down to flaws in teams, just trying to be a step ahead of your opponent. I believe that this is more mind games, then anything. Some people, more inexperienced quite literally, after a mental beat down, get really shaked up and start throwing out random things. Sometimes, it does actually work, even though they don't know what their doing. This, I believe, is glorified guesswork.


Anyway, what do you guys think?




-SevveN

 

[locoghoul]

ugh, turn off the bold. It also depends on the opponent, some risk too much and overdo the double switches for example. Some will switch all the time and some will eq with snorlax against the lucario (like me)

 

[familyguyman]

Well, it's technically guesswork, ie: guessing what your opponent will do without a guarantee but the percentage of correct guesses is a skill.

Obviously if both players are trying to maximize their percentage of correct guesses, you'll get into mind games and such but at the end of the day one player will have more correct guesses than the other and, over time, that is how skilled a player is.

So basically, any prediction is guesswork but consistent, accurate prediction is skill.

 

[Huntofthelion]

99% of people with percentages in their signature made them up!


On that note, prediction is guessing. You can be good at guessing, or make educated guesses based on knowledge, but the fact remains you're still taking a guess. I guess the skill aspect of it is being aware of all the factors, and knowing how to balance risk with reward. You need to be able to see when making a guess is more trouble than it's worth, or when it's needed to get you out of a hole. I think that would fall more under planning ahead though, but it should be a factor when you're making your 'predictions'.

 

[Ninjadead]

Prediction isnt a skill, if you mean just predicting if someone is going to switch or use a certain attack, then that's just assumption and guessing.

 

[mtr]

I feel obligated to answer this, since I'm the one who started this "meme", if you can even call it that. 92%/8% is the classic percentage thing, and I think the place where it started is when people claimed that 92% of people had started listening to rap/pop music, and only 8% of people liked to "headbang".

Back to this thing. Other users apparently agree with me, so I'll expand on the point. Let me bring up a highly ubiquitous real-life example.

It is a standard Ubers ladder match. The teams are aggro vs. stall. The aggro player is using Kyogre, Garchomp, and Scizor, who synergize with each other very well in terms of typing, making them a favored combination on many aggro teams. The stall player is using Forretress, Blissey, and Giratina-A (he's playing classical stall), the backbone of a defensive line.

Both teams have Stealth Rock on the field. The Pokemon on the field are SpecsOgre and a StandardBliss. SpecsOgre is locked into Surf (which Blissey switched in to take), while Blissey is outside KO range of the next attack. Your opponent is likely to send in Scizor, and if Scizor uses U-turn, you suffer residual damage on the switch to Forry/Gira-A, not to mention that your opponent once again has the advantage, forcing another switch and collecting more damage from Stealth Rock. If Blissey stays in, it just took 62%, which kills it in this example.

In Ubers Stall, you need to work fast and contest the aggro player in terms of momentum, or you will lose. Any chance you have to use Forry to get off a layer or a spin, you do it. So since you're predicting your opponent will send in Scizor, you switch in Forretress, because then you just managed to use Rapid Spin or Toxic Spikes for free.

And you do that, but to your dismay, your opponent keeps Kyogre in and uses Surf again, which turns Forry into a sodden pile of sediment. Now think about what would have happened had your guess succeeded: you would have been hailed as an excellent predictor and you would have stopped a good amount of the aggro player's momentum. But since you failed, you may have lost the game (although its entirely possible to win without Forretress against an all-out offense team, just harder).

Where is the logic in the aggro player attacking Blissey with Surf if it is outside KO range, knowing that it has Softboiled? Where is the logic in the stall player switching in Forretress while a special threat is on the field? The players could be 2 random newbies, or EUM and Jib, but the minigame is the same. To say that a prediction like this has skill involved is like saying that the prisoner's dilemma is a game of skill rather than guessing. Draw a diagram out and see for yourself.

Some might say that you can tell if a player is a conservative or liberal duelist, and predict accordingly, but such is rarely the case. Player's are so often governed by the gravity of the situation they're in, and they all just seek to make the best moves they can, be they conservative or liberal. If you guess very liberally, you will suffer the results of false positives, like having a ScarfGon use Earthquake on a predicted Heatran (you're revenging a Mixmence) while a Gyarados arrives instead. If you don't predict at all, you will see things like the same ScarfGon using Outrage as the same Heatran switches in and destroys you with Dragon Pulse. Therefore, every decent player, knowingly or unknowingly, acts both conservatively or liberally in the space of a given match.

 

[Critical_Hax]

Its both. Prediction is nothing but logical guessing. Nevertheless, the best players have what would seem to be an innate skill, which allows them to always predict correctly. It does take some work to think through what your opponent is thinking, so I guess it can be called a skill

 

[Wyzra]

I've thought about it a lot in my previous thread and come to the conclusion that making the right prediction is highly situational. But there are a few basic principles. First, check the risk vs. reward of the move you're going to do. Compute the expected reward of each move. By doing this move, what's the worst thing your opponent can do in response? What do you gain if your opponent does what you predict? Don't try a flashy prediction if you lose from it. You shouldn't always act in response to what you predict your opponent will do, because what if he doesn't?

Put yourself in your opponent's shoes: if your opponent has something that entirely walls your active Pokemon (which in turn threatens his active Pokemon), then under typical circumstances he's very likely to switch that thing. It's a low risk move for him. His Chansey will always come in against your very threatening Raikou (pre-ban). If you have a high reward play against it, then you should do it. Also, reason what information your opponent knows and doesn't know, based on how he's playing. Try to understand his every move as part of his strategy against your team. There is a large psychological factor here.

Finally, know some statistics. First, understand the metagame, e.g., how likely it is that Primeape is scarfed in NU (very). Also, how many Sucker Punches will this guy take against my Stealth Rocking before he finally decides to use another attack (usually two).

EDIT: On the other hand, of course, there are situations that are just pure guessing. But usually, if you look at the expected reward, you can determine the best option.

 

[Wyzra]

Where is the logic in the aggro player attacking Blissey with Surf if it is outside KO range, knowing that it has Softboiled? Where is the logic in the stall player switching in Forretress while a special threat is on the field? The players could be 2 random newbies, or EUM and Jib, but the minigame is the same. To say that a prediction like this has skill involved is like saying that the prisoner's dilemma is a game of skill rather than guessing. Draw a diagram out and see for yourself.

Here I personally believe the stall player exposed himself to a very unnecessary risk, but I won't go into the details. But as for the prisoner's dilemma, there is an equilibrium. You should defect, because regardless of your accomplice's decision, defecting is the superior option. I think the situation comes up like this a lot in Pokemon as well.

 

[bearsfan092]

Well, when battling, you must consider two things.
1. What are the potential logical moves my opponent will make? What are the odds of each move occuring?
2. What is the potential payoff or loss of choosing a certain course of action?

Actually, I just had a battle that demonstrates this thought process. It goes like this:

Opponent switches in Magnezone
Player's Milotic uses Hidden Power Electric for negligible damage.

From the information I gave you, it's plain to see that Milotic is in trouble. In a simple 1 vs. 1 situation, the most apparent move from Magnezone is Thunderbolt, which is quite capable of denting Milotic from Magnezone's impressive Special Attack stat. However, I neglect to mention this particular occurence earlier in the same battle.

Opponent's Gyarados used Dragon Dance!
Player switched in Milotic!

This sequence of events precedes the first sequence of events listed in this post. As anyone can tell you (especially those who played the early DP metagame), Gyarados has fierce sweeping potential to the point where many players try to make Gyarados sweeps the central focus of their teams.

With this information now in hand, we have altered our decision making process. The opponent can do one of two things (barring overprediction or a double switch, which cannot be easily accounted for):

a.) Use thunderbolt on Milotic
Response: Switch out Milotic to a ground type or electric resist. I carry a Flygon on this team, so this move seems like a viable one.

b.) Use Explosion on Milotic
Response: Switch out Milotic to a ghost type to avoid the boom. If you are like me and are a wifi battler, then you know you don't have access to the Rotom formes. In this particular situation, I have a Gengar waiting in the wings.

So, it's clear we have to get Milotic out of there, but the question lies whether to go to Flygon or Gengar. Scarf Flygon will die to Explosion, removing a key source of revenge killing. Gengar will be dented/die by Thunderbolt, losing some sweeping potential/revenge killing.

If you consider only the Pokemon currently on the field, you would guess that Thunderbolt is coming up. This is because your opponent would seek to preserve his Magnezone to fight another day. However, we are considering the Pokemon in reserve, namely, his Gyarados. We can guess reasonably that the opponent wants to remove your bulky water that has HP Electric that way Gyarados may set up a Dragon Dance or two and proceed to sweep. He will want Milotic dead right now, allowing the door to be blown open for a sweep. The odds have now shifted to him using Explosion, prompting us to switch to Gengar.

Player switched in Gengar!
Opponent's Magnezone used Explosion!
It doesn't affect Gengar!
Opponent's Magnezone fainted!

And now by playing oddsmakers, the opponent is now down one pokemon (and probably a revenge killer if it's scarfed) and our bulky water is free to stop a Gyarados sweep.
---------
Now getting back to the question whether it's skill or guesswork. I believe that this is a skill. Having a basic and rudimentary grasp of game theory is a skill in my book. Guessing correctly on a consistent basis is a skill. These same skills apply everywhere. For instance, as a safety I read the gapping of offensive linemen in football to have an idea whether the play will be run or pass and whether it will be inside or outside. It will sometimes be wrong, but more often than not, I will be put in position to make the play, which most people would call a skill.
Tl;dr: I guess what I'm saying is that good guesswork is in itself a skill.

 

[mtr]

Its both. Prediction is nothing but logical guessing. Nevertheless, the best players have what would seem to be an innate skill, which allows them to always predict correctly. It does take some work to think through what your opponent is thinking, so I guess it can be called a skill

The best player's can't all have an innate "skill" at prediction. Stathakis claimed that he doesn't like to predict, even on offense. And many top players use heavy stall, which innately minimizes prediction in favor of planning.

Instead, I would argue from playing [and mostly losing to :(]top players is that many seem to have a bond with their team. Lol, that sounded retarded, let me be more specific: their teams are incredibly fine-tuned, and they know exactly how they work. They will often use specialized sets like Mixdance Rayquaza or Specially Bulky Mewtwo or defensive Manaphy (in Suspect) to fulfill a given purpose, while many lesser players don't have such a fine knowledge of how their teams work and how they should be optimized. They don't use gimmicks: one thing that characterizes a great player is that their sets almost always work.

And they don't think in terms of beating Pokemon, they think in terms of beating strategies. For example, what's the point of having Hippowdon to check Lucario if an offensive player can just sac a Mixmence to get it down to <43%, killing range for Close Combat? Why have a Blissey if Scarftran can explode on it, opening the door for Zapdos? But what if the defender uses an RT Gyara to provide a redundant check on such strategies? And thinking in more sophisticated terms, what if the defender uses Forretress and blasts Heatran/Lucario/Infernape with an unexpected Earthquake.

Pokemon is not a game played in the abstract, it is a game played by humans, and anything can happen on the field of battle.

Here I personally believe the stall player exposed himself to a very unnecessary risk, but I won't go into the details. But as for the prisoner's dilemma, there is an equilibrium. You should defect, because regardless of your accomplice's decision, defecting is the superior option. I think the situation comes up like this a lot in Pokemon as well.

Well, it is a risk, but if the risk isn't taken, the stall player is likely to lose. I mean, if residual damage inflicted on both sides is equal, the aggro player wins. The whole point of stall is to make your residual damage higher than the opponent's. And if the opponent gets in Scizor on Blissey, the defender has a few options.

1) Keep in Blissey, though if you are under 62% (as in the example), you probably die. Or a crit happens, and you die at 100%. And if you don't die, you lost 6% of your health, and the other dude might have a scary Stallbreaker Giratina-O or Physical Mix Palkia on the field for free, which isn't cool. More to the point, what do you gain from Softboiled? You actually lose 6% of your health after leftovers, and the opponent has another Blissey counter on the field.

2) Go to Forretress, at which point the Scizor U-turns, weakening your Forretress, which diminishes its usability, and forces you to take more residual damage as you have to run from a Forretress counter.

3) Go to Giratina-A/Giratina-O/Groudon, which saves Forretress, but you don't actually gain anything from this, as the aggro player just goes to a counter and you have to switch again.

The stall player might be able to afford this dilemma a couple times, but he needs to get the ball rolling on his side by using Forry, or he loses.

Against Kyogre leads, I like to switch to Blissey to absorb their attack, then move right back to Groudon. This usually catches their physical attacker-often Scizor, off guard and nets me the Stealth Rock I need. It works 99% of the time, because no right minded person would keep a Scarf/Specs Ogre against Blissey while it’s Sunny.

Almost the same situation I was talking about. What if the other guy does predict? It's happened to me before.

 

[Coronis]

Ah this.

I too have been seeing it in a fair amount of sigs, and have thought about it more than once. Prediction is first and foremost, guessing. However, there is a fair amount of skill involved. My old teacher used to say "If you don't know the answer have an educated guess." this actually is pretty accurate for this topic. Nobody actually knows what their opponent is going to do, so you base what you are going to do on what you know.

Let's say you have switched your Heatran in on your opponents Skarmory. You have been in this situation before, and the player switched in Blissey both times. Of course, since the player has done this before, you would assume he will do it again, and explode. Or, let's say you know this is a stall team, but have only seen his lead, which has fainted and his Skarmory. In this situation you would base your decision on what you know about Skarm, and stall teams in general. It is extremely likely that your opponent has a Blissey waiting.

Ah, but what if your opponent predicts you will explode to get rid of his Blissey? Then you have lost your Heatran for pitiful damage.

Prediction is guesswork sure, but educated guesswork, or as some call it: skill.

 

[Wyzra]

http://www.youtube.com/watch?v=TUee1WvtQZU

It's like this.

 

[Steven718]

Did I put the poison in your goblet? Or did I put the poison in my goblet, knowing you would switch them? Or did I know you would know I put the poison in my own goblet, believing you would switch them, and put the poison once again in your goblet?

There's really no way to know for sure. Choosing the correct goblet, is that luck, or skill? When many people claim to have lost in pokemon, they blame it on "luck" (usually, meaning hax moves) I believe a much larger portion of the outcomes are based on luck than that.

 

[diinbong]

And the moral (shown by video) is to build up an immunity to getting disbundled when you lose.



No, but in all honesty, I agree with skillful guessing. I don't think the "predicting" itself actually takes skill, just having knowledge of the metagame and of your opponent's skill/recent desicions to make choices. Weighing in all possible outcomes divided by their chance of happening are important too.

 

[MrIndigo]

Prediction is guesswork, directed and educated by knowledge of the metagame and observation of what decisions your opponents make. It is for this reason that switch-forcing and moves like U-Turn are a valuable strategy. In addition to netting extra switch-in damage, you get to see how your opponent behaves to certain threats.

Although this is not a Prisoner's Dilemma problem (People always misuse that term; I'll explain if necessary), there is still a Nash equilibria in the end whereby a mixed strategy (i.e. sometimes you switch, and sometimes you don't) is most effective, with the exact percentages of how often you should switch or not switch being calculable from the damage percentages you expect to get.

Believe it or not, there is actually a right answer to these things, it's just hard to calculate on the fly.

 

[d2m]

http://www.youtube.com/watch?v=TUee1WvtQZU

It's like this.

This is precisely it.

You can attempt to enter your opponent's frame of mind all you want: sometimes you're a genius and sometimes you cackle maniacally right before keeling over.

 

[VulturesRow]

These arguments that say that there is no skill involved in prediction is akin to people who claim it takes no skill to play poker. Yes its guesswork in the end. But your knowledge of the game, your appraisal of your opponent's tendencies, etc., all play into your prediction. Are you going to get it right every time? OF course not, no one is saying that. To go back to the poker analogy, the player is going to weigh all the factors and then make what is the best decision based on those factors. Is he always going to get it right? Nope but it is still the best play based on his analysis of the factors. Same goes for Pokemon battling.

 

[Cathy]

Prediction is obviously a skill. However, you need something to go on to make a prediction, so in certain situations, such as toward the start of a battle against a new opponent, you have no choice but to make a blind guess, since you have nothing to base your decision on. As you gain information about and the opponent's tendencies and his team, you can make predictions rather than blind guesses.

 

[Kachalski]

I had this happen to me recently:

Stormkid5 switched in Latias!
??? switched in Gyarados!

This was a while into the fight, and Gyarados had taken some SR damage. My oppenent knew by now that I was Scarfed, and had Thunderbolt as one of my moves. I knew he had Flygon. Based on what I knew, Draco Meteor would KO either of them.

??? switched in Flygon!
Latias used Draco Meteor!
It's super effective!
Flygon fainted!

Now, this is an example of prediction, but can we also call it commen sence? If I know he is likely to switch, but I can KO his Pokemon anyway, is that really prediction.

 

[Coronis]

No, not really..

 

[Aldaron]

This problem has a very simple solution, but it comes from laying out a premise of what "skill" is. In DPP, with the enormous number of threats in addition to various Pokemon being able to threaten in many different ways (see Tyranitar, Salamence, Lucario, and Latias to name a few), I feel that battle prowess is demonstrated not by 1 on 1 battles, but specifically how a battler and his team manage to handle various teams over a long period of time.

That's "skill" regarding the overall metagame. But this topic is talking about "skill" in battle, and I hope that you'll see that this idea of long term management applies once again.

A lot of you seem to be separating "skill" and glorified guesswork; I ask "why" Glorified guesswork as a phrase obviously has negative connotations, so I'll use assumptions based on information gleaned in battle, and influenced by time in battle. "Predicting" your opponent will never be a 100% art; that much, I hope is obvious. Since it cannot be 100%, I hope you can see that "prediction" as a "skill" then becomes how close you can get to 100%. In this sense, your ability to make guesses is a skill in itself. So the "simple solution" I talked about is that predictions are a result of skill, which is the skill of making educated guesses. Of course, I will also show cases where in spite of any and all efforts, a "prediction" is merely "glorified guesswork." Note that this is situational, and just because there are some situations where a prediction might not be more than an arbitrary guess, that doesn't mean prediction isn't a skill.

I'll describe a bit by what I mean by information gleaned in battle. This has various aspects, but the most obvious is first "in simple terms, what is the skill level of your opponent" If he, in the early game, is pulling all obvious moves, you can use this information to turn a 50-50 switch into a confident "he is most likely to do this" switch. If your opponent is skilled, this becomes much more complicated, and ironically, more guesswork than before. 2 skilled players will know that the other is skilled, and will know that the other is capable of making a clever switch. This, of course, introduces a back and forth mindgame of various degrees which eventually leaves knowledge of style almost meaningless. That's why, if you'll notice, so many high level matches in DPP turn out to be influenced by team matchup (and which is why many of us are pulling for Bo3 tourmaments to reduce the effects of team matchups). So, in terms of using opponent's skill level as information, this works best against players with a tendency for the "Safe" or "obvious" play; as you face more skillful players willing to take risks, this becomes more and more guesswork. Of course, to reduce the unfortunate direction back in the "glorified guesswork" direction, you use the next (second) available information: opponent's team.

General knowledge of team type and pokemon teammate trends along with whatever 3,4, or 5 pokemon are revealed can allow you to make educated guesses regarding what is left over and what he might do. This specifically is paramount in the "long term thinking" factor that you'll see many good advance players talking about. If you "predict" that a Latias is the last pokemon on a team, the skillful move would be to reduce risky situations for your Latias counters, and make moves with the knowledge that your opponent has a Latias waiting in the wings. This doesn't have to be a specific Pokemon; for example, if you are facing an offensive team and note that the opponent hasn't used a Steel like scizor / salamence / metagross yet, you can "predict" with your own Salamence in spite of not definitely knowing that he has it. Knowledge of team also helps reduce guesswork among 2 skilled players against each other; regardless of how 2 players would choose a move depending on playstyles, teams and team strategies inevitably influence a player towards certain tendencies, and a skilled player will be able to recognize this.

A third example and the final I'll bother wasting your time with lol is very closely related to the first 2 (in fact some might argue that it is just a subset one of the first 2) and quite "obvious." That is the information that YOU have given to your opponent, both regarding your own skill and regarding your team, but more your own team. Depending on what strategy your team has shown the opponent and what Pokemon have been revealed, you can further reduce the "pure guesswork" or "arbitrary" aspect of a "prediction."

Going back to my "I'll use assumptions based on information gleaned in battle, and influenced by time in battle," you'll notice I mention time in battle. This is referring to, in very rough terms, the early, mid, and late games. In the early game, with an offensive against a defensive Pokemon, it is generally advisable for the defensive Pokemon side to "use the safe play" because at this time in the battle, your ability to reduce the arbitrary nature of a guess is significantly low and as a defensive Pokemon it's general job is to be at high health. For the offensive Pokemon, it is generally advisable for a "random play," because the defensive Pokemon will rarely be able to significantly damage the offensive. Where this becomes problematic in the early game is when an offensive pokemon and an offensive pokemon are pitted against each other. With little information readily available, this once again boils down to a back and forth of mindgames of various degrees until a "prediction" is a random move, and unlike the offensive pokemon vs. defensive pokemon, not a random move of "skill." Again, this tends to boil down to how effectively a person has built his team and team matchup rather than prediction skill. The mid to late games, however, provide an opportunity for a user to use all of the information he has gathered and assumed to try his best to reduce the arbitrary nature of a guess as much as he can.

So basically, if you recognize that battle prowess in DPP is how a battler and his team do over a long period of time against various teams ("threat management" + "ability to successfully accomplish a strategy"), you can make the argument that "skill" in specific battles is using as much information available from assumptions on opponent skill, opponent team, and information given to the opponent + appropriate knowledge of the time in battle to reduce the arbitrary nature of a guess as much as possible.

So to answer this thread...of course prediction is a skill

 

[Olliert]

IMO, predicting what your opponent is going to do is not the skill involved with selecting what move to use. The skill is using the move that is best for that current situation, weighing up the possible outcomes and selecting the move with the greatest gain if it were to go right or the lowest loss if it were to go wrong. For example, if you have a CB Tyranitar out against a Latias locked in to HP Fire, you may therefore 'predict' that the Latias will switch out and therefore you may use pursuit. But if you know the opponent has a Lucario waiting for an opportunity to set up and your Lucario counter is gone then selecting pursuit is obviously the wrong choice. Here the player has weighed up the possible outcomes and I believe there is some skill involved. In this instance, the player may use Earthquake or Aqua Tail in case of Lucario switching in. Granted, this won't do much to Latias but the loss here is far less than if you were to let Lucario set up.

 

[Objection]

I think when it comes to predicting, the first thing you need to know is the skill level of your opponent, and unless your opponent is a renowned battler, you have no idea of how (un)skilled he or she is. In this way, it is best to use scouting moves, especially U-turn, as much as possible early on. This way, you get an idea of whether your opponent is a noob, hardcore or somewhere in the middle.

 

[Masamune]

Prediction comes with experience.

I find that the more skilled a player is (better knowledge of the metagame, knowing the common strategies and so on), the better at predicting they are. I guess what I'm trying to say is that prediction just comes with being a skilled player.

On Shoddy, it matches you up with the people closest to your rank. You'll notice that as you start off, if you have any skill at all, you'll basically rampage through everyone, but as you move up the ladder, your opponent's prediction 'levels' will begin to rise...and so on

Prediction isn't a skill, and this is shown by the fact that if it was, even the lowest skilled players could do it, as they could 'study' how to predict as a standalone skill, but as they can't predict...

/rant