I need help understanding prediction.

novelty

Hmmm, there is a problem I've been facing since I started properly competitvely playing this game, and that problem is prediction, well a certain type of prediction, it's kinda hard to explain, so I'll give an example.

Example

I have a scarf infenape, my opponent has a celebi and a heatran, with the right predict I can get a kill, I seem to guess in these kind of situations, but always seems to get outpredicted/out-guessed whatever.

Many of these "mispredicts" have cost me the match, so I really think it is important to ask this question.

Help please!
 
I don't know if this can be taught... prediction is based on a) Knowledge of the opponents team, b) The play style of your opponent, which you should try to pick up on, or c) Experience.

Battle more, and you should get the hang of it.... that's all I can possibly say. It's not something I find can really be 'taught' effectively...
 
Prediction isn't really taught, it's more learned. The more you have battled, the better off your prediction will be. Knowing your opponents team and how they play can also help you predict better. =)


Now, not everybody can predict correctly 100% of the time; it's just not possible. But the more you play, the better off you will be.




and it's not really a dumb first post. :p
 
Yeah, I kinda understand this speak of "Knowing your opponents playing style" but how does that help you figure out if he's switching to Heatran or Celebi.

Oh and thanks for the posts guys ^_^
 
you think about it and make decisions that involve low risk but high return. for instance, you have scarfape out against celebi. what you'd do, is u-turn, which will hit celebi 4x super effective. if he switches to heatran, you do piddly damage, but at least you get to go to a counter for free.

I find myself making some predictions, but usually I'll do something that has low risk. i.e., if I'm trying to kill pokemon A, I might predict the switch to pokemon B and use a move that kills pokemon B, but still does decent damage to pokemon A. generally, in battles, you try to make the situation so that you can win without predicting too much, as prediction is risky.
 
I know in long battles I try to switch up the style of my switches. Early in the match when neither opponent knows each other team the style of play of the majority of players is pretty straightfoward (for the average shoddy battle player). In my opinion, shoddy battle ladder players tend to be more conservative early in the matches and if it is Heatran versus Celebi and they haven't seen the rest of your team they will 'probably' just go with Flamethrower to scout. Later in the game they might start outpredicting and you can really catch them outhinking themselves. Someone accused me of "way to not predict" - I just switched up my play style to stop what they were doing. They had seen my whole team and knew the counter I tried to bring in to their Mamoswine, and they were hoping I would bring Skarmory in again so they could run Blizzard, but instead I just sat put with my Heatran out to take the Blizzard easily and Flamethrower it to death. Another trend I notice from people is they get really conservative with their switches when they go down a pokemon or two early. When their pool of pokemon gets limited they seem to go more conservative and reliable with their switches in fear of losing even more pokemon and deepening the hole they have already dug into. Of course, some people will go really aggressive when they go down a couple pokemon and really try to outswitch/outplay you to regain that edge. It definitely takes a read of your opponent and your own evaluation of the situation.
 
You can't really teach prediction, it is something that you get from experience. Risk vs. Reward is a big part of prediction and you'll only ever learn this from just playing a lot of people. Also, you are going to be outpredicted (or you'll make a wrong prediction) sometimes, but if you are just beginning to learn how to predict the opponent it's probably best to use "safe" prediction before you become more skilled at it.

Let me explain what I mean (this actually happened to me about 5 minutes ago so it's fresh in my mind) I have SD Lucario (Crunch, Close Combat, Extreme Speed, SD) out against a Blissey. I also know my opponent has a Choice Scarf Gengar. I could Close Combat and go right for kill on Blissey or I could use Crunch. What is Blissey going to do to me? I already know it's moveset and it doesn't have T-Wave or Flamethrower so I'm perfectly safe here, so I reasonably predict the Gengar and OHKO with Crunch. If I were wrong and Blissey did stay in, nothing would happen to me. This is what I meant by making a safe prediction. Now, if I was in the same situation except my opponent had a Tyranitar in place of Blissey I would be less likely (although I would judge the opponent and predict what they would do) to go for the Crunch since Tyranitar threatens and OHKO with Earthquake. Safe prediction is probably the best way to play if you are still becoming acquainted with prediction, once you become better at it then you can start to go for more "risky" predictions based on your opponent.
 
You really have to know your opponent to predict properly. If he tends to sacrifice stuff, then you should take advantage of this and kill what he left in. However, be wary, as usually they are sacrificing because:

A) they have no answer to yuor pokemon, and have to revenge kill it.

B) they want to get in a sweeper unscathed and sweep you.

If it's the latter reason, it's probably best to set up on their sacrifice (I love the move substitute for this reason) so you're not completely screwed afterwards if they get in something powerful. If he's one of those people who switches a lot, by all means, predict a switch. However, sometimes it's better to stay on the defensive and heal yourself or set up entry hazards. Entry hazards tend to limit switching and make prediction easier.

In the case of the Infernape question, I would most likely Close Combat, unless the Celebi has Thunderwave. Worst case scenario is it could heal itself, or grass knot you for crappy damage.
 
Sometimes it helps to simply switch in your Infernape and then switch out to one of your walls just to see what they bring in. The next time you bring your Infernape in, you can then predict that the opponent will switch in the same counter. Granted this doesn't always work but at the very least it can help you see your opponents team. Remember, scouting with moves like U-Turn, Protect, and Substitute can be invaluable at times.
 
Get in your opponents head, I make noob moves on purpose ALL of the time simply because I know based on chat, posts, or previous play that they are overpredicting or just predicting that I'm not a total noob in general. I've left Skarmory in against Infernape to drill peck because he used a grass knot predicting a switch to my suicune for example. It's something that can't really be taught, you just need practice.

Also keep in mind that humans will learn from their mistakes, and you must do so too. In my skarmory killing Infernape example, I had switched to my Suicune to avoid a flamethrower hitting my skarmory a few turns before that. he switched out, knowing that I was running a more special defense oriented resttalk suicune set and that it would survive anything he threw at it, but Infernape would not survive an incoming surf. He switched in Gyarados, attempting to set up, I switched in Skarmory to WW him out (he was running the 3 move DD set rather than the bulky taunt version, I knew this because I saw him using both waterfall and ice fang earlier in the match, and no taunting gyara set carries both of those moves), I did so and he switched in Infernape (from the WW) for a third time against SR. At that point the noob's move would be to leave Skarmory's slow arse in and attempt to kill the injured Ape with Drill Peck while the move using more 'prediction' would have been to switch into my suicune, the only bulky thing I had that resisted a flamethrower.

Needless to say, I didn't make the 'smart' move and it payed off. If I had only been predicting what the logical move for him to make against the pokemon on the field was, I probably would have lost suicune, who was pretty much the only thing stopping Infernape from hardcore anal rape of that team in the first place. You have to get in the enemy's head.
 
To be honest, prediction is overrated. Watch goofball, or jrrrrrrr play sometime. They hardly, if ever predict, but both are top tier players. What you need to learn is to just make the best move for the situation your in. Like Slakethesis said, in your example, I would U-turn. You either do massive damage to Celebi, or force Heatran to switch out immediatly by sending in something it can't take. And if they switch something else, you get to scout that as well. There's no downside to that.

There are really very few situations in which you will need to truely "predict/guess". Good play comes from making the best move for the situation. Often times this looks like good prediction from the sidelines. You have CB Weavile vs a mid HP Blissey, and you Ice Punch, catching Salamence for the OHKO. Sure, looks like good prediction, but Ice Punch would've KO'd Blissey anyways, so it was really a no risk situation. Remember, you don't always have to hit SE to KO a pokemon. This is another thing that leads to bad "prediction". I suggest running damage calculations during battle (IRC is great, quick calculator if you know your stats).
 
There aren't rules for prediction, actually if there where it wouldn't be prediction. You have to improve your skills at your best (experience).

Other than that, my best tip is, try to use the information you got wisely.

For instance, in that nape vs celebi scenario:

1-If he stays what do I need to use to kill him?
2-If he switches out, what pokes are likely to come in? What are my options? (remember that switching is an option).
3-Does he think that I'm skilled enough to outpredict?
4-If there has been a situation like this before in this match, how did he predict? How did I predict?
5-Does he takes risks or is he conservative?
6-What does he think I'm thinking?
7-From experience, how does he act under pressure?

With all those questions you should be able to draw a picture of the situation, and then do what intuition dictates to be correct.

You will eventually understand that prediction has little to do with luck. Also don't predict when it's not needed.also don't play randomly. For instance:

You've got skarm vs nape. You've got a safe nape counter but you guess he's gonna nasty plot and try to brave bird. Don't push luck and play it safe. Only predict when you're at a safe position or when not trying to predict will lose you the game.
 
IggyBot said:
To be honest, prediction is overrated. Watch goofball, or jrrrrrrr play sometime. They hardly, if ever predict, but both are top tier players. What you need to learn is to just make the best move for the situation your in. Like Slakethesis said, in your example, I would U-turn. You either do massive damage to Celebi, or force Heatran to switch out immediatly by sending in something it can't take. And if they switch something else, you get to scout that as well. There's no downside to that.
Click to expand...

I agree 100% with this.

But for your example in the OP, if you didn't have U-Turn your move there really depends on how the rest of the battle went. If your opponent made a lot of obvious switches and never tried to "out-predict" you, you'd probably be alright to use Close Combat expecting Heatran. Usually in that scenario though it's less of a prediction and more of a straight up guess based on the risk/reward of each move. If Heatran is the more threatening pokemon and you have lots of other pokemon to take on Celebi then going with Close Combat would probably help you out more. And vice versa. Sometimes your guess will cost you the game and lose you your Infernape at no cost to your opponent but hey that happens. You won't win them all but usually you can find solace in knowing you at least made the best move you could based on the situation.
 
General guidlines for prediction:
1) When it's Dark, Play it Safe: make obvious, safe plays early in the game. This will give you time to scout, as well as observe your opponents movesets.

2) Be Predictable: Don't try to predict an opponents move until you know that your opponent knows what your going to do. There was an example given earlier on about player A switching skarm to cune out of Player Bs flamethrower twice, and then on the third time drill pecking for the ko. Player A knew that Player B knew that Player A would always make the safe switch to cune no matter what because he did it twice already, and showed no signs of stopping. Rule of thumb: If you repeat the same safe move twice, the third time is the time to predict.


3) Situational Plays Should not be Repeated: As stated before for safe plays like switching to a wall or using an attack that's super effective, the rule is safe move, safe move, predict. But for riskier moves like Taunt and substitute, you want to do it safe, predict, predict, safe. So if I taunt or sub my opponents Blissey, and then we end up switching out and back to the same pokemon, this time I'm going to Beat on Blissey/start to stat up, because I know that my opponent knows that I have taunt/substitute, and therefore will want to get a free turn by using Seismic Toss(or whatever damaging move blissey uses now). Here it's a toss up, but it's usually the safer route to expect the Blissey user to continue to fear taunt/sub, so keep on smashing/stating up.

4) Be Self-Aware: When deciding what move to use, be aware of what playing style you have displayed to your opponent. Remember, your actions up until now(this turn) determine what your opponent thinks youre going to do.
 
Prediction is the game of the gambler.

Personally, I think you should try to minimise your loses, as opposed to maximising your wins. Turn the problem around 180 degrees, so to speak.

Instead of thinking: "what is my opponent going to switch into?" or "Is my opponent going to switch?", look the problem from the other end: "What does my opponent think I am going to do?". If the answer is "He thinks that I'm going to use an Ice Beam against his Salamence", then you can take appropriate action to out-predict him.

But that's just my opinion.
 
novelty said:
Hmmm, there is a problem I've been facing since I started properly competitvely playing this game, and that problem is prediction, well a certain type of prediction, it's kinda hard to explain, so I'll give an example.

Example

I have a scarf infenape, my opponent has a celebi and a heatran, with the right predict I can get a kill, I seem to guess in these kind of situations, but always seems to get outpredicted/out-guessed whatever.

Many of these "mispredicts" have cost me the match, so I really think it is important to ask this question.

Help please!
Click to expand...

Prediction is all about playing fate, gambling your hand. You evaluate your options and then hope the decision you made is correct. It is unfortunately entirely based on luck unless you can read your opponent and even then you still need to hope they follow the strategy you think they will. Just last night I was reading how abused Wobba is (henced why he's banned) but came up with a possible counter. It however requires the person not to realize what I have in mind or it could be recountered, although there is a certain level of full proof to it.

Basically it all comes down to luck, sometimes you have it and sometimes you don't.
 
You must log in or register to reply here.
Back
Top