Something that is often overlooked when discussing prediction is the concept of measuring the risks and rewards of a move. This article will look at the basics of measuring the risks and rewards of moves, strategies for minimizing risk or maximizing reward, and Pokemon who skew the risk/reward relationship.
There are two ways in which you can measure risks. You can measure the risk of your opponent's immediate moves, or measure the risk of moves your opponents could take that would hurt your team or your overall strategy the most. Both of these have scenarios in which they can be applied. However, deciding which approach to measuring risks in a given scenario is also key.
In many cases, you will have a good idea of what your opponent will do. If a Metagross comes in on a Regice, you can expect it to use Meteor Mash. Before moving, you unconsciously take the risk/reward into account. If you stay in and he uses Meteor Mash, Regice is dead. That is a big risk. However, if you switch to Suicune and he uses Meteor Mash, little damage is done, thus eliminating the risk factor. If you are playing conservatively (in this case intelligently) you will take the switch to Suicune. This analysis gets more interesting once you consider Player 2's position. You have just brought in a Metagross on a Regice. You have a Zapdos on your team as well. You can be fairly certain that your opponent will switch his Regice to his Suicune. Analyzing this, you can be almost certain that your opponent will switch to Suicune, as it would be a great risk for him to keep Regice in. This is an example of the first way in which you can measure risk. Your opponent's move choice is almost definite, so you can use his lack of options to switch to Zapdos.
This prediction is not cyclical because of the lack of risk for Player 1. He is forced by his opponent to switch to Suicune, but his opponent possibly predicts that and switches to Zapdos. He does not Ice Beam with Regice, however, because there is no risk in his switch to Suicune. Regice can continually wall Zapdos, and Suicune can continually wall Metagross. Player 1 does not have to worry about prediction games at this point, because there is very little risk to playing his opponent at face value.
Medicham has come in on Regirock. As player two, this scenario should be analyzed in the second method: considering the most dangerous outcome. You are not sure whether the opponent's Medicham will use Hi Jump Kick or Rock Slide. Rock Slide will kill Salamence, and Hi Jump Kick will kill Regirock. If you switch to Salamence as your opponent Hi Jump Kicks, you will survive and can sweep his team after just one dragon dance. If your opponent uses Rock Slide as you switch to Salamence, you lose. If your opponent switches to Gengar as you switch to Salamence, you also lose. There are two worst-case scenarios in this situation: you take a Rock Slide from Medicham or your opponent switches to Gengar as you switch to Salamence. Both of these losses stem from making the switch. The reward is greatest if you take the small risk that Medicham Hi Jump Kicks (resulting in a Regirock death). If you Rock Slide from Regirock, you force the win no matter what your opponent chooses.
This scenario is very specific, but is chosen from an actual battle as it exemplifies how risk/reward can be far more important than prediction. And it is also a very nice transition into the next section of this article.
Playing conservatively is most feasible when your team has an advantage over your opponent. In this style of play, you take care to minimize risk at every single turn. Say your opponent has no Jirachi counter, a far more common instance than I would have thought. If your team has few weaknesses to the opposing team, you should play extremely conservatively. If you play with a team with three or four walls or tanks, this should be no issue. Simply switch in a Pokemon that can completely stall whatever Pokemon they switch in. Only one prediction is needed in the entire match. After your opponent is struggling to get through your tanks, a switch to Jirachi on a Pokemon such as Blissey or Milotic will swing the game permanently in your favor. This strategy can be quite boring, but quite effective when you know you have a team advantage over your opponent.
In every battle, try to find the one Pokemon that the opposing team has the most trouble countering. If your team has an advantage over your opponent's, then move cautiously to minimize risk until you are certain you can get that Pokemon in. Before every turn, carefully consider both the most likely move of your opponent and the worst-case scenario. If your team is matching up well, however, then worst-case scenarios will be largely irrelevant. If your team matches up poorly against your opponent, you will have to ignore worst-case scenarios completely as you struggle to get your most-damaging Pokemon in. The worst-case scenario is hardly worse than the average case, in which you are worn down or beaten by a Pokemon for which you have no counters. By manipulating the most likely move of your opponent, you can create opportunities for your most damaging Pokemon to switch in and attempt to equalize the match.
Certain Pokemon have a huge impact on risk/reward analysis. Dugtrio is possibly the greatest of these. If you are using a Celebi, you may switch in and out a few times to see if your opponent has a Dugtrio. Once you are aware your opponent has a Dugtrio, your risk/reward analyses are skewed for the rest of the game. Dugtrio forces you to consider worst-case scenario options continually when thinking about switching in Celebi (worst-case being your opponent switches to Dugtrio at the same time). This can cripple your defense against Pokemon you may depend on Celebi to counter, such as Swampert. Dugtrio's presence, even without usage, forces you to play with far greater risk no matter what strategy you choose. If you choose to go to Celebi, you are taking the large risk that Dugtrio switches in, for the small reward of forcing a Swampert switch. If you choose not to switch to Celebi, you are not using your usual minimal risk defensive strategy for that mud monster.
Magnezone operates much the same way. Sometimes people might fish for Magnezone by Roaring unnecessarily with their Swampert. I encourage you to put Magnezone in to get roared, as that will greatly enhance the usability of physical attackers. The knowledge that Magnezone is waiting in the wings will greatly decrease your opponents' urge to block your Lucario (Swords Dance) with Skarmory. He will be forced to, at the minimum, alternate physical walls in order to keep you guessing when he will bring his Skarmory out. Unlike Dugtrio, however, Magnezone has a very limited usage. He is extremely useful against Skarmory, obviously, and to a limited extent against Pokemon such as Gengar (presuming you can switch it in while Gengar does not use Substitute). Using Magnezone is another bet, seeing as it is riskier to use than Zapdos, but the potential rewards are far greater. The great effectiveness of both Dugtrio and Magnezone lies in creating a sense of enhanced risk in the opponent. It greatly increases the risk of using specific very common Pokemon, and that opens up the defenses of teams.
A Pokemon doesn't need to have a trapping ability in order to have a great impact on an opponent's sense of risk. Choice Banders are easily the greatest example of this, especially the ridiculously powerful ones such as Slaking, Medicham, Metagross, Salamence, and Mamoswine. Pokemon such as these, with no completely safe counters, force your opponent into situations where he has no opportunities to minimize risk. The best he can do is guess at what move you have chosen, and choose an appropriate counter. Prediction is key in using these, but even guessing right only occasionally will have an extremely large impact on the game. These Pokemon with just a few hits open up the opposing defenses for previously walled Pokemon. Versatility and surprise can be an important factor in risk analyses towards the beginning of the game. A Choice Band Gengar lures a Celebi switch, only to do huge damage with a STAB Shadow Ball coming from a pumped attack stat. Surprise lets the opponent think they are minimizing risk, only to be placed in a handicap. After the surprise is gone, however, the surprising Pokemon (such as choice band Gengar) is often placed at a disadvantage to its lack of power and now lack of versatility. Thus it is often advisable to fake a normal set until such a time when a surprise can open up the game.
The most important analysis to be done is to compare teams. First, figure out which team holds the advantage. Second, figure out which Pokemon on both teams can do the most damage or change the course of play. If your team has the advantage, play conservatively while minimizing opportunities for your opponent's game changer to switch in. Play conservatively until you get a sure-fire opportunity to switch in your game changer. If your opponent's team has the advantage, you will need to use surprise, predictions, and most likely a bit of randomness in order to regain some momentum. By making yourself unpredictable, you can attempt to force a switch which allows your game-changer to enter play.