Players are not supposed to be predictable. If your opponent is being predictable, he is simply not a good player, and it is a pokemon's potential at top level play that determines its tier.No, my point was only that any decent player is trying to out predict his opponent. Your exaggeration is what we call a 'straw man”. I never implied something so extreme. Nobody that makes all his predictions based on a single move is is a “decent player”. Still, since this upset you so, a “good player” (as you put it) will gather as much information from the beginning and throughout the match as possible. They will try to make a distinction between what information is relevant and what information must be disregarded and/or ignored. I gave examples of this in an earlier post, but I guess you were to fired up to just ask for my “proof”. No good player is going to allow themselves to be predicted with utmost certainty, but to say that they cannot be predicted is a joke. How does anyone play professional chess, professional poker, professional GO? I think such “patterns” produced by players are called tails. Even if it were the case that players were unpredictable wouldn't that mean that the entire meta is a series of 50/50's, and Aegislash just forces some special variant? I mean if that is the case then we can go ahead and throw that argument out, can't we? Atleast, someone can make a distinction between the 50/50's of all these unpredictable players and Aegislash. Of course, I wouldn't want to “straw man,”, so I will give you the benefit doubt and assume that you either did not mean that the meta is completely unpredictable or that you have in mind how Aegislash scenarios are relevantly different.
There is no advantage to having a pattern in your predictions, but it comes with the nasty disadvantage that your choice can be outplayed easily. Good players know that it is best to choose randomly (against another good player who could otherwise adapt to their patterns, anyways), even if it's not with 50/50 odds it should still be random.
And not everything is 50/50 odds. An extreme example is Scizor without Superpower vs Heatran. Both of you know heatran will predict Scizor to switch, fearing a fire move, and that will give Heatran a free turn to set up Stealth Rock. Yes, the Scizor player could predict this, and instead switch out using U-turn to get some damage in. But this is where the Skill comes in, where you can analyze the risk and reward. Scizor is better off hard switching almost all of the time, because it's worth being predictable if it means you guarantee Scizor's survival. It is not worth doing 8% damage to Heatran, even if you know you will probably win the prediction, because it is not worth the risk of losing Scizor. So that situation is more like a 99/1, because it's worth being predictable 99% of the time when the reward (keeping Scizor) far outweighs the reward U-turning (almost negligible chip damage).
Aegislash, vs contact attackers in particular, forces true 50/50s because the risk and reward for both players are very similar. If Aegislash's opponent attacks, they either will do big damage to Aegislash or give Aegislash the next turn free. If Aegislash's opponent doesn't attack, they either take big damage from Aegislash or that turn was free because Aegi used King's Shield. So the risk and reward for both options are very similar (free turn or big damage).
I made a more complete analysis a few pages back.
Also idk why you compared to Chess and Go, in those games you have full information about your opponent's choices by the time you make your move so there is no need to guess. Poker is more comparable to Pokemon, there is a big element in guessing and while risk/reward factors in, it is best to randomly bluff some times and not bluff other times so your opponents can't tell when your hand is as good as your bets indicate.