Jeffrey ‘Trump’ Shih is one of Hearthstone’s most popular content creators, and with good reason. But not all of his efforts are universally appreciated. His card reviews have gained praise for their bold nature, but also infamy for their inaccuracy. Frequently, he has rated swathes of cards incredibly pessimistically, and been shown to be catastrophically wrong. These unfortunate errors gained meme status. Is this fair? And why do Trump’s supposedly unique methods fail so spectacularly?
Trump has responded to his critics with good grace. As he has reiterated many times, he is aware that he sometimes looks silly. But to Trump, this is a small price to pay. He’s deliberately attempting to avoid what he characterises as the overly safe predictions of other streamers.
The root of the problem, in Trump’s eyes, is that other streamers view cards in a vacuum and then make vague assertions. This, according to him, leads to the non-committal assessments and lack of real insight. In the end, what matters for a cards viability is how often and how successfully it is used, not how good it is in an impossible, hypothetical Hearthstone that somehow assesses its raw individual power level.
Trump’s solution is to not just try and predict cards; he tries to predict the entire meta. When he rates a card 1 star, it does not reflect upon its potential; merely that it won’t see play. This could be down to meta, class viability, or availability of supporting cards. This is a far more interesting approach, as it is a holistic analysis rather than focusing on a single piece of the puzzle. But it’s not without its serious flaws.
“1 star: Warlock won’t see play”
Take the example of Warlock in the Knights of the Frozen Throne. Trump rated all but one Warlock class card at 1 star out of 5. This included what we now know as huge power cards for Warlock, like Bloodreaver Gul’dan and Defile. The only exception, Despicable Dreadlord, which was only given 2 stars. This massive blunder is part of the reason many mock his assessments.
Crucially, Trump called many of these cards ‘good’ or ‘great’, just considered Warlock as a whole to be unplayable. This is a major problem with Trump’s methods; he makes bold assertions about the the meta, which is arguably far far harder to predict than simply rating cards. And if his meta assessment is off, all of his ratings are massively off.
Predicting the meta, an impossible task?
In theory, what Trump seeks to do sounds doable. Trump’s attempt at meta prediction is similar to that done about the arena at ADWCTA and Merp’s lightforge.com. Unfortunately for Trump, Constructed is almost impossible to predict compared to the arena. The endless complexity of deckbuilds interacting, countering and developing with each other is as chaotic as the weather. And if a single powerful card or deck flies under the radar, it can completely throw out all calculations.
In short, Trump is attempting the impossible. If his predictions are reliant on accurately foreseeing the meta, then they will never be right, and if they are, it will likely be by coincidence.
But surely this is still better than the ‘in-a-vacuum’ card reviews of other streamers? Well not quite.
Trump’s secret ratings
The style of card rating Trump disapproves of, ‘in-a-vaccuum’ power-level assessment, is actually still necessary for Trump’s reviews. In order to try and predict, say, how powerful a given class is, he needs to look at the power level of the cards it is given. So Trump is essentially pre-rating cards using the exact method he criticizes, only not telling his viewers!
Of course, you can try and glean his thoughts from what he says about the cards, but the lack of a clear rating obfuscates his both his assessment and any mistakes he makes in the process that affects his final rating.
Ambitious but not visionary
Trump’s method is definitely interesting and different. But it’s not anything more innovative than a standard meta prediction spread out among cards, an its flaws deeply limit Trump’s usefulness as a card reviewer. His meta predictions are never accurate, and because he does not fully explain the background ratings that result in the final ratings, his viewers are left confused and his mistakes seem more egregious.
Perhaps a better option would be to keep his ambitious card review method and add another rating system to represent the ‘in-a-vacuum’ power level decision he makes in the background to both help explain his mistakes and temper his more extreme decisions.
Images courtesy of Blizzard Entertainment via Hearthstone.gamepedia.com. Title image from Trump’s youtube channel