Since 2007 The King of Kong hit theaters, the back and forth race for ever higher scores on the arcade classic Donkey Kong has drawn inordinate attention from competitive gamers. Last week, Wes Copeland tried to end that battle once and for all with an incredible new world record score of 1,218,000 points.
Some press reports have referred to this achievement as the world’s first “perfect” game of Donkey Kong, implying a score that will never be surpassed. That framing has been driven in part by Copeland himself, who said on Facebook that “this will be my last record score… I don’t believe I can set a game higher than this.” A breathless post from the score-watchers on Donkey Kong Blog calls it “a score we can confidently say, even if it’s not final, ‘this will never be beaten.'”
Yet there is still some reason to believe higher Donkey Kong scores are technically feasible, even if no one in their right mind will be able to do so anytime soon.
Point ceilings 101
Donkey Kong is fundamentally different from more deterministic games like Pac manwhich has a concrete score ceiling of the 3,333,360 first achieved by Billy Mitchell in 1999. In Donkey Kong, the highest possible score depends heavily on the random movement of the numerous enemies in the game. A few more barrels together for multi-ton jumps, or a few more flaming barrels wandering close to Mario’s hammer, can mean the difference between a mere good score and a record-breaking score. (Donkey Kong Forum goes much deeper into state-of-the-art point-pressing strategies).
However, that arbitrariness hasn’t stopped people from trying to calculate a theoretical maximum score. In 2010, then-champion Steve Wiebe said that 1.15 million points “probably where it would be realistic to [maximum] score.” In 2012, the score watchers on Donkey Kong Blog named 1.2 million points as the “Current prevailing wisdom among the game’s top players” for “the practical ceiling” of the game. The fact that both have “ceilings” now broken should show the “perfect” Donkey Kong score is a moving target and that Copeland’s 1,218,000 may not be unbeatable.
In a 2015 FiveThirtyEight piece, DonkeyKongForum stats guru Jeremy Young mentioned the very specific 1,265,000 as the game’s theoretical ceiling, though he couldn’t explain the math behind that number. In contrast, a detailed analysis by Twin Galaxies referee Robert Mruczek found a “reasonable theoretical maximum” of just over 1.3 million, given ridiculously perfect luck and impeccable execution. Discussions of tool-assisted runs suggest scores of over 2 million points could be possible with tedious frame-by-frame analysis and massive state-driven manipulation of in-game random number generation, which would never be possible for an unassisted player.
Better to be lucky than to be good
Theory aside, a score higher than the current 1,218,000 should be humanly possible. For proof, check out this analysis of Robbie Lakeman’s previous world record game, which scored just over 1.19 million points. Early in that run, just past level 6-6, Lakeman was “on pace” to score a whopping 1.26 million points, before the vagaries of chance and skill brought his score back to more reasonable levels. If Copeland’s lucky (and skillful) performance of 71,200 points in the sixth leg could somehow have been repeated over and over for the rest of the run, his score could easily have been over 1.3 million.
But even without ridiculous luck and execution, Lakeman’s score shows that Copeland’s goal can be beaten. The real key to Copeland’s new record was the fact that he didn’t die at all during the game’s first 155 levels – an incredible feat for anyone, let alone someone who did risky, point-pushing moves for over three hours. made. This allowed Copeland to craft the final barrel stage for over 10,000 points per life while deliberately sacrificing three of his extra Jumpmen, leaving him one life to reach the “kill screen” just two levels away. A game that combined the luck of Copeland’s run with the deadly skill of Lakeman’s could easily break 1.22 or even 1.23 million.
Even if higher scores are technically possible, we are definitely reaching the point of diminishing returns Donkey Kong push point. For top players, breaking the record is now just a matter of aligning the stars for a perfectly focused run to coincide with an incredibly lucky streak of in-game randomness. The record will only be broken if someone who plays perfectly also gets the in-game equivalent of a dozen coin flipped “heads” in a row (or if someone manages to find an improbable feat to break a currently unknown score- inflating strategy)
That daunting prospect is daunting enough to make many competitors stop trying. Lakeman, who has been embroiled in a high-score battle with Copeland for months, said on Facebook that he is “not wasting his time” trying to outperform the new high-water mark. “I’m not lucky,” he wrote. “Good enough, but not happy enough… I won’t beat it.”
Of other high-level players who may be willing and able to take on the new record, Donkey Kong Blog lists only Dean Saglio, the first player to break the 1.2 million mark in 2013. However, Saglio’s use of MAME emulation to make his mark caused quite a bit of controversy, with many claiming that the more responsive keyboard controls gave him an unfair advantage.
Even if Copeland’s current score ends up being beaten, Donkey Kong Blog is probably right to get the Donkey Kong race with high score effectively over, almost ten years later King of Kong sent it into high gear. “Copeland’s run is a practical maxout in every way,” the site wrote. “That’s what we wanted to see, and it exceeded expectations. This is a world record that has nothing to complain about, and so much to praise. The theoretical possibility of better is one thing; conjuring up the insane and blazing fire of motivation will take it to to realize that it is another.”
List image by Donkey Kong Blog