CAMBRIDGE, MA—During a press conference Thursday at the Massachusetts Institute of Technology, leading members of the scientific community confirmed that despite decades of research, the best available theoretical models still cannot predict more than one upcoming Tetris block.
While they have long possessed the technology to accurately forecast and display which of the seven distinct Tetris shapes, or tetrominoes, will appear next, scientists told reporters they’ve so far had no success determining what any of the blocks beyond the immediate successor will be.
“Modern supercomputers have now reached speeds of 30 quadrillion calculations per second, but even with all that processing power, we’re no closer to solving this problem than we were a generation ago,” said MIT professor Michael Haemlin, who has studied tetromino sequencing since 1984. “Right now our most sophisticated equations can correctly predict two upcoming blocks just one out of every seven times, which unfortunately is no better than the result achieved by random guessing.”
“So for years, all we’ve been able to do is rotate each descending shape 90 degrees at a time, control its lateral movement, accelerate the speed of its descent if desired, and simply hope for the best,” Haemlin added.
According to researchers, the ability to predict two, three, or four incoming shapes would be “nothing short of revolutionary,” allowing blocks to be stacked in more optimal patterns in order to achieve the goal of eliminating horizontal lines as efficiently as possible.
For example, with enough advanced warning of an arriving I-shaped piece, scientists said they could in theory prepare the Tetris matrix in such a way that up to four rows are cleared simultaneously, with many stating this was one of the ultimate goals of their rigorous stacking trials.
“Over the past 30 years, we’ve developed a much better understanding of how blocks fit together,” said Dr. Florence Edelman, the designer of a well-known Tetris experiment in 1993 in which a perfect slot for an L or zigzag block was created under carefully controlled conditions, only to be closed off when a series of ill-fitting square pieces appeared instead. “But without a working predictive model, our entire field of study is at an impasse. Indeed, reaching unpassable standstills is well chronicled in trial after trial.”
“Some of the world’s most brilliant scientific minds have dedicated their lives to unlocking this mystery, yet for all their knowledge and expertise, it’s still anyone’s guess what will happen five or 10 seconds into the future,” Edelman continued. “Once that last bit of uncertainty has been eliminated, we believe our interactions within the Tetris matrix will be far more successful.”
Edelman stated that many aspects of Tetris are indeed predictable, noting that workers in her lab had discovered that each falling piece is consistently composed of four individual tiles; that the speed at which descents occur increase in direct proportion to total time elapsed; that pieces with the same shape also share the same color; and that while the color of a piece is not particularly important, its shape is of vital significance.
Foreknowledge of those shapes, she explained, could lead to a breakthrough phenomenon she described as “a perpetual Tetris” of unlimited duration.
“While this remains entirely hypothetical at this moment, there exists a theoretical point at which the elimination of bottom rows occurs with such speed and efficiency that there is always enough room at the top of the matrix to accommodate new pieces,” Edelman said. “This would create a fluid and sustainable Tetris that could exist indefinitely—perhaps forever. It would of course still require continuous human maintenance, but it would be a monumental step in the right direction.”
“For now, though,” she added. “We’ll simply keep trying and see how far we are able to progress.”