• Rhythm games. With frame exact reaction times, it is trivial to get a perfect score. The games are also fixed length and there is nothing that can be done to change the length of the game
• Many puzzle games where knowing the solution renders gameplay trivial, such as Sudoku. If you know the solution, a speed run just a matter of entering it as fast as possible.
• Many racing games, particularly those where there is an obvious optimal route; the precision can trim off a fraction of a second compared to expert players, but for casual viewers the difference is largely unnoticeable.
• Some top-down or platform games where gameplay is largely "run forward and dodge obstacles", such as Front Line.
• Interactive novels and adventure games. Most such games is largely composed of dismissing text and navigating menus. There can be some merit if the fastest solution is hard to find, but in general there is only one way to beat them.
• Quiz games, watching someone enter answers is not fun. Unless the answer recognition routine is abused. Abuse of glitchy input detection can make for interesting runs (such as with Family Feud or Brain Age)

• Microgame collections, comprised of courses of a fixed number of fixed length microgames.
• Party games, we all know that they are just disguised minigame collections.
• Rhythm games. As mentioned before, the songs have fixed lengths.

• Intentionally boring games such as Desert Bus.
• Many board games such as chess.

• Games without goals If there is no goal, there is no clear way of comparing runs.
• Games intended for small children No, Color A Dinosaur and Where's Waldo don't have gameplay with any meaningful technical content. They are great candidates for April Fools' Day submissions, but no more.
• Online games We require repeatability and that is impossible without controlling the server. We also wouldn't allow any homemade server software.

If you want examples of games that you shouldn't pick, you can take a look at some examples of bad game choices.