Why must games stand in their own way? When we pay homage to classics of the past, that does not mean that we must forget the lessons of modern-day design. Xeodrifter, like Axiom Verge, looks just fine: you can slap just about any aesthetic style on a Metroidvania core and be fine with it. But almost all of its secrets are endurance trials: they’re not particularly hard to discover, they just require too much backtracking to justify. If there were a fast-travel option, that might at least reduce the frustration, but once a player has cleared an area, there’s only so many times you can ask that player to clear it again before the game becomes visibly padded. Exploration’s fine, but you should never reach a point in a game in which you know what you need to do, but lack all motivation to do so. (The lack of any story or twist in Xeodrifter may exacerbate the situation.)

Teslagrad, at least, is on more solid footing–which is ironic, since you’ll use magnetic forces to navigate its electric tower, rarely touching the ground. Each area spirals off the central spire, so while you’ll have to re-explore some of the earlier rooms, it doesn’t take all that long to do so. Moreover, each new ability can be used to cheese through earlier, potentially difficult puzzles, so while you may have to walk the same corridors, the path won’t be anywhere near as laborious. Best of all, the collectibles in Teslagrad are each tied to the plot–the scrolls are what fill in the backstory of the ruined kingdom, and players must earn at least fifteen of them to reach the finale.

In essence, then, I propose a few rules:

• Keep the focus on the exploration, not the traveling. The goal is to have players constantly doing new things in the environment, so let the computer automatically return players to any point, with the implication being that if they’ve gotten there before, they could do so again, so why force them to do so?
• Make the reward worth the journey. That is, don’t just hide extra health and missiles. Some of that’s fine, but Axiom Verge and later Metroid games also squirreled away lore and weapons, and that made filling in the map more of a necessity than a chore.
• Harder doesn’t mean stronger. This one’s universal to all games, but in general, you should be doing more to make a game difficult than simply increasing the damage an enemy deals while reducing the damage it takes. The one thing Xeodrifter nails is that its boss gets increasingly difficult; the problem is that there’s only one boss, which you continually encounter in various pallets. Teslagrad, on the other hand, has encounters that change up their patterns, and each puzzle builds upon the previous one, combining tools, as opposed to Xeodrifter‘s monotonous way of using tools to simply block off new areas.
• If you’re going to imitate, innovate. There’s nothing wrong with paying homage to classics of the form, but I can’t think of a single thing Xeodrifter does that other games haven’t. Sure, you don’t often gain the ability to swim by turning into a submarine, but using a charged shot to break down barriers or jumping in and out of the foreground? These are borrowed devices that must do more than simply exist. Again, Axiom Verge found ways to subvert expectations of classic Metroid weapons, and Teslagrad created its own magnetic mayhem.

Ultimately, keep the player experience first and foremost. If you’re going to make them work, really ask yourself why, and if it’s going to be worth it for both you and the player.