Post by hartlaboratories in Gameplay Discussion

Viewing post in Gameplay Discussion

So one suggestion for game play I have is to even out the difficulty variance, and to have it ramp more at the low end. In the forest having an easier first couple hundred feet/meters seems like it would make it more approachable to new players, but overall the variance over the entire game is very high. I feel like a huge amount of the current difficulty variance (seed luck aside, which is a separate thing) is a result of homing enemies.

The spawning algorithm does not seem to take this into account, so sometimes you make have 3 homing enemies (pink spiders and large pink fly's) spawn at once early on and follow, which is very demanding so early for new players, or you may only have 1 per screen or less at higher levels which makes it fairly simple for experienced players. Right now a runs difficulty curve can be all over the place and sometimes decrease as time goes on.

Something that could maybe be done is to spawn homing vs non-homing enemies independently so they are not over or under concentrated on spawn. Also possibly tracking the total number of homing enemies on screen and limiting it based on progression seems like it would help even out the enemy variance.

benedictfritz7 years ago

In the forest having an easier first couple hundred feet/meters seems like it would make it more approachable to new players

Agreed

The spawning algorithm does not seem to take this into account, so sometimes you make have 3 homing enemies (pink spiders and large pink fly's)

It does take it into account, but not in a way that is as effective as it could be. Right now every enemy has a weight. We spawn zones of enemies, and each zone has a weight sum. Homing enemies have a much higher weight than non-homing. The idea being that ideally a single homing enemy (or other more-difficult enemy) spawns in the zone, and then another high-weight enemy can't fit under the weight limit, so the rest of the zone gets filled with smaller enemies.

In reality what happens is we get bad spawns when homing enemies happen to *just* fit into an enemy's weight-sum so they are the only enemies in that zone. And of course two or three homing enemies with no filler enemies is MUCH harder than one homing enemy with a commensurate amount of filler enemies.

All this to say that we are trying, it's not working as well as it should, and it should definitely be improved! Limiting the number of homing enemies per zone is definitely one approach on the table right now, and should probably be the system since it is what the weight system is attempting to do right now, and failing quite often.

All this to say that you're right!

hartlaboratories7 years ago (2 edits)

Ah cool yeah that all makes sense. I guess the problem is non-linear difficulty of multiple homing enemies. 3x pink spiders is more than 3x as difficult as 1, but it is any homing enemy that effects this not just individual types. An exponential weight multiplier would probably do something. Also that they will persist and follow from the previous zones. Tracking homing enemies close to the player and including it in the new zones spawn weight to reduce it etc. Give you a hope of running away when under pressure from them with a slightly easier zone.

benedictfritz7 years ago

Tracking homing enemies close to the player and including it in the new zones spawn weight to reduce it etc.

We spawn the whole world at once so that we can have consistent seeding and eventually have daily runs.

The exponential weight multiplying is really interesting!

itch.io

Viewing post in Gameplay Discussion