Though Starseed Pilgrim may seem simple with its simplicity of graphics, sound design, and mechanics, its starseed design space is fairly complex. It’s not that there are many rules to learn for each seed, but that these rules interact with each other in large range of ways. Understanding the game design of Starseed Pilgrim by solely considering this design space is a difficult task that is too abstract and conceptual for most people to truly grasp. Fortunately, we can combine an abstract analysis with concrete scenarios to better help us understand how design decisions and rules ultimately affect the way we play. The previous Starseed Pilgrim strategy analysis is excellent in that it takes each seed and considers the possible useful strategies based on the main, goal seeking sub-goal one might have playing an actual game of Starseed Pilgrim. Explaining these strategies in words is particularly useful because we can learn to internalize the reasonings and logic behind the strategies to better aid our own ability to play well. The only downside to such a strategy analysis is that it’s very wordy; after all it attempts to address as many practical emergent gameplay scenarios of Starseed Pilgrim’s gameplay as possible.
This addendum seeks to take the information presented in the Starseed Strategy Analysis and compile the data into numerical charts for the purpose of better understanding how Starseed Pilgrim is designed, how randomness and probabilities affect play, and to help us consider exactly what changes, if any, we might make to the probabilities of random elements to ultimately improve the play experience. In this case, any “improvement” suggested will merely be a degree of balance fully articulated to the same level of specificity found in the Starseed Strategy Analysis. In other words, I’m not suggesting a vague idea of what a “better” version of Starseed Pilgrim would be; rather, I’ll use the research to accurately depict a version of the game in ways that any player of Starseed Pilgrim can understand.
Before getting into Johnathan’s research, I wanted to get an idea of how a new player of Starseed Pilgrim might rank the usefulness of each seed. The point of this assessment is to get a clear idea of what a player might think who is inexperienced with the wrinkly design space of Starseed Pilgrim but experienced in the rules of play and general strategies. The chart reflects the rankings of each seed based on simple categories. The higher number the rank number, the more useful the seed. The rankings for average vertical movement, horizontal movement, max and min area growth is based on the idea that more growth is better. The vertical movement column only considers upward movement. Notice the % results at the end. Without knowing too much about specific strategies, a beginning player might consider the red, green, and orange seeds to be the best as they are the most useful most of the time. The % column is a simple way to compare seed value based on a much more complicated evaluation.
These charts are based on Johnathan’s research and how he ranked the seeds by their usefulness when moving horizontally, up, and down. After being Starseed Pilgrim and collaborating with other players who have done the same, Johnathan based his research on what I refer to as expert or high level play. Notice how purple seeds are valued as the most useful for an expert and not as useful to a beginning player.
A key point in Johnathan’s analysis is that moving horizontally has many more strategic advantages than moving up or down. To reflect this idea, I took the data and added a weighting to each column. Because horizontal movement is the most useful, I multiplied each ranking value by 3x. Then I applied a 2x multiplier to the upward movement rank column. Notice how orange becomes the most useful seed at 19.25%, which can be interpreted as the orange seed would be the best seed to use for ⅕ of all scenarios in Starseed pilgrim.
When considering the “reach” strategies, touching plantforms to stars/key blocks in the fleeting world, the values in the table shift.
The raw data, which treats all sub-objectives [moving up, down, horizontally, and reach] as equal, tends to value the purple and red seeds higher than orange.
Consider that a given seed has a 1-in-8 or 12.5% chance of appearing in the player’s seed queue excluding cases where that same color seed is the previous seed and when the previous 6 seeds are non-pink seeds. Recall that there are rules in place that prevent the same seed from appearing in the seed queue back-to-back; and that a pink seed will always be given after a min-drought of 6 non-pink seeds. Because of these rules, pink seeds actually have a slightly higher chance of appearing in the seed queue at 14.29%. Using these percentages as a basis, we have a simple way of interpreting the data in the charts above. If a non-pink seed is valued at 5.71%, like with the blue seed, yet you’ll get this seed 12.5% the time, then most of the time you get a blue seed will be in scenarios where you will not have a good use for it. Likewise, if the orange seed is useful 19.59% of the time and you will likely get this seed 12.5% of the time, you’ll likely need the orange seeds more than you’ll get them.
In my gameplay analysis on Starseed Pilgrim, I explain that the uneven or wrinkly design space of the seeds is the most interesting part of the game. The design space is uneven because properties that determine how the seeds interact are not applied to all the seeds. For example, ice and red seeds have the ability to override other blocks. That property is applied only to ¼ of the total seeds. This uneven properly is the main contributing factor that makes figuring out how to best use Starseeds in a given scenario so challenging. As is evident in Johnathan’s analysis there are many branching “if-thens” to consider when placing each seed. Unlike Tetris, where each piece is equally useful for filling in 4 squares of space in the play field and the challenge comes from properly fitting these pieces together, there are many scenarios and challenges in Starseed Pilgrim where certain seeds are useless. Wasting red and blue seeds is common in the high level strategies detailed by Johnathan. Setting green seeds out of the way is also common.
Consider what the gameplay of Starseed Pilgrim would be like if the probability of getting each seed matched the data table above. The most any seed increased in probably would be orange by about 6%. And the most any seed decreased by is blue by about 6%. All the current arrangement of seeds in the seed queue would still be possible, the biggest difference would be that players would waste less seeds trying to play well according to any sub-objective they pursue.
There are two sides to this gameplay balance problem. As suggested above, we can change the probabilities of the seed queues so that the current gameplay strategies feature less waste and more useful seeds. Or we can consider changing the types of challenges in the gameplay so that the less useful seeds becomes more useful. There are many ideas discussed on the “Beyond” section of this site that would give the player new mechanics and strategies and ultimately shift the relative value of each seed. From sculptures, to floating seed pickups, to folded fleeting world challenges, all the details of these ideas can be found in the Beyond section. And some of the ideas will soon be playable in the expansion of Starseed Pilgrim so you can try them out yourself. I’m particularly curious to know what players think of the beyond-ideas before and after playing them.
by: Richard Terrell @KirbyKid
March 3, 2014