A bunch of random things to discuss this week…
First and foremost, we have another alpha player survey online. There are questions about skins/emotes, unit balance, the armory card-flipping minigame, and more. Please fill it out; your responses are extremely helpful!
We collect a huge number of statistics on Prismata units—how often they are bought, how often players win with them, and how balanced they are. In this article, I wanted to share a few of our findings.
Player 1/player 2 balance
Across all ladder games played since our last balance patch, we found a player 2 winrate of 51.3% (ignoring draws), which is close to our goal of 50%. This winrate is relatively consistent across all skill levels, with a slight increase to about 51.5% among tier X players.
This is down from over 52% a few months ago. The improvement is likely due to a number of balance changes that helped reduce player 2 advantages in certain sets. (Note, however, that we also made adjustments to units like Odin and Drake that previously favoured player 1; the adjustments to player 2-favouring units like Cynestra and Tesla Coil simply had a greater effect.)
Though we still have some work to do and plan several additional balance adjustments in the future, we’re nonetheless very pleased with these results. For comparison, the red vs blue balance in League of Legends hovers around a 55% split in favour of the blue side in top level games.
The most-bought 1-supply units
Here are the purchase frequencies for each of the 1-supply units in Prismata. These numbers give the chance that a single player will purchase the given unit, and are obtained by averaging the player 1 and player 2 purchase frequencies of that unit over all games in which that unit is available.
|Percentage of the time that a unit is bought by a given player when it is available:|
There weren’t too many differences across skill levels, but we did observe a few trends while studying the play of elite Prismata players. In games where both players were in Tier X and had ratings over 1800, players bought:
Fun Fact: The most frequently-bought non-base unit is Doomed Drone. Combining the purchases of both players, an average of 17.9 Doomed Drones are bought in each game in which Doomed Drone is available. In games where both players are rated above 1800, this increases to 23.9 Doomed Drones per game. Players rated over 1700 buy at least one Doomed Drone in more than 99% of games in which it’s available.
Case Study—Vivid Drone
Normally when we see statistical imbalances this large, we study the unit to death and ultimately uncover opening problems or other issues. These issues then prompt us to apply balancing changes to reduce the advantage that a unit might provide to one particular player.
However, Vivid Drone is a special case.
After a tremendous amount of research (observing Vivid Drone games, analyzing Vivid Drone sets, and delving deep into the statistics), we discovered that player 1’s advantage in Vivid Drone sets is almost entirely due to player 2 having more opening choices, and consequently choosing a suboptimal choice more often than player 1. In fact, the winrate gap shrinks away almost entirely when only considering games featuring top Tier X players.
I’ve spoken to this issue on Reddit on prior occasions, but today I’ve prepared a video showing some examples of player 2 Vivid Drone openings. If you struggle with Vivid Drone as player 2, it should provide some helpful tips:
Now, as I mentioned on Reddit, the balance of units like Vivid Drone is tricky from a design point of view. If we could adjust Vivid Drone’s balance to help out player 2 and bring the winrates closer to 50/50, it’s quite possible that we would inadvertently create a situation where player 2 had winrates above 50% after selecting the right opening. We don’t want that either!
Often, when balancing a unit, we have to choose between:
- Balance that leads to a 50% winrate.
- Balance that leads to both players having equal winning chances if they both play optimal openings.
Units like Vivid Drone, which offer drastically fewer opening choices for one player compared to the other, are precisely those units for which these two methods of balance lead to substantially different results. Not everyone is going to choose the optimal opening, and players who choose the wrong opening drag down the average winrate for player 2 quite substantially.
So what’s the plan? Right now, our strategy with Vivid Drone is to wait a little longer to see if the community can adjust and learn to play a bit better as player 2 in Vivid Drone sets. We believe that we’ve given both players enough tools to do well in Vivid Drone games, and that it’s up to our players to make the most of them. I hope the video helps a little bit!
Statistics are one of the many tools we use in balancing Prismata, but ultimately, our goal is to make Prismata the best competitive strategy game it can be—not necessarily for every statistic to be 50%. The most important input we use in balancing the game is your feedback, which is why you should go FILL OUT OUR SURVEY right now if you haven’t done so already!
See you next time!