Recently as part of my continued education four classmates and I formed Team Skeleton Cocoons. This team consisted of one designer, one artist, one producer and two designers including myself. Our fist assignment as a team was to create one prototype a week for three weeks. For our first prototype it was decided that we would make a two player dinosaur building/fighting game named Build-O-Saurs. The game was divided into three sages. In the first stage players would compete as archeologists digging up dinosaur parts and sabotaging each other. Then in the second stage the players would chose which parts they would use to build their dinosaur. Finally in the last stage players would take their created dinosaurs and us them to fight against each other.
For this project I was assigned to handle parts balancing and QA while the other designer tackled level and combat design. We were able to get a build done after the fist few days and contacted acquaintances of ours to come over and test. We were able to get around 8 people to agree to be testers which we considered to be a good sized pool for early development. Then a snow storm hit. Because of the storm our eight testers quickly decreased to two. We were unable to find anyone else willing to test because of the storm and if we waited for it to pass it would be too late to make any changes before the deadline. So we decided to go forward with the QA and closely examine any feedback they gave us. Upon testing one thing was made painful clear. Build-O-Saurs was horribly unbalanced.
The way bones worked in Build-O-Saurs was that the player was given a base skeleton by default and could upgrade it using collected bones. These would modify the dinosaur's stats which were split up into three categories health, attack and speed. Bones were of four types arms, legs ,head and torso. Each type of bone had three versions increasing one primary stat substantially while giving minor increases to the other two. Each piece of different types would increase stats in different ways with the torso giving the greatest boast and the arms/legs giving the smallest boast. When I had done my initial balancing I had assumed (foolishly) that each player would collect one piece of each type. However as we found during testing that is often not the case as it is possible to collect multiple of one type and not have enough to craft a full dinosaur.
I discussed this with the other designer and it was decided that we could modify the gameplay in time to make sure that each player could get a part of each type. Instead it was decided that I was to make each piece balanced enough so no player would be at a disadvantage no mater what pieces they and their opponent had.
To begin this task I fist set some rules to guide my balancing. First the arms/legs must be weaker then the head and the head must be weaker than the torso. Secondly balancing would be measured in hits it would take for the maximum stat dinosaur (a dinosaur with all attack parts) to kill the minimum dinosaur (a dinosaur with only a set of speed focused arms/legs) and visa versa. I was my goal to get both of these numbers within 20 of each other. I determine that this would provide an advantage to a player with the better character while still allowing the weaker one to have a chance.
To first accomplish this I reassessed how my stats relate to each other. I had already done a great deal of work to make sure that speed and attack were balanced to each other 1 to 1. So I turned to health to be the key change. After few rounds of in sheet testing I determined that setting health to be a 4 to 1 ratio to attack and speed was the correct ratio to get matches to the length we desired.
After balancing health I moved on to the bigger issue of balancing piece types. I started by reducing the values of the head and torso while upping the values of the legs and arms. However with the guidelines I had established it proved impossible for the two dinosaurs to be evenly balanced. This is because no mater how close I got the numbers together the larger dinosaur would still be substantially more powerful than the smaller one.
Realizing the purely stat balancing would be a useless task I looked at the availability of the parts. My fist instinct was to make the more powerful pieces rarer in order to compensate for their value. However this simply exasperated the problem and made victory rely far to heavily on luck. The solution i finally found was to go in the opposite direction. Powerful pieces would be the most common while weaker pieces would be rare. This would ensure everyone starts off with a good base while enabling smaller parts to tip the scale. This enabled parts to have a impact on game play while not putting the weaker player at a insurmountable disadvantage.
Comments