Rogue Based Immitation Baseball: 7drl Game Jam Postmortem
Jon Trainor, Anthony Duggan, and I had a lot of fun participating in the 7-Day Rogue-like Challenge last week. Unfortunately both Jon and Anthony experience debilitating illnesses over the week so we weren't able to finish our admittedly ambitious goal in time for the Sunday deadline. So instead, we're planning to continue working on the game and make it available elsewhere. On the bright-side, this means we will have a much more balanced game for you to play. In the meantime, here's a minor postmortem on how the game jam went and where we stand now.
Here's the summary from the design doc we're using:
1) Fielding Phase: Originally after each hit we would switch to fielding view, and a simulation of all the fielders would be depicted. When one of the fielders got the ball, the player would then have to decide who to throw it to. This was taken out because of bad ROI, and because there really weren't meaningful decisions to make. There is generally only one, fairly obvious, correct person to throw the ball to in a baseball play.
Instead, player behavior will still be simulated in the background based on those fielder's stats and the trajectory of the ball, but not depicted. Instead, the results of the play and hints about the fielders stats will be conveyed in text. We though this would fit the rogue-like genre better anyway.
2) Multi-Turn Pitching: The Batting phase lets a pitch play out over several turns, allowing the batter to adjust their aim based on their perception and reaction stats. Originally the pitching phase also took place over several turns with the player getting to adjust the balls trajectory. We decided this was inappropriate for the theme (because it that's not how throwing things works), would be complicated to present to the player, and only gave an illusion of gameplay. If you could control the trajectory midair, the optimal strategy would always be to be as erratic as possible to confuse the be batter.
Instead, pitchers will now choose a pitch type and and aim their pitch. They'll have corresponding stats for each pitch type. This emphasizes understanding the stats and matching pitchers to batters, which is more of the kind of game-play we were going for.
Once the jam started, Anthony and Jon got to work getting the core backend systems working. It was slower than expected (which is always expected). Anthony spent some time learning the Haxe toolkit and Flixel game library, while Jon spent a lot of time setting up the data for things like items, events, seeds name generation.
I went to work making the concepts and assets for the visual presentation, as well as designing the flow of the game phases. Here is some of the art we have.
We also already have some great retro sounds and music from Graham Southern. I especially like his minor-key rendition of "Take Me Out to the Ballgame" for the enemy teams.
Were do we go now?
From here on out, our work will be getting the presentation right and doing lots and lots of play testing to balance all these stats. Anthony and Jon are working hard to get a prototype working and I'm making more concepts to help guide them. We expect to have something to share soon.
Anyone interested in play testing?
Here's the summary from the design doc we're using:
Summary
Rogue Based Imitation Baseball, takes the great American past-time and gives it an RPG twist with turn based decision making and Rogue-like randomization. The player must make it through an entire simulated season against a procedurally generated league without losing a single series.
Gameplay
The player alternates between Manager Mode, where they analyze players stats, equip items, and set their rosters, and Player Mode where they play as pitchers or batters in a turn-based battle of wits. They must leverage their strengths and out-smart their opponent.
Mindset
The player is focused and analytical. They are trying to glean clues about players on the opposing team from anywhere they can (behavior, flavor text, etc), and make decision that maximize their chance of making successful plays. They should also be thinking long term
We had a couple long pre-jam meetings to discuss the game design. Jon had a very robust concept going in, so our meeting consisted mainly of fleshing out the feature he had imagined, and pairing down to what features would offer the kind of game play we were looking for and offer the best return on investment. Some examples of features we decided to exclude:Target Audience
The target audience is nerdy fans of the Rouge-like genre and fans of ultra-retro text-parser graphical action adventure games. It should be especially appealing to fans of baseball since the game is designed to incorporate legitimate real world baseball strategies.
1) Fielding Phase: Originally after each hit we would switch to fielding view, and a simulation of all the fielders would be depicted. When one of the fielders got the ball, the player would then have to decide who to throw it to. This was taken out because of bad ROI, and because there really weren't meaningful decisions to make. There is generally only one, fairly obvious, correct person to throw the ball to in a baseball play.
Instead, player behavior will still be simulated in the background based on those fielder's stats and the trajectory of the ball, but not depicted. Instead, the results of the play and hints about the fielders stats will be conveyed in text. We though this would fit the rogue-like genre better anyway.
2) Multi-Turn Pitching: The Batting phase lets a pitch play out over several turns, allowing the batter to adjust their aim based on their perception and reaction stats. Originally the pitching phase also took place over several turns with the player getting to adjust the balls trajectory. We decided this was inappropriate for the theme (because it that's not how throwing things works), would be complicated to present to the player, and only gave an illusion of gameplay. If you could control the trajectory midair, the optimal strategy would always be to be as erratic as possible to confuse the be batter.
Instead, pitchers will now choose a pitch type and and aim their pitch. They'll have corresponding stats for each pitch type. This emphasizes understanding the stats and matching pitchers to batters, which is more of the kind of game-play we were going for.
Once the jam started, Anthony and Jon got to work getting the core backend systems working. It was slower than expected (which is always expected). Anthony spent some time learning the Haxe toolkit and Flixel game library, while Jon spent a lot of time setting up the data for things like items, events, seeds name generation.
I went to work making the concepts and assets for the visual presentation, as well as designing the flow of the game phases. Here is some of the art we have.
We also already have some great retro sounds and music from Graham Southern. I especially like his minor-key rendition of "Take Me Out to the Ballgame" for the enemy teams.
Were do we go now?
From here on out, our work will be getting the presentation right and doing lots and lots of play testing to balance all these stats. Anthony and Jon are working hard to get a prototype working and I'm making more concepts to help guide them. We expect to have something to share soon.
Anyone interested in play testing?
