Gameplay Balance for 1.29: Part 1 – Fixing Gameplay
Hello everyone! If you missed them, we’ve already written two articles that should start to warm you up for the massively awesome and super-improved patch coming down the pipe. One was about general coarse tuning, and one was about the huge Tutorial improvements we made. There might also be another article soon about some visual art improvements we’ve been making.
For a timeline, we plan to get this patch submitted this week, and usually iTunes takes about a week or so to review changes.
Today’s post will be a bit of a deeper dive into some of the issues we’re fixing with balance. This is only part 1: we’ll be releasing a second part detailing our metagame changes later this week (edit: it’s released!).
Master Mode Problems
The exploration into our problems came from watching a lot of Master Mode play. Our best players were playing at ranks 12 all the way to as high as 15. This isn’t a problem in and of itself, but what was a problem was the way players were getting so high. In version 1.28, it’s optimal to play Auro very defensively. In other words: you should basically never take more than a couple points of damage, get a few kills, and then get to the exit as quickly as possible. Repeat as many times as necessary to win.
This pattern of play results in two realities, both of which I consider to be very bad:
- The matches are much too long, just in terms of time
- The matches are pretty low-tension.
Master Mode Auro matches were actually sometimes painful to watch – top players who should be doing the coolest stuff were just doing similar, almost rote patterns to avoid taking damage and get to the end of the level.
In short, it just wasn’t working. And while you might say, “well, that’s just Master Mode, it’s not really a big deal if only the best players have that experience”, I disagree.
Firstly, we don’t want any players to have a bad experience, and there is no reason why we’d exclude our biggest fans from that. But also, if the Masters are discovering a pattern of optimal play at their level, that pattern actually is optimal at all levels of play. The fact that the lower-level players haven’t consciously realized that this pattern is optimal does not mean that it’s having an impact on the way they are playing. In other words: while the problem becomes most visible in Master mode, it exists at all levels of play.
To clarify: this is not some design-wonk esoteric thing that affects a tiny percentage of players. This is fixing fundamental balance issues.
So the first step was to solve this problem of defensive play. We did this by re-arranging how the scoring system works. Most of you probably know that in the current version, there are four “combo meter” levels: 1 pointers, 2 pointers, 3 pointers and 4 pointers. Each level has its own number of “pips” – segments that make up the full ring:
- 1 pointers: 2 segments (technically, it was 1 big segment, plus zero segments counts as another)
- 2 pointers: 8 segments
- 3 pointers: 4 segments
- 4 pointers: 2 segments
In short, this arrangement was… actually kind of all wrong. Firstly it allowed players to “bottom out” way too easily. Once a player actually hits that bottom of the 1-pointers, they have no incentive to get monster-bumping again – it’s optimal, at that point, to pick up all the candies and Power Tiles you need and go to the next level / start the next battle at your leisure. This is huge problem #1.
The other big problem is that the difference between the “good points” (3 and 4) was too significant compared to the “ok points” (2). That’s +50% and +100% better, which ends up making the game pretty swingy. If you can get a huge Abomination or Gale combo going in an area, your score gets super-deformed quickly. It’s impossible to balance around that – it just makes the potential score ranges too large.
Defensive Play Solutions
The first of two major changes we’ve implemented is that we changed the combo meter points/pips arrangement. The point values are now 1-point, 3-point, 4-point and 5-point. This means that 4 and 5 pointers are far less significant – +25% and +40%, respectively, a huge drop from the previous 50/100. So that means that, while getting those glamor points is still of huge value (+40% more points is really nothing to sneeze at), we won’t have as much of a massive score deformity issue as we had before. We should be able to balance around this.
It also means that from the “ok” pointer, the “penalty” pointer is now even more severe (33% of the “ok points”, down from 50%). And speaking of making the penalty more severe, the new “pip count” for the combo meter is:
- 1 pointers: 8 segments
- 3 pointers: 8 segments
- 4 pointers: 6 segments
- 5 pointers: 4 segments
The 4 and 5 pointers might seem to have a lot of segments, but we’ve added a new rule: at these levels, each turn depletes your pips twice as fast. This makes it pretty difficult to maintain 4 pointers for very long (you basically have a max of 3 turns living there) and usually about one, maybe two turns at the 5 pointer.
You might notice that the total number of segments, from top to bottom, increased from 16 to 28. In addition to this, we also made each kill give you 5 segments (up from 4).
Overall, these changes “slow down” the movement across the spectrum. It’s now a little bit harder to go from the bottom all the way to the top, and it’s much harder to completely bottom out.
One thing we realized is that if we “cap” the number of stages somehow, that means that “high point values” are really valuable. If you only have, say, 30 monsters total that you’re going to face, then you really, really need to make sure you’re not getting 1 pointers, or even too many 3 pointers.
For the number of stages, three feels right to me. We want game lengths to be as short as possible – everything should be as short as possible – but with fewer than three stages we feel like the Barrier system actually isn’t even working correctly.
We played with the idea of just literally ending the game when you hit the level 3 teleporters, but ultimately we decided to go with a “soft cap” instead. Level 4 exists, but it is too hard. I’m talking like “holy crap, are you kidding me” hard. The idea for level 4 is, you’re going to die here, but before you do you might be able to eke out some points. In fact, if you play excellently, you could even squeeze out a solid 10-20 points out of this level since the monsters are so densely packed. But not much more than that, and often no points at all will be made on this level.
So that means you now really have to make stuff happen on levels 1-3.
In our testing, these changes are… TOTALLY AWESOME. The game is now super tense. Optimal play went from:
“hang back, aggro one monster at a time, never take any damage, and eventually win after a 25 minute game”
“DIVE into the fray, aggro too many monsters, and then use every trick in the book to kill them all quickly. Almost always this involves taking damage. Then as quickly as possible aggro another huge pack of monsters. Repeat!”
Dilly-dalling really hurts now, and this is what we’ve wanted all along. Of course, we understand that new players aren’t likely to play that way, so we are taking measures to ease them into it (we’ll talk about that in Part 2). And make no mistake: 1.29’s Easy Mode will be easier than 1.28’s! Don’t think that we’ve simply made the game harder. We’ve really just made it better, although in that process it has become hard (at all) again for master players, who previously had more or less solved the game.
Before we finish, I have to give a shout-out to our amazing beta testers who helped make these changes over at the Dinofarm Forums: Nachtfischer, Vivafringe, Senator, Major_Shiznick, Lemon, Disquisitor Sam, Blox and a few others. Thanks, you guys – we really couldn’t have done it without your help.
In Part 2, we’ll talk about the lower end of the spectrum and what changes we’ve made there, as well as changes for the metagame. Thanks for reading!