Current Public Version: 3.02
Current Developer Version: 15942

3.5 Blogpost #8: Various System Changes


Greetings! We're here today with another blogpost detailing changes for the next version of Project M, 3.5.


As in our 6th blogpost, today we'll be going into changes to the game's mechanics. These are changes that, while on their own may seem small and not significant enough for their own blogpost, when combined should make a big difference in gameplay. Similar to our Demo 2 blogposts, these are going to be focused primarily on competitive play. We hope to put out some more universally appreciable content in the coming weeks, so stay tuned!


Shield inputs have received a lot of love this time around. The stats on shields have been improved in every version of Project M to date, and 3.5 will not disappoint on that front. We have also finally managed to match shield pushback to Melee stats, which means that weaker hits will have a bit more pushback and stronger hits will not have as much pushback as they do in 3.02.


On top of this, Shield Automatic Smash DI, a significant aspect of shield play in past Smash titles, will return in the upcoming version. This means that when your shield is angled, you will move in the held direction a bit when your shield is hit. This allows for greater control over the situation after the hit, letting you move forward for a punish or move backward to create a little more space between you and the aggressor.


Top: No Shield SDI. Middle: SDI forward. Bottom: SDI backward.


One of the most requested changes, separate inputs for individual buttons, has been implemented, allowing a command to remain usable even if you are holding the same command on another button. This means that players can now wavedash with R while still holding the L-trigger (or vice versa), bringing back this comfortable input option from Melee. Other notable combinations are using two jump buttons for more flexibility on instant double-jump inputs, as well as allowing the Grab button to work even if Attack is currently held.


We have also looked at previous techniques that seem to benefit some characters more than others and thought about how to make them more available for everyone. Footstooling has grown to become a powerful, if situational, option for continuing combos and landing low-percent KOs on unsuspecting opponents. In 3.02, footstools were made to go through shields to make it more useful for the cast as a whole, but it became clear that this heavily benefited a small subset of characters without much to show for most of the others. To remedy this, the ability to footstool through shields has been removed, but footstool heights have been adjusted on every character to allow for a greater balance of pressure and follow-up potential throughout the cast while preventing the guaranteed shieldpokes available in 3.02. This makes its use more accessible and intuitive while toning down the most drastic advantages.


Left: 3.02 behavior; Footstool goes through shield. Right: 3.5 behavior; Dedede Footstool Jumps but DK remains in his shield


The Dash Attack Canceled Up-Smash (often abbreviated as DACUS) exists as a way to offer extra range and movement speed to an Up-Smash. In 3.02 and previous builds the input window was tied to a character’s pre-jump squat timing (a largely unrelated attribute), where slower squats had a wider window and quicker squats had a smaller window, or even no DACUS window at all. We have decided to implement a universal two-frame DACUS window for the whole cast of characters, opening up new tools and making more sense from a general design perspective. Though there is now a globally consistent input window, the distance each DACUS travels is still primarily based upon a character’s initial Dash Attack movement, with minor adjustment on a per-character basis as needed. Additionally as a result of the independent button press coding, DACUS now has the same window whether the Up-Smash is inputted with Attack or Grab, a welcome addition to those of us who do so with Z.

With so much talk about individual character balance, this is a breath of fresh air that should move the game forward in a way that any character can benefit from. It is our hope that these changes will enhance the fast-paced experience you have come to know and love. We have always tried to push for smooth and exciting gameplay that is fun to watch, and we aim to bring that to the next level in 3.5!


As always, stay tuned to our website and social media outlets for further updates. Sometimes, not even we know when a blogpost will be posted!

-The PM Dev Team

3.5 Design Blogpost #7: Trimming the Fat


Another update for you from the Project M Development Team! We've been working hard on our next public release, and we'd like to talk with you about our design philosophy for Project M version 3.5.


As we have previously mentioned, we are aiming to tighten up the game to meet our project's design goals. In general, we are aiming to increase the depth of the game, while retaining the accessibility that the Smash Bros. series is well regarded for. At the moment, we believe our cast to possess tools that are a little too excessive in terms of flexibility, recovery prowess, punishment and more. This means that we are going through the cast and toning down tools and/or traits that may be seen as superfluous or detract from the game's complexity.


Our goal is to provide the player with enough tools to create a highly nuanced and effective playstyle, while providing opponents enough counterplay to punish mistakes to a high degree, necessitating thoughtful, challenging and engaging interactions with each other. We want to make sure that with skill, control, knowledge and creativity, a player can achieve phenomenal results and feel well rewarded as a result.


To provide you with an insight of the methodology we have been using, here is a general list of criteria that we are using in our cast-wide overview:

  • Revising moves with static knockback curves and making them scale more with percentage. This is intended to create a much more dynamic combo game that requires knowledge of percentages, fall speed/weight, stage positioning etc. and allows a player to develop with more time, knowledge and experience of the game.
  • Balancing cooldowns with power/start-up/duration/utility.
  • Tools that mitigate positional advantage (e.g. circumventing juggles, edge guards).
  • Tools that provide excessive burst movement without appropriate risk.
  • Moves that hit at unintuitive angles.
  • Moves with SDI/hit-lag multipliers on them that circumvent the strength of Directional Influence (barring grounded multi-hit moves).
  • Recovery potency (as previously described in our earlier blogposts).


Note that while these are general guidelines, expect some leeway to be made while considering what makes something acceptable.


As previously mentioned with our recovery blogpost, please note that this stems from design, and not necessarily balance. Our overview of the cast is a global one, and characters will receive due consideration regardless of current perceptions of their viability. As such, it is important to consider your character's changes within the scope of our goals for version 3.5, as opposed to viewing their changes in a 3.02 vacuum.


We are pleased with how 3.5 is shaping up and are very excited for its release!


There is more for us to share soon!
-The PM Dev Team

Turbo Contest Results


Hey everyone, on the 3 month anniversary of the Official Turbo Mode Competition, we are proud to announce the winners. Thank you to all who participated and enjoy the show.