Early Access #1 - Metrics and Feedback

1. Overview

Recap Video: Sparkball - "SO FREAKING FUN" - Early Access Weekend Results

From June 30th to July 3rd, Sparkball held its first ever Early Access Weekend in an effort to prove out the depth and viability of its core loop.

We had 3,200 total players, averaged ~1,800 DAU each day, and peaked around 300 CCU, maintaining a steady ~150 CCU for nearly the entire event. Our servers were live for approximately 72 hours, here’s the stats:

• Top 500 Players: 35 Game Average (12 Hours)

• Top 1000 Players: 22 Game Average (7.5 Hours)

• Top 50% of Players: 15 Game Average (5 Hours)

• Top 20% of Players: 29 Game Average (10 Hours)

• Top 10% of Players: 44 Game Average (15 Hours)

• Players (1,717) that played at least 3 games played an average of 14 Games (4.5 hours)

Sparkball is so fun and addicting that our players played a game in Alpha with only 8 playable heroes, zero account progression, numerous gamebreaking bugs, and imbalanced matchmaking a remarkable amount in the short time it was available. Many described similar experiences to the tune of “I haven’t had this feeling since I first played League of Legends 10-15 years ago” or “I logged in to check it out for fun, not expecting much, and ended up with 30 hours played”.

2. Streaming and Watchability

Sparkball garnered a large amount of organic interest solely due to its watchability. Over 60k hours of Sparkball were watched in a 3 day span with only ~$8,000 in influencer/marketing spend. This includes being picked up by top LATAM streamer YoDa who noticed it on a friend’s stream and decided to play for 3 hours straight (for free!).

This reach even extended into Japan (despite zero localization and no servers in Asia). Streamer Shinji played for several hours with 250 ping (and went UP in viewers), including a super fun dev match to cap off the weekend which culminated in our favorite clip of all time: https://youtu.be/pA8lErcD2HM

3. Player Feedback

Raw Data (including long-form player comments) - Sparkball Early Access - Feedback Form Responses.xlsx

4. Key Takeaways

• Hunting for the perfect match: There is absolutely nothing like a truly competitive match of Sparkball. When teams are balanced, everything clicks and it’s the perfect blend of teamwork, individual fighting, and ball scoring. With the disconnect bugs, uneven matchmaking, and flood of new players (i.e. fixable issues), these matches only came 1 in every 10 or so, but players continually queued up solely due to the rush when you got one. As we hone in and get this down to 1 in every 3 or 4, we’re confident you won’t be able to stop playing.

  • Anecdotally, I had a competitive match early in the weekend where every round went to overtime in this amazing back-and-forth swing, teams were perfectly balanced. My hands were sweating, heart was racing, adrenaline pumping. We ended up losing, but immediately after the game, nearly all of us went to Discord with some form of “OH MY GOSH THAT WAS INSANE”. It was hands down one of the top/most unforgettable gaming experiences I’ve ever had in my life and I believe the others could tell stories of similar experiences.

• Mass Appeal: Similar to the Invitational, we saw overwhelmingly positive feedback from players of all genres. Even players who historically don’t play games of this type find something to love, with one player stating, “I hate mobas, brawlers, and rocket league. They all felt like ingredients for something better but by combining them you have made a delicious cake.” Funny enough, our ONLY response below a 3 for “is it fun?” is from a player who exclusively plays casual games (and still played 30 matches of Sparkball).

• Going Global: English was not the first language for approximately 46% of our player base for the weekend. While this places an obvious importance on future localization, this shows us that users were willing to play Sparkball despite not even being able to read ability descriptions/tutorial text. (Playing Lost Ark in Russian vibes!)

• Depth: The primary concern about Sparkball was similar to that of games like Battlerite and Omega Strikers, does it have the depth to sustain players beyond just an initial week? Our games per player metrics state a resounding yes, but in addition, we surveyed users if: on release, they could see themselves enjoying Sparkball for a long time over similar competitors. Results out of 5: LoL (3.9), Rocket League (4.2), Battlerite (4.3), Omega Strikers (4.4), Fangs (4.4). These results come from users who main/have thousands of hours played in these titles (LoL especially) so it paints a clear picture that we have potential to dethrone the kings.

• Fun even when you lose: We had multiple cases of players with a ~25% win rate who still played 50+ matches. Given the average match time and never feeling helplessly behind, players continue queueing regardless of match history. Our Art and VFX Directors (Alvin Lee and David Hall, who both worked on LoL but rarely played it) were both bottom 10 in MMR after 50 games and we had to actively tell them to go to bed because they were having too much fun and wouldn’t stop playing.

• Character Attachment: By the end of the weekend, players were already one-tricking and making memes of their favorite characters.We also held an event promoting the two team captains, Glo vs Roo. The attachment/love around something as simple as “what team are you on?” was amazing to see.

5. Lessons Learned

• Tech Issues: Sparkball needed another few weeks in the oven to properly iron out tech issues, namely around matchmaking, reconnecting, and game crashes. Crashes included a server crash that would end your game the moment before you score to win. Reconnecting resulted in countless 5v3 or even 6v2 games. That said, players repeatedly stated that if this was any other game, they would have quit in frustration, but they continued wanting to play Sparkball regardless.

• Matchmaking: We allowed 4-stack teams in the interest of letting players play with their friends. We also announced a small tournament for the top 50 players coming out of the weekend. In addition, in order to keep queue times low, our matchmaking rules were fairly lax. Top player 4-stacks quickly rose to the top and subsequently got matched with lower tier/new players, completely ruining their first game experience. Deep into the weekend we restricted 4 stacks above a certain MMR which led to DRASTICALLY higher quality games, but it’s clear we need a clear focus next event on prioritizing competitive play or fun/new player growth, not both. We’re pleased that so many players stuck around despite getting stomped their first few games and firmly believe that retention/feedback would have been even better with a more reliable/easier new game experience.

• First Game Experience: The average win rate across our players who only played 1 match is ~25%. This also doesn’t account for the high possibility that their first match was a 5v3 as well due to bugs. This unfortunately leads to lower metrics as those players justifiably quit early, but we do see this as a solvable problem with not only the matchmaking/bug fixes mentioned above, but also rolling out players vs. AI mode.

• Timing: Early Access took place the same weekend of July 4th, LoL’s new game mode release, Ibai’s streaming event, Twitter outage/throttling, and Theorycraft’s Loki embargo lift (which did draw comparisons). We didn’t expect massive reach, but both player and influencer availability time was severely limited. We reached out to a few players who had a high number of games played on Friday, but didn’t play again the rest of the weekend and every response was directly related to unavailability.

6. Audience

Player breakdown by region largely followed the amount of influencer engagement in that region, namely LATAM/Japan. The weekend featured zero localization/translation, meaning ~50% of our players couldn't even read anything happening yet still played a large amount.

While we knew we’d do well in LATAM, the feedback from this region largely was focused on the futbol aspects, while the US was focused more on the combat elements (with Japan sitting somewhere in the middle). Our favorite description of the game came from YoDa who described the game at “FutLoL” (Futbol + League of Legends), but we’re thrilled that the game has something for everyone to latch onto.

Worth noting that many users asked specifically about playing this on mobile and we believe the transition from PC to Mobile for Sparkball given the game type and control scheme is an easy win for the future!