About a month ago we overhauled the main gameplay map in Revn, “City Map”. This overhaul was shipped in Update 0.6.8.4614. We analyzed matches for the month prior to Update 0.6.8, and compared this data to an analysis of matches after Update 0.6.8.
Our goal for Revn is to have matches that last about twenty minutes. This was one of the first metrics we analyzed. Prior to 0.6.8, the average match was 26 minutes and 10 seconds. This usually meant that games felt drawn out and that ending the game was a relief instead of an intense final showdown. After Update 0.6.8, the average match lasts 18 minutes and 16 seconds. Not only did we reduce match times by nearly 8 minutes, but we are much closer to our goal of 20 minutes. Players have recently described the final minutes of a match as “climatic”, which is certainly a good thing!
Average Metal per player is another useful statistic to look at. Before we reworked the map, the average was 8644 Metal per player per match. This is enough to get all of a player’s items and abilities to Level 4 in most loadouts, but not quite enough to get to Level 5 with every item. The average threshold for Level 4 is 7,950 Metal, while the average threshold for Level 5 is 11,750. This seems to be a good target, because players will almost reach maximum power, but they will always have a small handful of upgrades they can continue making in the final minutes of the game.
After Update 0.6.8, the average Metal per player dropped to 7031. This is 1613 Metal lower than the previous month of data. It also falls below the range of 7,950 – 11,750, meaning players probably aren’t getting to Level 4 loadouts as often as before.
Before the update, 44% of player-match data showed zero kills. This means out of all the players in all matches for that month, 44% of them didn’t get a single kill. After introducing the map changes, we saw a 7% increase in overall players who got at least one kill per match. The statistic was lowered to 37%, which is still pretty high. We expect that as players become more skilled in the game, this stat will decrease.
Another contributing factor would be low team sizes. Our average playtest attendance is 3 players per team. With one Engineer per team, we can expect 1 out of 3 players to focus on building and providing support to his or her team. It wouldn’t be unheard of for an Engineer to go an entire match without securing a kill. As the player population increases, and we see more 5v5 matches, perhaps the statistic will naturally drop to around 20% (one out of five players).
Why is this number so important? We think that the game is more fun when more players are getting kills. It can be disheartening to watch a single player dominate the match.
We’ve also seen a 1% increase in the average number of kills per player per match. Its not much, but its something. However, if we ignore data from players who didn’t get any kills, and we look at the average number of kills per game per player, we’ve gone from 3.1 kills to 2.8 kills. This indicates that players who previously dominated matches are dominating a little less now (lower killstreaks, more even distribution of kills across the teams, etc).
Overall, this has been a great update for the game! We’ll continue monitoring match data to improve Revn, and sharing our findings with the community from time to time.