Making a Better Ranking System for Halo 5

I was surprised when I heard Halo 5's matchmaking implemented a ranking system with skill groups. The system allow for public players to battle against opponents of similar skill, to have an incentive to focus on team play, and to sport gaudy icons on their profiles as bragging rights. However the matchmaking system has been subpar. 343 said it would prioritize quality of matches (matches with evenly skilled opponents) over brevity of queues, but this has fallen flat as better players get queued against fodder constantly.

The major problem is the ranking system's lack of factoring in one's individual performance within matches. It only accounts for a player's win and losses. This lack ensures that the system does not accurately measure a player's skill. It allows for bad players to be carried to higher ranks despite a lack of ability and for bad teammates to harshly detriment good players. Lastly it gives high ranked players the same reward no matter if their opponent is a similar rank or a lower rank. Such a system is unfair in the way it matches people up, rewards people, and estimates their skill.

A better ranking system for Halo 5 works as follows. Wins add matchmaking rating (abbreviated MMR from this point on) while losses subtract MMR. In order to rank up, one must gain enough MMR to pass the threshold into a higher rank, ranking up and acquiring a neutral MMR (meaning that one loss should not immediately drop the player back below the threshold and derank him). In practice, a win will determine that a player will be awarded MMR, while individual performance determines just how much MMR will be awarded. For example, a player who won and played very well will gain a lot of MMR while a player who won and played miserably gains very little MMR. Losing a match functions similarly, with a player's performance determining just how much MMR he will lose. Factoring in individual performance allows for a larger pool of data related to a player's skill thus better measuring said player's ability.

The criteria for determining performance is the most important part of the ranking system. A good ranking system measures the impact a player has on the completion of a game type's objectives. Thus the system rewards impact players and punishes players who avoid the objective.

Commendations are the best possible measure of a player's individual performance related to the completion of a game type's objectives. They already exist as a superficial reward system within Halo and their use as a metric to influence one's rank is natural. But in order to value play that completes a game type's objectives, there should be tiers of commendations that net more MMR than others. High tier commendations will net the most MMR and are only awarded to impact players playing the objective. In Halo 5's Capture the Flag game type, several commendations are awarded for actions within the overall rule set of capturing the enemy's flag and defending one's own flag.

First tier CTF commendations in descending order of impact are capturing the enemy flag, killing the enemy flag holder, and returning the team's flag. First tier commendations involve directly completing the game type's objective and thus have the highest MMR reward based on their impact. Second tier CTF commendations in descending order of impact are multi kills and spree kills. Second tier commendations do not directly complete the objectives, but make them easier for one's teammates to complete them. Thus gaining second tier commendations means that a player impacts a match more indirectly by keeping opponents dead and respawning. This tier of commendations nets less MMR to push players to either kill an absurd amount of opponents or to play the objective and directly influence the outcome of the game. Third tier commendations would net no MMR as the game is rewarding players for being in the fray without actually completing any of the objectives directly or indirectly. Distraction is an example of a third tier commendation.

The ranking system should be more fleshed out beyond this, such as giving a higher weight to wins against higher ranked opponents and lower weight to wins against lower ranked opponents. But the most important aspect of a ranking system is to make sure that performance is factored into wins and losses. Impact players should suffer lower penalties to MMR on losses and gain more MMR on wins. Such a system takes in much more data regarding a player's skill, making it more fair and accurate in determining ranks. My proposal creates better quality matchmaking between more evenly skilled opponents, leading to more fun and less frustration in competition.

The Analyst's Resume of Play

The Analyst's Resume of Play

Game analysts and critics need exposure to high level play in complex games. Contact with skilled play opens up one's palette to the possibilities within the medium. But more importantly, it allows writers and video makers to pass on the appreciation of a game's zenith to their audience. During my late teens and early twenties, I've played moderately well in public matches of Street Fighter 4, spent 100 or so hours mastering PvP in Dark Souls, and soared to the upper echelons of public play in Counter-Strike: Global Offensive. My experience with these games exposed me to something every gamer should seek: high level play and great amounts of execution skill (i.e. one's ability to execute one's intention in game, such as hitting a headshot from across a map or performing a combo with a very limited frame window).

Analysts should seek to master movement, aiming, and countering opponents to better understand and appreciate games. Such mastery is paramount to those who seek to pass on this appreciation to an audience. The only way to discuss gameplay well is to be literate in high level play. Without said literacy, analysts will pass on the shallowness of their understanding to their audience. The possibilities of gameplay explored and discussed in games will be severely limited. And if everyone is satisfied with more shallow games, then there will be no demand for more deep and complex games.

There's more to games than just the mechanics, but it's the meat. Without a good palette, we end up with videos like this from PC Gamer, in which the presenter has no idea how to discuss the CSGO gameplay beyond stating that a player has good intuition or that they don't reload because they are really skilled. He routinely makes obvious or shallow statements that do not fill in the audience about what is happening on screen.

PC Gamer seems to be making a push towards more serious content recently. They just published a Top 10 CSGO Player article from Lurppis, ex pro and former writer for hardcore Counter-Strike site hltv.org. PC Gamer has a more casual audience than hltv.org, but the right person could still bring good analysis to the level of a general audience. Traditional sports viewers cannot stand obvious or shallow commentary from their analysts and casters. We shouldn't stand for it either.