Halo Wars: An Enduring Monument to Frustration

That this game was mishandled from the first cannot be for one moment doubted: one has only to look at how few players remain on the servers even at peak hours. As someone who has been heavily invested in this game, or more exactly, in the online multiplayer (I count some 2500 games to my record across 5 separate tags), I believe I am well within my right to draw attention to outstanding negligence on several key aspects of this game.

In the first place, this game’s single greatest defect is in fact no defect of the game at all–I refer to the matchmaking system which was implemented here, and for whatever reason, never fixed nor updated. Now 85% percent of my games have been in the 3v3 Standard Hopper, and it is precisely here that these failings are most apparent. My single biggest complaint is the absolutely arbitrary formula wherewith players are matched up and divided into teams. The result is that approximately one in every three games is effectively moot afore it has even begun, owing to the fact that the teams are, from a perspective of player skill, severely unbalanced. Thus it is that one often sees three high ranking players pitted against three players entirely new to the game, or what is even more offensive, one high player hobbled by two inexperienced bunglers who proceed to ruin the match for the entire team. I cannot understand why the matchmaking system, which utilizes a strict skill rating (whereof the actual rank is not indicative, being itself only a calculation of how many total points the player has accumulated; this quota has also been lowered repeatedly because the developers neglected to split the experience rank in Deathmatch games from those in Standard matches) should constantly and reliably produce such atrociously unbalanced games. My suspicion is that the low population at all hours forces the system to take the first comers into the hopper and organize them into a game, yet it still is inexcusable that the teams are nonetheless split according to (or so it seems) sheer caprice. Would it really have been so hard to devise a simple algorithm which averages the TrueSkill levels of all involved and generates teams based on the best possible distribution of this skill, so that, for example, two rank amateurs would be on opposing sides and not kill in the bud any and all chance for a fair game by being placed on the same team together? Or again, to separate three highly skilled players by pitting two against the one, the former being obliged to suffer the weakest player in the game as their third teammate? Such is not at all unreasonable. At any rate, this glaring problem was never redressed in the least, and I hold the developer in contempt for having designed such a bad system.

The second issue of concern rests with the TrueSkill system itself. This annoying little formula has caused players no small grief over the years wheresoever it has appeared, and the reason for this is that it is not at all transparent. I have seen the complicated mathematics that go into its calculation, and one is almost at a loss to make any sense of it at all even with the equations before him. Why this system was chosen is quite strange–I feel a fairer system in which it is not possible to de-rank (only to drop to the bottom of the highest established rank) would have been expedient. This would have spared many a player the frustration of losing his rank when his teammates abandon him twenty seconds into a match or when he is forced to quit a game due to intolerably bad connexion. On top of this, as Wars is a game at which many play today without much concern for rank, it means nothing to them if the other players happen to lose standing as a direct result of their indolence or impunity. And the crowning irony here is that the ranking system is nonetheless entirely defective, as I have already explained above. Moreover, the phenomenon of becoming ‘rank-locked’ has been well recognized and much derided, to the extent that the Leaderboards have had to be reset on multiple occasions. This shows a defect in the formula of TrueSkill itself, which is further exacerbated by the continuously low population on the game altogether. Why the duffers at Microsoft Research could not be bothered to correct this problem with their ranking system is beyond me.

Thirdly, I must voice a grievance on the fact that habitual quitting in this game was never addressed in any capacity. Why a simple system which banned from matchmaking for 15 minutes any player who repeatedly quit his games during a period of, say, three hours, could not be implemented, is really evidence of lethargy on the part of the developer. The excuse that the production studio was shut down shortly after the game’s release would have passed muster had the developer not promptly reformed itself and continued supporting their last game in decent capacity. A great deal of sense would have been found in replacing any player who resigned or disconnected within the first 60 seconds of the game with a low level A.I. bot, identical to those already programmed into the game’s Skirmish mode. This bot would have been competent enough to erect good bases and produce useful units for his team, and could have been intelligently directed by means of the flash function, again similar to the Skirmish mode. Such a straightforward measure could have alleviated the strain of having game after game to compensate for absent teammates. I remind the reader that the most popular hopper by far has been and remains 3v3 Standard, and this is where these annoyances obtain the most.

(continued in the following post)

A word will be given on 3v3 itself. It was a gross oversight on the part of the multiplayer developers that 3v3 was given only three maps on which to be played. One gets sick very fast of having to go again and again to Exile or to Fort Deen. In the DLC package only one 3v3 map was added: to the best of my knowledge, this content is still today not free to the playerbase, and the odds that one will ever get to play on this extra map are something like 1 in 36.

Moving on, I wish to turn now from the shortcomings of the online system to those of the game itself. Ensemble, having had a good deal of experience with developing RTS games all throughout its history, did a fine job of balancing the units and of correcting imbalances and exploits in post-release patches. However, the worst omission of the code itself is to be met with in the pathfinding for larger units. Everyone is familiar with the irksome business of trying to get a column of tanks to navigate a narrow ridge or turn about face in combat with good speed. Vehicles were apparently designed to be employed either in low number or over very great spaces, for they regularly pitch into one another and block the way of their fellows. They also have extreme difficulty maneuvering around both friendly and enemy units. This is infuriating because, as any real tank commander can attest, movement is half the strategy involved in deploying artillery, and here one cannot get his tanks to proceed over open ground without them stalling at every degree of path adjustment. Even more bothersome is the condition that all obstacles on the map extend as it were a few meters in excess of what the eye apprehends they do, so that one aims his vehicles in such a way that they should just avoid a particular object on the map, and then returns to find them shuffling in place about the very obstacle.

I am not versed in game code and neither do I claim that it would have been a trifling matter to fix these issues, but more should have been done, perhaps to the effect that certain units could occupy the same space as one another and clip over so that they would be able to move more smoothly. Thus aesthetics would have been sacrificed a little to the purpose of better gameplay, and I think this is a fair trade, as one hardly cares anything what his units look like in the middle of a pitched battle. Aerial units do not suffer from this shortcoming in code, and as a result they are unbalanced versus the vehicles, because they proceed evenly over their ‘terrain’ and can even pass over and into one another, whereas the vehicles are obliged to reverse and swivel in place because their pathfinding is poor. The Scarab is an almost absurd example of this, for despite its obvious ground clearance it cannot step over a single group of infantrymen which come not up to the joint of its foot.

Several other gripes remain. The Arbiter character will, as is well known, sometimes glitch through the map and disappear when engaging enemy units with his rage power; this is especially common when he attacks aerial units, to the degree that a frequent tactic used against him involves pulling him over intraversable ground and forcing this glitch. Also noteworthy is the glitch where a downed base shield will instantly recover despite being under fire.

It took a great while for Robot to repair the Cryo-glitch, but in fairness the utility of this exploit was low.

The DLC for this game was creatively conceived, yet it fell short do to a lack of playerbase support. Unfortunately, these additional modes for which one paid money are nigh inserviceable.

Not displaying the each player’s Trueskill beside his experience rank was a lame oversight that should have been remedied early on.

I could go on, but these constitute the bulk of my objections. I have thoroughly enjoyed this game when at its best, but I have likewise had to endure a tremendous deal of frustration for issues extraneous to competitive play itself. It is pure fancy to imagine anything further will be done in the way of support for a game that is so fallen out of favor, but I felt I had to give voice to these points, lest they go neglected. I am disappointed that Robot did not do more for the finer points of what was certainly a worthy effort to convert Halo into an RTS experience.

While many of those things can’t be fixed unless an entirely new game was made, you do have some points. I believe people from Ensemble who now work for Robot said that they wanted more time to work on Halo Wars, but Microsoft wanted it out on a certain date. Ensemble basically had to rush finishing the game.

It is really a shame that this game will not likely be getting a sequel wherein these things could be improved or overhauled.

C’est la vie.

Today I won a 3v3 game in matchmaking and de-ranked.

Lovely thing, this Trueskill system.

I’m surprised this game continues to stay afloat after so much time.

> I’m surprised this game continues to stay afloat after so much time.

It is very suprising. Most thought it would be dead by now