User:G-Mang/An Explanation of Competitive Preferences

From Team Fortress Wiki
Jump to: navigation, search
This is a copy of a Steam Forum topic I made in March '09. It's archived here because forum topics are deleted when they've become inactive for too long.

This post is about the reasons why competitive games generally have Nocrit enabled and Nodmgspread on. Along the way, it also explains why competitive games tend to follow certain patterns in terms of format and team composition. It's meant to be informative, not inflammatory or controversial. If this post seems familiar, it's because parts of it are derivatives of a reply I made to a topic some time ago.

An Explanation of Competitive Preferences

Preliminary Clarification

Before I go into why nocrit and nodmgspread features are used, I thought it appropriate to address what the features actually do, because I've seen a number of misconceptions about them.


Nocrit is a server cvar (basically a server option) performed with the console command "tf_weapon_criticals 0". Originally, it was accomplished with a third-party modification, but it's now done by a built-in cvar that Valve added to TF2 on February 28, 2008. It turns off all random-chance criticals, but none other. This means that weapons like Backburner and K.G.B. still have crits when appropriate, and the Kritzkrieg will still grant crits to the healed ally in full. Perhaps worthy of mention is the fact that it is quite true that this setting makes it so having no random crits on a weapon is no longer a cost at all (which could be seen as a "relative buff" to the non-random-crit weapons: blutsauger, medigun, kritzkrieg, backburner, knife, sapper, sandvich, and sniper rifle).


Nodmgspread is a server cvar that eliminates the random damage of weapons with the console command "tf_damage_disablespread 1". Similarly to Nocrit, it required a third-party mod to implement in the past (known as TFTrue), but as of the February 2, 2009 patch, it's a Valve-supported configuration. As you may or may not know, weapons do not have completely predictable damage: each weapon has a range of damage it can randomly deal in any given hit. For any given shot fired, the base damage is randomly chosen in the range of this: "weapon damage - 10% < x < weapon damage + 10%". The Nodmgspread cvar works by reducing that range to 0% in both directions (or, at least, it's supposed to, though it apparently still has a tiny range of a couple percentage points), and then applying any ranged damage falloff and splash effects when the hit occurs. This isn't really a buff or nerf to anything in a completely objective sense—it just averages things out. Some argue that guaranteeing middle damage is a nerf or buff for specific weapons; for example, if a weapon could one-shot-kill certain classes with random damage on, but not with random damage off, it could be argued that such a change gives the weapon less potential, even if it deals equal damage overall. With damage spread having been changed from 25% to 10%, however, the effect of Nodmgspread is relatively minor in terms of class balancing.

Reasons for the Competitive Format

There are two main reasons why Nodmgspread and Nocrit are used in competitive games. The first is the “standard competitive format” used for match gameplay. While there are plenty of leagues that break out of the standard format, it is by far the most common in TF2 competitive gameplay (with some slight variations by league), and there is logic behind it.

The standard competitive format is a six-versus-six match with class limits. In almost all such matches, each team is only allowed at most one medic and one demoman, with other classes generally limited to two slots each. Each team stays in communication through microphones in an external voice chat program (such as Ventrilo or Mumble). A twelve-player game of six-versus-six is half the size of default TF2 gameplay, which has twenty-four slots, and the initial reaction by most is that twelve players is too small a game. The fact that the teams are small, however, is very much intentional. There are a variety of reasons for this.


The most objective reasons for 6v6 class-limited gameplay are the logistical ones. Perhaps the most obvious of these is that it's easier to get six people dedicated and working together on a team than a larger group like the default twelve. Getting a team much larger than six players to show up regularly for scrims and matches, stay involved in strategy discussions, and remain happy and harmonious with each other is rather difficult, and most who've done it will agree that it can be hard enough even with only six. When team sizes get larger, like in 8v8 and 9v9 highlander matches, teams usually have to start using backup players or ringers (backups are players that don't practice with the main group, and ringers are players that aren't even on the team roster), or worse: sometimes they just can't get enough players to play at all and forfeit. When teams have to resort to backups or ringers, or forfeit completely, the game is less fun for everyone.

When a league has smaller teams, it also means the league has a greater number of teams because smaller teams are easier to form and maintain (and obviously take up less of the player population each), which helps to give league gameplay more variety and playtime. It's easier to get practice scrims against an opponent of your skill level on any given night when there are more teams in a league to choose from. It also means that if there is a problem with a team and they break apart, there are a variety of groups that those players (and players new to leagues altogether) can choose from to find a team that suits their schedule and preferences.

In competitive games, communication is very important, and having team sizes larger than six or so can prove difficult for coordination. If team sizes were the TF2 default of 12, working together in a Ventrilo or Mumble server would be extremely complicated because of the number of voices present and the area that the team would cover on the map. Alerting teammates of danger, explaining game plans, and coordinating ongoing plays are much, much easier to do when the team is a single group that can stay together and work on the same goal. Recognizing who's talking and whom they're talking to is increasingly difficult when you add more players to a game, and while it's possible to still be a coherent unit in large teams, communication will usually be hindered, which makes teamwork and strategy less effective and therefore less interesting and fun. When more people are talking at once, it's harder to get a picture of where your team is and how it's doing as a whole.

Having more players in a match would also mean more server stress and fees. When more people are on a server, it increases latency and hardware work on the server and clients. While probably lower in the list of significant reasons for why teams should be small, it's very true that having more player slots means the server owner needs to pay more to run the server, and it's generally in a league's best interest to make running a league-compatible server as easy as possible, for obvious reasons. Similar technical considerations arise for the players themselves, who might have frame rate issues or crashes in firefights when player counts get too high.


Having smaller teams with class limits has consequences outside of logistical considerations—as well as game flow and balance—that are desired in competitive play. One of the simplest is the idea that if there are fewer players, each player matters more. Competitive players want to test their skill and have it impact the game, and if the game was large, their effect on the game would be diminished. If a team is smaller, then each player's strengths and weaknesses greatly contribute to the team's strengths and weaknesses, and this generally makes gameplay more exciting for individual players.

When teams are smaller, it also makes it so unique strategies will be more interesting and impacting. When your team has the default twelve players, you can have one of each class and still have enough room for three duplicates. When teams are only six players, though, each class choice holds a serious player-slot cost, has a large effect on the team's strategy, and very much changes the tactics the enemy will have to use. When changing your team build has such strong gameplay effects, it makes team composition and strategy more interesting, for when teams get too large, counters for counters can be in teams at all times, and curveballs can't really be thrown.

Having class limits has the metagame benefit of letting players focus more on their favorite classes. If the games had no class limits, everyone would need to be well-practiced with a majority of the classes to be on-par, because there would be times in certain maps when any number of players could be called upon to play any of the classes because it would be optimal for the situation. For example, everyone would need to be a good scout so the team could win with scout rushes, and everyone would need to be a good demo to be able to lock down a cap point most optimally. While it may take more skill to be able to play any class well, it would make competitive play too hard to get into for most players, as the amount of practice time needed would be unreasonably high.


The standard competitive format provides very different gameplay compared to the typical public game, and these differences are usually desired. Probably the most straightforward difference is that the smaller teams make gameplay move around the map more. Pushing is much easier if you only need a couple kills advantage to make the enemy retreat, and gameplay is more interesting if such pushes are fast-paced. Aggressive, dynamic games are just generally more exciting, and at the very least, they help prevent stalemates, which are never a desired result.

This complements the fact that having a small team makes players stick together as a group and move as a single unit, which makes coordination and synchronization more impacting on the match. The smaller crowds also give more free space for players to experiment with new movement tactics and route exploration, because there aren't campers and patrollers watching every portion of the map.

Class limits keep the game less hectic in small games. In a game where teams are smaller, if one team is packed with a specific class, the other team is going to lose unless it was specifically countering that class ahead of time. The game would shift from that of using teamwork and coordination to that of making team compositions that would be the most unreasonable to fight against. By ensuring that teams will pick a variety of classes (generally at least four, because of medic's limit) for their six slots, it also ensures that the game won't run into situations where the only solution to a problem is simply “change your team build completely”, which isn't particularly fun to play in or plan for in a game like TF2. Class limits let teams counter each other's compositions without completely overriding each other's compositions, which keeps gameplay adequately consistent and dynamic simultaneously. There's enough variety to keep it interesting, but enough of a standard so that breaking away from the norm will have a significant impact on the situation. Nobody wants to play medics and demos vs. medics and demos every night.

Logic Behind Nocrit and Nodmgspread

The Format

With the reasoning for the competitive format understood, we can begin to see why competitive games would disable crits and random damage.

For one thing, the smaller team sizes and class limits make crits and damage spikes much less necessary. Whereas a public game will require you take down as many as 12-16 players (sometimes spawning faster than normally), including a number of turtles (several camping heavies, engies, and demos), the competitive format greatly reduces this obstacle. The waves of heavies and camping engineers you see in public games just aren't really there. You might see a sentry nest with a heavy next it it at most, but that will generally go down with an uber. While your opponent could resort to two engineers and two heavies, doing so would completely cut down their capability to actually fight off spam. If all you need to break through the densest defense is an uber, you don't really need crits or damage spikes to begin with.

Not only are crits considered unnecessary, but it probably wouldn't be too far to say that in 6v6 games, they can be overkill. A public game might be 12v12 or larger, and in these situations, getting a double- or triple-kill with a crit explosive can be a good advantage, but it certainly won't be a game-breaker. In the small teams of the competitive standard format, however, a double- or triple-kill crit explosive can give your team a sudden extreme advantage, where your mob goes from even with theirs to outnumbering it 2-to-1. Regardless of your skill level (unless they're extremely polarized), your side would have gained an immense advantage that came out of nowhere and become effectively unstoppable.

In competitive games, advantages and disadvantages grow exponentially, and things like crits and lucky damage spikes can ruin otherwise solid, skill-based strategy with a single lucky hit. Something as minor as an enemy getting a crit in a scout duel against you could change everything. That crit would give him the advantage that kills you, which could leads to that scout getting a clean path to your flank, which would give him the chance to rush in and easily kill your medic, oftentimes resulting in losing a control point, which could easily lose you the game. In competitive games, these domino effects, when presented, will actually go through in their entirety in many cases, unlike in public games where lack of communication and homogeneous skill prevent it from happening, and crits (and, to a lesser extent, random damage) would cause these unpreventable shifts in advantage to occur too frequently.

These uncontrollable shifts produce a strongly undesired result: crits, as an extreme random factor, and damage spread, as a minor random factor, undermine teamwork and skill. They turn the game from "offense's teamwork and skill trying to overcome defense's teamwork and skill" to "offense's teamwork, skill, and luck trying to overcome defense's teamwork, skill, and luck". The luck makes the other two parts matter less because it can override differences in the two within the timespan of an individual game. If one team is clearly better than the other, but the other team happens to get a lucky crit to the medic once or twice more than the better team, the worse team will probably win out of that advantage unless they happen to be much, much worse than the better team. Damage spikes from random damage spread has a similar effect, but to a smaller degree.

Because of the competitive format, gameplay commonly revolves around the team medic. In this scenario, crits make gameplay less fun. If your medic died in a single hit by an adventurous scout or unexpected rocket, any uber he was holding onto would suddenly be wasted, and any healing your team needed would be withheld for about 30 seconds. In a public game, this is okay for a number of reasons: a single uber isn't usually enough to completely change the tide of a game, most players can't make full use of their ubers because of communication problems, and there can always be more medics on your team. When both medics have fully charged ubers ready and your medic is killed out of the blue, you can pretty much kiss making a strong push or safe defense goodbye for at least a full minute or so (which matter a lot in matches).

Competitive Metagame

Beyond the format, there are tendencies and out-of-game considerations that take undesired consequences when crits and damage spread are turned on.

Perhaps the simplest undesired consequence of crits and damage spikes is that they make exciting duels and tests of skill end prematurely. If two scouts are dueling, one side could have very low hit points, but if it crits the other, the combat that was previously interesting and testing suddenly ended because one side got luckier. The game becomes less fun when combat is resolved with a figurative roll of dice. While this was just a single scout duel, the aforementioned domino effect of the advantage pushing in competitive gameplay could easily make the duel mean much more than just having a team size advantage, such as the winning scout getting to sneak around and cap a fifth point, ending the game.

Out-of-game, there are clear reasons for avoiding random shifts of advantage. Unlike in public games, the result of a competitive game could have large implications, such as a league title or even monetary reward, and critical hits and damage spikes can cause match results that are essentialy unfair. If a team that would have otherwise lost for whatever faults it possessed happened to win a big match because of random advantages, the other team would feel cheated—and quite reasonably so—for while both teams had an equal chance of getting crits and damage spikes in the long run, the fact is that the games are too short for this to really even out statistically. It's not unlikely at all that one team would get one or two more lucky multikills or medic picks than the other, and such an advantage alone could probably get them the victory in a relatively even matchup.

Players want big game changes to happen because of the players' choices and skills, not because of random game mechanics. Competitive players generally find it less fun when luck gets in the way of strategy, prediction, and honed technique.

Closing Thoughts

While some believe that playing with crits and random damage requires more skill, such a viewpoint is hard to defend realistically. When you play with crits enabled, you acknowledge their presence and perhaps act with some extra caution and more harassment tactics, but neither of these are particularly desired traits, and aside of them, behavior doesn't change much to accommodate chance. You can't control when you get them, and there's no way to predict when one will get flung at you. Crits and random damage simply add unpredictability, which might spice up pubs, but in matches, where teams are small and advantages high compounding, they make it so the strategy and coordination you've been practicing for months can be trumped by a lucky grenade. While you statistically have the same likelihood to crit and get damage spikes, in practice, within the time frame of a single game, there's a good chance that one team will simply get luckier, and such results would have further reaching implications than just a round victory, and it's not fair to the team that lost.

For reasons already explained in detail, competitive players enjoy the standard competitive format, and because of both the format and the competitive scene's metagame, the general consensus among those competitive players is that critical hits and damage spread are undesired in organized play. The fact that they do so does not make their style worse or better than vanilla Team Fortress 2 public gaming.

Competitive gameplay and pub gameplay are very different. While a casual gamer might enjoy crits and random damage because they make things happen, a serious player usually wants only his own choices and his opponent's choices to be what make things happen. Competitive players don't disable random effects because they get jaratey about dying or can't adapt due to a lack in skill. Their reasons are legitimate and pragmatically realistic.


I've had this post in a document of mine, adding bits and pieces to it, for quite a few weeks now. The bulk of it was written before the scout update came out, so it doesn't really address any of its changes specifically, though a lot of the reasoning behind nocrit and nodmgspread could be easily applied to the Sandman.

I'm aware that it's very long (so much so that this epilogue made it exceed the character limit), and I don't expect it to be fully read by anyone, but if you're going to speak out against competitive play, I encourage you to give it a perusal. I fully expect someone to post a straw man's argument in response (in before). I also wouldn't be surprised if people disliked it simply because it was long, but I figured a text block was better than incessant bickering back-and-forth in every topic that mentions crits or competitive play. Whether or not I really want to debate it further to begin with, actually, I'm starting to doubt. Don't expect further ranting from me, and don't ask for a tee el dee are, because if I post one, people will just read and argue against the brief instead of the actual points.

In the end, I think I hope for this to become a reference for people who don't want to keep fussing over things we've been over hundreds of times. When someone trashes competitive gameplay or tries to argue about it with faulty or oversimplified points, I hope they will be presented with this and see that, while possibly disagreeable, competitive gameplay is not unreasonable. Many players sincerely find serious play more fun than casual play, and they shouldn't be faulted for that. They join competitive groups because public games don't give them enough complexity, not because they want to change TF2.