tempest in a teacup

the pointless musings of a strange recluse

TF2, according to the experts

One of the things I do to pass the time when I’m bored is read the official Team Fortress 2 forums on Steam. A lot of different types of players post there, but the most interesting posts are the ones from the people who play the game competitively. Particularly since when you try to hold TF2 up to competitive standards, it starts to show some flaws.

There are a few issues that tend to come up again and again – I’ll try and go over the most important ones.

Damage Calculation

TF2 is different from a lot of other shooters in that weapons don’t do a fixed amount of damage. Instead, the damage a weapon does is a number within a range. The damage ranges are different depending on the distance between the shooter and the target. For instance, the Scout’s scattergun does 85-105 damage at close range, 10-40 at medium range and 3-10 at long range. In most cases this isn’t an issue, but it is rather an issue for certain characters. One example typically brought up is the Demoman – his sticky bombs have a very wide damage variation (actual numbers can be viewed here). This isn’t an issue for Demomen who use their stickies to trap choke points since they typically use multiple stickies in such a situation. However, at high levels of play Demomen tend to rely on tactics like midair sticky detonation, and this is where the unpredictability of the damage starts to bite. Arguably, medium range direct combat should be the Soldier’s forte rather than the Demoman’s, but I see a lot more Demomen being used in this capacity than Soldiers, probably because they can get similar results without being hobbled by the Soldier’s small clip size and long reload time.

Another argument that also comes up fairly often is that the use of damage ranges rather than fixed numbers also makes the outcomes of fights more unpredictable and lessens the impact of individual skill (a point which comes up fairly often).

Hit Detection

This mainly has to do with hitscan weapons (like pistols and shotguns). Apparently visual indications of damage (like blood) don’t always correspond to actual damage being done. So a Scout who’s hopping around like mad firing his scattergun can see the blood particles, but it doesn’t mean he actually hit him. This, needless to say, is pretty silly. I’ve only seen one thread on the topic (and it’s dropped off the front page of the TF2 forums) so it’s hard to say how common this issue is. I certainly haven’t seen it affect me all that much, with the exception of long-range shotgun blasts and maybe Engineer/Scout pistol spam.

Lag Compensation

This is somewhat related to the point on hit detection. Unlike its predecessors Quake III Arena and Unreal Tournament, Valve’s Source engine netcode has extensive lag compensation. In games like UT, the player sees his action execute with a delay that represents the lag he’s experiencing. In other words, if a player clicks his fire button, the game actually fires the weapon only after the server has been notified and has acknowledged this action. Under high-ping situations this results in a noticeable delay for every one of the player’s actions.

The Source engine does away with this by keeping track of state on the server and then ‘rewinding’ the state when an action notification arrives. It then checks to see where the target was when the command was given and then updates its state accordingly. This has the effect of being more charitable to people with laggy connections, but it can result in some odd results for people not handicapped in such a manner. For instance, a laggy Sniper can headshot someone who appears out in the open to them, but by the time the server figures out the person has been headshot, that person has retreated behind a wall – yet dies anyway.

Competitive players seem to argue that it’s easier to deal with the lag by adjusting your timing than the unpredictability of lag compensation – this is arguable. In any case, it sounds like they just want an option to turn off lag compensation at the server level, which is a reasonable option that doesn’t affect the game for regular players.

Critical Hits

This is probably the biggest thing that the competitive community has complained about (and already addressed). Critical hits are high damage attacks that for the most part occur randomly (with a few exceptions – all Sniper headshots are criticals, all Spy backstabs are criticals, and any attack while being charged by a Kritzkrieg are criticals). It’s possible to increase your chance of getting a critical hit by racking up kills, subject to a cap of 25% (which gets reset after you die).

Needless to say, the competitive community balked at this randomness being introduced into the game, and as a result pretty much all competitive games are played with crits turned off. However, as far as I know most competitive leagues still allow the Medic’s Kritzkrieg as it is deterministic (100% crits no matter what), and can be a game-changer under certain situations.

Again, much like the use of damage calculation, crits have the effect of lessening the impact of individual skill – a good Scout can still be taken down by a mediocre Soldier if he gets a lucky crit. Having been the beneficiary (and target) of several lucky crit rockets, I’m pretty sure that removing crits was the only way to go for competitive play.

All that said, Valve has catered to the competitive crowd in their numerous updates – Arena mode in the Heavy update was targeted at them (the characteristics of this mode make it ideal for competitive play), and more recently they removed the setup time from Granary, one of the most favoured competitive maps, which was a change that the competitive community had been asking for for a long time. On top of that, in spite of all the flaws I’ve mentioned (and doubtless a few others that I’ve missed), TF2 is still pretty successful as a competitive game (although not nearly to the extent of the old faithfuls Counter-Strike and Quake III Arena).

I’ve only recently gotten back into first-person shooters after a long absence from the genre (before I got my current desktop the newest one I had played was probably the original Half-Life), so reading about what makes high-level FPS play tick is pretty damn interesting to me. I hope this little overview was at least slightly interesting to you as well.

On an unrelated note, in spite of my misgivings I’ve gone ahead and preordered Sonic Unleashed for the Wii. I intend to use it as a gauge of whether the PS3 version will be worth my cash come December, although of course the $10 discount coupon I got from Amazon for my next video game purchase didn’t hurt either.

1 comment

One Response to “TF2, according to the experts”

  1. […] The ability to turn off random damage at the server level (something which I alluded to in an earlier post) […]

Leave a Reply