What Is Schwa?

From Valve forums.

I thought this was important reading and some of the more hard-core players might appreciate it. For those who don't want to read the whole article I have bolded the parts I think are very important to the average player.

I think it important to note that critical hits are still somewhat unknown. I believe there is some variable that increases critical damage based on how long you have been alive and how many kills you have that round. The tester was unable to incorporate all those variables into his testing, and it should be assumed that the crit values are the "lowest" possible values.

Late Edit:
Remember, all the testing above has been done by one player shooting one other player in a non-game environment. However, TF2 could have another possible answer to the damage variation dealt by weapons. The more I play the more I think this explanation might have some serious merit.

Warning: Reading the following may ruin your enjoyment of TF2.

From another source:
-Aaron

Sirusblk

Very good read, thanks

__________________

KnifeHeavy

Nice info and good testing!

__________________

Sirusblk

I sweat I think the wrench is a one hitter and I think the baseball bat crits more often.

Rate of Fire wasn't taken into account also. Without it the numbers mean next to nothing in my opinion. Suppose you can do 100 damage in one hit but it takes you 5 seconds to do so. Lets say another gun can do 20 damage in 1 second. They both come out the same.

It's a lot like saying what weighs more, a pound of feathers or a pound of bricks.

__________________

MagnaCentipede

Yes and no. Rate of fire is important for some weapons and not others. Sascha cares about fire rate because it's a hitscan that you trace an enemy with. Shotgun fire rate is less important because you might only get one or two shots on someone before they catch cover, and at range, the number of projectiles that catch will be increasingly random. Demo is all about the power of one hit because you generally try to trap with it, and you can't reliably lay multiple hits on the target.

"Suppose you can do 100 damage in one hit but it takes you 5 seconds to do so. Lets say another gun can do 20 damage in 1 second. They both come out the same."

100-damage rocket with five second flight time vs. Revolver doing 20, firing each second.

At time 0, the rocket fires and the revolver does 20. At time 5, the rocket strikes for 100, but the revolver fires its sixth bullet, totalling 120.

It's not just damage per hit and fire rate, as hitscan weapons do Damage on Time 0 while physics weapons are delayed.

__________________

sordavie

How much does the sniper rifle do on automatic?

__________________

Chewy

Interesting stuff... and now I have an answer to when I worry about why I suck so much in a round - because it's dolling out random damage!

__________________

Sirusblk

If that AI stuff is true it might be because you're too good

__________________

Tarenth

I think that in some cases you can actually force a crit to take place even if the random crits are still random.

Basically I've found that if you use your weapon in a way intended you will tend to score more crits than if you used your weapons mindlessly.

Flamethrowers seem to throw streams of crits at PB from behind or the side while only occasionally from the front. That seems to promote the 'hide in corners and take people by surprise' approach suggested.

Sniper rifles ALWAYS produce crits with a full charge shot to the head which is notable since if you don't have a full charge you won't make a headshot kill in most cases.

Melee weapons throw crits more often if you take someone from behind and aim for the head than if you charged at them from the front. They crit more often if its while they're doing something else (shooting, sapping, deploying, etc) than if they were just standing around doing nothing.

'Bullet' weapons tend to crit more if you aim for the head than if you shot at the body.

Last, and the most odd, the heavy minigun appears to crit more if you're stationary, elevated, and crouched. I stood on a platform and fired down into the masses for a few seconds before crouching and the moment I was planted crits seemed to bloosom all over my screen from every single hit made.

__________________
My sanity is not in question...
It was a confirmed casualty some time ago.

Sirusblk

A lot of people don't realize how good crouching is, especially for a heavy. Whenever I run into another heavy with a medic, it doesn't matter if he fires first, I crouch and fire and always come out on top.

__________________

Avs

I thought I would just clear a few things up so nobody gets the wrong idea.

These are in response to the comprehensive weapons testing done by Promii

1. Shotgun weapons (including Scattergun) are based on pellets as well as range, and rely on each other. Taking this into account the damage is quite constant.

2. Rockets deal normal damage against turrets, range does not affect its damage. Rockets will rarely hit at long ranges, so its actual damage is 0. Otherwise its damage is varies around 50 average at furthest range. This range is about 4000 units, which is something like 75 feet on screen. I have not had the time to test a range further than this, but I believe it wont be necessary because few maps will actually have shootable distances further than this that have hit value.

3. Butterfly knife CAN crit. But thats not the point. Never swing your knife at a target from the front. Even a crit hit with a knife or most melee weapons will NOT kill someone.

4. Sniper rifle headshots do not do infinate damage. The backstab does. Unless your rifle is fully charged, you can headshot people and not kill them.

Most of these tests are conducted in ways that do not reflect actual gameplay. Take the rocket for example. Also, his grenade launcher damages are off by a fair amount. It is well known that grenades have no diminishing returns. Yet the range of 65-125 don't make sense. And when matched up to more testing, the range of damage is closer to 90-125. I believe that in testing he lowered its damage to better reflect his diminishing return findings, or based it off of indirect hits included.

Melee Weapon ranking, by max damage potential as Valve intended per class:

Butterfly knife *
Bonesaw
Wrench
Fireaxe
Boomerang Knife
Shovel
Bottle of Rum
Baseball Bat
Fists

DPS: Damage per second. This doesn't matter at all as unless you have a flat rate of hits per second it cant be calculated. Therefore the only thing you need to worry about are direct hits and or area effects (explosions). The only time DPS comes into play is when you are on fire. A touch of fire (set on fire without extra damage at max range) will do around 100 damage to you if not left unhealed (ie. removed via medic, or health pack). It deals 4-6 damage per second, ticking at half a second each time. Promii later talks about how the baseball bat is a great melee weapon as it swings faster than most melee weapons. While it does swing faster this melee weapon is a poor choice due to the alternatives one has as a scout.

Crouching while Heavy: Crouching does not make you beat all other heavies. It also does not increase critical rate. I don't really know what is making you beat these other heavies besides the fact you have a streak doing this. The fact that crouching restricts all movement for you should mean instant death in any firefight unless also supported by others which is likely to be the case.

Ok so now that you've probably memorized those base weapon damages. Consider three other factors that apply to weapon damages. These factors change how these weapon tests actually work in game.

A) Team momentum factors.
B) Personal momentum factors.
C) Domination/Nemesis factors.

A. By now you should all know that the team with more CPs spawns faster. Little details like this make scout rushing the middle point worth its weight in gold, and are valid strategies. There is a theory however based on this momentum that affects weapon damages. When a team has the advantage (more flag captures, more capture points) they also gain a slight damage advantage. However to which gamemodes this applies, and whether it is total wins or per-round basis, I do not know. An example would be granary, where one team has 1 CP, and the other team has 4. Not only does the team that is winning spawn faster, and spawn closer, but they also have say, an extra small amount of damage per CP advantage. Because there are two other factors that deal with weapon damages my testing for this became inconclusive. If however this factors in, my estimate is that it is no more than a 20% bonus damage at any time. I believe this is to compensate the offense for the supposed advantage defense would have. But then not all maps are made equal, which makes this point very debatable.

B. Personal momentum does however factor into damage rates. Random crits are dependent on your own merits, and are not as random as many assume. The more "on a roll" you are doing, the more crits you will see. Marduke says: "From my tests, the base crit % for primary weapons is about 5~%. I tested the Rocket Launcher (3.75%), Scattergun (7%), and Sticky Grenades (3.9%).

That's right. About 1-4 crits per 50 shots. It really is not as bad as people make it our to be."

In anycase, the higher your score is raises your crit chance slightly. I believe this is based on a ratio to others on the server so it is kept in check. I also believe personal momentum affects your damage potential, while game time possibly affects overall base damage for all players. This may have been intentionally designed to avoid stalemates during last minute offensives. Also for CTF maps, there is no round time. This may point to a internal clock that checks how long the round has been going as I believe it applies to all game modes. My testing (30 minute rounds, with new late-joiner for time dependent damage vs score damage + time dependent damage) shows that there is an increase either in base damage or higher average damage by raising the potential maximum damage.

C. The Nemesis effect. This is based on the factor that the game attempts to help out players that are doing poorly against specific other players that are dominating them. One guy puts it:

"I found I often died in one hit to demoman grenades and soldier rockets playing as scout, but my friends did not. I was owning as scout, and my friends were not. Then we tested it on a lan, and we COULD NOT get a grenade to one hit a scout *until* we restarted the server and let the scout dominate the demoman. Then the first grenade he fired was a one hit kill (no crit required). Now given the randomness of the damage system, I feel that this deserves much more testing, but the system also seems to take into account how much of a threat someone presents, and how a team as a whole is performing. In summary, I think that there may be some massively complex "fungineering" going on in the background in terms of programming."

Via testing, I see that you do deal slightly more damage against nemesis targets but criticals are not involved in this process at all. My guess is the game makes a check against the targets affected on hit before applying damage. But then again this is traded off mostly because the enemy player is likely to hit you more times than you hit them, and fire off the occasional critical shot more often than you. The reason why his may not be very apparently or trustworthy at all is because of the other factors that affect damage. It may not even take nemesis for this to occur. This might be happening for a 1:1 ratio per player, and is mostly masked due to factors like range, weapon type, healing buffs, so on and so forth.

Examples:

Player base damage is 100 average. 5% critical rate. 0-0 score, round just started.

Late game: Player base is 100 average + ~15 for CPs held + ~20 for leading scoreboard + ~ 20 based on time. This means he has a potential to deal 15+20+20 maximum damage ontop of his base damage on next shot, averaging 100 at point blank. Add more for nemesis, etc.

I don't believe weapon damage ever decreases, but I do believe the floor on weapon damages stay constant. So a rocket at max range of 4000 will deal 50+potential damage instead of straight up 50. This answers why a demoman's grenade can deal more than 200 damage, but rarely does so. The base demoman's grenade deals an average closer to 110 right at the start of the game however. This makes it very easy to chunk a scout shortly after the initial game moments when scouts lose initial capture status.

Comments on all this analysis: I just want to say that numbers don't matter. People have endlessly debated all the stats for weapons and randomness but largely miss the big point. In TF2, you have to hit your target to make it apply. If you play unreal tournament you know what I am talking about. Even area damage weapons like rockets are largely ineffective without hits that are close or direct. Its obvious that all other cases the target was not at 100% hp. TF2 is unique in the sense that its important to have a good understanding of the basic concepts as covered in the above without needing to rely on set weapon damages such as in Counterstrike where certain guns will kill in one hit headshot, or be more accurate in X case.

Does this solve the great mystery of TF2 damages? Not really, but it gives you the idea that valve wanted to players to have without having to drudge through the whole damage scenario. This is why people bring it up, but the majority really don't care and only see criticals as being the main culprit to look at. If you ask me, even with all these mechanics in place, a good player will still beat out a bad player because of one this: the ability to aim. Either way you look at it, being able to aim will ensure you kill people period, or vice versa, dodging projectiles. No amount of damage bonuses will make that difference.

McQueen17

For the first part, crouching (as mentioned in my other post about dodging) decreases the size of your hitbox. As for the heavy-gun hitting totally random the size of your hitbox does matter. So it becomes at least explainable why a crouching heavy owns standing heavies as long as they both take the same aim and don't move.

*phew* The second part quoted saved a lot of writing from my side, Avs. You're perfectly right. All you wrote before MIGHT be right (people like to be sure about such things), but they seem to get so damn totally random as soon as you try to apply them to a game of 10 vs 10 that NO ONE (did I ever mention that I lack to find the function of bolding something?) should ever rely on such things. As for my part I don't think it is even in the code to have such possibilities. But I don't know anything about coding anyways.

Sirusblk

I like the idea of critical hits, the seemingly random aspect means players can no longer rambo. It's better to play it safe than risky most of the time.

This is the greatest part of the game that makes me turn to this more than say Counterstrike where my aim sucks horribly.

__________________