Using the list yttrium made:

Scout

• Baby Face's Blaster: downside of reduced speed is too easily negated by the boost. Maximum boost offers too large an increase to mobility. The boost reduction is negligible and easily overcome again.
  
• Soda Popper: +reload glitch allows fast rate of fire and more dps than stock which negates the only drawback. Increased mobility from jumps increases survivability in a fight too much. Boost is charged by running which does not require additional effort on the part of the scout.
  
• Flying Guillotine: (in my opinion) the burst damage from use in conjunction with the sandman stun is too high. Stats from use without the sandman seem alright to me (but admittedly I rarely play with or against this weapon when it is used alone)
  
• Pretty Boy's Pocket Pistol: Increased vulnerability to fire is not commonly applied, the scout is naturally more able to avoid potential sources of fire damage. Increased health gives the scout a massive advantage in fights/1v1s. I don't think the lack of fall damage is all that broken. No other serious drawbacks compared with stock pistol.
  
• The Atomizer: additional jumps on demand greatly enhance mobility and survivability. The cost to health is either not high enough, or easily overcome through intelligent play (or the use of other unlocks). This weapon seems more balanced than others on this list.
  
• Wrap Assassin: I do not know of a use/glitch that would make this weapon overpowered, damage reduction makes it feel severely underpowered (but again I rarely use it, or play against it, so I could be wrong).

Soldier

• Beggar's Bazooka: Rate of reload allows it to match/surpass the fire rate of stock. Never in a situation "without rockets loaded". Beyond this, the burst potential is very high, the inconsistency to aim is negligible and easily accounted for through play style adjustments. No other particularly relevant drawbacks.
  
• Liberty Launcher?: Feels fine as is to me. I differ to the judgement of other soldier mains though.
  
• Battalion's Backup: Buff feels acceptable, not certain how 'spammable' the buff is in practice though - potentially an area of issue. The increase health coupled with the increased resistances when the buff is used give massive advantages to fights/bombs/1v1s. No obvious downsides beyond the lack of shotgun/gunboats. This weapon feels more balanced that others on this list.
  
• Reserve Shooter: Clip is not reduced enough to noticeably affect gameplay - only really noticeable for players who have difficulty with ammo management. Detection for minicrits has always been buggy, allowing for increased damage evening against enemies not obviously 'in the air'. No other meaningful downsides.
  
• Disciplinary Action: Utility, specifically around mid fight timing, forces a team to run it or risk losing mid. Beyond mid fights I don't think it is overpowered, maybe in certain transition situations (or for making other classes more viable in transition). No serious drawbacks compared to stock.

Pyro

• Rainblower (with new server cvar to disable pyrovision): as long as glitches involving pyro vision aren't an issue this probably doesn't require changes. I'd love to see someone make a strat specifically around the AoE of the taunt kill.
  
• Powerjack: Health on Kill and increased movement speed. Health on kill balanced by the increased damage vulnerability (might need to tweak some of the values), movement can be used in transition without risk of any drawbacks. If used for the utility no drawbacks when compared to stock.
  
• Lollichop (with new server cvar to disable pyrovision) (banned in all other 6s but not ESEA?): Same as Rainblower.

Demo

• Ali Baba's Wee Booties
  
• Bootlegger
  
• Loch-n-Load
  
• Persian Persuader
  
• Ullapool Caber: maybe it is because of my time in HL, but this weapon doesn't bother me. Maybe the negatives do not offset the potential to burst in a small AoE. Point blank burst damage might be too high (though you are trading your demo for a low change of killing their medic, provided their medic is paying attention)

Heavy

• Brass Beast: In a defensive setting the movement penalty is negligible and the damage increase is effectively free damage. No other meaningful drawbacks.
  
• Natascha: I actually don't mind the movement penalty. Value tweaks around this and damage reduction. Current damage reduction is easily negated by coordinated play.
  
• Dalokohs Bar: Free, temporary health. Temporary nature is completely removed when the item is used only for defense. No real drawbacks (besides lack of alternative secondaries... but heavies rarely use their secondaries on defense). Loadout can be changed after the effect has worn off or the defense is over.
  
• Sandvich: An additional source of health on demand. Hard counter to any trade/suicide play that only damages the medic. Health can be given to any class, or used to self heal to full health. When used with good positioning or team support the temporary vulnerability is negated

Engie

• Frontier Justice: Free Crits that do not require a change in gameplay. Drawbacks are negligible in competitive play.
  
• Pomson 6000: Drains Uber/Cloak. Player Penetration. Damage potential. Limitless ammo. (what where the drawbacks of this weapon again?!)
  
• Widowmaker: Infinite clip provided that you can aim.
  
• Short Circuit: Rate of Fire and ammo cost remove any skill required to effectively use the weapon. Completely shuts down and class that uses projectiles. When used in coordination with team play it is not feasible to focus down the engie with secondaries while ignoring the rest of the enemy team.
  
• Gunslinger/Minisentry: Denies mobility of the other team (damage denies scouts, knock back in the air denies explosive jumps). Does not require that the engie 'watch' his gun allowing him to do other things for his team. Low build cost makes it spam-able. Range and FoV mean that outside of full nests it is categorically better than stock sentries.

Medic

• Quick-Fix: The weapon encourages very passive play. Heal rate is a direct counter to uber that also builds faster. In 6v6 there are not enough sources of damage to focus down the medic/heal target effectively. Lack of knock back and capping while charge is active is easily abused when focusing on objectives.
  
• Vaccinator: Lack of varied damage types in 6s greatly reduces current drawbacks (though the decreased heal rate is still fairly significant). With intelligent play greatly nullifies 50% or more of the incoming damage without a need to switch types.
  
• Solemn Vow: Direct upgrade from stock (that said, so is the uber saw in most cases). This weapon feels far more balanced than others on the list.

Sniper

• Sydney Sleeper: rewards less skilled play. When used in combination with team play makes it very easy to get high value picks. Biggest issue is that it allows for equivalent impact for less skill.
  
• Cleaner's Carbine: (nearly) free crits. drawbacks are negligible in competitive play (?). No other meaningful drawbacks in a competitive setting
  
• Darwin's Danger Shield: Combination of health and damage resistance negates low charge head shots. Massive advantage in SvS fights. No meaningful drawbacks.
  
• Jarate: Free mini crits for the entire team. Spy is not as viable, team support negates survivability cost

Spy

• Enforcer: 2 shots most weak classes. No meaningful drawbacks. Rewards low skill play.
  
• Spy-cicle: Is this weapon broken in 6s? Fire immunity removes a hard counter at low cost (given viability of various revolvers). Greatly increases survivability of spy. Rewards low skill play.

There are a few instances where not running the whip is actually more viable. A perfect example is process. If your soldier is consistent with the fast rollout you can deny the health kit (and potentially kill the enemy demo). That said, this only really applies to maps where the soldier has the option of a 'fast rollout' that offers some immediate use that can offset the fact that the rest of your team will be a second or 2 slower to mid. Maps without an option like that, it remains a must use so long as the other team is running it.

I think the point that makes me doubt that these kind of strats would be as effective as people think (at least at higher levels... in the lower ranks I expect teams with multiple medics and heavies would be stupidly strong... but that would not last as the average skill of players increases through the ranks) is that you aren't just adding this additional medic or demo, you are removing one of the other classes as well.

Sure a second medic would be strong in terms of more uber potential and increased heals. But the extra medic is also removing a lot of mobility and damage potential. Meaning that if the other team is able to effectively focus down players during off uber pushes, then they would be at a fairly big advantage. Additionally you would be losing the versatility of the more expendable classes... sending in a scout or soldier is a lot less effective if you don't have another one available.

Same for a second demo. You are gaining more area control (which a sniper can already give you more easily than another set of sticky traps) and more general damage. But it is costing you in terms of 1v1 engagements which the demo is terrible at.

Then there is the question of what you trade. Do you remove the roamer, and create the need for more heals across the entire team (roamer is viable because of the ability to self sustain, a second demo would not be so heal independent). Or do you drop a scout, and lose the ability to capitalize on the increased damage output.

I don't for a moment think that we will not see these kinds of compositions come out for specific pushes or holds. And I expect as they are used valve will make adjustments (either to the class or to the format) that help to balance things out. But I do think that in terms of general usefulness these compositions are going to bring weaknesses that can be exploited... and that is sort of what makes a competitive game interesting. It would stop being a matter of who plays the 2 scouts, 2 soldiers, 1 demo, and 1 medic style of tf2 better. We've seen that game, and the answer is Froyo (and has been for a very long time). And it would start becoming a question of which team can more quickly recognize and effectively exploit weaknesses in the other team.

So long as every class still has a counter, and that counter remains viable, I think that the game will remain interesting and dynamic.

Heavy already dominates at low levels in the current competitive scene (both in 6s and in HL). Most sub silver HL teams in UGC focus the majority of their heals on their heavy at the expense of heals to the scout. This is because mechanically, heavy is very easy (the 'skill' required to play it effectively at higher levels has almost nothing to do with the mechanics of heavy and everything to do with positioning, timing, and general decision making). I doubt valve will care if Heavy is OP at low ranks, so long as it is more balanced at higher ranks (even just in certain situations).

As for valve buffing classes because the meta uses them as specialists rather than generalists. It is possible, but I was encouraged by the fact that they wanted to model their match making around existing formats. I think the devs at valve are intelligent enough to understand that in a competitive format not every class is going to see equal use and that alone does not mean there is a balance issue. My hope is that they will look at the data they collect intelligently to consider the effectiveness/balance of a class or weapon in regards to the intended purpose of that class or weapon.

Final note, while it is interesting to read all of the change suggestions in this thread. I agree with the first post that suggesting changes is not the way to help valve balance weapons. Maybe, as the dialogue between devs and comp players continues that will be possible, but right now it would be more effective to create a comprehensive list of core problems with weapons WITHOUT suggesting possible fixes to those weapons.

I'm not trying to theory craft what the strongest composition would be without class limits (though I am still inclined to believe that, at least in the current world of tf2, scouts are too strong for good teams to cut them in favor of another soldier or even another medic. I've no doubt that without class limits we would see more dynamic class changing, the creation of pocket strats that use unexpected combinations, and things of that nature. But at the end of the day, I think the current spread is likely the most ideal composition you could have for a generalist role).

At the end of the day, valve is going to want simple and easy to understand. Creating a whole system of coded logic around class limits, and when/how often you can off class. That would be a nightmare to code, and worse it would create potential obstacles to entry for new players.

I can't pretend to know what valve would do. But speaking as a programmer I could not imagine anything but the most simplistic of class limits/weapon bans making it into the first iterations of the match making system. After that, maybe more complexity can be added if the system proves ineffective, or if the meta that evolves is too static/stalemate-y to create good, interesting games.

Honestly, I think it is far less complex than all of this discussion about class limits and stuff like that. When MvM was released there were no class limits, no weapon bans, and no established meta game. But now not only is there an expected class composition that is rarely deviated from, but there are accepted weapons and even an accepted order of upgrade purchases for a lot of the classes (though there is more deviation accepted in upgrades than in the other areas).

When it was new people learned this through experimentation, then from getting matched with people who had more experience, reading online guides, or watching streams. People learned the expected classes, they learned the accepted weapons, they learned the rolls that each class fills, and they learned the most effective order of upgrade purchases. And such, there is always the occasional new player who doesn't understand. Often they are kicked outright, or someone takes the time to teach them. Or they keep trying and failing until they learn in the same way that everyone before them learned.

And sometimes there is the experienced player who pulls out a less used class because they have discovered a way to make it viable, and often they are still expected to perform or their team will ask them to switch (or, more often, they will switch or leave on their own).

I think 6s can work without hard coded limits on class and team composition. I think that the same basic reasoning the allowed this format to refine itself to the current meta will still apply (mobility, more than anything else, wins in this game right now - ability to kite, to chase, to dodge, to quickly change position in in a hold or transition). Add to that the stream integration, the easy access to strong examples of the strength of that format and casual players will gravitate towards it.

If you want to be certain, cue with teammates, or friends. Stomp matches with the familiar classes and team composition. Players will, by nature, emulate the things they see working in streams or the things that they are beaten by first hand.

I'd like to believe that the meta that has existed for so many years already is not just the product of arbitrary bans and class limits. And I'd like to believe that regardless of how broken some of the weapons in tf2 currently are they will not be over powered enough that 6 random players can be expected to beat a group of 6 already veteran players adding up together.

The amount of entitlement in this thread is amazing. Losing LAN sucks, but at the same time now Invite teams get a chance to break even on winnings from finals. AND we have some actual support coming from valve. Sure the finals will be online, but hopefully they will also be featured in the main menu of the game. Imagine what would happen if we put on an amazing show (and considering the production quality of the last LAN that wouldn't be hard to do) and start pulling in even a fraction of the casual player base as viewers.

LAN is gone this season, but that doesn't mean it is going to be gone forever. It means that this season we really push for exposure, we work with valve, we make our game appealing to watch and more known to casual players so that they can watch. Then, next season, we take those numbers back to ESEA and see about getting our LAN back (maybe with a bigger prize pool as well if the support from valve starts to take off in big ways.

Its a business, lets prove that TF2 still has a market.

Or you know, we could look at Valve finally offering direct support... and decide that after 7 years of keeping this scene alive without any of that we've had enough. That seems like it would be a waste to me.

No, match making works in CSGO, and literally everywhere else, because people like to win. They find winning fun. It works because the 'meta' isn't so fragile and pitiful that it can be completely broken by a few trolls running silly strats all the time. It works because while the 'pocket strats' and stupid stuff might work occasionally it is the exception, not the rule.

If 6v6 is such a weak meta that a handful of engies running mini sentries and short circuits (or whatever other cancer class/loadout you can imagine) is enough to beat a coordinated team of 6 intelligent players then I'm sorry, we've failed as a competitive format. As over powered and broken as some of the weapons in tf2 are right now (and if valve supports match making we WILL see balance patches) there is no weapon that is going to turn the tide of a game so completely as to create the world that you are imagining.

And if a strat that differs from our current meta does appear, and does prove successful, then people will start running it... because people like to win, they enjoy winning. And if that strat is so dominant that nothing works against it, then valve will be able to balance it. Because valve doesn't have to worry about every single weapon in all of tf2. They will have data, they will see what is used in match making, they will see stats like wins, and they will hear a much stronger voice from the tf2 community (given that in game match making will have a larger player base) and they will fix those weapons.

CSGO and DotA work because the players stopped screaming about the end of the world every time something changed and realized that things tend to balance themselves out naturally.

So, you mean it could be exactly like lobbies and centers are right now.

News flash for everyone: at the lowest ranks people are going to do stupid things. Run stupid classes, play stupid. They are new, they don't know better, and they will either learn and (as a result) rank up, or they won't care and stay at about that level.

Match Making isn't going to replace outside leagues, tournaments, or other official forms of competition in the exact same way it doesn't in dota and CS:GO.

If you are expecting Valve to make and run a fully functioning league then you are going to be disappointed. Period. Nothing that they can feasibly do is going to make players, who don't know better, immediately start playing the exact meta that you want them to.

As a medic I don't play in centers unless I know people on my team because I have played centers before where my pocket spent the entire game playing engineer. Or where no one on my flank said anything. I've played in UGC matches where teams ran a heavy to mid for every single mid, and Open matches against essentially full time snipers and pyros. It happens now, it is going to happen more in match making and that really shouldn't be an issue.

Exposure to streams and tournaments that showcase 'correct' play, a desire to rank up, and the natural quality of stupid strats to fail more often than not (especially against intelligent players), is going to weed out a lot of the nonsense and let the 'cream rise to the top' so to speak.

That was my point.

I'm not suggesting that things won't change. People are going to try out some crazy stuff, and some of it will work. But the point I was trying to make is this:

2 medics is powerful, but those medics, alone, don't contribute damage, so if the other team does counter them, the team is collectively weaker. A heavy, with even just a single medic, is strong. But moves slowly, and is hard countered by a sniper. A sniper, is strong, but does not have much health and is weak at close range.

Without weapon bans all of these other classes are more viable, but I still think (and I could definitely be wrong) that each composition would have more inherent weaknesses either in the form of hard counters from the other team or an inability to move quickly for timing critical pushes. I don't think that 2 soldiers, 2 scouts, 1 medic, and 1 demo will always be the best option. Even in our current system there are plenty of times when off classes are just better. But I do think that that class composition will prove to be the most reliable. And the other compositions will be too vulnerable to counters to be run as often.

To catch a team off guard, sure, that is what a pocket strat is for. But run it repeatedly against the same teams and it will get shut down.

I imagine that valve will quickly realize the need to limit the number of classes allowed for things like Medic or Demo. But beyond that, I really don't think that a lack of class limits would break 6s. Even now, after the recent demo nerf, I'm not certain you'd need to limit the number of medics and demos either.

2 Medics might have the advantage of more up time for ubers, but medics are useless in non uber fights. So that composition would be punished heavily in transition plays.

More than 1 demo could be dangerous, given how much damage that class can put out. But the demo also requires a high level of protection given how useless it is at close range. So again, I suspect that the composition would be fairly easy to exploit.

I expect we'd see a lot of 'pocket strats' built around interesting class compositions or individually talented players who can make less common classes viable. But I also expect that the meta of 2 scouts, 2 soldiers, 1 demo, and 1 medic will still survive, given that it has been proven to be the most effective composition across the largest range of situations.

Honestly, I think 6v6 makes the most sense because it fits better with the size of teams used in other eSports. 9v9 is going to be a pain to do in match making. Nothing wrong with HL, it is fun, but logistically it is a nightmare.

As for 6s, I expect that there will be far fewer limitations in whatever valve makes. And I don't think that is an issue, so long as valve is also actively working to balance weapons. What makes the most sense, to me, is limits on a handful of classes - such as 1 medic, 1 demo, maybe 1 heavy (but I imagine that more than 1 heavy would have some fairly serious drawbacks), and probably 2 of anything else. That is what 6s is right now. People could run full time snipers and pyros, or whatever else they wanted. They don't because after years of playing this format we've discovered that what matters most is mobility and burst/splash damage.

There is a meta in games like LoL and DotA, not because the players are forced to play that way, but because the players have discovered that that way of playing is the most effective. And it is still possible for them to play something different and unexpected. And that makes it fun to watch, fun to play, and appealing for sponsors to support.

If 6v6 with 2 scouts, 2 soldiers, 1 medic, and 1 demo is THE most competitive version of this game then trust that that is what the meta will gravitate towards - people will want to win after all, and if that works best they will make use of it. And if it isn't the most competitive version of this game then expect something different.

I looking for a medic spot for the coming season. Haven't had much luck with 6s teams staying together in the past, hoping that this season will be different. I've played part of a season in cevo before the team died, and part of a season in open before the team had some fairly serious roster/internal issues. I've played a lot in UGC, both HL and 6s.

I can scrim often, I don't mind playing late. I'm EST and I do work, but if I know what days will be scrim days ahead of time I can pretty much always make the time.

If you're interested in trying me out add me here: http://steamcommunity.com/id/_In_Sanity/
(if you have a really low steam rank and your profile is private I may not accept unless I can tell from your name that it is comp related)

I've been playing for a while, and I'd like to think that I have my basics down. But I'm new to ESEA, had part of a season in CEVO Open, but the team had to disband part way through the season. Now I am on an Open team for the coming season.

I think I can improve the most with my positioning, and my play during transitions. I'd be looking for a mentor, preferably IM+, who could help me bring my game to the next level, but it would be awesome to get some general advice for my team as well. A lot of medic, in my experience, is heavily reliant on how the team is choosing to play.

If you've got the time, or even just some passing advice, add me on steam: http://steamcommunity.com/id/_In_Sanity/