Tourney leagues pretty much have to run on whitelists unfortunately, they need to have the semi-same version of the game as the league goes on for many reasons. Letting people just pick/ban whatever they want for each match is unbalanced in the grand scheme of the tournament because different matches are playing different versions of the game. I won't get into the theory behind it. Think of it like how captain's mode in DotA 2 doesn't have new heroes in it. Unless you want the matchmaking system to have static whitelists which would be useless you're not going to get them to work the same way as the leagues.

Well, firstly, aside from stock weapons, which just can't be banned, having preset bans completely destroys the point of the entire venture. For example, the wrangler is essential for engineer play, so it might find a way into a "permanent" banlist. It also needs a little nerf. So, in these lists, it would probably end up with a high proportion of bans. Valve would see this and nerf it appropriately. Thus, with a small amount of inconvenience, a long-term problem (too much sentry power/whatever) is solved. If this wasn't the case, it would never see change.

This brings up a similar point- specific weapons have different reasons why people don't like them- so (following on from earlier, with each player having an "offline" system of picking their own consistent banlist, which I think is a great idea) having a checklist asking why people don't like a weapon would help Valve make the appropriate nerfs, rather than changing the weapon in completely the wrong way, like making the Wrangler only able to target people long-range, or something.

To clarify, I think when you make your personalised banlist, you should have a series of boxes to tick that ask why you think the weapon shouldn't be used. These could specify "Does too much damage," "annoying," or "Gives player too much health" for example. They could be changed slightly for each weapon, so that it makes sense, because "Too much damage" doesn't apply to Bonk. Then, Valve would have very clear details about what weapons the majority of people don't like (maybe taking into account time spent in the PickBan HL games, so that it makes sure that people who are completely new don't get too much power, in case of a misinformed opinion), and their job would be easier, so we'd get more precise results.

#92 Why not just have checks next to each of the item's stats and then an extra "bugged" one?

I'm really happy to see that a lot of people are starting to see the actual point of the pick/ban system, which is to force adaptivity and metagame diversity. With that in mind, here's my proposed system for the pick/ban concept:

Beat me to it, but yeah, this to start. However, the system I'm thinking of imposes a more Dota-esque pick/ban style that goes as follows: Each player starts with one ban in either Primary, Secondary, and Melee. They get to choose in what order they would ban them in. The drawback to this is that if you intermix the banning and picking, then players have to think about what they want banned most from a class before the opposing player picks it for use. So the final system would look like this:

Start with scout. 1 Ban on either a Primary, Secondary, or Melee. Switch to the other scout, he picks a ban. then move on to the next class. Continue on down the list, continuing to do this with each class until the first round of bans over. Then starts the first picking round.

Starts with the scout again, where he may select 1 weapon from any category to be picked and allowed for use that game by whichever side picked that item, meaning that the opposing side can not pick and or use an item that has already been picked by the opposing side. This applies to all classes and in all loadout slots. Any items left over from the pick/ban system are auto-banned for that lobby, but for valve, are recorded as a different statistic.

This same system continues for 3 rounds of pick/ban. At the end, you have a pool of weapons that are not necessarily the weapons they are used to playing against or with, but at the same time, are forced to adapt to and deal with.

One very important thing to note is that the only weapons that I think should be able to slip under both the "if they're not picked they're banned" clause and the "if I picked it you can't use it only I can" clause are the stock weapons, as well as a very selective list of unlocks that would be decided by the top levels of tf2 HL comp, through methods such as an official vote, in which only the leaders of any teams in platinum in UGC that has been in platinum for a couple seasons, would be allowed to participate in. This selective list of unlocks, such as gunboats and kritzkreig, which these top HL players would be able to agree upon these weapons being a wholesome good to the game mode, would also not be auto-banned if they didn't get picked, in addition to all stock weapons. These weapons would also be available for use by either side, whether they are picked or not. This would not protect these weapons, however, from not being banned. All that would happen is whenever one of these weapons is banned, the player on that class on the opposite team from the person who banned it, would get an extra pick on that item slot in order to be able to recover and still have diversity mid-match on that item slot.

I have tested this system with some friends, and have found it to be extremely successful once the match gets started. The only problem I ran in to is how long it takes to actually get started, since you have so much picking/banning to do. However, if this entire system were completely automated, I believe it would only take minutes to complete this process from start to finish. The best way to enforce this would be with another Dota-esque concept, which would be a time limit per ban/pick round.

TL;DR: This pick/ban system has the potential to spice up the TF2 competetive scene, and should be embraced with vigor. In the future, if this system were to be enforced in a 6's lobby system and maybe potentially league play, it could lead to some pretty big upsets solely due to getting, for lack of a better term, "out-drafted".

I think leagues banning new weapons (maybe even bugged ones) is an "acceptable" difference and wouldn't really throw off a transitioning lobby-er that much.

#95 They also need to keep items banned from the start banned the entire time.

Well, that's one of the potential boxes, yes. The point is there are many to choose from.

I think I graduated to this thread, this may have been mentioned but:

The way I understand it, Valve wants feedback on what works and what doesn't correct? Meaning if something is banned over and over again, it will be nerfed/fixed/looked at. My suggestion is:

1) Start with the UGC whitelist, focus first on all weapons currently banned and let players start to pick/ban from those items (only) in pugs. When people consistently ban the ones universally disliked (should be all of the already banned HL ones) those will be the first on the fixing table at valve.

2) Then talk to experienced HL players about weapons that are annoying/could possible be bannable and begin the process again for round 2.

3) Keep going

There should be a bottom up approach. Begin on equal footing where everyone can agree X weapon should be banned, then start to get into the muddier waters. Basically this allows you to focus on groups of unlocks at a time, starting with the most disliked and moving up. That should greatly reduce the chaos of the whole process.

Valve should then see that certain weapons universally disliked (and thus voted out) need a look. Conversely, if all of a sudden player keep voting in something HL bans, then maybe HL leagues need a second look. Am I understanding this correctly?

We played a pug with this where all weapons were allowed and then each team got some time together to pick weapons to ban. Each team got 5 bans on non-stock weapons, we then took it in turns to say what we wanted banned. If the other team said something already on our list before we said it then we'd pick another weapon to ban.

People should be officially uploading stuff to the feedback results thread soon.

Keep in mind that typical TF2 rounds last maybe 10-15 minutes. A public lobby / matchmaking system is not likely to have the same people stick around for a full match of 3-5 rounds. If we have 2-3 rounds of pick/bans with each person having a mere 20 seconds to deliberate, that is 6 full minutes per round, or 12 minutes for 2 rounds of simple pick-banning. That's why I suggested a faster automated system in my post.

Just played an HL pug with a ban system, thought I'd share my thoughts.

Also I should preface this by saying that this is the only HL pug I've ever played so I have nothing to compare it to (and thank god it was on a 6s map or I'd have done even worse).

http://logs.tf/45234

We did eight bans, with each team picking four. I think we did 1-1-1-1-2-2, but I wasn't really paying attention cuz I was eating Arbys.

The bans were:

Razorback
Jarate
Wrangler
Flare Gun
Pomson
Phlog
Gunboats
Dead Ringer

The very first thing I noticed was me getting anally destroyed by the enemy spy because I'm a shitlord who can't handle good spies without a Razorback apparently. As I waited to respawn after being backstabbed for the 6th time in one round, it occurred to me that this ban/pick system might actually work better than I first thought. It's not really that different than, like, SSBM picks or something.

You attempt to restrict your opponents strongest players by relegating them to weapons they may not be as effective/familiar with. My team called bans on Jarate and Razorback immediately because the other team's sniper was HL plat, under the assumption that I wouldn't be able to control him. They were right, pretty much. Anyways, it gave our spy the option to focus their sniper a lot more than he could without the bans.

I expect to see weapons like Wrangler, DR, Jarate, and Vita Saw to be the most frequent bans. This is because these weapons aren't skill indexed. Nobody is 'good' at throwing jarate, or aiming the wrangler. They're basically ever present features of the match, that are good for one player and bad for eight. So they'll be very common bans, for the sake of the other eight players. The only exception to that rule is the DR, but with in it's current state it functions as a non-skill indexed weapon in a lot of situations.

---

Despite all of the above, it didn't result in any exciting, game changing strats that I could see. This might have been because it's KOTH, but I think it's because we spent all our bans on the shitty, badly designed, and annoying weapons, with two notable exceptions being the Flare Gun and Gunboats (which were mostly for experimental purposes anyways). Banning stock weapons is pretty much out of the question, because it assumes that all players have alternatives when they might not.

If weapons like Jarate/Wrangler/DR/etc were to get some appropriate changes, I think then we might see some more exciting strategies.

How about this I have complied a complete list of the unlock weapons below. I would like to compile feedback data, Before this lobby thing gets implemented.

All you have to do is pick the 18 weapons that you would like to see banned.

http://kwiksurveys.com/app/showpoll.asp?qid=164684&sid=qdn4r4wyvdtr8dx164684&new=True

Why would you ban weapons that arent skill-indexed? Why would you waste a ban on something like that?

I would just ban the stuff my team isnt good at dealing with. Barring that, Id ban stuff that the other team "needs", but my team doesnt necessarily have to have. Id only ban the vitasaw if i thought my medic would really benefit from the other medic melees. If he doesnt, why bother?

Because it inhibits the Scout, Soldier, Pyro, Demo, Heavy, Engineer, Medic, and Spy. By banning this one weapon, every single other class on both teams becomes more important. Because Jarate isn't skill indexed, it basically becomes a permanent fixture in any game the Sniper is running it. Banning it is banning the fixture, and giving everyone else some more breathing room.

Same for the others

Dead ringer banned until krits do full damage.

each class, from each team selects 1 weapon from each load out slot; then all weapons picked cannot be used by either team. As for the order of selection red scout - blu scout, blu soldier - red soldier . . and so on

-----

All stock weapons cannot be ban.

There are no permanently ban weapons

No hat bonus sets

Each team can undo any one [more?] pick of their opposing team, at the end of the pick-ban selection ]

-----

This system would allow for Valve to see which weapons are ban more frequently, and possibly see which weapons give an unfair advantage when used; as well as potentially allowing them information on adding weapon to the white-list of always allowed weapons, and the possible formation of a black list of perma-ban weapons

Not to mention this method has VERY competitive aspects to it.
-You can pick weapons you do not want to fight against
-You can pick weapons your opponent is better with
-You can ban items in such a way to enforce a strategy or method of game-play
-You can play mind-games by banning certain things
-You can add more variety to the game by being able to un-ban one weapon

This allows Valve to see every weapon being used, the the amount to which it is being used, while still creating a pug system which would bolster the competitive scene. i posted more on this on Sal's how to get in to comp lobbies thread - read more there if you have nothing else to do [ http://teamfortress.tv/forum/thread/8952-how-to-get-to-in-game-comp-lobbies/16 ]

As I understand these are Valve's priorities:
1) Make a highlander lobby system accessible to pub players.
2) Get lots of data about which weapons need fixing.

Having a perfectly "balanced" whitelist is irrelevant to both of these goals. It may turn out that all of UGC's banned weapons are consistently banned, but they don't need to be banned in every game.

To make these games very accessible, I'd say just let each player ban 1 weapon from their own class. The ban would apply to both teams. It could be a simple check box and the ban could be blind. Overlaps would be allowed (so both medics might ban the vita saw or hippocrates, for example).

This system isn't complex and lacks some finesse, but it should still show trends when analyzing large sets of lobbies.

There's no need to ban any "obviously" OP items (set hats, for example) because they'll come out in the data [and the goal is to tweak these items, anyway, not simply disallow them from comp lobbies], and well regarded items (such as gunboats) will not likely be banned because a soldier wouldn't want to prevent himself from using them.

Why not just a simple nominate system? E.g. all items are allowed by default but if a player feels that a weapon is overpowered/unbalanced they can nominate it for a vote by the players in the lobby. It could have a simple popup screen similar to when you find an item with an option to ban or allow. Once a player is happy with the weapons they have nominated they can "ready up", and the process continues until everyone is ready. If someone trolls by nominating every item in the game they can surely be removed by a votekick?

A player can nominate any item as many times as they want (although repeat nominations are not allowed). This accommodates for people who have a) aren't familiar with the system in place or b) item x being overpowered had not occurred to them/they had forgotten about.

However, this system doesn't yield any data regarding which items people consistently want to use (the proposed "pick" function). But surely statistics regarding in game item use of the remaining weapons will give enough data for them to work from, since it is a "play to win environment" people will pick the "best" items rather than the most fun ones like they would in public.

I feel like we should get some "great guide to what we're thinking about" together (even if it's heavily summarized) if we're going to do this very well. I'm giving it a shot, but I know I'm not the best at these things since I literally just list stuff.

Matchmaking concept:
These are my original thoughts.
- People may sign in to play Highlander (in the future, perhaps and/or: 6v6 and "esoteric" formats like 4v4 and 8v8) as a starter exclusively or also as a sub/backup (which will put them into active lobbies/games which have lost or kicked a player matching their settings)
- People may sign in to whatever classes they want to predominately play (keep in mind that locking people onto classes is a naive idea, just keep classlimits in place and have some sort of grief reporting system if necessary)
- People may sign in to what kinds of map they want to play (CTF, Payload, A/D, "small" 5CP [like gullywash and badlands] or "large" 5CP [like foundry and well])
- People are matched according to whatever the best criteria are available to judge them by: performance in what level of matchmaking game, presence of league medals in their inventory (since it's the closest thing to an official steam-side tournament link -- fix me!), hours spent playing TF2, etc. If this is a bad idea, by all means, ignore it!
- There are no captains, players are placed into a lobby once their match is made and they've accepted it. Matches are started simply by everyone connecting fully with a small pre-game period like DotA 2's hero selection system to sit in the lobby, during which weapon voting can be done.

Weapon voting concept:
There are a bunch floating around, but I'll do my best to summarize the main ones.
- A nominate/vote system, working similar to some servers' map voting. When you get into the lobby, people are asked to nominate weapons they want to ban, which are put to vote. This would probably provide data and expand meta roughly evenly.
- A raw democratic system, where people vote ban/keep on each item in the game at the start of each match. There several ways to handle the threshhold for banning and to handle tiebreakers. This would probably provide data and expand meta roughly evenly.
- A save/ban system that emulates DotA 2's general idea of a pick/ban system, but with some kind of dynamic that changes depending on how same-y the banlist is -- if everyone in queue is treating the same items as first-saves/first-bans and second and so on (making it seem static), the length of the voting system would expand; if it seems like people are treating it random/arbitrary, it would shrink.
- Some system that starts with a list based on the general state of whether people ban-or-not each item, and then gives one of the above systems to deviate from that list. Would especially work with the nominate/vote system, and is similar to what ESEA's process already [socially] does. This would probably provide good data more than expand meta, but depends heavily on what other system is used with it.
- A rotation system where every class gets X guarenteed items and Y "extra" items that are rotated through/scrambled through the item pool. Some number of items is somehow banned from this basis item pool. This would probably expand meta more than provide good data.

^ Great summary, just to add in, I think the other big question would be whether there will be an elo/ladder based system so that people can create proper teams and play for a ranking. If there is such a system, I would definitely use the pick/ban system far more frequently as it offers an incentive to keep playing, and my "team" can keep working together as a group.

I don't think the majority of 6s players understand how little some of these unlocks actually affect the game in terms of HL. From my 1 season of HL the only things worth banning are
- All the FULL sets
- Pocket pistol (and possibly atomizer)
- Wrangler (if not the rescue ranger), take 1 or the other, it's stupid playing against both
- Beggar's Bazooka
- Huo-Long Heater
- Vita-Saw
- Enforcer
- Reserve shooter

Vanilla weapons should never be banned and the bans shouldn't total 18 weapons. Maybe something like 5 for each team. When you start to take those 7 out of the game (similar to UGC whitelist. Also take note that people refuse to use reserve shooter out of pride and other similar weapons.) it plays normally and removing more weapons usually does nothing but screw over other players. If Robin wants quirky plays it won't come from banning every weapon that isn't vanilla, it will come from removing what destroys the game and allowing everything else. They are all side grades and the majority aren't even worth the swap.

Ex. You won't see any of the top plat snipers use sydney sleeper over stock/machina, you're just asking to lose to the other sniper if you do. Jarate makes them move into the fight and most snipers I've seen like to sit at the other end of the map. A close sniper is an easy sniper to kill.

I support this 9v9 lobby system but I think we need some of the well known HL players to chime in instead of speculating with 6s players that only have HL pugs experience or think it's a glorified pub. It's a whole different experience.

But these bans are neutral. If a weapon would be used if you didn't ban it, and it would be just as good for either team, then banning it helps or hurts both teams the same amount, so it's an irrelevant/wasted ban for that team. Teams should ban weapons that the opposing team would make better use of, not weapons that helps both teams the same amount.

Just because two engies are about equal with the wrangler, doesn't mean they'll be about equal without it. The idea is you'd learn how to play without "normal" unlocks then ban them so the other team is stuck with a loadout they're not good at.

It swings both ways. If my engie had shit dm, I'd ban the gunslinger.

Quote of the thread, folks.

From playing in a pick ban system yesterday and thinking about it for a while, I think the biggest problem facing a TF2 pick/ban system vs. one from Dota is that unlike in Dota, people in TF2 don't all have access to the same things. Unless the game were somehow tweaked so you could "borrow" weapons, not everyone has access to every weapon in the game. This poses a few problems.

1. It prevents the banning of stock weapons. If you don't have any alternatives (which new players often don't, or people such as myself who don't keep unlocks for every item slot in my inventory), then it's impossible to ban say, the rocket launcher since the soldier might have no other primary weapon to use. This means that using the stock in most situations, which is often the best option anyway, will likely become the most common playstyle.
2. Some weapons are straight upgrades, or pretty damn close. For example, I was using the Vita Saw in the bans pug I played to see how it would affect gameplay. I don't believe the other medic had one or if he did he wasn't using it. If you look at the logs http://logs.tf/45234 we had about the same number of deaths and we were both using the kritz, but I had WAY more opportunities for uber than my opponent. That's just plain unfair.

The pick/ban system was fun and I think it's a good idea, but it has some fundamental flaws that require fundamental changes in how we view TF2 weapons in your inventory.

@ 116:

1) You conclude that because you can't ban stock weapons, using default will become the most common playstyle. How is that a problem?

2) The fact that some weapons are straight upgrades means that they'll be banned more often, flagging them as in need of revision. Again, how is this a problem?

all I want is in-game 6s
qq

But it's not good for both teams, it's good for both snipers. Balance wise it's even, you're right. But banning it gives the other 8 players on your team more room to work with.

I hope to see bans like this appear frequently, because fundamentally broken weapons like these are worse than OP ones like the Enforcer, IMO.

I would hope that given the interest the TF2 team has expressed in balancing items, that if the stock weapons are becoming clear favorites in the system, that other alternatives will receive buffs of some manner and hence encourage players to keep at least a few extra weapons.