The only time you get screwed on a push is when you begin to bleed players or lose your medic without winning the fight. You can push in on what appears to be no advantage, and if you do this fast and with proper buffs, you may very well catch them building uber slowly/not buffing properly. Even if it was actually even, you have the liberty to roll the dice in terms of DM once on their point, then if you lose the fight, as long as your combo gets out and you don't slowly drop players one at a time, you can back up behind your point and meet with your spawns for another 6v6 before they cap it.

I think that a halftime break would just slow things down unnecessarily. No matter how bad your newbies are, you can fix them to a playable level by saying "don't jump in spawn" or "heal the demo first". I always just take the time while people are getting in the server to talk rollouts and what people's options are on mid, and I rarely encounter someone who's completely unable to rollout after that. After that, you can easily keep up will calling simple things like "push here" "back up" "scouts come in" etc... whether or not your newbies execute these things well, I think being in that environment is the best way to learn and stopping the game in the middle to talk about them missing some of the finer points seems unlikely to fix anything. I prefer to just play the game then talk about it/answer questions after.

Yeah, I thought you were talking about the old thing where you could stop it entirely. Didn't even know that it paused during weapon switch, not sure if that still happens.

I admit that I haven't checked if flashing a bunch of people then switching weapons slows the drain rate, but I don't think it does anymore. At most, it might reset multiple flashes to the base drain rate, but I doubt it. Either way, it's not very practical because you almost always need to be ubered yourself while flashing, and the weapon switch + switch back delay is more than enough time for someone to react and shoot you. In the past (up until early 2008 iirc) your uber would just not drain at all if you switched off your medigun, so you could keep someone ubered for like 30s, but the medic wouldn't be ubered because you lose uber instantly and don't flash when you switch weapons (and thus it was rarely useful).

You lose 0.5s for a single flash. That is to say, if you switch targets once then switch back, if you have 0 delay between breaking your beam and reattaching you lose 0.5s (which you don't, so in practice flashes cost more like 0.6s). If you're keeping 3 players fully ubered the entire time, then you're doing 8 flashes.

One misconception you have is that 3 players = 225%. It's actually 200% (4=250%, 5=300%, etc).

TF2 constantly checks throughout the uber to find the drain rate -- once a flashed player is no longer ubered, your drain rate returns to normal. Specifically, the game checks for the number of players that are flashing and you're not currently healing (note: your drain rate increases if you uber 1 person then break the beam as if you had just flashed someone). In your hypothetical situations, the drain rate is 100% when you pop on the pocket, 150% for 1s while the roamer is flashed, then 100% after. If you give the beam to a third player, it increases to 150% for their flash (or 200% if the roamer is still flashing) then goes back to 100% once no-one is flashed.

Perhaps an easier way to look at it is that you have 8 seconds of uber + 1 second flashing at 0%, and every time you flash you lose 0.5s.

Here are some more details:

That's a fake article.

http://www.youtube.com/watch?v=EGjhg9Hb-GA

If you find a team willing to do this, good for you. Though to me, it seems like it would be easy enough to pick up scrims above your level particularly early or late in the night when there's not much around, then get just as much out of it by watching the demo instead of finding a team that will give up time that they could be using to improve.

If your issue is not knowing what to look for in the demo, you can either get one person to help you with that, or just ask yourself these general questions:
-were our scouts buffed on mid/before we pushed?
-were our scouts with our combo when we have numbers advantage?
-did someone just fight a 1v1?
-could we have pushed faster?
-did we overcommit?
-did we engage properly uber vs uber / uber vs none / none vs uber / none vs none?

VODs are up here:

IM vs Watch: http://www.twitch.tv/sigmasin/b/345907809
dtp vs 6C: http://www.twitch.tv/sigmasin/b/345919455

will also be on my youtube later: http://www.youtube.com/user/SigmaArcSin

Option A: Shove them in spawn with the uber and cap before the uber is done.
Option B: Kill them with uber and cap when they're dead.

Many people attempt to do B, but they fail because the other team uses spawn. When you're pushing any last with an uber advantage, you must be aware that they might cycle through spawn and you must be ready to cap as soon as they do (rather than standing outside of spawn and putting rockets on a heavy sitting on a resupply cabinet).

also doing dtp vs 6c now

I'm going to cast this match tonight: http://play.esea.net/index.php?s=stats&d=match&id=3079221

Going live when it looks like it will start: http://www.twitch.tv/sigmasin

If it doesn't get played for some reason or ends quickly, maybe I will cast one of the other games.

sizin' with cyzer (cyzin' with cyzer?)
bannin' with b4nny
taggin' with tagg
lannin' with lansky
shruggin' with shrugger
shadin' with shade

There's solutions on that page though, so I didn't link it.