Gank squads have three primary advantages: High-burst damage, coordination, and numbers. They rely on the fact that they're better prepared than their victims, at least at the outset of the engagement.
If you can catch them attempting this early, an aggressive warding policy and a decision to shut them down can go a long way. However, this level of coordination is hard to pull off in a non-premade game. Individually in your lane, however, something similar can be accomplished: Aggressively ward the entrances to your jungle, where they're likely to come from and play conservatively. Since they aren't farming, they must get kills in order to scale. If you are premade, and you notice multiple members often missing from lane, aggressively ward jungle entrances and try to catch that squad unprepared.
Obviously, the other problem is their burst. Because burst damage is typically magic damage (at least, until the game gets out of control), a Hex Drinker can go a long way. Any magic resist, however, is a big plus. Since they have to be aggressive, having sufficient magic resist can bait them into overextending into a tower dive. Assuming you have an equal number of teammates, that plus the tower damage can turn the fight in your favor.
Even if you die, your team should try to use the fact that they saw 3 enemies in your lane and try to make the most of it, be it being aggressive in lane knowing the jungler can't gank right now, taking a dragon/their blue or pushing down a tower.
So if your are absolutely positive you are going to die, try to spend as much of their time as possible, giving your team more time to do something else elsewhere.
The ultimate point being that after a few failed attempts, the gank squad won't be strong enough to be as big of a threat.
First of all, let's clarify the question. This situation will only happen during online tournaments. During offline LAN events, players have infinite RP and IP as well as every champion purchased. During online tournaments though, players use their standard LOL profiles. As a result, there is a decent chance that some summoners will be missing heroes. When that happens, a placeholder champion that never gets actual play like Eve is chosen instead and replaced with the intended hero later on.
As for why they don't just pick the heroes as soon as they are available? In the current meta, the top and middle lanes are solo lanes. Since solo lanes are 1v1 battles, having to face an enemy who counters you can be too much to overcome. As a result, teams try to pick these two lanes last, in order to avoid getting countered.
Best Answer
There are several different things that determine good or bad picks in League Tournaments or 5v5 Team rankeds in general.
TL;DR Answer
Bad picks are picks which:
In-Depth Answer
First of all:
The tier list
As most people know there is a tier list for solo queue. But since you're asking about the OGN/LCS games I will take a bigger look into the 5v5 Tier list.
The Team tier list isn't too different from the soloqueue one but it's not that relevant. The soloqueue tier list focusses on the potential of a champion carrying a game alone. The Team list focusses on the team(fighting) ablitiy of a champion. You also have to keep in mind that the champion tier list itself for 5v5 isn't that relevant just because there is another list of team compositions and strategies. This brings us to the next point:
Team compositions
The team comp is the most important factor in team rankeds/tournaments. Currently i would say the most used comps are:
Wombo Combo
The wombo combo team is the most used teamcomp within unexperienced players since it's easy to follow. Most unexperience players will just think "Hey if we win teamfights we win the game so we'll just pick amumu,kata,mf,leona,renekton". This might be true but in higher level play the wombocombo can be easily countered since it really relies on teamfights and going as 4 or 5.
Pro Strong teamfights, can turn a game instantly, sieging against such a team is really risky
Con Needs to teamfight, weak when alone, actually the team that get's countered the most.
Exaple champions Leona, Amumu, Miss Fortune, Ziggs, Orianna
Strongest phase Mid-Late
Movementspeed comp
The movementspeed comp became more popular during season 3 especially with the support and sivir rework. It's one of the stronger but harder to pull off team comps since you have to time right. It's also more or less a teamfight composition but you aren't dependent on teamfighting as much as a wombo combo team. Your strenght lies in structure control and the ability to engage and/or disengage fast. You could for example go as 5 to mid when only 2 people guard the turret, take it down and use your movementspeed to get away again
Pro Very fast, Good teamfights, Strong disengage, quck turret kills.
Con Sieging/Defending isn't that good usually, most comps lack cc in supporter or toplaner.
Example champions Sivir, Lulu, Karma, Sona, Mundo, Shyvana, Udyr
Strongest phase Mid-Late
Poke comp
The poke comp is also another rather old and commonly known composition. It's basically just a bunch of champions that can sustain themselves and/or the team while poking the enemy down and forcing them to go back. This means that this team excels at Objective control (have you ever tried to do baron against nidalee/Jayce??) and sieging but it's greatest weakness are of course teamfights. This comp is pretty strong against movementspeed comps but it usually gets wrecked by a well played wombo combo team.
Pro Great poke, great sustain, Best sieging potential, Best objective control
Cons Cannot teamfight, movementspeed reliant, usually low amount of hard cc and disengages.
Example champions Nidalee, Ziggs, Jinx, Jayce, Varus, Gragas
Strongest phase Midgame
Global ultimates
Well as the name suggests, these are mostly champions with global or semi global ultimates. This comps strength lies in surprise ganks and splitpushing while its biggest weakness is it's champion pool. There aren't really many champions with global ultimates and most of them got nerfed since this method was an almost guaranteed win in earlier tournaments. They don't really get countered by any of the other comps but they also don't really do that well currently. This is mostly because of all the champ nerfs.
Pro Split pushing, good objective control, ability to teamfight anytime from anywhere.
Cons Champions are weak.
Example champions Nocturne, Shen, Karthus, Pantheon, Twisted Fate
Strongest phase Midgame
Catcher comp
This comp focusses on catching champions off guard and/or taking down high priority targets like an adc quickly. They usually have lots of single target burst and one or two assassins that can get in and out quickly. Their biggest strenght lies in the early game and it forces the enemy to stay in a group since they can't really risk staying alone and getting caught. Their weaknesses are teamfighting and their dependency on snowballing. If you screw up early game with a comp like this the game is pretty much over.
Pro Target lock down, Control of the enemy team, Splitpushing.
Con Teamfights are hard, If you can't kill the high priority target the teamfight is over, Really risky to use in high level play.
Example champions Lee sin, Kha'Zix, LeBlanc, Elise, Thresh, Leona
Strongest phase Early-Mid
We-Play-What-We-Can-Play-Best comp
This comp usually leads to the "we lost because of bad picks" comments. It's usually playing champions that people are the strongest with but ignoring the real composition of the team. While this may work out for solo queue it's usually a bad thing to pull off in team rankeds/tournaments. The only cases it really works is when you pick a guy that play a champ well and you build the team around it (Overpow on kayle for example). The problem is that your main champions will get banned in tournaments and very high elo teamrankeds since you exactly know who your opponent is.
Pro You can play what you are best with
Con Your champions can be banned, Rest of team has to build around your champ.
Example champions Insert your favorite champions here
Strongest phase Insert strongest phase of your favorite champion here
Of course there are some other comps but these are the most frequently used ones that can really be categorized
Now to the last point:
The enemy team
In Challenger team rankeds and tournaments you will know who your opponent is. Thus you are (or should be) informed of their main champions and their main strategies. When picking or banning champs on such a level you can't just go for the "Oh i'm going to ban kassadin because he's OP" tactic but you have to pick and ban AGAINST the enemy team. This is also really different from soloqueue. While in soloqueue you should follow the order:
Pick what I can play best > Pick what matches the team = Pick what counters the enemy
In High level premade play you have to go for:
Pick what counters the enemy > Pick what matches the team > Pick what I can play best (Of you should be able to play the champion you're going to pick decently)
Remember that you can ban away the main champ of their player with the smallest champion pool. This offers a great opportunity to win a game. Also if you don't fail completely with a certain champ that the enemy is good with you can pick it away from the enemy. When you don't follow that order this often leads to these players saying "We had bad picks". Because they knew "Player x is going to pick champion y and we got wrecked by it the last 3 games but I will try to Pick a counter to it". And instead of taking the champ away from this player they just tried to counter it.
A closer look into Viable champions
Now back to the different tier lists. As already mentioned the composition or synergy of different champions is much more important than the position in the tier list for team rankeds but this doesn't mean you can play Heal+Revive jungle soraka just because it matches your team in a certain way.
First of all we have to keep in mind: The tier list is not perfect It's actually just a list written by pro players based on statistics and personal experience. Also if you only play God-Tier champions in soloqueue you won't jump from bronze to diamond in one day. This list only describes the potential of a perfectly played champion and not the percentage of games actually won by this champion (If you want plain statistics check out op.gg). Of course this also means that certain champions that might be really strong most of the time can be absolutely useless against a special teamcomp.
In general it's of course safe to pick champions that are higher in the tier list but there are some exceptions to this (Talon for example is a terrible Team pick but is great for soloqueue). On the other hand there are many champions that are lower in the tier list but can be really strong in team rankeds (Take shen for example). You can basically play every champ if you know that it will work. The tier list and meta games are usually created during tournaments. Of course the patches etc are another factor but have you ever seen a toplane mundo before the popular guys played it? Mundo was a champion that wasn't touched for like a year or so and the mastery changes weren't that extreme. He was a viable pick before but no pro player picked him in major tournaments. Surprise picks also happened this season which will work out if you know how to play this champion right (Teemo pick in the TSM vs C9 for example). So another thing that can be a reason why people are talking about bad picks is either because they tried the surprise pick (High risk, High reward) and lost, or if they lost against it because they sticked to their usual playstyle and/or tier-list.
Preparation
Now this only counts for tournaments. You can prepare to a certain degree in high level Teamrankeds by knowing all the teams in your elo but you will only be able to exactly know what or who you are against the next (few) game(s). The preparing the team for the upcoming match is mostly job of the team coach. The players only have to put the theory into practice (sure sometimes they have to learn certain champions but when it comes to picks & bans it's mostly theory). I personally won two tournaments with my team because our enemies didn't prepare as much as we did. And by preparing I don't mean "Hey i'm gonna check out his lolking and then I know what champions he picks". No. I mean analysing the playstyle of every single player in the enemy team. You have to find out how they handle different situations. CLG is a good example of a team that relies on a good but risky strategy. They almost always play lategame comps. So what should you do? You pick a team that wrecks the enemy early so there will be no lategame. Finding out favorite champions of players is an important thing too of course but the playstyle of a player will tell you more than his match history. You have to watch previous matches, previous tournaments and maybe even scrim with them. As soon as you know how they play you can start counterpicking and counterbanning. I don't really know how much of a problem too little preparation is in LCS but I think even there you will sometimes lose because of it.
Also always remember that not kills, gold or baron win the game. Those help you winning but the actual goal is to kill the Nexus (sometimes I feel like even pros forget this). No matter how many kills you get. When your toplaner is always mid to gank, top will get pushed and maybe even to the inhibitor/nexus which will cost you the game. And there are only two results: Win or lose.