well I have noticed some sandbagging players who were lowering their rating on purpose, but it's very easy to catch them 
so I don't really think it's an issue, these guys have a short life
also, probably the way rating is calculated in teams need to be fixed, the average is a bit unfair since the level of a team is much closer to the higher rated guy level
Sandbagging in Teams is cheating. Sandbaggers will be banned. If a low rated player joins your match, and then resigns, disconnects, or clearly loses quickly. Please report them.
Partnering (by invite) with a sandbagged player is also cheating. Don't do it.
Said top player @trinhlan has been removed from the leaderboard for teams, so it’s clear that some action has been taken. They are also clearly cheating in FFA by teaming up with partners or multies.
I have met people in FFA who resigned on move one. I thought it was a connection issue or whatever. Now I'm starting to think that those people are sandbaggers. After the game, I met that same person (who resigned) in the next game, guess what he did? He resigned. Sadly, I forgot his username, so I cannot report him.
@ignoble, I like what you said.
There should be some automation of this process. For example, it is easy to collect the stats (on the server side) about those who disconnects often at the beginning of the game. And notice those who used to bee 1700, but now is 1200. And those who lost 20 games in a row. Then a moderator can take a look and decide if it is a cheater or not in every particular case.
One fix could be to use the highest rated player in each team into the calculation instead of the average rating. I think that would be fair, but I imagine playing with random partners would become even more frustrating that way.
well I think it should be an weighted average
es. player 1 1500 player 2 1200
average (weighted) = (2*1500 + 1200)/3 or (3*1500 + 1200)/4
I don't think a weighted average would solve the problem, but only make the problem a bit more mild. Still it sounds like potentially a good compromise though.
But overall, as long as the cheaters get punished there might not be a real issue. I like what @ignoble said 
we should do it so that if your teammate resigns or time outs, then you take over their pieces (as someone suggested this elsewere)
then if your low rated opp resigns to sandbag, you might still lose
Treating sandbagging as cheating, and simultaneously creating incentives for players to manipulate their rating (in other words to sandbag), creates a funny situation. The incentive is the huge increase/decrease of one's rating after every single game. It is not realistic to assume that a person's playing strength was increased or decreased by 50 points after a 30min game. But most online chess sites make such unrealistic assumptions, just because most players love to manipulate their rating. Make the rating changes realistic, something like +/-3 points max for every game, and watch the players leaving in masses.
That said, the current implementation of ratings at 4-player chess is close to the extreme of the spectrum. You play 1 game, and the system assumes that you played three. So the rating changes are usually 3 times bigger than the already unrealistic global standards. Something needs to be fixed here, IMHO.
I'm glad that said player has been taken off the leaderboard. And sorry about making a public accusation, I should know better. As far as I know this was the only player who did this. She pushed it to the limit, playing with a user rated 500! Imagin how many games you have to resign to end up with a 500 rating! Such a complete disregard for all the others, what a shame.
Also, same user is on top the leaderboard in Ffa, and there has to be a trick. It's impossible to get to 2100+ in the current state of affairs, because there are no other 2000+ players to play with, and it's impossible to win all your games unless... Unless you have multi accounts, and you know how to change your IP, and you seek many games at the same time and often enough you get paired with yourself.... There must be some kind of trick like that. This player is not that good. Wraku89 got o 2050 and it's a miracle, but the man is a GM, and a really powerful player. but 2132 at this point in time? No way, there is no way.
I know all the other players on top of the leaderboard, Rezacz is just a really strong player, so are wraku, a1t19, Dragon, terjanq, shineko... 1so cheating in team mode isn't a widespread issue, as it can be in ffa.
I think Bad_Dobby_Fischer's idea is a good one. In team, if a player resigns, disconnects or times out, the remaining player could take over his pieces. That would resolve pretty much all of the frustrating situations that happen in team play besides uncooperative partners
I didn't actually know the player did this, so my post was just directed on how it is possible to cheat. Not meant as public accusation. I'm surprised it was as extreme as 500 rating partner.
I also think Ne2 is correct in his observation on the FFA leaderboard. Since teams was released you can not play with as many high rated players as you could before, so getting high ratings is really hard. Players like me with 2000+ rating is just not playing FFA. I would not be surprised if I would drop below 1900 if I started to actively play again.
A couple of thoughts.
I to think the rating for (high low partners) system needs to be changed/adjusted. The reality (for me) is if your a low rated player in teams and the higher rated player is using arrows showing you what to do for your move your kind of fool not to take the experts advice (*never said it was fun). In which case the other team is not playing an average of the ratings but the top rated player. That should weight much more then the lower rating players rating. (don't mind giving the lower rating player full points as that averages itself out quick enough but the top player needs to have an adjusted down + score IMO.
On another note... I think just because you are a low rated player in normal chess it stands to reason you will be a low rated player in team or 4-player chess is a leap. Sure most people will fall into this grouping but the game is different enough that someone that may not have the best skills (GM) in normal chess can use some political talk, some duck and cover options, or understand the dynamics of the 4 people better than the black and white regular chess game. So I would caution that thinking all players must be about same in score is a dangerous and quite possible wrong assumption in some cases.
The player who @Ne2willdo is referring to has asked me to team up with her in an FFA game which I said no to. After that, I have spectated on some of her other games, and she has blatantly asked other players in the chat to team up also.
As Ne2 and @Martin0 have said, it is almost impossible to reach 2000 in FFA now, after the teams mode is out, since most top players are playing teams now. Yet the said player somehow reached 2132 in the past week, even though she wasn’t even on the leaderboard before that. Strong players like shineko and Collisionofminds, who were 2000+ before are now in the 1800 range.
It’s good that the said player has been removed from the leaderboard for teams, and their recent antics in FFA should also be worth investigating.
On another note... I think just because you are a low rated player in normal chess it stands to reason you will be a low rated player in team or 4-player chess is a leap. ... So I would caution that thinking all players must be about same in score is a dangerous and quite possible wrong assumption in some cases.
You are right regarding the FFA mode. I have a blitz rating of 1650 (classic chess), but managed to get to the top 10 of 4 Player Chess FFA. The same with Martin0 (his blitz is 1700+). However, most of the FFA top players have 2000+ blitz rating. (Wraku89 2500+, Fukadzume 2200+, try2see 2000+).
The Teams Mode seems to be more sensitive to the classic chess rating of the player. It does not mean a 1700+ player could not be in the top 10 list. But it is much more difficult to get there.
One fix could be to use the highest rated player in each team into the calculation instead of the average rating.
I like this idea.
This has occasionally been a problem in bughouse too.
Using the highest-rated player only works if the higher-rated player controls the team. If his partner doesn't follow his advice 100% (or if he doesn't tell his partner what to do for every move), that won't be an entirely accurate way to measure things.
Today in the chat @Ne2willdo said that one of the top players is cheating. I asked how exactly. @Ne2willdo did not answer yet, but Martin0 said:
"Currently rating in teams is calculated based on average rating, so one easy way to cheat in teams is to have a partner that has sandbagged his rating."
I did not know about this cheating way before. I was sure that Teams Mode is protected from cheating. Now I am not sure.
So, if one player has a rating of, let's say, 1700 and his/her teammate is 900, their average rating is 1300. So, if they play vs 2 x 1200-1300 guys they will earn noticeable amount of points. If that "900" account is a second account of the same player or his/her friend (who is actually 1700 too) then it is definitely cheating.
If both teammates have 1700 and play with their real (not secondary, "zombie") accounts then in a game vs. 2 x 1200-1300 guys they get only +2 or +3 points for a win. This way they need to win 40 or 50 games to become 1800.
Any ideas how can it be fixed?
Maybe something like "You can invite a player to a rated game if the rating difference is less than 300?" But this does not fix the case with random teammates. Set an auto-range +-300 in matchmaking for random teammates?