There are people on chess.com who simul (play both boards of a bughouse partnership). It's not clear whether simuling itself is allowed under chess.com's rules. Let's assume for the rest of this post that simuling is allowed. I want to discuss simul ratings, which is an interesting sub-issue.
Simuling would seem to open the door to rating manipulation.
It is much easier to fix games by losing than by winning. In order to win you must play very well... only a tiny edge can be gained by partnering people with complementary bughouse styles or exclusively playing with underrated partners or overrated opponents in the random pool. Having an engine feed moves is more difficult in bughouse than chess because computers are not significantly better than humans at bughouse. On the other hand, in order to fix games by losing, all you need to do is hang a few pieces or walk into checkmate or feed a piece that mates your partner or simply hit the resign button.
If you have access to two accounts, losing on one account can give you an advantage on the other account. In theory you could throw games to yourself or to an intermediary who in turn loses to your other account. Bughouse has an interesting wrinkle because when your partner is rated very low, that allows you to gain more points each match. Therefore, if you can make one account very low by losing intentionally, you can then partner that low account and make your other account very high.
To get an average team rating you add the ratings of both partners together and divide by 2. So a 1000 + a 3000 has the same average team rating as a 2000 + a 2000. This means if a person is simuling and his actual simul strength is about 2000, he can get one account's rating up to 3000 by getting the other account down to 1000 before partnering. This is a problem.
A rating is supposed to be a measure of your strength at a particular chess variant. There will be fluctuations up and down, but in theory if you are overrated (your rating exceeds your strength) then playing random opponents will lead to losses which push your rating down toward your actual strength. Similarly, in theory if you are underrated (your rating is less than your strength) then playing random opponents will lead to wins which push your rating up toward your actual strength. This makes ratings tend to be reasonably accurate if there is no cheating involved.
Accurate ratings are good for a number of reasons:
1) They help you see your own strength and notice whether you improve over time.
2) They help your opponents forecast how the match will go and know what they are getting into. They may play different openings or strategies vs. a very strong player than a very weak player. (For example, not worry about leaving open tactics that a weak player is unlikely to see.)
3) They help potential partners forecast how much help their partner will need and whether he is likely to listen, likely to win on his board, etc.
4) They help the community at large figure out who is better than others; and therefore who might be a good person to ask for advice -- or even seek paid lessons.
When a rating is wildly inaccurate, all of those things can go very wrong.
Simul ratings are an unusual animal though. If you are playing both accounts, your simul playing strength should be the same for both accounts. But if one or both of the accounts also plays with a partner sometimes, then there may be a discontinuity because simul strength may be different than strength with a partner.
For example, suppose I am 2200 strength with a partner but am 2000 strength when I simul. What should the ratings of my accounts be if I have 2 accounts?
In my opinion there are two ways of arriving at reasonable ratings:
1) You have 2 accounts and they do nothing but simul. In that case, you would expect both their ratings to be the same, matching their simul strength. In my example above, both accounts would be rated around 2000.
2) You have 2 accounts. One of them does nothing but simul. The other one both simuls and plays with partners. In that case, you would expect over the long term that the account that sometimes plays with partners would tend to that strength and the other account would tend to 2x (simul strength) - (strength with partner). In my example above, one account would be rated around 2200 and the other would be rated around 1800, which averages out to 2000. Neither simuling nor playing with a partner would fluctuate the ratings that much because both are at equilibrium.
Even though #2 is at equilibrium, it still would make the simul ratings a little confusing for opponents and observers... the 1800 account would seem to be playing above its strength and the 2200 account would seem to be playing below its strength.
But an unscrupulous person could act differently. If I were to sandbag one account down to 500 by losing intentionally, then I could partner myself and get the other account to around 3500 just by playing at my usual skill level. This would leave a very confusing situation... the 500 player would appear to be playing much, much better than its rating and the 3500 player would appear to be playing much, much worse than its rating.
Unfortunately several people were doing just that. They were friends of mine; however, eventually I got so sick of it that I reported them. Several accounts were banned and several ratings were reset. In my opinion the ratings were reset way too low, which really punishes their future opponents rather than the offenders.
What I find really weird though is some (but not all) of the people punished for that behavior don't think they did anything wrong at all. They believe if the average of the 2 account ratings is about their simul strength, then the ratings are accurate -- regardless of whether one account is ridiculously low and the other account is ridiculously high. When I suggested playing with accounts with a similar rating, they said that was no fun and it is only fun to inflate an account to a ridiculous rating. They said they like the attention of people who think they are bughouse gods and ask them for advice. Basically they think manipulating ratings to make one account look like the best bughouse player in the world is okay because they will attract a lot of attention and use that attention to teach more people bughouse (the ends justify the means).
In my opinion, this is nuts. Rating manipulation specifically designed to inflate a rating well above skill level is cheating and tricks the community. It generally hurts the ability of the rest of the players to find fair games and know their opponents' strength.
I'm fairly confident over the years I have played more simul games than anyone else in the world. However, I have played most of them unrated on FICS and I have never simuled on chess.com. Therefore, this is an issue near and dear to my heart.
I've also probably lost a few friends this week by reporting them for rating manipulation cheating. I was hoping they would be mature about it, but they are mostly teenagers who don't seem to understand basic morality and etiquette.
I am interested in hearing other people's opinions about simul ratings.
Is it inherently wrong to manipulate a rating to something wildly divergent from your skill level?
Should simuling be allowed at all on chess.com?
If simuling is allowed, how should simul ratings be handled?