Variance

When it comes to tournaments, I think one of the most important factors is the size of the field. If we were to order all the formats from highest variance to lowest it would look somewhat like this:

Scheduled MTTs > MTSnGs > STTs > HUSnGs

Not sure where DoNs fit in, but I'd think they are more or less on par with HUSnGs, maybe even lower.
Interestingly, the order is reversed for cash games:

Heads Up > Shorthanded > Full ring

The blind levels certainly play an important role as well. The higher the blinds and the faster they climb, the more pressure is on you to risk your tournament with marginal hands. You have more time to carefully pick your spots when the blinds are low and rising slowly.

I think a good way to determine the variance of a format is to imagine what happens if you stack an opponent:

HUSnGs: You win, end of story.
STTs: Big advantage, but you have to keep using your stack to win, and risk losing or crippling it.
MTTs: You still have a long way to go, and will run into many situations where you have to risk your stack all over again.

Btw, cash games are always less variance-heavy than tourneys, so you might want to consider switching to those if you are looking for a low variance game. That was actually one of the reasons I made the switch a few months ago.

Never thought that HU's would offer the least variance of tournaments. I might be wrong, but I can't agree with that. Good analysis, though.

As for the factors influencing variance - that I forgot in the first post... multitabling is somewhere near the top.

Originally posted by Styr
Never thought that HU's would offer the least variance of tournaments. I might be wrong, but I can't agree with that.

I'm not entirely sure about that either, since I'm not really a HUSnGs player, but it seems to be the general consensus. It is also suggested in our article The 6 Keys to Winning Heads-Up Sit ‘n Go Play:

With only two players and 50% of the field paid, heads-up is a less swingy game compared to many other formats.

That makes sense to me, but then again, HUSnGs are quite loose and aggressive, so there is a big element of chance in individual hands (which is the reason HU cash games are more variance-heavy than full ring games).

You might also want to check out this blog: http://www.nsdpoker.com/category/poker-pro/ . It deals with variance in different poker formats in great detail.

Originally posted by Styr
As for the factors influencing variance - that I forgot in the first post... multitabling is somewhere near the top.

I don't see what multitabling has to do with variance. Could you explain?
It's true that it makes you play more tourneys in a shorter time, but shouldn't 1000 multitabled tournaments have the same variance as 1000 singletabled ones?

Harrier88's list is right but you need to consider winrates too since a high winrate game with high variance could leave you with less swings than a low winrate game with low variance.

I think the only way to deal with variance is volume.
For example, I play HU Hyper-turbo SNGs.
They are known to have big swings due to the hyper-turbo structure and small winrates.
But one is able to play at the very least 100 games a day, which makes it easy to reach a realistic Roi and expectation values. So I'm able to find out if I'm beating them or not quickly, instead of when I play a slow format that I can't put much volume in.

Also, the winrate is an interesting factor.
It does not change the variance or swings of a game, but the higher your winrate, less likely you are going to have a negative profit, which makes it easier to cope with the swings.

IMO, what should matter the most is finding a game you are willing to play and work at a lot.
If you are happy to put volume and study a specific game, why wouldn't you get past the swings and get profit?