How do you make a decision to play at any particular table online? The tables list stats like hands per hour, average pot, etc. Are those factors useful in determining which table to sit at?
Keep a Fish List
Depending on the level of traffic and skill at the site/stakes you are playing, often the best way to find a good table is to seek out a table that gives you position on a player who you know how to beat. This can be due to that player being a fish, or simply due to some knowledge you have of their game. To supplement this, I would recommend keeping a list of players you want to target in this manner. Update the list whenever someone makes an impression on you as easily beatable, and migrate the players you see online to the top every time you play with them.
Look For a Balance of Other Factors
If you can't find any of your "list" players, or the site that you are playing on does not support searching for players, then you can revert to other factors. What works best for you will be something to determine from your specific style/game. I typically play a tighter game, which means that I seek tables with a high VPIP, and a large average pot. Of the tables I find meeting these conditions, I will drill down further and look for the ones that give me position on a few players with full stacks, preferrably not regs.
The tables list stats like hands per hour, average pot, etc. Are those factors useful in determining which table to sit at?
Yes it's useful, but not as useful as keeping a fish list (as one of the other answer pointed out).
In addition to your fish list, another very important factor is your place at the table: even if you found one fish but have to sit with several good "regs" (winning "regular" players) at your left, the table may not be turn out to be that great for you may be outplayed by the regs (they'll try to keep the fish for themselves).
If I find a nit to my left, a good reg to my right and a fish anywhere, then it's much, much better than a nit to my right, a good reg to my left and a fish anywhere... The position where you seat is very important.
I never used a "fish list" so I describe what the stats could mean in a poker site and how can you decide which is better.
If a table have a high players per flop statistic, (over 30% at 6-max tables) it mean's that generally too much players want to see the flop. Specific to 6-max tables, the usual situation that two players are heads up, because of the possibility of strong hands preflop. The most common is when someoney is raising, and everyon is folding, so this stat wont go up. The second most common situation is a raise and a reraise, (it's a long story why) this situation change the stat the same way as everybody would folded for obvious reasons. So when this stat is ridiculosly high (for example 40-45%), you can be sure that at least one very bad player sits at the table. Even if very good players sits at that table, and you sit in and play tight you still can make a profit if you dont make big mistakes, because their initial hand range will be much more weaker than your, and this case, the cards will do their job instead of you. So, this is one important statistic I watch for when I want to chose a table.
The second is average pot size statistic. If this is too high (in 6-max games over 20 big blind amount is high) this means that players put too much money in to the pot with weak hands. Even if they are tighter, they put too much in then they sould probably, so you can make a nice profit here by value betting more. Actually, this is my primary variable I watch for.
The third is average VPIP. They are too loose preflop, playing too many hands (over 35% in 6-max is high) so they will have weaker ranges as I explained above.
This one is not that important, but you shpuld watch for the waiting list too. If too much players are waiting for the table, it's almost sure when your turn comes, you wont play with the same opponents who you wanted to play, because the fish(es) probably go broke until that, and all the other players went away already, maybe the whole table is totally different, so can be the statistics, compared to when subscribed to the list.