In which limits are discussed, and the reasons for them.
Fantasy baseball teams are often limited in the things they can do.
In some cases, these limits are intended to enforce a sense of verisimilitude in the game. The best example of such a limit is the rule that says a player can only be put into a positional slot if he's actually played a certain number of games at that position. The exact number of games can vary from league to league, and even season to season (20 games at a position in 2009 qualifies a player at a position, while 5 games in 2010 also qualifies), but without such a rule you'd likely see fantasy owners trying to play Ryan Howard at shortstop, an image which is either hilarious or despair-producing depending on whether or not you're playing in that league. Other examples include specific roster slots for specific positions (even though the league doesn't use defense in scoring, it means you still need at least one third baseman and a shortstop), and an overall active and total roster limit (to better simulate actual roster makeup, as well as prevent owners from 'hoarding' players with potential who aren't playing every day, or even in the majors just yet).
In other cases, the limits are meant to compensate for limitations in the structure of rotisserie-type games. For example, the same rule listed above regarding how players qualify for positions means that a player like Phil Hughes, who pitched primarily in relief in 2009 but is a starter in 2010, is extremely valuable, particularly if he's pitching well, since he can fill either a starter's or reliever's spot, depending on what your team is short of at any given moment. In fact, it's possible, if you can find enough players like Hughes, to have a pitching staff composed entirely of starting pitchers, even though nearly half are listed as relievers. Or, conversely, you could have a staff composed entirely of relievers, with some filling starter slots.
Why is that a problem? Because it bends the assumptions on which the league is supposed to simulate baseball. Rotisserie baseball, since it can't simulate head-to-head matchups with a clear winner and loser (though some fantasy leagues do use a more fantasy-football type head-to-head structure), are instead statistical races -- the team with the best stats wins. (As an aside: those of you who've followed Contrarian Bias may remember that 'rotisserie thinking', or the idea that if we lead the league in every stat category that it makes us a good team, was one of the things I railed against most frequently when evaluating real baseball teams -- but in a fantasy league where the only thing that matters is the stats, it makes perfect sense.) And regardless of which way you go, you can twist the numbers to benefit you.
In most leagues, counting stats outnumber rate stats. So if your pitching staff is composed entirely of starting pitchers while everyone else's is more conventionally constituted, you're almost certain to lead the league in wins, unless someone else gets really lucky. Starting pitchers simply get more decisions than relievers do. Likewise, such a team will likely lead its league in strikeouts, or at least come close -- while relievers tend to have higher strikeout rates per inning, starters throw around twice as many innings, so even at a lower rate, starters will tally up more aggregate strikeouts. (If you don't believe me, check out the strikeout leader board from any of the past seven or so years -- you'll see a few closers in the top 25, but most of the guys on the list are starting pitchers.)
On the flip side, you can really abuse your standing in the rate stats by going heavy in relief -- relievers, particularly closers and setup men, tend to have lower ERAs than starters, and have lower WHIPs as well. Leagues that score on those numbers will give up significant points to a team composed of guys who, by design, throw for lower ERAs and WHIP totals.
Maybe you'd find such a league interesting; the puzzle of whether or not my staff consisting of every closer in the NL Central battling your staff of every #2 starter in the AL East might seem fascinating. But to a lot of folks, it doesn't seem much like baseball, and 'seeming like baseball' was what attracted them to the idea of fantasy baseball in the first place. Lining up pitchers as if they were people dressed in Klements sausage costumes and sending them around the bases simply doesn't feel like the game they want to play -- which brings us back to the first point about verisimilitude.
However, the biggest limitations imposed on teams are limits on the number of games and innings they can amass in scoring.
Anybody who's ever followed baseball can tell you that, while a baseball season is 162 games long, it's much more than 162 days long. And since fantasy teams are composed of players from all over the league, it's not unrealistic to think that, given a large enough separation of teams and leagues, that you could swap players in and out of the lineup and get 175 games worth of hitting stats in a season.** As with the pitching example above, the problem becomes one of making the counting stats meaningful -- did you hit the most home runs because you identified good home run hitters, or because you got in 100 more at bats with each of your players over the course of the year? The TTFBL has different limits for position players and pitching staffs -- each singular position on the team is limited to 162 games worth of stats, and once you've reached that 162 games, you're done contributing from that position. For pitchers, since the distinction between starter and reliever is amorphous, the max limit is 1500 innings pitched. To honor the sanctity of rate stats, there are also minimum limits of 3950 at bats and 900 innings pitched.
** - If you're still skeptical about the value of simply playing more games, convinced that a decline in the rate stats would compensate for the slight increase you'd expect in counting stats, consider this: the current TTFBL leader has either 24 or 25 games played from every singular batting position on his roster, putting his club on pace to have no fewer than 156 games from each slot. Meanwhile, your humble bottomfeeder has only one slot with as many as 24 games played, and one with nearly half (13) of that. It's far from the only difference between us, but it's a real difference.
The limits on a team are what add interest and drama to the decision-making process: sure, you can add that streaking rookie free agent to your roster, but who do you waive, and does that guy get picked up by someone else as an upgrade? You can demote your struggling starter to the reserves, but do you fill the position with a lesser starter, hoping for the best, or with a reliever, changing strategies mid-stream?
Without limits, fantasy baseball wouldn't be nearly as entertaining as it is.