I doubt many people are saying that the 18-player teams have zero chance of winning the contest. The above statistics from last year show that the largest teams (24 players) had more than 4 times the odds of making it to the final 250 than the smallest (20 and 21 player) teams - and 3 times the odds compared to the 22-player teams.
Some features of the contest are significantly different this year: (1) rosters from 18 up to 30 players are allowed, and (2) the dollar scale for players is compressed with lower values for the best players and substantial reduction of the number of $1 and $2 players.
Does this mean that the larger rosters still have a better chance of winning this year's contest than the smaller rosters? Time will tell I guess. My gut says yes. If you can find "value-players" who are likely to significantly outperform their $ costs, then it's a good strategy to load up with a bunch of value-players and a few studs.
Look at the low-priced WRs with upside:
Mike Williams, TB - $8
Nate Washington, TEN - $8
Bernard Berrian, MIN - $7
Chris Chambers, KC - $7
Laurent Robinson, STL - $7
Josh Morgan, SF - $6
Naanee, SD - $6
Brian Hartline - $6
Louis Murphy, OAK - $4
Deion Branch, SEA - $3
Shipley, CIN - $3
Justin Gage, TEN - $2
For the contest, would you be better off spending $25-30 on a single stud receiver -- or spreading that money out over 5-7 value-priced players? Although some will disagree, I think the answer is obvious.

I had 22 or 23 players last year and lasted pretty late, but this year is very different, with much cheaper studs and much more expensive lower-end players for the value. The shorter rosters last year probably didn't have many of the great $1-4 players that were common. This year, there are very few such bargains, so it's apples and oranges.
This year, I have only 18 players, but I have 4 of the above value WR's to go along with Andre Johnson and Welker. I also have Rodgers, Chris Johnson, Ray Rice, and Finley, so went with the stud theory combined with value. With less depth, I may be more susceptible during byes than a 26-player team, but if I can survive the bye weeks and the top guys stay healthy, I think I'll have a better chance of winning than most 26-player teams (projected #29 in week 1). Hopefully the rest of the studs and value players will make up for the studs on bye.
To win, you need to get lucky either by having your guys stay healthy, hitting on some lower-priced guys who turn out to perform like studs, or probably both. Just because last year higher-player teams had a higher chance of lasting to the final 250 doesn't mean this year's the same and those with 18-player rosters are clueless. I think the numbers will look different this year because of the changes, but many 18-player teams will still get knocked out during the bye weeks and because of injury. I think there's a good chance the winner, or many of the top 100, will have shorter rosters this year.
Man, good luck in week 10 with both Rodgers and Finley out - and the competition getting fierce at that point.Two WRs frequently mentioned are Calvin Johnson and Wes Welker. According to the latest projections (with ppr), they are projected at 275.1 points and 202.6 points for a total of 477.7 points. They cost a total of $48.
For a total of $48, you could get a total of 1,302 points from the following:
Mike Williams, TB - $8 [194.8 pts]
Nate Washington, TEN - $8 [147.3 pts]
Bernard Berrian, MIN - $7 [162.9 pts]
Laurent Robinson, STL - $7 [165.4 pts]
Brian Hartline - $6 [137.5 pts]
Louis Murphy, OAK - $4 [163.3 pts]
Deion Branch, SEA - $3 [121.4 pts]
Jordan Shipley, CIN - $3 [101.3 pts]
Justin Gage, TEN - $2 [ 108.4 pts]
Plus you get (1) a substantial amount of diversification, (2) insurance against injuries, (3) better coverage of the flex position, and (4) reduction of bye week problems. Unless the projections are flawed, it's a no-brainer. It's harder the find outright value-priced players at QB (except for Anderson) and at RB (except for Foster and maybe Leon Washington or Thomas Jones or Fred Taylor), but there are a lot of value-priced TEs. My conclusion is that the most effective strategy is to use all 30 roster slots and pack your lineup with value-priced players along with some selected studs.