Survey: Cubans disappointed in Ozzie

Ozzie Guillen is viewed unfavorably by nearly half of Cuban-Americans surveyed in an ESPN-commissioned poll, but nearly two-thirds of Miami residents, and 56 percent of Cuban-Americans, think Guillen should keep his job despite comments about Fidel Castro that earned the Miami Marlins' manager a five-game suspension.

Only 27 percent of Miami-area Cuban-Americans view Guillen favorably, while 28 percent view him "extremely unfavorably," according to a survey of more than 900 adults in the Miami media market. The survey was conducted on Tuesday by ESPN and the Global Strategy Group and included nearly 200 Cuban-Americans. The overall margin for error is plus-or-minus 3.2 percent.

The survey found that nearly three out of four Cuban-Americans found Guillen's comments that he "loves" Castro and respects him for staying in power for as long as he has offensive. Half found the comments very offensive.

Despite that, 69 percent of sports fans and 62 percent of baseball fans said Guillen should keep his job. One-third of Cuban Americans said he should be fired, but 56 percent said he should remain the manager of the Marlins.

Guillen returned to Miami on Tuesday to hold a televised news conference, where he apologized and attempted to explain his comments, which were made to a reporter for Time magazine.

Guillden's apology was enough for 62 percent of sports fans, and 56 percent of baseball fans. However, only 43 percent of Cuban-Americans felt his apology went far enough.

In terms of the five-game suspension, the survey found that 29 percent of respondents felt the suspension was too light; 37 percent of Cuban-Americans felt it should have been harsher. A majority of respondents felt the punishment was fair or even too severe.

The survey also found that 26 percent of Miami residents, and 37 percent of Cuban-Americans, said they would attend fewer Marlins games because of Guillen's comments.

The automated survey asked respondents to use their touch-tone phone to respond to pre-recorded questions.