Though for all intents and purposes - this is a three person race, Governor Crist will appear ninth on the ballot. Because state law gives the partisans the top spots on the ballot, and specifically the first spot on the ballot to the party occupying the governorship, Marco Rubio will appear first on the ballot, followed by Kendrick Meek. Then will come the candidates from the other parties (Libertarian, etc.), before a "No Party Affiliation" candidate will appear. Governor Crist says "It's an issue, obviously."
So - is it an issue? Does ballot position or ballot design play a determinative factor in elections? Many have long suspected that the ballot design in Palm Beach County in 2000 - the "butterfly ballot" designed by Therese Lepore, was determinative in the 2000 Presidential election (the ballot may have confused some voters who intended to vote for Gore, but instead voted for Buchanan - and giving a margin of victory to Bush for the State of Florida and consequently enough electoral college votes for the win). There are other anecdotes as well, but what about systematic testing? Michael Alvarez at Cal Tech is probably the leading scholar on the issue. Alvarez identifies several different issues including name order, the amount of information provided (name, party, incumbency, occupation, etc.), ordering of races in an election (attendant issues with undervoting for later races), voting instructions, and language and images. In short, there are a host of factors related to ballot design that might change a few votes in an election. While most races are decided by larger margins - in close races, the design of the ballot could be determinative for the race. While the Governor is right that ballot design is an issue, whether this race will be close enough to be affected by the ballot design remains to be seen.