that's what I remembered:


In December 2002, Dowd told the New York Post that he had reliable evidence that Rose bet against his team but didn't include it in his 225-page report because of time constraints. He later backed off of those statements. "I was never able to tie it down," Dowd said. "It was unreliable, and that's why I didn't include it in the report. I probably shouldn't have said it. I was not trying to start something here."



http://seanlahman.com/files/rose/rose-faq.html


Then again, this is as close to betting against your team as you can come without actually doing so:


According to the Dowd report, which included a diary of bets that Rose made on Reds games and many others — it listed bets on 390 games over all, 52 of them involving the Reds, in a three-month period in 1987 — Rose developed a consistency of not betting on certain contests.

In particular, Rose stopped betting on Reds games that Gullickson started. If Rose bet on his team to win other games but didn’t bet on Gullickson’s games, he was sending a signal to the bookies he was betting with that he, as manager of the team, didn’t think much of his team’s chances in those games.



http://www.nytimes.com/2007/03/16/sports/baseball/16chass.html?_r=0

Not betting on your team to win when a certain pitcher is starting is damn near the same thing as betting against your team when that guy is pitching, especially when you're betting on the Reds to win almost every other game.