This could be cleared up if Butkovitz would release the results, which he won’t, but the likely issue with this poll is either a “likely voter” screen that includes only certain Butko-selected zipcodes, or there was some unprofessional push-polling going on. A PoliticsPA commenter who claims s/he was polled says this is exactly what happened - the caller tried to prime respondents with positive statements about Butkovitz prior to asking which candidate the person was likely to support.
How do we know it’s not a credible poll? Look at how low in the mix Anthony Williams and T. Milton Street are. Williams isn’t a popular character with this blog’s readership, but he’s been in the state Senate since 1998, he’s run for Governor, he’s got a famous political name, and he’s constantly talked about as the Mayoral frontrunner in the press. There’s simply no way he’s only at 2%.
Same deal with Street’s (in)famous name.
The other tell is the two-way match-up between Butko and Williams, which has Butko with a 2-1 advantage. There’s just no way. Dude is in his second term as Controller. He’s run two campaigns in extremely low turnout elections. Almost nobody has experience actually pulling the lever for him, he’s rarely making front-page news, and holds a boring office where nobody is actually familiar with the job responsibilities. I would be very surprised if his name recognition with likely voters topped 20%.
Again, Butko can clear this up by releasing the methodology, the weighting of the different geographic areas in their likely voter screen, and the rest of the internals. But until he does that, there’s no good reason for anybody to take this junk poll seriously.