
Slide from the State of the City Address (pdf of full slideshow here)
"Statistical tools and methods, like many other technologies, can be employed either for social good or for evil" (the American Statistical Association’s Ethical Guidelines for Statistical Practice).

Unfortunately for us, we’ll never know the difference: Darryl Huff’s classic How to Lie with Statistics is, at the moment, unavailable at the San Francisco Public Library.
The probability of this occurrence is… you figure it out.
The fourth slide in our fearless Mayor's State of the City PowerPoint Presentation featured the bold words: "Progress by the Numbers." Earth-shaking statistical analysis of the numbers after the jump.
We know that the statistically-deft Gavin Newsom would never use the numbers for evil, but as responsible citizens, mustn’t we inform ourselves, so as to be able to tell the difference between the two?
Let's look more closely at the much-remarked-upon State of the City Address (which doubles, presumably, as the State of the County Address). By our reckoning, approximately 52.5% of the 95 minutes were spent presenting numbers and statistics, thus proving the old adage that “90% of politics is half statistics” -- assuming that the remaining 10% of politics is three-quarters statistics (this statement has not been evaluated by the FDA).

At least the library has three different ways to digest Harry Frankfurt's slim volume On Bullshit: print, cd, and now, Windows Media eBook.




Sorry, Mac users, the NetLibrary audiobook version of On Bullshit just won’t work for you--see below.




I remember reading "How to Lie With Statistics" at City College!
Sure came in handy when reading those charts some odd years ago onboard muni (claiming they "improved")
Newsom bragging that Muni has improved its "on-time" performance misses a basic point: Muni substantially cut service a few years ago. So he's really mixing apples and oranges. It's a lot easier to have a high on-time goal when you're running fewer buses.
How to lie with statistics ... clear that Newsom read it well.
During rush hour or any traffic-causing event, Muni on the westside goes to hell. 50% on-time is probably being generous. 60+ minute waits for the 28 line are common. Taking the 29 line from the Sunset to the Richmond can take substantially longer than walking through GG Park. So, from my perspective, Gavin's presentation is 73.89% hot air.
It's a lot easier to have a high on-time goal when you're running fewer buses.
No, it's actually harder, because fewer buses means more people crowding on each bus and the "bunching" of buses that inevitably results.
Unless you're saying that it's easier to improve performance by cutting lines. This would be true if the most delayed lines were the ones being cut. It sure didn't look like that to me -- most of the lines eliminated last time (FY 2005-06) were very regular lines that no one happened to ride.
This post is really pathetic.
I'm sure Muni is cooking the numbers *somehow*, but they were probably doing the same thing two years ago.
If there's any new statistical manipulation going on, is it too much to ask SFist to look into it, instead of making a bunch of unfunny ad hominems?
And in a post that's supposedly about statistical literacy, I find it amusing that so many SFist readers have decided to dispute data with anecdotes. The plural of "anecdote" is not "data."
But I once was on an N Judah that was delayed by 1 hour!!! I can't remember what year that was, but I'm sure it's getting worse!!!
71.9% of my waking outdoor life is spent standing at a bus stop waiting for MUNI to show up. Gavin takes B.S. to a whole new level. Maybe if he left the limo in the garage and took the 45 or 28 to work he might actually have a clue.
sagitta100: I had a similar experience on the 29 when I used to go to City College from the Outer Richmond district.
The 29 was so slow (and the drivers were complete a-holes), I would take the 38AX express to downtown, and either catch BART to Balboa Park (fastest ride), or take the "K" to CCSF and still make it in less time than the 29.