Thoughts on the Behind the Myths Tour

The Behind the Myths Tour

TW and I went to the behind the myths tour yesterday afternoon. I was super excited for them to come to DC. Despite the fact that there are a ton of nerdy folks here in DC (I saw at least 10 of these shirts), I feel like a lot of tours/shows that are geared towards a this type of audience skip the south entirely. DC isn’t in the south, no matter what Jeremiah Dixon, Charlie Mason, or JFK think, but it’s close enough that I think it suffers from some perception. The idea that there isn’t a large enough market in the south for this kind of entertainment is silly and somewhat insulting. Atlanta is home to a number of venues no larger than the Warner Theater and one of the best engineering schools in the country yet nerdy/geeky acts seem to avoid it like the plague (TMBG are an exception to this). But I digress.

The show was excellent. More than anything, it took Jamie and Adam out of the television and made them real. Very few myths were discussed or busted. Instead the duo spent time drilling home Mythbuster’s core message: Anyone can and should question the world around them.

There were three things that stuck with me from that show. The first two happened during the Q&A sessions. First, Adam was asked about appearing at the Reason Rally hours before the show. Adam is a well known and politely (to me anyways) outspoken atheist. I sensed the questioner wanted Adam to rehash his comments to this crowd. Adam effectively deflected the question, realizing I hope, that this was not the venue for such a topic. The audience had not paid to hear that discussion and more than a few would have been offended had it taken place. I’m sure this wasn’t the first time a question like this has come up. The enthusiasm that Adam displayed when talking about the event was enough to show the audience that he cared about this cause.

The second was Jamie’s answer to a question about the infamous cannonball. After the question was asked, a number of cheers/chuckles broke out in the audience. I was impressed that someone had the nerve to ask Jamie about it in person. Jamie gave a thorough explanation about what went wrong and what factors caused the cannonball to ricochet out of the range(short answer, a brush fire and drought made the ground hard enough for the shot to skip). I kind of enjoyed his scolding of the audience. His response to the chuckles was to remind everyone that this was not funny and might have cost them the show. I was impressed with the level of analysis the team (I’m assuming everyone was involved) did to determine why this happened and how to avoid it next time. His response reinforced the idea that the Mythbusters team think of themselves as engineers/scientists who frequently learn more from failures than from successes.

The last thing that surprised me was the fight. I never expected to see Adam coldcock Jamie with a chair like that but damn. When the big man went down, he went down hard. I got this pic right before it Adam about to lay the smackdown on Jamiehappened. I was particularly worried about the kids in the middle. They were brought up to help with some thing or another but had to dive out of the way. I knew something bad was going to happen when Jamie bellowed at Adam from off stage. I’m assuming the crowd in Raleigh had a good show but things were kinda tense in the audience for a while.

Couple of random thoughts:

– Adam dancing the horn pipe to The White Stripes “Seven Nation Army” was strange. The between skit music was odd as a whole. Who ever was covering “Highway 61 Revisited” could not do it justice.

– Jamie bringing out Blendo as if it was Apollo Creed from Rocky IV was a nice touch. I wonder how many people under the age of 30 got the reference.


More on First Pitch Strikes

A couple of weeks ago I posted about plate discipline and in particular about taking strike one. I wanted the opportunity to look at the numbers in couple of other ways. I’m using the same data that I used for the last analysis.

Called First Strikes over the past 15 years

Per game, the average number of Called Strike Ones (CSOs) has risen slightly since 1997. In ’97, the average was around 19 CSOs per game and rose to 23-24 per game in 2011. At the same time, ERA (secondary axis) has dropped slightly as well (LG ERA came from baseball-reference). I haven’t tested the relationship here but it makes sense at as pitching became more dominant, batters may be more likely to get caught off guard on strike one.

CSOs Relationship to Offensive Outcomes

I wanted to look at how a player’s stats relate to taking strike one. The vast majority of ABs where the pitcher gets strike one usually end in the pitchers favor (see The Hardball Times for more info) but how does that impact a batter’s productivity.
Here’s the graph of wRAA and called strike ones over the 2011 season (min 200 PA as usual).

Despite the facts that the numbers say there’s a correlation (r= 0.371, p<.0001), I’m hard pressed to see it. The graph for wOBA looks much the same and has about the same correlation.

What impresses me is that taking strike one doesn’t appear to have any bearing on how well a player does. It certainly infuriates the fans but in terms of player value, it doesn’t appear to impact value.

One fact that I thought might be confounding this is the number of plate appearances a player has. It’s unfair to assume that a player with 250 PA has the same opportunity as players with 400+ to take strike one even if they’ve got close to the same wRAA. In order to adjust for that I ran the following regression:

wRAA=β1*Called Strike One + B2*PA

The results were pretty much what I expected. The estimate of  was -.0848 (p<.001) and the whole model had an r2 of .26.  Again, not a very strong indicator of any relationship between called strike one.  Using the same equation, I a slightly better relationship with strike outs (r2=.49).

Live blogging Duke v UNC 3/3/2012

Going to try this live blogging thing. Let’s see how it goes?:

Pregame: Haven’t paid a whole lot of attention to it but I’ve got to say, Digger looks old. He’s wearing an overcoat while his cohosts are wearing blazers. Can’t even see is matching tie. I assume it’s blue since his highlighter was but we’ll see.

As a quick reference, I dislike Dicky V as an announcer.  I admire his passion for basketball but I really don’t want to hear him for a whole game. Se la vie.

The crowd sounds good so far. In the FSU game, they died towards the end of the game. They need to keep it up all day.

Love the Lexus commercials. Clearly they know who’s playing the game.

Anthem dude is kinda scary. Like the fact that he’s got two mic’s in one hand but the mics seem to be reverberating.The sound isn’t that good. And……….please don’t hit the high notes…..oooh. Not good.

The Roy’s fashion choice tonight is really great.  The suit is conservative and the tie is an awesome paisley.  Good choice tonight. Normally it’s questionable. The pocket square is a nice touch

What is Erin Andrew’s wearing? It looks like she’s some kind of cult leader. The outfit looks like some kind of vestments.

Tip: Out of bounds………and Carolina gets it. Boooo. Zeller gets the odd hook.

19:30 Miles Plumlee!!!! (think he traveled there. Gotta be honest).

17:42- They showed Payton Manning. He’s better than Jeter who went to a game while I was at school. He has some reason to be there. But Dickie V needs to stop talking about NFL trade rumors.

16:00 And Duke is as cold as ice. This isn’t entertaining.  All I have to say is that I’m glad we’re not wearing the ugly grey unis.

Charlie Sheen Commercial: Umm……..odd. Best of those commercials   But still. I’d rather reenact some of Charlie’s parties and not Platoon.

13:00: 0-7 from 3. How about we try and get into the paint some. It’s like we think blue paint is lava and we can’t touch it.

Wonderful….TW is now chanting “Up by quad”. That’s not what I need. Going to get more beer. K needs to get the paddles out.

11:00 And we finally make a shot. Can’t block out or rebound but hey, making a basket is a big first step. Let’s try and make two.

9:00 The pace is much, much better. I feel a bit more confident that Duke can come back at this pace.

TW: “Up by Double! See, I learned some Duke cheers!”

8:43- I’ve always thought K was a vampire. He always looks pale in those dark suits but he always looks the same. Every season since I got there, he hasn’t aged a bit. Can anyone back me up on this?

6:26- Finally! That possession took for ever. For the size that we have in the front court, I can’t  believe that we suck so much in the paint.

5:30- The Plumlee’s mom looks angry. I think she’d be the second most scary person in the huddle, after K of course.

4:00- No idea what just happened. Trying to regroup here after TW put a huge plate of food in front of me. This “Goodnight Moon” Audi commercial is terrible. Will never buy an Audi now. Also, watching a gospel choir singing “Blister in the Sun” is weirding me out.

2:43- The officials probably should swallow their whistles a bit. I’m kinda tired of seeing their s dramatic interpretation of the charging motion.

And TW brings back the “Up by double cheers”. I really question why I brought her to games in Cameron.

0:58.5- Dickie V thinks we do too much review. I think he talks too much. Particularly about things that aren’t the game in front of him.

Half: Oy vey iz mir. Can’t shoot, can’t rebound, can’t defend. It’s going to be a long, fun second half and half time.

Who’s the bartender in that awful Mio commercial? I’ve never heard the dialogue before now but the animation always freaked me out.

So which school has more people that need tax help? Carolina? Duke? I figured the rich alumni from both schools know about off-shore tax shelters so who are they actually advertising to?

For a  second, I thought the 3pt FG% (18.2) was our overall percent. It can’t be that different. Thinking about trying to source the ingredients for this.

18:30- Kelly looks good when he pulls the head fake and drive. Why couldn’t he have done that for 3 years?

16:18: THORTON!!!!! (Update:2:00) Thorton, you missed a fast break lay up. Idiot.

Under 16 time out: Nice. This is type of game I wanted to see! Go Duke! Still not sure we’re going to win but I like this idea.

15:00- Dickie V is talking about the Bobcats. Again, shut up and talk about the game at hand.

13:20- Foul trouble mounting. I’d be ok if all of our good players fouled out and the last 5 min of the game had to be played by scrubs. We might close the gap in that case!

10:55- OOOOHHH!!!! NJIT won a game.

10:00 We’ve cut it to around under 20.  That sounds like progress, which depresses me even more. Where’s the beer?

8:00- CURRY!

7:11: We’re swapping Plumlees now. That’s always a good sign.

3:54- The Roy is jumping around on the sideline like a gremlin. I like that image. Someone, photoshop this!

3:00- When Duke doesn’t win this game, I”m going to feel good about how we didn’t give up and gave it  the old college try this evening. And by the old college try, I mean that we got some pizza and cheap beer and forgot about tomorrow.

1:00- My only defense to “Miles Plumlee looks odd” is “He’s not Chris Lang ugly”. Also, we’ve given up now. So that’s that. I’ve totally given up on this. Liver, sorry, but you’re taking the brunt end of this.

55.9- Dickie V is now talking about Tina Turner. God, I hate him. Hell will be narrated  by Dickie V. Him and Brent Mussberger.

48.9- OOH! Harvard won!

20.0: <In Sean Conery voice> Duke. Let it go.

Well that sucked. Going to pretend that I didn’t see this. I’ve been watching something awesome tonight. What have you been doing?


Making It Up As I Go

For one of my “research” projects, I need to generate some fake data. Specifically, I need to generate data on how attractive the 144,702 librarians working in US public libraries are.  Why you may ask? To quote Richard Needham  “For a well rounded education, you could try curling up with some good books, and bad librarians”.  Also, I’ve always wondered if having a more attractive library staff would lead to more usage.  Besides, TW and a lot of her friends are librarians. I expected this to be easy.  As I’ve found out, making up good fake data is really hard. Any idiot can generate a stream of random numbers in excel.  It’s making fake data  that looks and acts real that’s hard (and I’m not good at it).

Generating Attractiveness Data

Everyone should be familiar with the standard attractiveness scale that runs from 1-10 (with Bo Derek being, of course, a “10”).  I’m using that as the basis for my scale (the Kelwick Attractiveness Scale).  How to model that accurately thought? I settled on a normal distribution. Everyone’s favorite; easy on the eyes and the numbers (just like a good bad librarian). But in order to generate good values, I need a realistic mean and SD.  5 seems like the most logical mean so I went with that. What about SD? An SD of 1 would imply that 68% of the librarians would be between a 4 and a 6 and 95% would be between a 3 and a 7. That seems a bit restrictive to me.  An SD of 2 means that 95% of public librarians would be a 1 and a 9 but that seems to generous.  I don’t think ≈15% of public librarians are above a 7. Maybe 7-8 but not 15%. That’d mean roughly 21K pretty hot librarians and I don’t remember my public library being staffed by Parker Posey.  I settled on an SD of 1.33, which produces the following distribution:

Do I like that distribution? Maybe. The problem is going to be that I’m not going to get enough variation across libraries.  To do that I’m going to have to figure out how to distribute the librarians effectively.

Assigning Staff to libraries

Now that I’ve got 145K individuals rated, how do I assign them to a library? I could randomly assign them to each library but I’ve got a feeling that the NYC library system has more hot librarians than Dekalb County, GA.  My plan is give libraries in very urban settings a bit of a bump (say .25 ) to their average hotness and see how the distribution changes.

What about men? The best number I could find was 18% of public librarians are men.  Don’t believe that number but hey, it’s a starting point.  I’m trying to work out a better system for this.

Additional factors to include in my analysis

Glasses:  In the current climate, where being a geek is in, glasses are sexy.  And everyone knows serious librarians wear glasses.  How to model this? I chose a Poisson distribution with the expected number of glasses wears is a random percentage of the staff.

Mustaches: Manly librarians should have manly mustaches.  ‘Nuff said. Modeled this one the same way as glasses but only for male staff.

Other factors: I haven’t come up with any other factors but would be will to define some other ones. Pencil skirts? Tweed jackets? Let me know.

I’ll have the full write up out here in a couple of weeks.

Got a better way to measure attractiveness?  Got a better solution to generating some fake data? Drop me a line. I’d love to hear from anyone with a good idea.