I’ve been posting Olympic qualifying tournament updates at VeB, but I’ll copy-paste what’s I’ve written there to here to update things:
Game One:
Team USA won their first game: USA 9, Canada 3. The story leads off with Skip Schumaker’s 4-5 night (with a double). He also scored three of those runs, so he did the Cardinals organization proud. It would be nice if he tears up the tournament and joins the Cardinals with a hot bat to compete with So for the job of OF defensive replacement down the stretch.
Also noticed that the team’s 3rd-base and bench coach is Rick Eckstein—David’s brother and the hitting instructor for the gNat’s AAA affiliate.
Game 1 Boxscore
Game Two:
Team USA has a difficult time against Brazil, but wins 8-7 in extras. Skip went 0-3 with two walks.
Game 2 Boxscore
Game Three:
Team USA won their third game of the opening round today against a strong Mexican team, 13-5.
Skip went 3-5 with a double and must have reached on an error, because he didn’t walk yet scored four of the thirteen runs.
Boxscore here.
Current Stats
Through game 3, Skip’s numbers in the tournament:
7-13 with 2 2B, 2 BB, 0 K, 8 R, 0 RBI
LINE: .538/.600/.692
Two quality opponents coming up: Puerto Rico on Tuesday and Venezuela on Wednesday. Canada beat Venezuela today 7-5.
[UPDATE] More on Skip Schumaker’s game now that the recap has been posted: Both ends of the fourth belonged to Schumaker, as he stroked a leadoff double and scored on a Mike Kinkade single to push the advantage to 10-0. Then he made a spectacular diving catch at the wall to rob Mexico’s Luis Suarez in the bottom half.
Completely unrelated to the the tournament that determines whether the United States will play baseball in the final Summer Olympics in which baseball will be played, but the second week of classes have begun and I’m definitely going to stick out all three courses I’m enrolled in. It’s gonna be hectic, but they’re all very interesting—if I plan ahead a bit more than usual, it shouldn’t be a problem.
The first class is Formal Grammar Development with Typed Feature Logic, in which we’ll be building relatively wide-coverage grammars in HPSG using TRALE. Not to brag, but I have something of a knack for writing grammars. Unfortunately, until now at least, I haven’t known anyone who actually does it—everyone I’d known prefers to induce a stochastic grammar from several years of annotated newspaper text instead of writing it by hand.
The second class is The Speech Chain and covers the mathematical background needed to rigorously study phonetics and acoustics. So far, it’s been a matter of re-learning stuff that I’d covered in undergraduate physics classes. Today we covered harmonic systems—the Newtonian mass on a spring stuff that’s so exciting the first time you’re exposed to it. Eventually, we’ll learn how to do a Fourier transform with pencil and paper which leads into a joke I read today:
Person A) “What is this?” as she rubs her hand over an invisible level flat surface
Person B) “I dunno”
Person A) “It’s the Fourier Transform of this” as she gives him The Finger
I fully intend to share that bit at the next Ling happy hour.
The last class is Corpus Methods and Modeling for Phonetics and Phonology and looks to overlap a bit in content with the Speech Chain and also some material that I had to teach myself over the summer for a project at work. We covered signal-to-noise ratio today, and the prof mentioned the anechoic chamber at VFI where they can make true recordings with a SNR of 90 dB. I checked the most recent recording project I did for a client and the (roughly calculated) SNR was around 66 dB, which is very good. That’s cheating, though, since I run the signal through an analog compressor to remove low-amplitude noise, among other things.
Completely unrelated to anything preceding, but did you ever wonder where the Closed Captions on TV come from? I did, since I watch a lot of sports where the captioning is usually pretty unreliable (and always uninformative). Well I did and found the answer from the people who make them here and here. It’s a little disappointing that there’s so little technological wizardry involved. I figured at least there’d be some sort of a speech recognizer involved. Turns out there’s just a special keyboard that lets a trained operator type real fast, probably using a database similar to what a T9-equipped telephone uses. Ah, the keyboard uses phonetic symbols.
I’m sure that could be automated—also that it would be a fun project to work on.