In case you read the previous post about baseball scorecards and were eagerly awaiting the Mark 3, I am pleased to announce that it is ready to begin trials. In addition to recreating the basic form from the Mark 1 and Mark 2 in a more open file format, SVG, I have made several other changes, predicated on two ideas: implementing Tufte, and changing from traditional to modern measurement statistics:

• Eliminated some wasted space from the header
• moved the summary block down
• Changed the pitch count grid from an inning by inning count and total per pitcher to a grid. This is an experiment, and it's probably easier to print the scorecard, go to a game, fill it out, scan it, and upload it than to explain how it should work. The idea is that 1) the marks are the data; 2) it's easier to do totals inning to inning; 3) you can see at a glance how many pitches a pitcher is throwing per inning relative to a complete-game rate (12/inning).
• Changed some of the stats from traditional (At bats, hits, runs, strikeouts) to modern (On-base percentage, slugging, and what I think are the necessary raw ingredients - plate appearances, outs, and total bases). So when Jeff Kent (I still can't believe he's a Dodger. Or that he's hitting so well at 37.) draws three walks and makes two outs, his line OBP/SLG is .600/.000, which is more useful than his batting average, which would be 0-2 or .000. When J.D. Drew hits two singles, a double, and a homer, and grounds out to first, his line is .800/1.600, instead of just .800.
• Dropped the convention of recording RBIs, though this change is not visible on the blank scorecard
• Cosmetic changes to the at-bat box; removed the pitch count circle.
• Followed Tufte's ideas and made most of the lines gray, instead of black, so that data, which is recorded in pencil anyway, is more visible than the grid lines.

300dpi PNG, ready for printing: Mk3 Scorecard Front, Back.

Get the SVG source files.

The strategy of the Chinese government is to change the subject.
When complaints are lodged about the imprisoning of dissidents, the Chinese do not forthrightly proclaim “Indeed, we do put them in prison. We are justified in doing so. They are a threat to our security.” Instead they change the subject to “No country should interfere in the internal affairs of another country.” When America attacks China’s human rights record, the Chinese do not say “You are mistaken about our human rights problem, and here’s why.” Rather, they change the subject: “What about your human rights problem?”

All governments—all human beings—are guilty of this move, which in American parlance is called “spin”. But in China the technique has been reflexively applied for so long, it is now simply the default official approach to any awkward information whatsoever.—Ann Condi

How much longer until this is the norm in the United States? Or have we already passed that threshold? How do we measure? I think that even though the Administration is either there or headed there, that's not a direct comparison. I believe the change the subject syndrome applies to all levels and branches of the Chinese government, whereas I would still expect to get straightforward (or at least not deliberately misleading or obfuscating) answers from, say local governments or branches of federal agencies or judges or a district attorney. Not from politicians, but from bureaucrats. If this changes in the US, will "terrorism" be the excuse?

I've made several custom baseball scorecards, since the ones that come in programs are quite inadequate. Unfortunately, I made them in Coreldraw and Visio, so I couldn't use them on my linux computer even if I hadn't misplaced the original files. I decided to use SVG (Scalable vector Graphics) to create version 3. This is a standard XML file format which allows you to "draw" shapes by describing them in text. Instead of using the mouse to lay out shapes, I can just type in the dimensions I want. It looks like this:

<rect x=".16" y=".16"
        width="0.18" height="0.18"
        transform="rotate(45 .25 .25)"
        fill="white"
        stroke="#cccccc"
        stroke-width="0.01"
        stroke-dasharray="0.02,0.02"/>

That code creates a diamond bordered with a gray dashed line.

I have run into two main problems with this project. The first is that the SVG specs on relative and absolute coordinates for reused components are clear as mud and devoid of helpful examples. This means in effect that I'm doing a lot of stuff the Wrong Way, using offsets and other tricks which just make it harder to move stuff around and correct mistakes. I got to a certain point where I probably should have backed up, created a simple trial document, and mastered viewport coordinate system, effect of the viewBox attribute on sibling attributes, Nested transformations, and the tangled relationship between patternUnits = "userSpaceOnUse | objectBoundingBox", patternContentUnits = "userSpaceOnUse | objectBoundingBox", and viewBox. Instead, I just bulled on ahead to finish the darn thing.

The second problem is that, although SVG has been around for a few years, support for it is mixed. I tried four different programs and got four different outputs, wrong on four different ways. GIMP's SVG module botches the text utterly, putting it in the wrong place and orders of magnitude too big. GIMP also gets most of the lines wrong, though that may only be an artifact of going to a print-ready resolution and then shrinking back down for display. KSVG in Konqueror looks much better, but the text is still destroyed: it seems to round all text block locations to the nearest inch or so, so that the headings for innings 1 and 2 are superposed, as are 3-4, etc. Sodipodi does the text fine and the lines look good, but the diamonds which are the most crucial element of the whole thing are not rotated. Finally, squiggle, a java program which is part of Batik, whatever that is, seems to get everything right except that the text seems to alternate being a pixel too high and a pixel too low. Still, if I can cajole print-quality raster files from squiggle, that will probably be the solution. But a one-afternoon project is now, stretching out into its fourth session of work. Bah.

As a result, I had to use fresh xeroxes of my old scorecard for the Dodgers-Padres game last week. The Padres' stadium, (commercial naming sponsor) Park, is reasonably nice. If you come on the trolley, you have to walk around most of the back side, which looks like a blank mall exterior, to get to a gate, but once inside it's pretty. The commercial signage is a notch past excessive, but at least the players don't wear any advertising (aside from little logos that are practically invisible). The game happened to fall on Military night, and the Padres were wearing camouflage jerseys—grounds for protest if ever there were any. Aside from lots of froo-frah before the game, there was a very very cool parachuting demonstration, with an eight-man Navy parachuting team jumping into the stadium. The first one put his foot down directly on the target in the outfield. Others did tricks with smoke, joined up in pairs, and/or did this thing where they plummet towards the ground at 60 mph until well below the tops of the nosebleed seats before pulling up and abruptly not dying. Useful when being shot at, no doubt, but I'd hate to practice that much. Most importantly, the Dodgers won, although it was the last victory of an improbable 8-game winning streak.

The Ancestor's Tale, Richard Dawkins

A magnificent book. Dawkins adopts the narrative structure of The Canterbury Tales to a reverse chronological trip from modern humans to the origin of life on Earth. The book is a series of rendezvous with "Most Recent Common Ancestors;" at each rendezvous, species which are joining the pilgrimage tell their Tales. For example, at rendezvous 22, at about 530 million years ago (mya), we encounter 41 species of lamprey and 43 species of hagfish (lampreys and hagfish diverged around 480 mya, so the rendezvous, like all rendezvous, is with the trunk of a lineage). That means that, somewhere on Earth, our "240-million-greats-grandparent," some sort of jawless swimming creature with a notochord but probably not a real backbone, gave birth to children, and the siblings diverged (through geography, like swimming off in different directions, or maybe just through having an argument about inheritance and never speaking again, and one child is an ancestor to all vertebrates, and another is an ancestor to all lampreys and hagfish.

Literally an ancestor; they laid eggs that grew into more proto-vertebrates, and so on, with gradual genetic change over many generations, until there was another speciation split (proto-sharks vs other vertebrates), and so on. For any given period, such as 500 or 2000 years, almost certainly any creature at the young end, given a time machine and some breath spray, could mate with any appropriately gendered creature at the old end of the span and produce a fertile child, but as the years pile up, the chances of that child being fertile decrease, until eventually there's a certainly that it would be sterile (e.g., mules, ligers), and then a chance, and then a certainly, that they can't reproduce at all. And so we have a chain of creatures, about 240 million long, leading from our parents back to a proto-vertebrate that swam in the sea and had no jaw and not much of a spine.

(Clarifications: this doesn't mean that literally one animal is the sole parent of all vertebrates. As the proto-vertebrate species splits into two other species over a period of many generations, lots of different animals in that species can claim ancestorship on both sides of the split. But there must necessarily be one last ancestor who spans both sides. It can then be shown (buy the book) that almost all of those ancestors are shared among all descendents; if you look over a long enough time scale, such as 20 or 50 generations, the combined family trees of all survivors merge into a sort of braid. For example, most humans of European descent probably have some genes from King Richard I, just as they probably have some genes from almost every person alive 1000 years ago who has a descendent chain to the present.)

Each rendezvous presents some Ancestor's tales, and rendezvous 22 gives us the Lamprey's Tale, which turns out to be about hemoglobin. Human hemoglobin has four different types of globin proteins. Two are closely related "alpha" globins, and come from chromosome 11. Two are closely related "beta" globins from chromosome 16. The split between alpha and beta reflects a point in the past where, in one ancestor (literally a single creature), there was a transcription error somewhere in the germ line (e.g., in a cell that makes sperm or eggs, or maybe in an actual sperm or egg, or maybe in a very young embryo?) and a creature was born with two sets of genes for globins.

This creature survived, and bred, and eventually this mutation spread to all survivors of its species. (That spread, by the way, is inevitable in any gene pool given enough time, even for survival-neutral mutations. That is, if the creature was actually stronger or faster as a result of the mutation, then it had a good chance to spread that gene on its merits. But, even if extra hemoglobin genes don't provide any benefit, they will either become standard or disappear. Reason: Imagine a graph showing how many members of the species have that gene. At the top of the graph is 100%, and at the bottom, 0%. For a gene invisible to natural selection, the percentage of the species with that gene will move up and down randomly. If a creature with that gene has a lucky childhood, grows up big and strong, and gets a big harem, the percentage will go up. If a whole family group with the gene have a nice habitat near a volcano which erupts, the percentage goes down. It would wobble between the two extremes as long as the species survived, except: if it ever actually hits 0% or 100%, it can never come back. If ever all members of the species have a mutation, breeding along won't ever lead to a member of the species without it, so the mutation is now permanent. Conversely, if the last creature with the mutation dies, breeding along won't bring it back, so the mutation is gone. A similar or even identical mutation may pop up in the future, but the first one is gone forever. When you run the clock over geologic time, the chance that random perturbations nudges the percentage to either 0% or 100% at least once rapidly goes to certainty. Thus, any mutation either becomes standard or disappears. The only time two versions of one gene remain viable in the population over time is when they have comparative advantages; e.g., type A blood and type B blood each confer immunity to a different parasite or bacteria. Interestingly enough, the type A gene in other mammals is closer to the type A gene in humans than the human A is to the human B. So if you share a blood type with a chimpanzee, for that part of the genome you are more closely related to her than to any human with a different blood type, even a sibling.)

At some point, after the mutation become standard, the two copies of the alpha globin gene diverged, and moved to different chromosomes, and diverged again (there are actually "seven [alpha] globin genes. Four of these are pseudogenes — disabled versions of alpha with faults in their sequence, never translated into protein. Two are true alpha globins, used in the adult. The final one is called zeta, and it is used only in embryos.") When the dust settled, all members of the species had seven alpha globin genes and six beta globin genes, some of each disabled and others active.

The reason this is the Lamprey's Tale is that "Given that the split between the alpha cluster and the beta cluster took place half a billion years ago, it will of course not be just our human genomes that show the split, and possess both alpha genes and beta genes in different parts of our genomes. We should see the same within-individual split if we look at the genomes of any other mammals, at birds, reptiles, amphibians, or bony fish — for our common ancestor with all of them lived less than 500 million years ago. Wherever it has been investigated, this expectation has proved correct. Our greatest hope of finding a vertebrate that does not share with us the ancient alpha/beta split would be a jawless fish like a lamprey or a hagfish, for they are our most remote cousins among surviving vertebrates. They are the only surviving vertebrates whose common ancestor with the rest is is sufficiently ancient that it could have predated the alpha/beta split. Sure enough, these jawless fish are the only known vertebrates that lack the alpha/beta divide. Rendenvous 22 is so ancient, in other words, that it predated the split between alpha and beta globin."

I picked that Rendezvous and Tale at random; they are all deep and cool just like that. A lot of popular science books about biology and evolution tend to be collections of cool stories, but the power of The Ancestor's Tale is that it places all of the stories within a huge, sweeping narrative, so that the sense of drama and suspense (who will we rendezvous with next?) builds and builds. I won't spoil the ending (beginning), but suffice to say, it's not much like Genesis.

My one complaint is that, in a few places, Dawkins' conversational tone devolves to bitchiness about creationists. It's not hard to see how he could be tired with the fight: any sentence he writes which mentions any problem with any aspect of the entire theory and data of evolution is liable to be quoted as proof that all of evolution is bunk and even the scientists admit it. Beyond that, reading the book and seeing the astounding depth and complexity and the interrelated nature of evolution, fossils, molecular genetics, information science, plate tectonics, etc, I get the feeling that a creationist trying to argue with a working biologist is like a flat-earther interrupting 747 pilots chatting about the best route to take over Iceland en route to Malaysia.

Certainly this book moves up to my favorite Dawkins book (over Blind Watchmaker) and favorite book on evolution.

... the scientists wanted me to do my bit to help fix the terrible little statistic they keep hearing about, the one indicating that many more Americans believe in angels, devils, and poltergeists than in evolution. According to recent polls, about 82 percent are convinced of the reality of heaven (and 63 percent think they're headed there after death); 51 percent believe in ghosts; but only 28 percent are swayed by the theory of evolution.

Scientists think this is terrible—the public's bizarre underappreciation of one of science's great and unshakable discoveries, how we and all we see came to be—and they're right. Yet I can't help feeling tetchy about the limits most of them put on their complaints. You see, they want to augment this particular figure—the number of people who believe in evolution—without bothering to confront a few other salient statistics that pollsters have revealed about America's religious cosmogony. Few scientists, for example, worry about the 77 percent of Americans who insist that Jesus was born to a virgin, an act of parthenogenesis that defies everything we know about mammalian genetics and reproduction. Nor do the researchers wring their hands over the 80 percent who believe in the resurrection of Jesus, the laws of thermodynamics be damned.

No, most scientists are not interested in taking on any of the mighty cornerstones of Christianity. They complain about irrational thinking, they despise creationist "science," they roll their eyes over America's infatuation with astrology, telekinesis, spoon bending, reincarnation, and UFOs, but toward the bulk of the magic acts that have won the imprimatur of inclusion in the Bible, they are tolerant, respectful, big of tent.

...

According to a 1998 survey published in Nature, only 7 percent of members of the prestigious National Academy of Sciences professed a belief in a "personal God." ... Yet only a flaskful of the faithless have put their nonbelief on record or publicly criticized religion, the notable and voluble exceptions being Richard Dawkins of Oxford University and Daniel Dennett of Tufts University. Nor have Dawkins and Dennett earned much good will among their colleagues for their anticlerical views; one astronomer I spoke with said of Dawkins, "He's a really fine parish preacher of the fire-and-brimstone school, isn't he?"

So, what keeps most scientists quiet about religion? It's probably something close to that trusty old limbic reflex called "an instinct for self-preservation."

—Natalie Angier, Free Inquiry Magazine

[Cubs shortstop] Nomar Garciaparra took two steps out of the batters' box, then collapsed in pain.

...

``I've strained my groin before but I've never felt anything like this before,'' Garciaparra said. ``It's kind of a freak thing.''

...

``It's not good, it's a bad groin,'' manager Dusty Baker said. — AP Sports

Gus adds, "What's really interesting is that the letters in the name Nomar Garciaparra can be rearranged to spell,"

ram... groin... crap... aaar!!

Wow. Cardinal Ratzinger becomes Pope Benedict XVI. I know he was spoken of frequently as a possible, even one of the most likely successors to John Paul II. But I'm still a bit stunned to see it.—Josh Marshall

"NOBODY expects the German Inquisition!"[1] [2]

[1]: "Ratzinger's stern leadership of the Congregation of the Doctrine of the Faith, the modern successor to the Inquisition, delighted conservative Catholics ..." — Reuters

[2]: "NOBODY expects the Spanish Inquisition! Our chief weapon is surprise...surprise and fear...fear and surprise.... Our two weapons are fear and surprise...and ruthless efficiency.... Our *three* weapons are fear, surprise, and ruthless efficiency...and an almost fanatical devotion to the Pope.... Our *four*...no... *Amongst* our weapons.... Amongst our weaponry...are such elements as fear, surprise.... I'll come in again." — Monty Python

"Over the last 20 years, inflation of recommendations has paralleled the inflation of grades," says Stuart Rojstaczer, an associate professor of hydrology at Duke University. "Someone to whom you might have given a good recommendation 20 years ago, you now say is very good. Very good is excellent, and excellent is outstanding. And if someone truly is outstanding," he says, his voice trailing off, "well, I don't know what you say."

He once made the mistake of pumping up the volume in a letter sent to a university in Britain, where hyperbole is not the norm. The student was excellent; he called her "outstanding." The next thing he knew, he was the one getting called -- by the search committee. They wanted to know if the letter had been forged. "It was so hyperbolic in their eyes that they couldn't believe it," Mr. Rojstaczer says.

Mr. Leiter, the Texas philosopher, explains: "An English philosopher might write, 'So-and-so has done very fine work.' If that were coming out of Harvard, it would mean this person barely has a three-digit I.Q. Coming out of Oxford, it could well mean this person is one of the top three people coming out of the U.K."

— ALISON SCHNEIDER, The Chronicle of Higher Education

I recently spent a week outside Amsterdam running a training session for Greenpeace's various regional webmasters. In addition to spending time with so many bright, shiny world-changers, my trip offered a few highlights. One, I was able to meet my grandfather's cousin, Heinz Aufrecht, whose family fled from Berlin to the Netherlands in 1934 or 1936, shortly after my grandfather Werner Aufrecht fled to the US. I also caught an Einsturzende Neubauten concert at the Paradiso, very cool former church which is now a nearly perfect music club: great acoustics and two big wraparound balconies providing plenty of seating for those (like me) who would rather not stand jammed in a crowd for hours, peering around tall Dutch (Paradiso) or having drunk people lurch into them (Belly Up in San Diego). Also very cool was that the band started very close to the printed time on the ticket, which essentially never happens in my experience. In fact, what they did was even cooler: around starting time, the lights dimmed partway and roadies came out and turned on the two air compressors on stage. The conversations then competed with the chugging compressors until first one and then the other coughed and stopped, leaving silence and darkness as the band came out on stage ....

Instruments played included: PVC pipes, played with compressed air sprayed into ends or beaten like xylophone keys; spoked metal wheels in conical shapes, struck; kettle drum, drummed; electric guitar and bass, played normally or, once, with a golden vibrating dildo to the strings; synthesizer keyboard and Powerbook; big metal plate, drummed; power drill, applied to metal plate; 10 gallon tin can, e.g. an olive oil can, drummed; five ten-gallon cans, tied together like a cat-o-nine-tails, dragged across the stage and flung overhead; wrenches or tools, rattled and tapped; giant slinky, strummed; metal brick, drgged along metal plate; throat singing; chains and sand, moved around on steel bench; electric grinding tool, applied to steel plate; space blanket, shaken gently; long plastic pipe, blown like a horn; transistor radio, played (possibly a prop); spinning cellopane/plate/cup contraption, function and sound hard to discern; plastic jerrycans, beaten; face and mouth, sprayed with compressed air; pvc pipe, about (10cm x 2m), curved and strung, string beaten to produce tenor-like tone; stage, drummed; metal pipes, arrayed on the stage on bumpers and played by seated band members to produce a hex on a previous manager and record company; various sound control boxes, played with bare feet and used, among other things, to produce loops of previously played sound.

Baseball Prospectus breaks down the 47 free agent contracts signed over the 2004-2005 winter. The math is too complicated to explain without quoting most of the (subscription-only) article; basically they estimated how good the players in the free agent class of 2004-2005 were, and counted how much money they were paid, and then figured out which of those players were underpaid or overpaid, relative to the group. Other adjustments included discounting multi-year contracts by 5% per year and comparing the players to "replacement level" (the quality of player you can get by paying the minimum $316,000/yr salary) instead of to zero.

The three best deals:

1. Guzman, Cristian, signed a 4 yr/$15mil contract. Projected to win 14.7 games. By BP's estimates, underpaid by $17 million.
2. Drew, J.D., 5 years/$40 mil. Worth 29.1 wins, and underpaid by $14mil.
3. Eckstein, David, 3 years/$9 mil. Worth 10.2 wins, so underpaid by $13mil.

The worst deal:

1. Ordonez, Magglio, 5 years/$67 mil. Worth 16.4 wins; overpaid $32 mil.

Dodger fans will be interested to learn that Drew is projected to win more games than Beltre in the next five years, and at a substantial discount. (29.1 wins vs 28.2; $40mil vs $57mil). Overall, the Dodgers were the second most efficient spender in the market (Drew and Kent were both very good deals; Lowe and Perez were both mildly bad deals). The Mariners, Diamondbacks, and especially Tigers overpaid the worst.

Also interesting is that this market put the value of a projected win at $2.14 million, which is double the current estimated value of a win. That is, looking at the revenue and records of the different baseball teams, BP statheads have determined that winning one extra baseball game is worth (other things being equal) a million dollars in gate revenue, TV contracts, etc. Therefore, paying a player up to a million dollars per projected win is good economics. By that measure, almost all of this years free agents were overpaid. Which is not unexpected; baseball players are indentured servents for the first six years of their major-league careers, during which period they are severely underpaid relative to the money they generate for their employers. By limiting free agency to veterans, the owners have ensured that free agents are almost always overpaid.

As a senator, I worried every day about the size of the federal deficit. I did not spend a single minute worrying about the effect of gays on the institution of marriage. Today it seems to be the other way around. — John C. Danforth