The Value of Not Getting to the Point

I read somewhere recently, I forget where, that the purpose of people getting together for a conversation over a beer or coffee or lunch or dinner is that it the food and drink spare us from the burden of needing to have something to say throughout the whole conversation.

This was a revelation to me. All this time, I assumed that the primary purpose of lunch was lunch. All this time, I figured that I was just lousy at conversation because being an introvert made conversation awkward and laborious for me. For everyone else, conversation seems comparatively effortless. But it seems from this data that conversation must be harder for everyone else than I had assumed.

My oldest daughter is a freshman in college. She recently texted me and said she wanted to talk. I asked, what about? She got annoyed at me for asking.

I was clueless as to why. I guess the Dunning-Kruger effect applies to all of us, there’s always some area of life where we’re so incompetent we don’t even know we’re incompetent. This area, apparently, was one of mine.

She asked if we could just talk about something stupid. So I called her, and we talked about Donald Trump and the presidential race and stuff like that for a good long while. I didn’t ask about what was really bothering her.

Eventually, the conversation turned, and we finally got to talking about the thing she wanted to talk about. But that probably at least half an hour into the conversation. We segued slowly and organically from the stupid stuff into the real issue.

And this, too, was a bit of a revelation to me, that someone would not want to get straight to the point, that someone would need a nice long conversational warmup before they’d feel comfortable enough to be ready to talk about something more uncomfortable. I’m very much a get-to-the-point kind of person. I tend to say what I mean, or nothing at all.

Language is imprecise. Our feelings don’t always have direct translations into speech. It’s hard to explain what we feel, to say exactly what we mean. We have wants and desires and emotions, and we often try to rationalize those feelings. Those rationalizations are often logically incoherent. But it’s hard to see the incoherence of our own rationalizations because our points of view are so limited. And often (if we’re not falling prey to the Dunning-Kruger effect) we intuit that our rationalizations may be incoherent. So we’re cautious in what we say. We know that there can be social penalties for saying the wrong thing in the wrong way to the wrong person.

All this adds up to making the act of talking about something sensitive daunting. There is a vulnerability in speaking. That’s why our culture has all these rituals and conventions around conversation, like idle chit-chat and coffee and such: to build enough trust in the environment where we can feel comfortable enough to overcome the vulnerability inherent in speech.

I never fully understood this before. I feel like everyone else understands it, though, because they act as if they do. But if they do, it must be an intuitive understanding, a grokking, not an explicit fact that people state out loud. Otherwise, I probably would have heard someone say it explicitly sometime before in the almost 50 years I’ve been in this earth.

Having now finally come to this understanding, it occurs to me that perhaps this is the great flaw with Twitter, why everyone I know on Twitter seems to eventually run into a wall with it. The 140-character format pushes you to get straight to the point. There is no room for the idle chit-chat and sips of coffee and other conversational rituals that let us dance around the sensitive issues. Without these rituals that are built into real-life human-to-human conversation, the problems with speech that those cultural rituals are designed to prevent come flooding in.

There is so much hair pulling and teeth grinding about what people should and should not say online, and how they should or should not say it. And maybe all that hair pulling and teeth grinding arise because our online conversational cultures, and the technological platforms they reside on, have not had the time to evolve into something that works, the way that our real-life conversational culture has.

There are many, many more people who are clueless about how to behave in online conversations than there are people who are clueless about how to behave in offline ones. How I came to be the flipside of that, I don’t know.

And it also occurs to me that there is a value in stating explicitly the things that are mostly just intuited about human nature and human culture. I want to explore these sorts of things. There is a risk, though, a vulnerability, in stating these things. The people who intuitively grasp these things will feel as though I am insulting their intelligence by stating something so obvious it shouldn’t need saying. But it isn’t meant as an insult to their intelligence, it’s meant as an insult to mine. I need to say these things because I’m the one who doesn’t understand these things. I need them explained to myself.

Which is all a roundabout way of stating something that maybe could fit into a tweet: I plan to start saying things that aren’t obvious to me but may be obvious to others. Sorry if you fall into the latter category and I waste your time. Such is the risk of saying anything, ever. And sorry for the roundaboutness in getting to this point. I seemed to need it, for some strange reason.

Vulnerability, Trust and the Marshmallow Test

In God we trust; all others must bring data.

–W. Edwards Deming

The Marshmallow Test is one of the classic experiments in psychology. In 1972, Walter Mischel gave 600 preschoolers a choice of one marshmallow now, or if they could wait, two marshmallows 15 minutes later. When he followed those kids into adulthood, he found that the kids who could wait had much more success in life.

In 2012, researchers at the University of Rochester added a little twist to the famous experiment. They preceded the marshmallow test with a promise about some crayons that was either broken, or kept.

It turns out that almost none of the kids who were given the unreliable offer ended up waiting for the extra marshmallow. Why should they? They already waited for a promised reward that didn’t come. The rational thing to do in an untrustworthy environment is to take any reward that is presented. Fool me once, shame on you; fool me twice, shame on me.

As I wrote in Box 32 of my Forty-two Boxes essay, children are born trusting. They are totally helpless, they have no other choice but to trust. But then through life experiences, they gather data that can change that default setting of trust.

In my last blog entry, I quoted Mister Rogers saying, “One of the first things a child learns in a healthy family is trust.” This modified marshmallow test demonstrates is why trust is so important. Trust allows people to spend energy and resources now for a greater payoff later.

You are told if you do your homework now, in 10 years you’ll get to go to college, and in 15 years, you’ll have a great career. You are told if you work hard at your entry-level job now, and you will eventually get promoted into management later. But if you don’t trust what you are told, if you don’t trust these equations, if the data keeps telling you there’s a glass ceiling you’re unlikely to surpass, you’ll probably go do something else where the data tells you the odds of success are better.

This is why poverty, racism, sexism, and totalitarianism are so destructive. Not only because those things present roadblocks in and of themselves, but because they corrode the trust that a sacrifice now will be worth a payoff later. Both society as a whole and the individuals in it fail to achieve their potential, because people in an untrustworthy environment take fewer long-term risks, and receive correspondingly fewer long-term payoffs. The result is stagnation.

Mister Rogers on Vulnerability and Trust

In my last post, 42 Boxes, I spent 17,000 words trying to get around to the point that the first principle of human morality is that humans are vulnerable, and that the antidote to that vulnerability is trust.

In 1969, Fred Rogers (aka Mister Rogers) addressed Congress. Early on in his address, he came right out and said this:

One of the first things a child learns in a healthy family is trust.

It didn’t take him 17,000 words to say the same thing I did, because Mister Rogers was a saint and a genius, and I, in comparison, am a dim-witted blowhard. He later adds:

If we can only make it clear that feelings are mentionable and manageable, we will have done a great service for mental health.

The recipe for a healthy family, or a healthy society, is simple. Admit your feelings about your vulnerabilities, trust that you can talk about them, and you will be able to control these vulnerabilities in a constructive, not destructive, fashion. It’s not that complicated.

Forty-two Boxes



When you start looking at a problem and it seems really simple, you don’t really understand the complexity of the problem. Then you get into the problem, and you see that it’s really complicated, and you come up with all these convoluted solutions. That’s sort of the middle, and that’s where most people stop. . . . But the really great person will keep on going and find the key, the underlying principle of the problem – and come up with an elegant, really beautiful solution that works.

Steve Jobs


Beginning a story with a quote often implies that the rest of the story will say same thing as the quote, but with different words. This story follows that formula. The opening quote serves as a box within which the rest of the story is confined.

This story is not original. It says what Steve Jobs said in the above quote. It says other things that other people have also been saying for hundreds and even thousands of years. So why bother telling this story?

We tell stories because there are simple approaches that don’t address the complexity of the problem. We tell stories because there are convoluted solutions where people have stopped. We tell stories because sometimes the underlying principle remains, but the old, elegant, once-beautiful solution has now stopped working.

Sometimes the lock changes, and we need a new key. Sometimes we refuse a key from one person that we will accept one from another. Sometimes this particular key won’t work for us, but a different key will click the door open. And sometimes we need to try a different door entirely to get into that room.

We tell stories because we are human beings, endowed by our creator with the delusion of hope. We tell stories in faith, believing, without evidence, that communication will forge a key that unlocks something incredible and amazing.


I got mad at my kids recently for having a messy room.

It’s such a cliché, I know. In that moment, I was an ordinary parent, just like everyone else, easily replaced by a thousand identical others.

Although, that’s not exactly true. I had my own, different angle on the messy room story. I didn’t really get mad because their rooms were messy. I got mad because their messiness was starting to spread out into my spaces, the common areas of the house that I keep clean. I did not want my space to be a new frontier for their stuff to conquer.

Wait, that’s not exactly the whole story, either. I didn’t even get mad because their stuff was getting all over the house. I got mad because when I suggested that we go to IKEA, like a good Swedish-American family, and look for some solution for where they can put their backpacks and schoolbooks and binders and such, so that I can keep my spaces clear of their stuff, they laughed.

I got mad because they laughed.


Is a story a kind of technology?

The word technology derives from the Greek words for “skill/craft” and “word”. Since a technology is a set of words about skills, perhaps a story is the original technology, the underlying technology upon which all other technologies are based.

We craft our words into a story, to transfer information from one person’s brain to another person’s brain. The more skillfully we craft our words, the more effectively that information is transferred, retained, and spread.

The most celebrated technologies of our times, Google and Facebook and Twitter, are merely extensions of this original technology. They are the result of stories built on stories built on stories over thousands of years, told orally, then in print, then digitally, all circling back to their original purpose. They are ever more effective tools to transfer, retain and spread information from one human being to another.

Continue reading

The Long, Long History of Why I Do Not Like the Josh Donaldson Trade

Once upon a time, about a billion years ago, life was simple. Everybody lived in the oceans, and everybody had only one cell each. This was quite a fair and egalitarian way to live. Nobody really had significantly more resources than anyone else. Every individual just floated around, and took whatever it needed and could find, and just let the rest be.

This golden equilibrium was how life did business for a couple billion years. There was no such thing as jealousy or envy, and as a result, everyone lived pretty happy lives.

Then, one day about 800 million years ago, a pair of single-celled organisms merged to become the first multi-cellular organism in the history of the earth.

At first, these multi-celled creatures were just kind of like big blobs of single-celled organisms, and didn’t cause a lot of problems. Everybody was still kind of doing the same job as everyone else, even if they had organized themselves into a limited corporation of sorts. Most other single-celled creatures just figured they were harmless weirdos hanging out together, and ignored them.

They could not have been more wrong. For once the multi-cell genie was out of the bottle, Pandora’s box could not be closed, and the dominos began to fall. This simple change may have seemed innocent at first, but little did the single-cells know that they were the first creatures on earth to fall victim to the innovator’s dilemma. The single-celled creatures were far too invested in the status quo to change, and consequently ignored the multi-cellulars as irrelevant, and did not realize until it was too late that the game had suddenly shifted.

Continue reading

The Short, Short Josh Donaldson Trade Story Based on Platoon Splits

Ok, look, I told y’all with the Cespedes trade that you can’t analyze an A’s trade of a position player without breaking it down by platoon splits across the whole lineup. But did any of y’all listen to me? No. Y’all are still trying to analyze Donaldson vs Lawrie as if they are single players on single teams instead of two players on two platoon teams with other players on the team. So stop that.

Now look, I’m gonna make this simple. I’m going to assume that both Lawrie and Donaldson will be equally healthy, and they’re roughly comparable defensive players. They may not be, but this is a quick and dirty exercise here, so bear with me. And I’m just going to use OPS, so I don’t have to make this story as long as the other Josh Donaldson story that’s coming later today.

Let us begin.

* * *

OPS 2014/career, Josh Donaldson vs. RHP: .727 / .744
OPS 2014/career, Josh Donaldson vs. LHP: 1.007 / .953

OPS, 2014/career Brett Lawrie vs RHP: .760 / .760
OPS, 2014/career Brett Lawrie vs LHP: .595 / .713

See, Brett Lawrie is actually better than Josh Donaldson against RHPs. The difference is that Donaldson crushes LHPs, and Lawrie for whatever reason actually is worse against LHPs than RHPs. He was particularly bad in 2014. I do not know why.

So for the platoon team that plays 2/3s of the A’s games, the one against RHPs, the A’s lineup actually just got better.

* * *

So now we need to fix the 1/3 of the A’s games against LHPs.

Last year, one of the A’s primary 1B/DHs against LHPs was Alberto Callaspo. He was awful. The A’s have signed Billy Butler to replace him.

OPS 2014/career, Alberto Callaspo vs LHP: .518 / .729
OPS 2014/career, Billy Butler vs LHP: .847 / .912

So the A’s are losing about .400 OPS points by downgrading from Donaldson to Lawrie vs LHPs, but they get back about .300 of those OPS points by upgrading from Callaspo to Butler.

So now all Billy Beane has to do find that extra .100 points of OPS against LHPs, and the math works. Maybe it will come just out of the fact that most players don’t have reverse splits last their whole careers, and Lawrie will actually bounce back and hit better against LHPs in the future. If so, QED.

* * *

Disclaimer: the above analysis does not mean I like this trade. I do not like this trade. That (much longer) explanation is here.

10 Things I Believe About Baseball Without Evidence

Well, here we are. The Giants won another World Series, while the A’s flopped in the playoffs yet again. I’m not one of those A’s fans who hate the Giants, but it’s starting to annoy the crap out of even me to see the Giants always succeed in the playoffs, while seeing the A’s always fail.

The A’s have had 14 chances in the last 14 years to win a game to advance to the next round of the playoffs. They have lost 13 of those 14 games. If the playoffs are truly a crapshoot, the odds of this happening are 1-in-1,170. (So it’s not technically always — they could have gone on to lose the 2006 ALDS against the Twins, too, which would have made them 0-for-16, with an unlikelihood odds of 1-in-65,536. So if you want to look on the bright side, things could be 56 times worse than they are.) And in a crapshoot, the odds of the Giants winning 10 playoff series in a row, as they have now done, is 1-in-1024.

So if you’re an A’s fan who hates the Giants, and who believes that the playoffs are a just crapshoot, you’ve been struck with a series of unfortunate events that had literally less than a 1-in-a-million chance of happening.

Sabermetrics has come up with no good explanation for it except to say, well, these things happen about once every thousand times, or once every million times, sorry A’s fans, it just happened to be your turn to hit that unfortunate lottery, and it’s just bad luck. Oh, and you have a crappy stadium that’s falling apart and a team ownership and a local government who all seem too incompetent to do anything about it unlike those guys across the bay, sorry about that, too, gosh you guys are unlucky, tsk tsk tsk.

Which is just a deeply, deeply unsatisfying answer. If you have an ounce of humanity, you will reject that explanation, and ask the obvious question.

But Why?

And to answer that question, the sabermetrician dives into the numbers, and pulls some out numbers with some number-pulling-out tools, and finds nothing to report. Nope, no evidence here of anything, so it must just be bad luck.

To which I ask: what if the reason the number-pulling-out tools can’t find any cause for the problem is because those number-pulling-out tools themselves are the problem?

I have no evidence of that. But it’s something I believe might be true, even though I can’t prove it.

* * *

I have a number of these beliefs–or hypotheses, if you will–about baseball, but I’ve mostly kept them to myself because of this lack of evidence. What the hell do I know, anyway? Who am I to pontificate? And why bother spouting these theories when I can’t defend them with evidence? So I just keep my mouth shut.

But I got a little bit of self-confidence in my belief system when Robert Arthur of Baseball Prospectus took one of my hypotheses (that injured A’s in the second half of 2014 had begun cheating on fastballs, making themselves vulnerable to offspeed pitches) and found evidence to support it ($):

The overall pattern of changes is beautifully consistent with Ken’s theory…

It’s very satisfying to find that the data supports one’s theory!

But I didn’t just come to this particular hypothesis that Mr. Arthur investigated out of thin air. This hypothesis arose out of a deeper foundation of hypotheses that color the way I look at baseball. I want to put all those hypotheses out on the table now, lack of evidence be damned. And maybe someone (maybe me someday, if I ever find the time and energy and resources and willpower to do so, which hasn’t happened yet) will take those hypotheses and invent the technology needed to find the evidence to support it.

So let’s put it out there.

* * *

Belief Without Evidence #1. A technological Sapir-Whorf hypothesis

The Sapir-Whorf hypothesis, a/k/a the Linguistic relativity principle holds that the language that a person speaks influences the way a person conceptualizes their world. The obvious example of this is that people have trouble distinguishing between colors if their language does not have a word for that color.

To a certain extent, I believe this hypothesis. Being fluent in both Swedish and English, I know there are certain concepts, such as the difference between belief in an opinion and belief in a fact, where the Swedish language makes clear distinctions (tycka and tro) and English does not. English speakers spend ridiculous amounts of time arguing about these things, and Swedes simply don’t need to. It’s not that English speakers can’t conceptualize the difference between opinion and fact, but doing so is way more difficult in English, because the word “belief” in English is quite fuzzy, whereas in Swedish, the language makes it simply impossible to confuse the two.

I touched upon this in my essay in the 2014 Baseball Prospectus annual, that I believe a similar concept applies to the technology we use. The reason statistical analysis began to influence the way we conceptualize baseball in the 1990s is not because human beings suddenly became smarter in the 1990s. There were statistically informed people who suggested such analysis almost a century earlier. It happened in the 1990s because the price of the technology needed to perform such analysis had finally became reasonable.

The predominant technology we use to perform such analysis is SQL, which is the primary language used to query relational databases. SQL and relational databases are technologies which are built upon set theory. A set is basically an unordered collection of objects.

And this is where I believe that a technological Sapir-Whorf hypothesis applies to baseball. Practically all of our analysis of baseball statistics treats its data an unordered collection of baseball events: pitches, plate appearances, games, series. Standard baseball analysis (the public kind anyway, who knows what is being done inside these organizations) treats its data that way because that’s the way SQL treats its data. The available technology guides our conceptualization of the world. And that leads us to my second hypothesis:

Belief Without Evidence #2. Baseball events are NOT unordered

For any batter to hit a ball, the batter needs to predict where the ball is going to be before it reaches the bat. There are two different mechanisms for this prediction.

First, there is a conscious prediction. The batter may decide, consciously, based on some sort of rational analysis, that he is looking for a fastball down and in, and wants to swing at only a pitch in that location that he can pull.

But once the pitcher releases the ball, this kind of conscious prediction mechanism is far, far too slow to be of any use. At this point, everything is turned over to a much faster, subconscious, automatic system to predict the actual flight of the ball, and to send the muscles in motion to meet the ball.

My thoughts here are heavily influenced by Jeff Hawkins‘ book On Intelligence, which lays out a framework for how this automatic system in the brain works as a memory-based prediction machine.

Order matters in baseball, because this automatic prediction mechanism has a strong recency bias. (A conscious prediction might not have a recency bias if truly rational, but how often does a batter perform a purely rational analysis at the plate?) The speed, location and movement of the most recent pitch will affect the brain’s automatic prediction of the speed, location and movement of the next pitch. The more recent a pitch, the more it affects the automatic system’s prediction for the next pitch.

Pitch sequencing, therefore, is at the heart of the very sport of baseball, yet it is woefully understudied in current public analysis, because our tools, based on a foundation of unordered sets, are woefully bad at processing and studying sequenced events.

There is a whole industry now dedicated to the statistical analysis of baseball using these set-based SQL tools. But SQL does not have a recency bias clause in its syntax that you can apply to a query. Because these tools don’t handle the ordered data well, they basically ignore The. Very. Core. of the sport: the sequencing battle between pitcher and batter.

Let me say that again: statistical analysis (that we in the public are aware of) takes the most important element of the sport, and ignores it.

It’s like having Newtonian physics without relativity and quantum mechanics. There’s a lot you can do with Newtonian physics, but at the extremes, it begins to break down, because it is ignoring some deeper, more fundamental truths.

If you’re a team that relies on constructing its roster using such statistical analysis, what mistakes are you making by ignoring the most important part of the game?

Belief Without Evidence #3. All high-level sabermetric truths derive from lower-level truths about human biomechanics and psychology

And not vice versa. Things like platoon splits and home field advantage are not Constants of the Universe like the speed of light or the Planck-Einstein relation. The arise from more fundamental truths about human anatomy and psychology.

For instance, once I got in an argument in which I did not believe that Sean Doolittle pitched better to certain catchers than others. The stats did not agree with me, albeit perhaps with a small sample size. But my objection wasn’t to the numbers, adequate sample size or not, it was to the lack of any sort of underlying physical/psychological mechanism where this these numbers could derive from. Sean Doolittle throws 90% fastballs. What the hell difference physically/psychologically does it make what catcher is back there catching it? It’s the same pitch, no matter who is catching it.

I do not consider a sabermetric truth to really be a truth unless there is a biomechanical/psychological foundation upon which that truth can rest, and from which that truth is capable of being derived.

Belief Without Evidence #4. Pitches are paths between states in a Prediction State Automaton

First, a little explanation of automata:

Automata theory is used in computer science to study states. For example, you can look at baseball as an base/out automaton, where before each plate appearance, the base/out combination is in one “state”, and in another “state” after the plate appearance. There are rules that tell you what possible states you can be in before and after a plate appearance.

So, at the beginning of an inning, the baseball base/out automaton is in a {Nobody on, 0 out} state. After the first plate appearance, you will be in one of five possible states:

{Runner on 1st, 0 out}
{Runner on 2nd, 0 out}
{Runner on 3rd, 0 out}
{Nobody on, 0 out} (Batter homered)
{Nobody on, 1 out}

You can’t, after the first appearance, reach a state where there are two runners on or two outs. You have to go to an intermediate state first. There are exactly 24 possible states you can have in this automaton. Each state in this automaton is a two dimensional {base, out} object. And from any of these 24 possible states, there are a limited, finite number of possible following states.

The “automaton” then, defines the what possible states can exist, and the rules by which you can move from one state to another.

Got it?

OK, now to the thing I believe without evidence: I believe that before any given pitch, the batter is in some sort of Prediction State for the next pitch. After each pitch, the batter then moves into a different Prediction State.

I don’t have a clear belief on exactly how many dimensions these Prediction States have. Maybe the Prediction State has three dimensions it:

1. Whether to swing
2. When to swing
3. Where to swing

Or maybe these Prediction States are much more complex, combining the above three states with specific kinds of pitches and movements and locations. It may be expressed by something like this, for example:

{60% expection of fastball, 30% changeup, 10% curve;
80% outside, 10% middle, 10% inside;
60% down, 30% middle, 10% up;
70% in the strike zone, 30% out of the strike zone}

and then if the pitcher throws you a fastball on the lower outside corner for a strike, perhaps you move to a state like this:

{70% fastball, 20% changeup, 10% curve;
85% outside, 8% middle, 7% inside;
70% down, 20% middle, 10% up;
75% in the strike zone, 25% out of the strike zone}

Or whatever. I don’t really know as what the parameters for these Prediction States should be. Is it {pitch type, in/out, up/down, movement/straight, fast/slow} or some other combination of pitch attributes? I don’t know.

And to what extent are these prediction automata more or less universal, or does each batter have his own unique automaton with its own unique rules? Again, I don’t know.

But I do know that if I were to build a technology for analyzing baseball, this is where I would begin, right at the core of the game, the engine that drives the sport: what pitch the batter is expecting from the pitcher, and what happens when the pitch he gets conforms or deviates from that expectation.

In order to unite the quantum and Newtonian versions of baseball analysis, the biophysical and the statistical, any Grand Unified Theory Of Everything Baseball must, in my belief, have some way to handle the Prediction State of the batter.

Belief Without Evidence #5: The quality of a pitch is a function of its speed, location, and movement, and also of the batter’s swing and prediction state

There are a few pitchers, like Aroldis Chapman, who can throw a pitch with such high-quality speed that the location, movement, and prediction state are rather irrelevant. And there are some, like Mariano Rivera, who have such a combination of high-quality location and movement that the speed and prediction state don’t matter much. With pitchers like that, the batter can predict perfectly what pitch he’s going to get, and still not hit it.

But most pitchers do not possess such a high-quality pitch that they can be predictable and get away with it at the Major League level. They need to manipulate the prediction state of the batter in order to succeed.

The less a batter is expecting a certain pitch, the less likely he is to make good contact. But pitching is not just a function of being unpredictable: the pitcher must balance what the Prediction State of the batter is and the batter’s ability to hit it, with his ability to also throw a pitch with good speed, location, and movement.

The complex nature of that 5-dimensional object ( {speed, location, movement, swing, prediction state} ) is what makes baseball so fascinating from pitch to pitch.

So for each pitch, the pitcher wants to:

1. Choose a pitch the batter is likely to predict incorrectly
2. Choose a pitch the pitcher is likely to throw with good speed, location, and movement
3. Choose a pitch which will result in a suboptimal swing path, resulting either in a miss or weak contact
4. Choose a pitch which, if not put in play, worsens the batter’s Prediction State for the next pitch

Belief Without Evidence #6: The quality of an at-bat is a 3-dimensional function

Those three dimensions being:
1. Getting a good pitch to hit
2. Hitting a ball hard when you do
3. Hitting a ball hard if you don’t.

A good pitch to hit is a pitch that (a) he is successfully predicting, and (b) he can get a good swing on. Whether he can get a good swing on a particular pitch depends on what his swing path is.

And again, there are two kinds of predictions: the automatic subconscious one where the batter just reacts to a pitch, and a conscious one where the batter decides beforehand to look for a certain pitch and ignore all others. And the count plays a big role whether the batter can take an approach to consciously look for a particular pitch, or whether he should (with two strikes especially) just let his subconscious react to whatever comes in there.

On the subconscious level, the more the pitcher keeps throwing the same pitch, the more the batter predicts that pitch accurately, and the more likely the batter is to hit that pitch. When pitchers talk about “establishing the inside fastball” for example, this is what they mean: to change the Prediction State in such a way that an inside fastball becomes part of the Prediction State, and thereby necessarily reduces the expectation of a different pitch in the future.

Just because a batter gets a pitch he is predicting, does not mean he will hit it. Most batters have some kind of hole in their swing. Some batters prefer high pitches, others low. Some are vulnerable inside, and others can’t hit the outside pitch well. Some can hit a fastball, but can’t time an offspeed pitch. Others have a slow bat speed and struggle with fastballs, but feast on the slower pitches.

So for each pitch, the batter wants to:
1. Predict a pitch correctly
2. Swing at a pitch that if it lets him approximate his optimal swing path
3. Take a pitch if it would cause a suboptimal swing path (unless 2 strikes in zone)
4. Take pitches out of the zone to move to a better Prediction State for the next pitch
5. If in a 2-strike situation, make contact (foul or fair) on a pitch in the zone

Belief Without Evidence #7: SQL-reliant GMs don’t value the third dimension of #6 enough

In a vast sea of unordered pitches from an unordered group of pitchers, you will get a randomly-distributed plethora of good pitches to hit, so the numbers will all work out in the end. So you acquire hitters based on these vast seas of data, ignoring what the batter does with difficult pitches to hit, because in the long run, they don’t matter much.

But against a good pitcher on a good day who does not give you a good pitch to hit, what do those batters do? Do they hit a ball hard if they don’t get a good pitch to hit?

To me, the biggest difference between the A’s in the playoffs and the Giants in the playoffs is Pablo Sandoval. Because there may not be anyone in baseball right now better than Sandoval who does damage even when he does not get a good pitch to hit. He can turn pitches in the dirt, in his eyes, and/or six inches off the plate into a hit. He’s almost immune to prediction state manipulation by opposing pitchers. And Hunter Pence, though not as extreme as Sandoval, has similar characteristics.

The A’s simply do not pursue those types of players. Players like Sandoval tend to have low OBPs, because they swing at so many bad pitches. Minor leaguers with that profile flop far more than they succeed, so they’re a bad risk to take. But there are times, against a good pitcher on a good day who is simply not giving hitters a good pitch to hit, that it is valuable to have a player who often does damage even with a bad pitch to hit. And those times happen more often in the playoffs.

A technology that used a system of evaluating players in which high-level statistics of player value were derived from a low-level {speed, location, movement, swing path, prediction state} matrix would better identify the true value of such players.

Belief Without Evidence #8: Diversity is Good for Batting Lineups

This belief is related to the belief about the definition of the quality of a pitch, and to the belief of a biomechanical/psychological foundation to all of this. A lineup with too many batters with similar strengths and weaknesses can make it easier for a pitcher to settle into a psychological/mechanical rhythm and mow down such a lineup. A lineup that is diverse (some hit fastballs, some like it inside, or low, some slug, others make contact, etc.) makes a pitcher have to change his approach from at-bat to at-bat. That forces the pitcher to have to make a variety of quality pitches in order to win. It’s harder for a pitcher to win if he has to have multiple pitches working well.

So when I praised the Giants for having Pablo Sandoval, I did not mean that an entire team of hitters like Pablo Sandoval would be ideal. But having one or two guys like him in a lineup with some more patient-type hitters is a good thing.

Belief Without Evidence #9: A lineup without holes scores runs exponentially, not linearly

This is probably the easiest of my hypotheses to disprove. But I have the gut feeling that one guy who is an automatic out in the middle of a lineup can take a rally that might score five runs and drop that rally down to 0 or 1 runs.

I think we saw this play out with the 2014 Oakland A’s. At the beginning of the year, everyone in the lineup was healthy and hitting somewhat near or above expectations. The A’s were just killing it in the pythagorean win column, because they’d get a rally going and that rally would just keep going and going.

But then Josh Donaldson and Brandon Moss started having some nagging injuries, and Moss in particular became pretty much an automatic out for a month or two. Those five-run rallies, once plentiful, almost instantly disappeared. Every rally seemed to be killed by a terrible at-bat in the middle of it.

Almost every team has a hole in the lineup at any given time, someone who is slumping for whatever reason. So for most teams, run scoring appears to be linear. But in those rare cases when everyone is clicking at the same time, their run scoring graph turns like a hockey stick and shoots upward.

The A’s success early in the year depended on the lineup being holeless, and when holes appeared, the whole thing collapsed back from exponential scoring into linear.

Belief Without Evidence #10: A’s fans are magical elves

I’ve been playing in my mind lately with the idea that A’s fans are like the house elves in the Harry Potter stories.

We exist so that others may abuse us. The greatest triumphs of others often comes at our expense. We dress in ratty clothing (stadium). Yet despite this constant abuse, we are fiercely loyal to our master. We attack viciously anyone who dares attack our master. We perform magic (great stadium atmosphere) on their behalf, no matter how awful our masters treat us in return.

If ever we were given clothing (a new stadium) by our master, we would be free of our bondage. Some, like Dobby, desire this, but others would not know what to do with themselves with freedom and wealth. It would ruin the very essence of their being.

I used to be like Dobby, longing for the freedom that a World Series victory and/or a new stadium would bring. But now, I am beginning to feel that the other elves are right — that it is wrong to support S.P.E.W. and long for something that would destroy who we are.

We are meant to suffer, so that other wizards may have their glories. We are elves. Let us be that we are and seek not to alter us.

Much Ado About SomethingGate

A play will be performed this weekend which has the following characters:

  • A unethical journalist who befriends a disgrunted employee and seduces an industry insider to gain access to better stories
  • A troll who, out of envy for what he perceives as undeserved fame and attention, slanders an innocent woman, who is then forced into hiding
  • A group consisting entirely of men who believe the slander and amplify it
  • A feminist who rails against the privileges and advantages of such men
  • A completely incompetent law enforcement agency

A brand new play, pulled from today’s headlines? Nope, the play was written about 416 years ago by William Shakespeare. It’s called Much Ado About Nothing.

Encinal High School in Alameda, CA, is putting on the play this weekend. And maybe a hundred people or so will see it, since it’s a high school play, and the first performance will conflict with Game 4 of the World Series, with a Bay Area team playing.

Which is a shame, because the things these kids (including two of mine) have done to pull the current relevance out of a 400-year-old play is remarkable. And that Shakespeare wrote a play 400 years ago that drips with relevance to today’s headlines is equally remarkable. It deserves to be seen by more people than that.

So if you’re an A’s fan looking to escape the Giants’ clutches on our attention this week, or if you have a few spare hours Sunday afternoon before Game 5, please come out and see Much Ado About Nothing at Encinal High School. You’ll enjoy it.


Much Ado About Nothing
Encinal High School cafeteria
Alameda, CA
Saturday, October 25, 7pm
Sunday, October 26, 2pm
Adults $10, Students $5
Running time: about 90 minutes



I believe the evidence is clear enough to tell us this much: We were created not by a supernatural intelligence but by chance and necessity as one species out of millions in Earth’s biosphere. Hope and wish for otherwise as we will, there is no evidence of an external grace shining down upon us, no demonstrable destiny or purpose assigned us, no second life vouchsafed us for the end of the present one. We are, it seems, completely alone.

Edward O. Wilson

In Sophocles’ play Oedipus the King, the title character hears a rumor that he may not be what he thinks he is: the son of Polybus and Merope, the King and Queen of Corinth. Polybus and Merope deny the rumor, but Oedipus seeks external confirmation, and visits the Oracle at Delphi. The oracle ignores his question, and instead prophecies that he will kill his father and wed his mother.

Oedipus has no evidence he is not his parents’ son. He has no evidence to suggest he will eventually kill Polybus and marry Merope. But the latter is a much bigger problem than the former, so Oedipus ignores the first small problem and acts on the second, leaving Corinth forever, so as to avoid this horrible fate. He then proceeds to live his life as if he had solved his problem. And, of course, because this is a Greek tragedy, he hadn’t.

Rumors are not facts. Prophecies are not proven theorems. Yet it is not true that Oedipus had no evidence that he was not his parents’ son. He had the rumor. He had the prophecy. In a Bayesian sense, he should have considered the odds of his being adopted having increased from 0% before hearing the rumor and the prophecy, to what–1%? 10%? 25%?–afterwards.

The odds being less than 50%, however, the logical thing for Oedipus to do when faced with any given binary decision is to act as if the rumor was false. That’s the choice that gives him the best odds of succeeding, based on the information he has.


Hubris is extreme pride and arrogance shown by a character that ultimately brings about his downfall.

Hubris is a typical flaw in the personality of a character who enjoys a powerful position; as a result of which, he overestimates his capabilities to such an extent that he loses contact with reality. A character suffering from Hubris tries to cross normal human limits and violates moral codes.

–Definition of Hubris from Literary Devices

Is it extreme pride and arrogance to make the most logical decision? If so, then the human condition is tragic no matter what decisions we make.

If we choose with the odds based on the best information we have, we risk making a catastrophic decision because we lacked a critical piece of data. If we choose out of rumor and superstition and fear, we risk living a life where bad decisions compound themselves with every choice we make, and we end up living a suboptimal life.

The more successful we are, however, the more likely we are to make the catastrophic decision that results in a classical, Greek-style tragedy. With every successful decision we make, the less likely it is, in a Bayesian sense, that we are lacking that critical piece of information, and the more likely it is, in a Bayesian sense, that our decision-making process is sound.

If you have a decision-making algorithm, and you’re 50% sure it’s good, and then you test it, and it works, now you’re, what–51%? 55%? 60%?–sure that it works. Test it again and it works again, and the odds rise again. Eventually, if you reach the top of a hierarchy and stay there, you get really confident that you know what you’re doing. You’re the king!

Hubris, then, is the logical result of success. In every form of competition, somebody has to reach the top. The closer to the top you get, the more likely it is that you think your success is because of your knowledge and your decision-making process. The more you become certain that your data and your process are sound, the more you should logically make bigger and bigger bets based on that data and that process. And because of those bigger and bigger bets, the harder you will fall if and when it turns out that your data and/or your decision-making process was flawed.


But if you look at the impact those trades have on this particular team’s offense, it’s negligable. Offensively, the numbers tell us that losing Cespedes is no big deal.

Ken Arneson

If you look at Yoenis Cespedes statistically, there’s no real evidence that trading him would hurt the A’s very much. His numbers are mediocre, and easily replaced.

But looking back on the trade now, it feels like the A’s and their fans were focused on the wrong prophecy. The prophecy that a superstar ace pitcher was the missing piece to Moneyball. The significant rumor, the important piece of Bayesian evidence that we ignored was this: that the 2012-14 A’s team was not a product of Billy Beane’s genius. That this team played like complete and utter crap for five years, and then Yoenis Cespedes showed up, and it suddenly and immediately became good. That for 2 1/2 years, when Cespedes was in the lineup, the team played well, and when he was out of the lineup, the team played like crap, regardless of how well Cespedes was playing.

And then Beane, in his moment of hubris, trusting the logic and the data and the decision-making process that had made a best-selling book and a Hollywood movie of his life and had seemingly landed him in first place for 2 1/2 years, traded Cespedes away, and the team reverted immediately to playing like complete and utter crap again.

Could this Cespedes anomaly possibly, actually be real thing? No one can explain it. The fans don’t know why this Cespedes anomaly exists, and all the statisticians don’t know why, and Bob Melvin doesn’t know why, and Billy Beane doesn’t know why. There no evidence! It’s just rumor, innuendo, speculation, unfactual gobbledygook, completely illogical bullshit ex-post-facto rationalization.

But it’s there. It exists. It hurts to look at it. And it has all of us A’s fans wanting to poke our eyes out.

The gods hate us. They want to punish us for our pride and arrogance.

And you may say, gods are superstitious nonsense, that there is no evidence of an external wrath raining down upon us, no demonstrable cruel destiny or fate assigned us, no eternal Sisyphean existence vouchsafed us for the end of the present one.

And that’s true. There is no evidence for the existence of God, or gods. Except for the small, annoying, persistent rumor that at this particular point in time, we are here.

Why I Have Stopped Tweeting

Because my hands are full. Literally. In each hand, I am carrying a whipped cream pie. I carry these pies with me 24/7, one in each hand, which prevents me from tweeting. I shall carry this burden with me until I find the inevitable person who is wearing both Google Glass and an AppleWatch at the same time, at which time I shall throw these pies at said person for being such a pretentious twit. And then, having completed what I was sent here on earth to accomplish, I shall at long last be satisfied with my life, and I shall immediately thereafter ascend into the heavens.

The End.

(26178) 1996 GV2

(26178) 1996 GV2 is an asteroid in the asteroid belt between Mars and Jupiter. It is small, so small that scientists can’t tell how small it is, only how brightly it reflects sunlight.

(26178) 1996 GV2 is small, and also faint.But despite its faintness, (26178) 1996 GV2 has been observed 800 times since it was discovered in 1996.That’s enough observations to calculate that (26178) 1996 GV2 orbits the sun every 4.4 earth years.(26178) 1996 GV2 does not make for a particularly illuminating blog entry.Other blog entries get interesting things when trick-or-treating from old Random Wikipedia.Not this blog entry. This blog entry says, “I got a rock.”Nobody cares about rocks.Unless they slam into the earth and cause mass extinctions or something.Then the rock becomes famous.So this blog entry is to blog entries what (26178) 1996 GV2 is to celestial bodies.It will be ignored and forgotten, just sitting there like a boring rock among countless boring rocks, unless I do something to stand out, to make myself famous.And just as the more people there are, the harder it is for a person to stand out, the more rocks there are, the harder it is for a rock to stand out.Can you imagine what Wikipedia would look like by the time we master interstellar travel?By then, they’ll have catalogged every rock orbiting around every star in this quadrant of the galaxy.Damn near every Random Wikipedia article that comes up will be about one rock or another because there are so many of them.There are a lot of rocks in the galaxy, I’m sure, so there will be lots of rocks in Wikipedia.Maybe one day, though, rocks will get demoted out of Wikipedia, unless it’s a planet.Do you ever feel sad that a planet didn’t form out of the asteroid belt?I feel like we’re missing out on a potentially interesting planet, instead of a bunch of rocks.Although, if there had been a planet there instead of an asteroid belt, we probably wouldn’t have had those chase scenes in Star Wars.OK, I guess the asteroid belt is worth it after all.I wonder, if you sent a rocket ship randomly through the asteroid belt, what are the odds you would actually crash into an asteroid?A lot less than the odds 3CPO gave that’s for sure, our asteroid belt not that dense.It is more spread out.Blah blah blah blah blah blah

And that is why people who are as dumb as rocks turn into trolls when they get online. They think this is their big chance to be a star. Nope, we see you, but you don’t have any brilliant ideas of your own. You’re just a rock. Go disappear into a catalog, like (26178) 1996 GV2.


Today, Random Wikipedia wants us to study urease, which is:

an enzyme that catalyzes the hydrolysis of urea into carbon dioxide and ammonia. The reaction occurs as follows:

(NH2)2CO + H2O → CO2 + 2NH3

I don’t really want to think about how pee breaks down. But I have been thinking about chemistry a lot lately. Mostly because I’ve been auditing an online course called the Fundamentals of Neuroscience.

I’ve been hoping to learn some juicy stuff about how the brain affects human behavior, but so far, the course has consisted entirely of the low-level details about the electrochemical properties of neurons. It’s all about how differences in the ratios of sodium, potassium and chlorine ions inside and outside a cell can lead to the flowing of electric currents via various chemical channels in the cell membrane. It blows my mind that this random soup of ions could arrange itself this complex way so as to send signals around a living body to respond to stimuli. And that this random soup of ions could arrange itself to make neurons, which arrange themselves to create networks of information, which arrange themselves to create human behavior, which arranges itself to create communities and nations — it’s hard to grasp the entire scope of all this.

Of course, we don’t grasp the entire scope of all this. We may know little pieces of it, like understanding exactly how to chemically block a sodium ion channel through a cell membrane in order to block pain signals. But the people who understand that probably don’t understand how drugs that block sodium ion channels get distributed through impoverished communities and create addiction and crime and distrust of poor people among authorities and distrust of authorities among poor people and economic vicious cycles that perpetuate that distrust.

Some people may have useful theories about how the big picture fits together but don’t understand the details. Others understand the details and but don’t see the big picture. The scope is too big for one human brain to comprehend. You have to hope there’s some mechanism through which a network of human brains can gather enough pieces to figure out a functional system instead of a disfunctional one, from low-level literal chemistry to higher-level figurative chemistry.

[insert sudden awkward segue to a very different topic here]

…and speaking of figurative chemistry, it’s amazing how poorly the Oakland A’s have played since trading away Yoenis Cespedes. On the spreadsheet, losing him shouldn’t make much of a difference. But no matter how brilliant Billy Beane is, he doesn’t understand the whole system from ion channel to World Series Champion. Nobody does, or can.

Who knew that how pee breaks down has anything to do with the Oakland A’s? Very few. But pee is part of the process, even if it’s not part of your model. Maybe there is some ununderstood literal or figurative chemistry that Cespedes provided which is affecting the A’s play.

I know that for me, as a fan, the trade has pretty much ended up ruining the season for me, and it would have even if the A’s were playing well. Because Cespedes, for me, provided the identity of the team. He was Us. Cespedes’ value as an entertainer was unmatched on the team. Every time he came to bat, and every time a ball was hit to him, my attention perked up, just in anticipation of what he might do. Even in a dreary game, he provided a reason to keep watching.

That reason is gone now. Jon Lester is a great pitcher, but he’s utterly joyless on the mound. And knowing that he’s a two-month mercenary, gone after the season is over, makes it difficult to create any emotional attachment to him. He’s not My Guy. Cespedes was My Guy.

As a result of trading the heart and soul of the team, I find that I’ve become detached. To me, the 2014 A’s season has now become all about The Destination instead of The Journey. So I find myself not caring whether I watch a game or not, because all that matters is the result. If they win, the trade was worth it. If they don’t, it wasn’t. So wake me up when the playoffs start, if the A’s even get there. In the meantime, I’ll hiding off here to the side, blocking all my sodium ion channels, numbing myself to all the inevitable pain that is soon to come.

Committee on Trade, Customs, and Immigration Matters


Random Wikipedia sends us today to the Committee on Trade, Customs, and Immigration Matters, which is a subdivision of the Pan-African Parliament. The Pan-African Parliament was established in 2004, and is similar in scope and goals to the European Parliament, aiming for central banking, unified currencies and free-trade zones. Obviously, to establish free-trade zones, you need rules and regulations regarding trade, customs and immigration between countries. Hence, this committee, probably tasked to create an African version of the Schengen Agreement.

Back in 1988-89 when I worked as a translator at the Nigerian Embassy in Stockholm (shown above, with me in the open window), I would not have envisioned that Africa would have come this far in 25 years. But they’re about at the same place the European Union was back then. In 1989, it wasn’t called the EU yet; it was the European Community. There were economic subgroups like the EEC and EFTA, but no common currency. The Berlin Wall had not yet fallen, and as a consequence, Sweden and Finland were not yet willing to join such an alliance. The pieces were there, but it had not yet all come together.

Of course, there are some unstable countries in Africa, especially in North Africa after the Arab Spring revolutions. But Europe in 1989 similarly unstable when the Berlin Wall fell. It would have been really interesting to still be working in the Embassy to experience the Nigerian reaction to the Berlin Wall falling, but I left that job in June of 1989, and the Berlin Wall fell in November. My successor as translator worked there in interesting times, to be sure.


Wow, look at how serious those young professional translators looked back in 1989!

“Please! Spare me your egotistical musings on your pivotal role in history. Nothing you do here will cause the Federation to collapse or galaxies to explode. To be blunt, you’re not that important.”
–Q, to Jean-Luc Picard, in the Star Trek TNG episode, “Tapestry”

You know, sometimes I feel like I’m living the life of the version of Jean-Luc Picard who didn’t get stabbed in the heart by a Nausicaan in that episode of Star Trek– the one who didn’t become a famous captain, the one who lived life too cautiously, who didn’t take risks, who drifted in life with no particular plan, and who as a result ended up with a decent, but forgettable and unremarkable career. But then I think, wow, I worked in European diplomacy as Communism was falling, and I worked in Silicon Valley as the Internet was starting, I got involved in blogging as social media became a thing, I covered the A’s as Moneyball introduced the world to statistical analysis. I’ve witnessed a lot of history unfolding, even if I never was the one who captained any ships to glory. All those events probably would have rolled on more or less the same without my being there. We can’t all be a Jean-Luc Picard (primary version). It is the nature of hierarchies that most of us, at best, are lucky just to be a Jean-Luc Picard (alternate version). I’ve been lucky.

Book Peddler

Today’s trip down Random Wikipedia lane introduces us to book peddlers, who were ‘travelling vendors (“peddlers”) of books.’ I’m not sure why book peddlers warrant a wikipedia entry, when other door-to-door salespeople like broom peddlers or brush peddlers don’t.

I’ve never peddled books much, but I’ve peddled blogs plenty of times. I’m a lousy peddler, though. I admire a good peddler, but it’s not for me. There is a good reason why in my career I have ended up in engineering departments instead of sales departments. I just want to focus on making good products and leave the sales work to teammates who are much better at it.

I am beginning a Twitter exile, partly to devote some of the time that Twitter sucked away to my family, but also to take some of that time to get back to doing some blogging. I don’t seem to be able to both blog and tweet at the same time. For me, it’s either one or the other.

So in exiling myself from Twitter to return to blogging, the arises whether to peddle my blog entries over on Twitter, despite my absence there otherwise. The question boils down to this: why do I write at all? Is it for the social rewards of praise from others? Or is it for the reward of a job well done?

Twitter in its purest form provides the former, blogging in its purest form provides the latter. While I have, on occasion, created a well-crafted tweet, it is more a source of quick, easy, (though ephemeral) social rewards than a place where to get the satisfaction of a job well done. And while I have, on occasion, written a blog entry that provided the social rewards of being widely praised, most of the time, even the blog entries I gained deep satisfaction from writing have largely gone unnoticed and/or unfeedbacked.

And so an experiment: I’m going to quit peddling what I write, and I’m going to remove all analytics from my web site, and all comments, so unless someone takes the trouble to email me, I will have no idea whether anyone reads my stuff or not. Any peddling will come purely from the kindness of strangers, not from me. Is writing well its own reward? I guess I’ll soon find out.

Xenotilapia leptura

Xenotilapia leptura is a species of fish that lives in Lake Tanganyika in Africa. It is currently in no danger of extinction.

But this will change.

Lake Tanganyika is the second-largest freshwater lake in the world by volume. It is located in the East African Rift, which is being formed by the African tectonic plate splitting in two and drifting apart. Sometime in the next 10 million years, the split will become large enough that a new ocean will form between the two new plates.

What will become of Lake Tanganyika when this new ocean forms in the East African Rift? Will it be incorporated into the new ocean? Will the change be gradual, or catastrophic? Will the salt water from the world oceans suddenly rush into the lake? Or will the saltiness increase very, very gradually?

These are important questions for the future of Xenotilapia leptura. You cannot just plop it into a saltwater ocean and expect it to survive. It needs the saltiness to increase gradually, so that the species has time to evolve with the change.

* * *

It is amazing to think that we can see 10 million years into the future of some other species, but barely see 10 days into the future of our own. Every day a new startup company is born, or within an existing business a new project is launched, with a mission to invent some new technology that will change the world.

Ten days from now, Apple will announce something.


What rough automaton, its hour come round at last, slouches toward Bethlehem to be born?

Will we, as a species, have time to evolve with the change? Or will the change overwhelm us?

* * *

I have been on Twitter now for seven years. There have been murmurs now that Twitter has changed, for the worse, and people are dropping out. Frank Chimero:

Here’s the frustration: if you’ve been on Twitter a while, it’s changed out from under you. Christopher Alexander made a great diagram, a spectrum of privacy: street to sidewalk to porch to living room to bedroom. I think for many of us Twitter started as the porch—our space, our friends, with the occasional neighborhood passer-by. As the service grew and we gained followers, we slid across the spectrum of privacy into the street.

But perhaps Twitter hasn’t changed. It’s just a dumb new ocean, flooding in. We’re the ones who haven’t changed, who haven’t evolved fast enough to survive the new saltwater.

I’ve dropped out of Twitter three times this year. I don’t want to blame Twitter for it. I just have trouble adapting to the changing environment. Twitter for me has become like Fox News is for many senior citizens — it’s entertaining and informative, but it also leaves me bitter and angry and frustrated at the world, 24×7.

True, there are some things worth being angry about. But I can be angry about those things without Twitter. It’s the things not worth being upset about that’s the problem.

I’m just not very good at dipping my feet into that ocean in moderation. It pulls me down deep, every time. And as a result, I become a lousy husband and father in the real world, and my productivity plummets.

When I’ve taken breaks, the anger and bitterness leaves, and everything in my life gets better. I’m happier, the people around me are happier, and I get a hell of a lot more useful stuff done.

I’m taking a long, looooooooong break from Twitter this time. I’m not planning to come back until (a) the bitterness is gone again, and (b) I have a real plan for using Twitter in moderation. Until then, anytime I feel like expressing anything, I’ll do it here, on this blog.

* * *

So long, and thanks for all the fish.

The Yoenis Cespedes Trade

The Oakland A’s made a huge trade yesterday, sending their biggest name, Yoenis Cespedes, and a draft pick to the Boston Red Sox for Jon Lester and Jonny Gomes. They also made a smaller trade, sending Tommy Milone to the Minnesota Twins in exchange for Sam Fuld. Of course, the sports world was abuzz from the Cespedes trade, which stunned many.

A couple of things left me unsatisfied about the reactions I’ve seen of the Cespedes trade. One is an old idea, expressed in Moneyball back in 2002: you don’t try to replace Giambi/Cespedes with one player, you replace him with other players in aggregate across the roster. The other a newer idea: is that the A’s platoon so much, that you can’t just analyze A’s players as atomic units. You can’t just say X is a 5 WAR player and Y is a 2 WAR player, and X – Y = 3 WAR. You have to break them down into their platoon split components, because the A’s use platoons far more efficiently than is baked into most of these formulas.

For example, if you look at Jonny Gomes as an atomic unit, he has suffered a severe decline this year. He’s hitting .234/.329/.354 this year, a far cry from the .262/.377/.491 he hit with the A’s in 2012, and in no way close to being able to replace Cespedes’ production. However, if you break Gomes down into platoon splits, you can see that his decline is entirely against right-handed pitching, where he is hitting a godawful .151/.236/.258 this year. Against left-handed pitching, however, he is still hitting a very healthy .302/.400/.431. A’s manager Bob Melvin is a master at getting the platoon advantage for his players, so we can bet we won’t see much of Jonny Gomes against RHPs.

So what I want to see is an analysis that really looks at the A’s as two teams: one team against RHPs which plays 72% of the time, and another team against LHPs which plays 28% of the time. Let’s look at those teams before and after the trade, and see how much the trades affected those two teams, even if we calculate these things in a kind of quick and dirty fashion.

To do that, you need to project performance by splits, which isn’t easy to find. PECOTA has a Marcel-like calculation called “Platoon multi”. Dan Szymborski pointed me to a platoon projection spreadsheet he created for his ZiPS projection. So I took that pre-season projected data, and combined it with their 2014 performance in a spreadsheet, to create a rest-of-season projection. (Okay, that wasn’t so quick, so the rest of this will be kind of dirty. We don’t have to be precise here, we just want a ballpark understanding of what’s going on.)

There’s another complicating factor here, in that the A’s currently have three players who are injured: Coco Crisp, Craig Gentry, and Kyle Blanks. Plus, Stephen Vogt has an injury that prevents him from catching, but not playing 1B or OF. So we’re going to run one set of numbers assuming everyone is healthy, and another assuming these injuries. Here are the best-hitting lineups (not by batting order, but sorted by GPA, from best player to worst). We’ll make removed (traded or optioned) players red, and added players blue.

Healthy lineup vs LHP: (position,obp,slg)

Donaldson (3b, .373, .604)
Norris (c, .399, .519)
Gomes (dh, .380, .440)
Cespedes (lf, .332, .473)
Crisp (cf, .353, .411)
Moss (rf/dh, .326, .439)
Blanks (1b, .336, .407)
Gentry (lf/rf, .348, .361)
Lowrie (ss, .320, .395)
Callaspo (2b, .304, .324)

Bench: Fuld, Vogt, Burns, Reddick, Punto, Jaso, Sogard.

Estimated runs per game, new lineup: 5.266
Estimated runs per game, old lineup: 5.218

The offense improves vs LHPs, because Gomes is actually slightly more productive than Cespedes, thanks to his high OBP. The defensive effect is that Moss gets moved from DH into the outfield, because he’s a better fielder than Jonny Gomes, but not a better fielder than Cespedes.

Healthy lineup vs RHP:

Jaso (dh, .372, .452)
Moss (lf/1b, .333, .510)
Reddick (rf, .325, .458)
Vogt (1b/c, .328, .422)
Cespedes (lf, .302, .453)
Crisp (cf, .321, .417)
Lowrie (ss, .329, .395)
Donaldson (3b, .321, .404)
Callaspo (2b, .333, .351)
Norris (c, .331, .353)

Bench: Blanks, Gentry, Fuld, Sogard, Punto, Gomes, Burns.

Estimated runs per game, new lineup: 4.810
Estimated runs per game, old lineup: 4.841

Losing Cespedes against RHPs has a more noticeable effect. Gomes and Cespedes are equivalent players vs LHPs, but the gap between Cespedes and his replacement against RHPs, Derek Norris, is larger, and creates a slight loss of runs per game. It also shifts Vogt and Moss around defensively to get Norris into the lineup.

Injured lineup vs LHP: (position,obp,slg)

Donaldson (3b, .373, .604)
Norris (c, .399, .519)
Gomes (dh, .380, .440)
Cespedes (lf, .332, .473)
Moss (lf/dh, .326, .439)
Fuld (cf, .337, .378)
Lowrie (ss, .320, .395)
Vogt (1b, .275, .448)
Callaspo (2b, .304, .324)
Burns (cf, .318, .292)
Reddick (rf, .245, .411)

Bench: Punto, Jaso, Sogard.
Out: Crisp, Blanks, Gentry.

Estimated runs per game, new lineup: 5.023
Estimated runs per game, old lineup: 4.852

Yeesh, those are some atrocious OBPs at the bottom of the lineup with these injuries, because LH batters Vogt and Reddick are forced into the lineup against LHPs. Fuld is also a LH batter, but he has a weird reverse platoon split in his career; he’s actually been better vs LHPs than RHPs. Like with the healthy group, going from Cespedes to Gomes is a slight upgrade against LHPs; but the upgrade from Burns to Fuld is enormous.

Injured lineup vs RHP:

Jaso (dh, .372, .452)
Moss (lf/rf, .333, .510)
Reddick (rf/cf, .325, .458)
Vogt (1b, .328, .422)
Cespedes (lf, .302, .453)
Lowrie (ss, .329, .395)
Donaldson (3b, .321, .404)
Callaspo (2b, .333, .351)
Norris (c, .331, .353)
Fuld (cf, .311, .321)

Bench: Sogard, Punto, Gomes, Burns.
Out: Crisp, Blanks, Gentry.

Estimated runs per game, new lineup: 4.685
Estimated runs per game, old lineup: 4.708

The main effect here is that Fuld gets Cespedes’ at bats, and that Reddick can move back to right field. But without the Fuld trade to complement the Cespedes trade, Sogard would be getting Cespedes’ at bats, and you’d have an awful outfield of Moss-Reddick-Vogt with Callaspo at 1b. Yeesh. You’re going to lose some offense, but that defensive alignment would probably kill you. I suspect that avoiding that defensive alignment alone is probably justification for trading Milone.

So let’s take those estimated runs per game, and extrapolate them over 162 games, and assume the average split of 72% RHPs and 28% LHPs, and combine those two split-handed teams into one team again, leaving us with just a healthy team and an injured team.

Of course, the injured team is not as good as the healthy team, and will be scoring fewer runs than the healthy team. But to analyze the trades, we don’t need to know the raw totals, we really only need to know how much the trades change the run scoring.

The healthy team loses 3.6 runs vs RHPs in the trades, but gains 2.2 runs vs LHPs, for a total loss of 1.4 runs over a whole season. It’s practically no loss of offense at all.

The injured team loses 2.7 runs vs RHPs in the trades, but gains 7.8 runs vs LHPs, for a total gain of 5.1 runs over a whole season. Most of that gain is from playing Fuld over Burns (vs RHPs) and Reddick/Vogt (vs LHPs).

Let’s say these three injured players are going to miss one-third of the remaining games to play. Multiply that 5.1 by one-third, and the -1.4 by two-thirds, and what you end up with is actually a slight gain (0.25 runs over the rest of the season), albeit so small that it is practically a wash.

The trades felt like a shock to many of us. On the surface, losing Cespedes’s sexy bat hurts, and trading a decent starting pitcher like Tommy Milone for a fourth outfielder seems like a waste. In a vacuum, that is true. But if you look at the impact those trades have on this particular team’s offense, it’s negligable.

Offensively, the numbers tell us that losing Cespedes is no big deal. And if everyone is healthy, trading for Fuld is a waste, because he wouldn’t play. But not everyone is healthy, especially in CF, and so Fuld is essential to keeping the offense at the level it would be without the trades.

So basically, we can consider the offense a wash. Now we can move on to analyzing the effect these trades have on the A’s defense and pitching. But I’m leaving that as an exercise for the reader. I’ve done enough for today.

Interview with My Mom about Life During World War II in Sweden

Last time I was in Sweden in 2012, I interviewed my mom about her experiences living in Sweden during World War II. The 17-minute interview is embedded here:

In it, she talks about:

  • How her grandparents escaped from Norway during the war and came to Sweden
  • How they had to deal with shortages of food
  • How she tried to smuggle some black market pork on a train
  • What it was like to visit Norway after the war was over.

Fixing the Oakland Coliseum Fences (and Foul Territory)

Grant Brisbee has a fun series over on SB Nation where he ranks MLB stadiums by how well they make home runs look impressive. Surprisingly, he ranks the Oakland Coliseum 13th. It gets that high ranking because the various levels of Mount Davis provide a good contrast between a mediocre home run, and a towering one. When someone crushes one at the Coliseum, you can tell it’s crushed because it lands in the 2nd deck (down the line) or hits off the luxury boxes in center field.

That’s fine and all. I suppose it’s good that Mount Davis has some redeeming feature. But there are far more mediocre home runs than monster ones, and it’s what the current version of the Coliseum does to those wimpy home runs that I hate.

Hate hate HATE.

Really, there is nothing I hate more about the Coliseum than the placement of the outfield walls. Nothing. Not the troughs, not the sewage, not the crap we A’s fans have to take from other fans teams about the troughs and the sewage, not the 8th-inning Call Me Maybe, not even Mount Davis itself. I hate the placement of the outfield walls more than all of those things.

Except at the foul poles, there is no logic to the outfield walls at all. None. Look at the fence at any point between the foul poles. Why is the fence there? Why is it that height? No reason at all, really.

And worse than that, what really drives me bonkers about it is this: any EVERY point from pole to pole, if you hit the ball just barely over the fence, it DOES NOT LAND IN A SEAT.

Home runs should land in seats. Or if not IN seats, then OVER seats. Period.

* * *

Ok, Ken, you’ve been made Dictator of the Oakland Athletics for a day, and you can change one thing and one thing only. Give us your plan.

OK, I’m going to assume the A’s will sign a rumored 5-10 year lease extension, and are therefore planning to stay at the Coliseum awhile. This may be putting lipstick on a pig, but nonetheless, let’s make it a better place to watch a ballgame.

First of all, do you know why there is so much foul territory in Oakland? The story goes, as former A’s broadcaster Monte Moore use to tell, that the third deck had obstructed views of home plate because of its slope, so they had to move home plate further out than they planned.

I don’t know if that’s true or not, but let’s say that it is. Well, guess what? We’re not using the 3rd deck anymore. It’s (mostly) tarped off. So why is home plate still pushed out so far?

We’re going to put home plate back and the foul poles back to where they originally were supposed to be. Then we’re going to use the extra eight feet or so we gain to add some seats in front of the current bleacher seats. What we end up with is (a) an outfield configuration where, except for at the stairs, every home run lands in or over a seat, and (b) every seat in the main seating bowl is suddenly about two rows closer to the action, in a way that (c) shouldn’t cost ridiculous amounts of money to implement.

Here’s what it looks like with the new configuration in left field, and the old configuration in right field (click image for larger version):


Let’s look at this in more detail:


1. We’re moving the foul poles over about 6-7 feet, so that there’s only about 1 foot between the pole and the foul line seats. This pushes home plate back about eight feet or so, thusly:



2. The wall nearest to the foul poles is about 2-3 feet shorter than the seats, and begins to angle away from those seats as you move more towards center field. We’re fixing this. The walls go all the way up to the seats, and hug the seating section all the way. No more balls that land over this fence, but fall short of the seats. Compare the new and old corners:



3. We’ll get rid of that stupid idiotic ledge above the out-of-town scoreboard. With home plate being pushed about 8 feet back, we have room to add two or three extra rows of seats, and still keep roughly the same distance from home plate as before.

I don’t know if we keep a scoreboard there or not. If you give free wifi throughout the stadium instead, you probably don’t need it.

I cut and pasted Fenway’s Green Monster seats here, to show you don’t need to add seats identical to the other bleacher seats. There’s room for some creativity in this new section.



4. Centerfield is now about 405 feet from home instead of 400, but we’ve cut down on the foul territory quite a bit, so this may keep the amount of offense roughly the same as before.


* * *

Ahhhhhhhh, now see? That’s much better.

I’m sure you have all loved your Dictator for the Day, and Wish Long Life for your Beloved Comrade Who Brings Glory to the Homeland. Now please excuse me, I have some propaganda posters to go photoshop.

Projected 2014 Oakland Athletics Anagram Roster

There’s no way to be gentle about this: A’s General Manager Baby Nellie’s offseason moves have clearly weakened the A’s anagram roster for 2014. They have become slightly worse across the board, but some of his moves in the bullpen…well, I just don’t know what he was thinking.

Starting Rotation:

The A’s have lost the two best anagrams from their 2013 starting rotation: Bartender Snot and No Local Robot. Angry Nosy and Rat Mocks Zit are decent replacements to be sure, but are also both clearly a step down. Fin Jar GIF looks like odd man out, as acronyms are purely replacement-level stuff, even if they can be pronounced.

11: Pro Radar Jerk
54: Angry Nosy
57: I Melt My Moon
64: Fin Jar GIF
67: Daily Rants
??: Rat Mocks Zit


Ask the Pen: is there any better anagram for a reliever? No, there is not. And yet, the A’s just let him go for nothing. To ask the pen without him to match what they were with him is unfair.

It gets worse before it gets better. Swapping closer No Fat Burglar with Oh MJ Is On NJ is nothing but a disaster.

Trading away JV Errs Byline is addition by subtraction, but similarly wretched She Aces JV EZ is somehow still around.

On the bright side, there remains a solid young core led by Oldest Toenail. Greek Loungers may be the best A’s acquisition this offseason, and don’t overlook Banana Fodder.

With no options remaining, there may be no room for Fedora Groupie, so perhaps Baby Nellie can find a match for him with the Astros.

48: Okay Corn
60: She Aces JV EZ
61: Neat Odor
62: Oldest Toenail
65: Fedora Groupie
??: Oh MJ Is On NJ
??: Greek Loungers
??: Banana Fodder


The roster of catchers remains the same. Order Sinker is the best gamecaller of the group, of course. Pegs Hot Vent remains to fill in should either of the other two catchers need to go on midseason pilgrimages again.

5: Hajj Soon
21: Pegs Hot Vent
36: Order Sinker


Armload Seas gnip-gnopped his way to Texas last summer, so the A’s have replaced him with Tonic Punk. It’s a slight upgrade, to a mostly intact infield where even the weakest link redeems himself with a Star Wars reference.

7: Mean Fainter
8: Roid Jewel
10: Rat Brain Doc
18: Palatable Colors
20: DJ Han Solo Nods
28: Scarier Dog
37: Random Snobs
??: Tonic Punk


Grouchy Sin is out, Great Crying is in. You reap what you sow, I guess. Don’t forget that Random Snobs can play outfield if needed, which may leave no room to Erotically Ham.

4: Cisco Crop
16: Jocks Did Her
23: Erotically Ham
52: Eyes Second Pies
??: Great Crying

This is Ken Arneson's blog about baseball, brains, art, science, technology, philosophy, poetry, politics and whatever else Ken Arneson feels like writing about