Monday, July 13, 2015

A Different Way to Poll on Baseball

Shortly after I wrote last week's post—in which I used polls to calculate how popular each MLB team is—the annual Harris Poll on baseball was released—in which Harris polled the nation on how popular each MLB team is. Although our methodologies are different, Harris provides a great way to double-check my project.

According to the Harris Poll:
"...the New York Yankees continue the more-than-decade-long winning streak they’ve been on since 2003, coming in once again as 'America’s Favorite.' Another repeat – in this case one we’ve been seeing since 2009 – is longtime Yankees rival the Boston Red Sox coming in at no. 2 once again. Moving up one spot to no. 3 are the Chicago Cubs."
That's in pretty good agreement with my calculations, with one notable exception. My method (aggregating Public Policy Polling surveys of baseball fandom in 35 states) found that the Atlanta Braves were comfortably the most popular team in the US, with 22,573,607 fans, followed by the Red Sox at 17,749,160, Cubs at 17,504,648, and Yankees at 14,793,886. However, my analysis suffers greatly from the fact that New York is not one of the 35 states we have data for, so Harris is very probably right that the Yankees have more fans than the Red Sox and Cubs.

But what about the Braves? That's the truly glaring discrepancy between our counts. In the Harris Poll, the Braves are all the way down in sixth place—although until last year they had never performed worse than third in the annual poll. I'll blame both Harris and myself for this disagreement. I think Harris probably ought to smooth their data a bit more over years, since historically the Braves are clearly closer to America's favorite team than they are to fifth runner-up. But I may also be giving too much credit to the huge bloc of potential fans that is the American South. In the South, less of the population are baseball fans than in other regions; we know this anecdotally (merely whisper the words "college football" and the region will throw a spontaneous pep rally) and from the Harris Poll itself, to which 25% of Southerners responded that they follow MLB, compared to 34% of the East, 36% of the Midwest, and 38% of the West. The Braves, of course, derive much of their numeric advantage in my calculations from the sizable population of this large region that they have all to themselves (as far as MLB is concerned). I could believe fourth place nationally for the Braves.

This year, Harris has the Los Angeles Dodgers and Detroit Tigers in fourth and fifth place. My fan counts agree that those teams have large fan bases, but not that large. One thing that could be holding them back in my calculations is if they have bigger national followings than they receive credit for. Because of the Yankees', Red Sox', Cubs', and Braves' reputations for having fans all across the country, PPP usually asks about them in every state it polls. However, there's a limit to the number of teams it can poll on before the poll gets unwieldy—usually eight teams per state is the limit. It's tempting to think that the Dodgers, Tigers, or other teams like the Cardinals could pick up a few million if PPP were able to ask about them nationally.

That's the advantage of the Harris Poll: it conducts a single survey, at one snapshot in time, across the whole country—to be exact, 2,200 adults nationwide (including 700 who follow MLB) surveyed online between June 17 and 22, 2015—so it avoids that problem. But I also have a few gripes with it. Principally, it falls short for me because it doesn't provide actual percentages, like most polls do, for each team; instead, it just ranks the teams from most to least popular. The original goal of my project—what drove me to do my own calculations using PPP—was not just to know teams' relative popularity, but specifically to get hard numbers for how many millions of people each team has in its corner. I've emailed Harris to see if it has specific percentages and is willing to share them, and I'll update this post if I hear back.

A second gripe is that, despite an initial sample of 2,200 adults, the Harris Poll still has far too small a sample for a national poll for which there are 30 possible answers. (Between the 35 state polls that my analysis uses, PPP surveyed 28,101 voters about their baseball preferences.) Simple math tells us that the average fandom percentage in baseball must be 3.33%—which also means that, for teams like the Yankees and Red Sox that we know exceed that number, there are also teams even smaller. Among Harris's sample of 700 baseball fans, a 1% haul would be seven respondents. That makes the numbers pretty sensitive to year-to-year variation and also likely puts a lot of less popular teams within the margin of error. This year, I would single out the Tampa Bay Rays as a likely victim of small sample size. Harris ranked them at #16 this year, up from a tie at #24 last year. That's a huge bump for no clear exogenous reason, and 16th is dubiously high for a team that has struggled so much to attract fans. Here I have more faith in my own calculations, which found that the Rays are probably the least popular team in baseball.

I don't mean to be harsh against Harris—I'm in favor of any pollster that asks about baseball! Any data added to the pool of consideration is valuable; I just mean to point out limitations in how we should interpret it. One place where I do value Harris's data, quite highly, is in the other questions it asks about baseball—and how those questions break down in the crosstabs among specific demographic groups. For instance, I was glad to see that a whopping 80% of fans approve of the new instant replay rules. That really puts a hole in the argument that instant replay takes away from the history and integrity of the game—especially when you look at the crosstabs. Instant replay is actually more popular among older demographics, with 75% support among Millennials, but 83% with Baby Boomers and 87% (!) with fans over 70 years old.

Then there's the big question for the folks over at 245 Park Avenue: what percentage of the adult population follows Major League Baseball? This year, that number dipped to 32%—the lowest Harris has ever found. Just last year, that number was 37%, and it was 41% as recently as 2009. I'm generally not a believer in the idea that baseball is dying, but the sport does face demographic challenges that the poll's crosstabs point to. Among age groups, baseball fandom was lowest among Millennials (29%); among income brackets, fandom was lowest among those making less than $35,000 a year (27%); by level of education, it's least popular with those who didn't go to college (26%). Baseball is a sport for the wealthy, and MLB would do well to lower economic barriers to both attending games and playing them—especially for young people. Because this is a political blog, I would also be remiss if I didn't mention that baseball fandom is largely party-blind: 35% of Republicans follow MLB, 34% of independents, and 32% of Democrats.

It's worth noting that Harris's wording to that question—"do you follow Major League Baseball or not?" is presented in "opt-in" format, whereas PPP asks about fandom in more of an "opt out" way. The result is that twice as many PPP respondents (78% in total) claim to be baseball fans—or at least have a favorite baseball team, which is the technical wording of the question. Simply put, we know that's too high. That's certainly a major limitation of my exact fan counts—they include millions of fair-weather or bandwagon fans who don't know a curveball from a changeup but who simply have pride in the hometown colors.

But then again, who are we to decide what it means to be a fan? Someone who follows MLB every night with an MLB TV subscription and three spreadsheets doesn't necessarily love the team any more than a guy who cherishes his annual tradition of hitting the ballpark bar with his three best friends from college. There are many different expressions of fandom, to be sure, but they're all equally valid. It's possible that Harris and PPP are just counting different things. And just like you can freely choose how you root for your favorite team, you, dear reader, are free to choose which poll's definition of fan fits your worldview.

Tuesday, July 7, 2015

The Severalest Fans in Baseball

A few weeks ago, we learned that the US Supreme Court would take up a case deciding the meaning of "one person, one vote." Thankfully, in Major League Baseball, we have no such restrictions. Fans can vote dozens of times for the All-Star Game, with the result this year of 620 million ballots cast—a record.

Around this time of year, and with the Final Vote well underway, I always wonder which teams have the biggest voting blocs behind them. (For example, are there really that many Royals fans out there, or is something funny going on?) The exact number of fans who swear allegiance to each MLB team is the holy grail for marketing executives, advertisers, statisticians, and baseball psephologists alike; hard numbers are elusive and ever-changing. However, a number of sources exist from which we can estimate fan-base population; there may not be a baseball Census, but it has its own American Community Survey equivalents.

In exploring this same question last year, I sifted through a few of these datasets, from Facebook data to the annual Harris baseball poll, before settling on state-by-state polling data from Public Policy Polling (PPP). PPP, run by my friend and huge sports fan Tom Jensen, is known for administering its political surveys with a side of quirkier questions: voters' favorite Great Lake, their concern about falling into sinkholes, and, of course, their sports allegiances. In my 2014 post, I used PPP's data to estimate how many millions of fans rooted for each team.

A year later, it's time for an update. PPP has polled more places, and it has released more up-to-date data in others. As of July 6, 2015, PPP has polled 35 states on their baseball preferences, accounting for 79.0% of the US population. I collect its findings and input them regularly into this Google spreadsheet. Multiply the percentage of respondents by each state's population, and voilĂ : raw numbers for each fan base.

Here in 2015, in the four-fifths of the country that we've snapshotted, the largest fan base belongs to the Atlanta Braves, with over 22 million fans. The smallest, at barely a million, belongs to the poor Mets, so at least there are fewer of them to be miserable. Here are the numbers for all 30 franchises:

Team Fans Team Fans
Atlanta Braves 22,573,607 Minnesota Twins 4,372,998
Boston Red Sox 17,749,160 Cleveland Indians 4,324,696
Chicago Cubs 17,504,648 Milwaukee Brewers 4,261,376
New York Yankees 14,793,886 Kansas City Royals 4,135,297
San Francisco Giants 10,990,204 Pittsburgh Pirates 4,004,612
Texas Rangers 10,094,530 Arizona Diamondbacks 3,998,691
St. Louis Cardinals 8,093,190 Oakland Athletics 3,703,872
Detroit Tigers 7,707,862 Chicago White Sox 3,486,142
Los Angeles Dodgers 7,459,833 San Diego Padres 3,299,274
Houston Astros 6,531,263 Miami Marlins 2,480,189
Los Angeles Angels 6,378,978 Tampa Bay Rays 2,417,083
Seattle Mariners 5,846,564 Washington Nationals 2,167,109
Philadelphia Phillies 5,303,911 Baltimore Orioles 1,594,533
Cincinnati Reds 5,122,065 New York Mets 1,283,038
Colorado Rockies 4,460,263 Toronto Blue Jays* 0

*PPP only polls in the United States; the Blue Jays probably do have only a handful of American fans, but obviously they don't belong at the bottom of this list given their millions of fans north of the border.

(Careful readers will notice that these numbers are dramatically larger than the estimates I published last year. That's due to a slight but crucial methodological change from last year: using total population—specifically, 2014 estimates from the US Census Bureau—as my multiplier, not the number of voters. Last year, I went strictly by the sample that PPP tested: usually registered voters, occasionally likely voters. Because our democracy is really sad, this looked at a much smaller pool of fans. This year, I decided that the share of Red Sox vs. Yankees fans in Connecticut really probably wouldn't be that different among registered voters than among the total population, so I took the liberty of assuming the poll spoke for everyone in the state. This gets us a lot closer to our goal of describing the fandom of every American.)

This methodology has its limitations—most glaringly, the 15 states (plus Canada and the District of Columbia) that are not included, simply because PPP hasn't polled on baseball there. While we've made progress from last year's analysis, adding three new states, the Braves, Cubs, White Sox, Nationals, Orioles, Phillies, Yankees, Mets, and Red Sox are still likely to be undersampled based on the missing states (shaded in red below). That leaves the Tampa Bay Rays as the likeliest smallest fan base.


Still, it's exceedingly clear that the Braves, Red Sox, Cubs, and Yankees, in some order, have the most fans in baseball. This could bode well for Brett Gardner or Xander Bogaerts in the Final Vote. Keep an eye on the excellent county-by-county vote maps that MLB puts out during Final Vote week to see where baseball's electoral clout lies.

Tuesday, June 23, 2015

A #CBG2015 Roundup

This blog has been taking a little Christmas vacation—Christmas for baseball-and-politics lovers, of course, being the June Congressional Baseball Game. As you've hopefully noticed from following my Twitter feed, I would never go dark during such a critical period—I've just been active elsewhere. For those who may have missed it, though, here's a recap of my Congressional Baseball writing out there in actual, respected publishing outlets:
  • To bore everyone to tears get everyone psyched in advance of the game, I wrote a sabermetric scouting report for the game over at FanGraphs, whose meticulous and intelligent number-crunching has inspired me, for the past few years, to keep a similar record of advanced statistics for the Congressional Baseball Game. In developments that may or may not have made me squee, the piece (and my database of stats) were picked up by the likes of the New York Times and FOX Sports.
  • Don't worry, old school readers, I poked my nose outside my spreadsheet long enough to watch the game too, which I attended and covered for The New Yorker. I talked to a Republican player you may have heard of—Rand Paul—and President Obama showed up to the game as well, giving me the opportunity to write about presidential figures throughout the history of the game.
Democrats ended up winning this year's tilt, 5–2, extending their seven-game winning streak but by a narrower margin than their superstar-studded team is used to. Despite the loss, Republicans should feel really good about their standing going into next year, as they've finally found a pitcher good enough to keep them in games (Rep. Mark Walker of North Carolina). I learned a ton writing about the game this year, as I do every year; it seems like there's an infinite amount of information, history, and fun angles surrounding the game. I hope you enjoy reading about it as much as I did writing about it.

Monday, May 25, 2015

Who Will Win the 2015 Scripps National Spelling Bee?

(Note: This piece is intended to be a lighthearted take on a fun event; those looking for serious bee analysis should look elsewhere.)

There isn't a sporting event I look forward to more every year than the Scripps National Spelling Bee. (Maybe the American Crossword Puzzle Tournament.) Lest you think I'm being sarcastic, try tuning in this year—it airs Wednesday and Thursday on various ESPN affiliates, with the finals in primetime at 8pm Thursday. It's nothing less than the purest form of human drama, as the best of us—our children—grapple with the worst: fear, crippling expectations, fainting (yes!), crushing defeats (no!), and, of course, the maddening phonetic rules of the English language. The tics, the techniques, the aha moments, the dulcet tones of Dr. Jacques Bailly—these are what help make the bee surprisingly spellbinding television.

Of course, for one* child a year, the pressure forges something special; the lights go up, their shoulders untense, and they are named national spelling bee champion. (*Except last year, when co-champions were controversially crowned for the first time in 52 years.) Two hundred and eighty five contestants will walk into the Gaylord National Convention Center this week, but so many are just there for window dressing. Close bee watchers know the prize will come down to one of the few favorites, many of whom are already well known in this universe; there's Vanya Shivashankar, whose sister won the 2009 bee; Gokul Venkatachalam, who finished third in 2014; Tejas Muthusamy, who turned heads with a phenomenal rookie performance last year; and a short list of others.

So who should you bet on this year? (Note: Please don't bet on how smart a kid is.) It's all about who has tended to win the bee in the past. Thanks to the spelling bee website's official speller roster, complete with bios and a statistics page, we can compare each participant's profile to those of past winners. Obviously, the bee is a competition for who is the best at spelling, not at who checks the most demographic boxes. But just for fun, we can paint a pretty good composite portrait of a bee winner, based on the following characteristics of the past 17 champions:
  • Race. We might as well start here, since everyone else does. I shouldn't have to tell everyone not to use racial stereotypes in predicting a bee winner, but the fact remains that Indian Americans dominate the bee. Thirteen of the past 17 champions, including the last eight in a row, have been of Indian descent; the remaining four were white. How you interpret that is up to you.
  • Gender. Eleven of the past 17 winners were male, while six were female. However, four of the last seven have been girls. (Fun with arbitrary endpoints!) Again, please be reminded that these are not causal relationships and that this whole exercise is tongue-in-cheek. For what it's worth, the list of 2015 bee finalists actually features more girls (146) than boys (139).
  • Age. As you might imagine, the older, the better. Fourteen of the last 17 champs were eighth-graders, the oldest you can be to qualify for the bee. (This is a much higher proportion than the share of eighth-grade participants, which this year is 41.5%.) The remaining three were seventh-graders, so it would be pretty shocking to see anyone in sixth grade or below take home the trophy.
  • Experience. Closely related to age is a contestant's prior experience on the national level. All but one of the past 17 winners had been a national bee finalist before, making it quite likely that we've seen 2015's champion on that stage already. The more times you've been to the national bee before, the more it helps, although second-timers do just fine, accounting for six of the 17. There are 57 returnees this year, including 17 on their third, fourth, or fifth try. (If you're looking for a shortlist of favorites, you could do a lot worse than those 17, who are all also in seventh grade or older.)
  • School type. Perhaps surprisingly, champs have been most likely to go to public school. Nine of the past 17 were enrolled in public school, six were enrolled in private school, and two were home-schooled. However, when you compare this to the overall list of contestants (at least this year's), that's actually an underrepresentation of public-school kids (who are two-thirds of the overall pool) and a dramatic oversampling of private- (23.9%) and home-schooled (4.6%) students.
  • Hometown. Finally, there is pretty good geographic diversity among the last 17 champs: three Californians, two New Yorkers, two Texans, two Hoosiers, a Floridian, a Missourian, a Kansan, a Coloradan, a Minnesotan, a New Jerseyan, a Pennsylvanian, and an Ohioan. Nothing really to go off here.
So, if you subscribe to past as prologue, you'd expect an Indian American eighth grader who has been to the bee before to emerge as the winner on Thursday night. Of course, when you watch the bee, it's worth remembering that all these kids are already winners, even if they're eventual losers at the bee: virtually guaranteed to go on to elite schools and careers, they're sure as heck smarter than all of us!

Monday, May 4, 2015

The Big, Bad, Colorful Chart That Explains State Constitutional Offices

Next election, spare a thought for the poor constitutional officers. The statewide elected officials not titled “governor” are little more than an afterthought for most election watchers, but, as I’ve argued before, they shouldn’t be: they make more policy than Congress these days, and they’re surrounded by every bit as much drama. (For an example, look no farther than the incredible fallout over Missouri Auditor Tom Schweich's suicide.) Oh yeah, and they often rise to become more than just constitutional officers.

Basically, we should keep better track of them—but, in fairness, that’s hard to do. Every state has different constitutional officers and different ways of choosing them, making the constitutional-office picture much messier than, say, the U.S. Senate. To solve that problem, I wanted a source that laid it all out visually—so I created this giant chart, also embedded at the bottom of this page. (A huge assist for this goes to Ballotpedia, where I researched all this data.)

The chart provides info on every constitutional office in the 50 states: which states have which offices (and what they’re called—a frequent local quirk); whether they are Democratic-held (blue), Republican-held (red), independent-held (yellow), vacant (gray), or nonpartisan (white); how they’re selected; and, if elected, when the next election will be. It lists the biggies—lieutenant governor, attorney general, secretary of state, and treasurer—but also the underappreciated of the underappreciated: labor commissioners, school superintendents, and mine inspectors (well, mine inspector, singular)—all the "individual dudes" in state government. (Corporation commissioners, public utility commissioners, railroad commissioners, and elected members of other statewide boards aren’t included.)

The chart illustrates some really important lessons that constitutional officers can teach us. First and foremost, it is proof positive of the Republican stranglehold on state government these days; the GOP has the edge in the partisan breakdown of every single constitutional office (viewable at the bottom of the Google doc). Democrats do OK in comptroller (5–4) and insurance commissioner (6–5) races for some reason, but they get clobbered when it comes to agriculture commissioners (11–1) and, strangely, labor commissioners (3–0–1). It’s also fascinating how all partisan superintendents and land commissioners elected in presidential years (four of 'em) are Democrats, yet all partisan superintendents and land commissioners elected in midterm years (10 of 'em) are Republicans. It really goes to show how constitutional offices can serve as weather vanes for which way the political winds are blowing.

I hope you'll play around and learn a little more about these forgotten offices, and keep your eyes peeled to this blog for more coverage of constitutional offices in this off year.

Friday, April 24, 2015

Hate the New Metal Detectors? Petition MLB

I've been through MLB's new metal detectors several times now. Once I had my wallet in my pocket and it didn't go off. Once I had nothing in my pockets and it did go off, and the wand the security staff waved over me proceeded to spaz out willy-nilly. Every time, though, it's been at least a mild delay and a nuisance.

Commissioner Rob Manfred claimed the other day that MLB hasn't heard any complaints from fans or teams about the enhanced security, which seems really hard to believe given that all we have heard are complaints. To me, the most infuriating thing about the metal detectors is how they don't actually make anyone safer. These aren't airport-quality detectors, and even in my two-week experience with them, they've failed to consistently detect anything. The kicker, though, is how bags still bypass the detector and get only a mild rummage or feel before they're cleared to enter the park. The only real argument for the detectors is that they protect MLB against a lawsuit—the most perverse of arguments, and one that only resonates with maybe 100 people who work at 245 Park Avenue.

I'm of the strong belief that it's in everyone's interests, including theirs, for MLB to remove the metal detectors and go back to the security measures that have worked perfectly well (rate of terror incidents at ballparks: 0%) for years. If you feel the same way, I hope you'll sign this petition that was started to ask MLB to do just that. The petition reads:
Whereas:

-- 2015's tighter security measures have led to major delays and fan inconvenience;
-- Experts agree that the enhanced security does nothing to actually make fans safer;
-- The new searches are overly invasive and a violation of fans' privacy;
-- The security checkpoints are major failures of customer service and turn fans off from MLB (baseball fans do not want to be treated like they are at an airport);

We ask MLB to rescind the new requirement that all fans must go through airport-style security screenings before entering a ballpark.
Please share the petition widely on your social media of choice, and thanks!

Sunday, April 12, 2015

The Arte of War (on Drugs)

Arte Moreno is displeased. After a panel rightfully declined to suspend his left fielder, Josh Hamilton, for a recent drug relapse, Moreno's Los Angeles Angels issued a statement not supporting Hamilton, but rather castigating him for his disease. It smacked of sour grapes from the loser of what was essentially a proxy war between the owners and the union—a literal inside-baseball dispute. Still, it should have been over at that point—Hamilton had won, and he was destined to return to play for the team for 2015.

Moreno wasn't willing to leave it there. Without any provocation from Hamilton or the union, on Friday he lobbed a second grenade into the fray—this time from him personally. "I will not say that" Hamilton will ever suit up for the Angels again, the owner proclaimed, asserting that the Angels held a clause in Hamilton's contract annulling the deal in the event he turned back to drugs. The union immediately returned fire: such a contract provision could not exist (and, according to NBC Sports, does not exist) and could not supersede the collectively bargained Joint Drug Agreement. In other words, no matter how much they'd like to, the Angels can't void their deal with Hamilton over his personal failings.

The conventional wisdom is that Moreno is motivated by one thing: cash. The Angels inked Hamilton to a five-year, $125 million deal before the 2013 season, and they've regretted it ever since; the injury-plagued outfielder has hit just .255 with 31 home runs and 3.0 WAR in the two years hence. If Hamilton had been suspended by the drug-enforcement panel, or (obviously) if the Angels were able to void his contract, the team wouldn't have to pay him to continue to hit like Garrett Jones for the next three years. That's a powerful motivator, but to the extent Moreno is setting the tone for the whole organization, I don't think it's what drives him. I suspect it's a matter of pure morality to the staid Angels owner.

Moreno is a private man, but one of the few things we know about him is that he is a dyed-in-the-wool conservative. Born in 1946 and coming of age during the 1960s, Moreno voluntarily enlisted in the US Army at the height of the Vietnam War. This was likely a man with nothing but disdain for the anti-war or counterculture movements; all the ingredients were there for Moreno to develop a deep hatred for drugs and those who used them.

Two years ago, Hamilton may have been a former drug addict, but he was also a poster child for religion and conservative values bringing redemption and leading to a healthy, clean life. That might have appealed to Moreno when he took the risk of committing nine figures to him. But now that Hamilton has fallen from that pedestal, Moreno may be less willing to forgive. (People are likely to react even more nastily than usual when they feel betrayed—say, denying your employee a physical space at his place of work.) It may also explain Moreno's seemingly irrational behavior of insisting he can sever ties with Hamilton even when there is pretty clearly no legal basis for doing so.

Moreno wouldn't be alone in such an anti-drug crusade. Once the young Arizonan made his fortune in the billboard industry, he became a commensurately generous donor to the Republican Party. According to Influence Explorer, over the years Moreno has given:
  • $8,200 to Rep. Matt Salmon (R-AZ), who has supported drug tests for employees—and making them ineligible for unemployment insurance if they fail and are fired;
  • $5,000 to presidential candidate Mitt Romney, who cultivated a reputation as a tough-on-drugs governor by increasing drug penalties on Bay Staters and funding schools that drug-tested their students;
  • $4,500 to former Rep. Bud Shuster (R-PA), who wrote in 1992 that the War on Drugs was working;
  • $2,300 to presidential candidate Rudy Giuliani, who as New York City mayor oversaw an exponential increase in the number of marijuana arrests;
  • $500 to former Senator John Ashcroft (R-MO), who told Larry King he wanted to "escalate the War on Drugs" when he was the freshly appointed attorney general in George W. Bush's administration.
Of the Hamilton situation, Moreno himself has said, "It's not about money." With this record, maybe we should believe him.