Tuesday, May 29, 2012

MLB.com Article on Hitting Streaks

MLB.com, the official website of Major League Baseball, has a new article on hitting streaks, with a focus on the recent offensive outburst by Josh Hamilton of the Texas Rangers. The article quotes a variety of sources, including yours truly.

Saturday, May 26, 2012

College Softball's Turang on Torrid Hitting Stretch

The NCAA women's softball super-regionals are taking place this weekend, with teams trying to advance to next week's Women's College World Series. One of the hottest hitters in the postseason, if not the hottest, is Oklahoma junior outfielder Brianna Turang, whose team has already qualified for the World Series. As shown in the following chart, Turang has hit .750 (9-for-12) in five postseason games (three in the initial regional round and two against Arizona in the super-regionals).You may click on the graphic to enlarge it.


The left-hand part of the display tracks Turang's batting average during the regular season. It first lists her batting statistics at the end of preconference play (i.e., as of the close of play on March 16) and then updates the statistics on a roughly weekly basis, coinciding largely with Big 12 conference play (even after conference play begins, however, there are still some nonconference games sprinkled through the schedule and these games are included in the stats). The data come from OU's softball news releases throughout the season.

As can be seen, Turang's batting average varied within a fairly narrow range -- either a little bit above or below .330 -- over the final two months of the regular season, ending up at .328. The right-hand portion of the display shows her game-by-game postseason statistics, documenting her cumulative 9-for-12 performance.

The relevant statistical question is the following: What is the probability of someone with a .328 regular-season batting average (i.e., long-term success rate) getting a hit in 9 or more of her 12 postseason at-bats? The answer, according to this online binomial-probability calculator, is .0034 or roughly 3-in-1,000.

Raising her game in the postseason is nothing new for Turang, however. Her regular-season batting average a year ago was .318. According to OU's news release for this year's super-regionals: “In 2011, Turang led all Sooners in postseason play with an average of .533 off a team-leading 16 hits…”

Monday, May 21, 2012

Spurs' on 18-Game Winning Streak (Combining Regular Season and Playoffs)

With their win last night to sweep the L.A. Clippers 4-0 in a second-round NBA playoff series, the San Antonio Spurs now have a combined regular-season/playoff 18-game winning streak. The Spurs won their final 10 games of the regular season and then swept Utah in four in the first round of the playoffs.

San Antonio's hot streak goes back even further. Sitting at 26-13 after play on March 9, the Spurs closed the regular season on a 24-3 run to finish the abbreviated 66-game NBA season with a 50-16 record. Adding in the eight playoff games, San Antonio is now 32-3 in its last 35 games.

The Spurs' roster has an aging core, led by 36-year-old Tim Duncan and 34-year-old Manu Ginobili. Perhaps the shortening of the NBA season from the usual 82 games to 66, due to the owners' lockout, has helped conserve the energy of the Spurs' players. A year ago, the silver and black appeared to hit the wall after compiling a 57-13 record through 70 games. The Spurs went 4-8 in the remainder of the regular season (including a six-game losing streak) to finish at 61-21, and then exited in the first round of the playoffs at the hands of the Memphis Grizzlies.

Tuesday, May 15, 2012

Anecdotes Aside, Trouncing Opponent in Game 1 of NBA Playoff Series Usually Presages a Series Win

Watching the closing minutes of last night's 119-90 Oklahoma City Thunder thumping of the L.A. Lakers in the teams' opening game of their NBA playoff series, I was reminded of two other Game-1 blowouts in NBA playoff history.

In what is known as the "Mother's Day Massacre" of 1982, the Boston Celtics routed the Philadelphia 76ers 121-81 to open that year's Eastern Conference finals. Yet, Philly took the series in seven games.

Three years later, the Celtics opened up the 1985 NBA Finals with a 148-114 thrashing of the Lakers, the "Memorial Day Massacre." Again, though, it was the team on the losing end of the massacre that ultimately prevailed, as L.A. hoisted the championship trophy.

Judging by these two precedents, therefore, this year's Laker squad didn't seem to be in such bad shape, after all. Rather than go purely by anecdote, however, I decided to investigate all opening games of NBA playoff series since 1982 (i.e., roughly the past 30 years) in which a team won by 25 or more points, to see how often that team ultimately lost the series. (I only examined 4-out-of-7 series; the NBA used to have 2-of-3 and 3-of-5 opening rounds.) The Wikipedia's year-by-year NBA playoff summaries supplied the necessary data. Here is what I found.


Well, what do you know? From 1982 forward, there have been 27 playoff opening-games in which one team triumphed by 25 or more points (excluding this year's Thunder-Lakers game, as the series is incomplete). And in 25 of those 27 series (93%), the team that won easily in Game 1 went on to win the series. In other words, the two counter-examples I initially came up with (the 1982 Sixers-Celtics and 1985 Lakers-Celtics series) were the only two exceptions to the larger trend.

Apparently, the outlook for this year's Laker squad is a lot worse than I had thought!

In thinking that a team that gets blown out in Game 1 has a good chance to win the series, I appear to have succumbed to what is known as the "availability heuristic." Because the 1982 Sixers' and 1985 Lakers' comebacks against the Celtics were so dramatic, they stuck in my mind (i.e., remained available) and evidently led me to overestimate the frequency of overcoming an opening-game disaster.

The late Amos Tversky, a co-developer of hot hand research, also launched the research on availability with his longtime collaborator Daniel Kahneman.

Thursday, May 10, 2012

Josh Hamilton's Four-Homer Game

With Josh Hamilton of the Texas Rangers on Tuesday night joining the club of sluggers who have belted four home runs in a single game, I decided to examine Hamilton's and all previous four-homer games in detail to see what we might learn about the continuity and timing of such outbursts. For example, when a player has homered four times in a game, have the four blasts tended to come in four consecutive at-bats? Against the same pitcher each time? If not the same pitcher, has the throwing arm at least been consistent across pitchers faced? You get the idea.

Hamilton, of course, is best known for overcoming drug-abuse problems to bring his beautiful bat swing -- reminiscent, in some observers' eyes, of the main character in "The Natural"  -- to the major leagues. In the 2008 Home Run Derby during the All-Star Break, Hamilton socked 28 homers in a single round. Thus, as rare a feat as hitting four home runs in a game is, one had the feeling that if anyone would do so in the near future, Hamilton could well be the person.

My starting point was this list, from Wikipedia, of the 16 major-league players to hit four homers in a game. For each instance, I then consulted Retrosheet, an online archive of MLB box scores and play-by-play sheets. Box scores routinely list all home runs hit in a game by hitter, pitcher, and inning. The accompanying play-by-play sheets supplement the box scores by listing in narrative fashion what happened on all at-bats in the game, thus documenting what the hitter with four home runs did in all his other at-bats (if any) during the game.

For the 11 batters who had four-homer games from 1950 to the present, the chart below thus documents what they did in each at-bat of their big game. For three batters whose four-homer game occurred in the 1930s or '40s, Retrosheet had box scores only, so the non-homer at-bats could not be charted. Finally, for Bobby Lowe (1894) and Ed Delahunty (1896), neither box scores nor play-by-play sheets were available, so they do not appear in the following tables.

Here are the charts. You may click on them to enlarge them.


Here are some findings from the charts:
  • Among the 11 players for whom the full scope of data was available, four (Carlos Delgado, Mike Cameron, Mike Schmidt, and Rocky Colavito) homered in four consecutive at-bats, with no other kind of outcome intervening.
  • Of the same 11 players (excluding Delgado, who had only four plate appearances), three (Hamilton, Shawn Green, and Joe Adcock) had additional hits (singles, doubles) beyond their home runs, without ever making an out.
  • Among the documented outs (pink squares), five were via fly balls or pop-ups, only one was a ground-out, and none were strike-outs. Thus, these hitters were getting the ball in the air with great regularity.
  • Hamilton (3 of 4), Delgado (3 of 4), Green (3 of 4), Cameron (3 of 4), Mark Whiten (all 4), Chuck Klein (all 4), and Lou Gehrig (all 4) hit their homers predominantly or exclusively against pitchers who threw with the opposite arm relative to the handedness of the batters' swings. To really know if there's any significance to this statistic, we would need to know the percentages of opposite-hand match-ups in all at-bats and all home-run at-bats in MLB generally.
  • Among the 14 players whose home-run innings can be documented, all except Schmidt, Colavito, and Pat Seerey got off to a quick start by homering in the first or second inning.

Wednesday, May 02, 2012

L.A. Kings Dominating Their NHL Playoff Series

The big surprise of the National Hockey League playoffs thus far is the largely dominant play of the Los Angeles Kings. The Kings qualified for the playoffs as the eighth (and final) seed in the Western Conference, but easily dispatched No. 1 seed Vancouver in the first round, four games to one, and are currently leading No. 2 seed St. Louis two games to none in the second round. Other than during a possible letdown after winning the first three games against the Canucks, the Kings have hardly been challenged.

I created the chart below to illustrate how many minutes, cumulatively, the Kings (in purple) and their opponents (in light blue) have held the lead in the two series (you may click on the graphic to enlarge it). Links to box scores of all the Kings' games are available here. To walk through an example, in Game 1 vs. Vancouver, the Canucks scored first 4 minutes and 17 seconds into the contest, thus coloring the chart blue. The Kings then scored 13:31 into the game, tying things up and returning the tracker to white. L.A. scored the next goal, 16:33 into the second period, bringing purple coloration. And so forth.


As circled in red, Vancouver led for only 9 minutes and 14 seconds total in the first three games (180 minutes) of the Kings-Canucks series. Now, two games into the Kings-Blues series -- both played at the home of the Gateway Arch   -- St. Louis has led during only 7:42 of the 120 minutes played.

The Kings and Blues resume their series tomorrow night in Los Angeles. Fortunes can change in a hurry in hockey, especially with the seemingly random bounces of the puck. For now, though, the Kings seem to have the right formula.