Numbers for players in the AZL, GCL, Pioneer, and Appalachian League are now in the database.
Added a “Fielding” table to the bottom of every batters page that contains games played and games started per position. Also the positions someone plays* are shown at the top of each player page. Nothing very exciting but these changes will allow us to add several features in the future.
Changed the minimum number of plate appearances batters faced to qualify for the hot and cold sections of the reports section to one PA for every day the report covers. Before it was one PA per game appeared in.
*Positions a player plays are figured out by having either 30% of their career games at a position or playing there five games in the current season
“Small Sample Size”
We hear these phrases thrown around every April and May as a way for fans of last place teams to justify their team’s losses and to detract from the first place team’s wins. But how often does the first 30 games represent the likely outcome at the end of a season?
Below are the standings as of May 1, 2012 in relation to how they finished the season.
Some of these differences can be explained by players getting hurt (Matt Kemp in LA or Troy Tulowitzki in Colorado) or franchise altering trades (Boston). For the most part, what we notice is that is that teams who have low expectations going into the season but start hot (CLE, TOR, LAD, NYM) aren’t all of a sudden world beaters and will come back to where they were expected to finish. Occasionally, we see a 2012 Orioles or Braves situation where a team enters the season without high expectations, plays well early on, and just keeps chugging along. In these two scenarios, we saw an “us against the world” attitude in Baltimore and a “Win one for the Chipper” drive in Atlanta. Certainly, a season-long emotional ride can take a team to the next rung up the ladder. Many teams who entered the season as favorites simply played good baseball wire-to-wire (TEX, TBR, NYY, DET, STL, WSN, CIN, SFG). Some teams, sadly, are what they are (SEA, KCR, MIN, PIT, HOU, MIA, CHC, SDP). Lastly, some teams have the courtesy to let their fanbases know really early on exactly what kind of team they are. Kudos to the White Sox, Nationals, Braves, and Phillies for playing not 1 percent different over the last 85 percent of the season compared to the first month.
I want to focus our attention towards Milwaukee and Anaheim. If your team enters the season as playoff contenders or World Series favorites, and they start the season slow due to bad babip bounces or short-term injuries, fear not. Every team goes through these banged-up babip-induced slumps; yours just happened to have it’s slump at the beginning of the season. That team you’re chasing that is playing at a 115-win pace is probably not the best team in history, just as your team probably isn’t going to lose 100 games. That ace pitcher you gave a six-year contract to in December, unless he’s hurt, is probably going to be fine. If your superstar is coming off an MVP season and is having a slow start, and he’s not hurt, don’t swear off wearing his jersey. Rather, go pick him up in your fantasy league. He’s going to be fine.
Injuries happen. They happen a lot. We’re watching highly-specialized athletes with over-developed fast twitch muscles perform at top speed every day for months on end with nary a day off, who also happen to run around with metal pegs on the bottom of their feet. Blame your team’s training staff if you want, blame the Baseball Gods, blame yourself for wearing the wrong hat or the wrong jersey or for not eating your pregame meal, but the real answer is that Baseball Happens.
Baseball Happens. It will continue to happen. Balls will bounce for and against your team all year long. Sometimes in streaks. The last bad bounce is not at all related to the next bounce being for or against your team. The whole point of needing thousands of data points is that data doesn’t normalize until the coin has landed on heads and tails enough times to warrant a conclusion. Fandom begets making conclusions more hastily than is necessary. Sports are emotional, irrational and unreasonable by nature. At some point one of those bad bounces could cost your team a game, a series or the season. Be upset. Get ticked off.
Then think back to the Angels and Brewers. It’s a long season.
If you visited the site over the weekend, you noticed that we were forced into database maintenance mode for the last couple days. There’s two reasons for this, the first lesser reason is that there was a power outage at my apartment while I was in the process of making the fixes, dragging the whole process out for an extra day.
But the main reason for the downtime was an attempt to fix a bug with suspended games that’s plagued the site since day one. Our parser that updates stats does not handle suspended games well. sometimes things will update fine, sometimes a players totals will get updated but their game logs won’t, and sometimes the game just gets ignored all together. This is a major bug that unfortunately is very hard to track down since the way the files that we base our stats off of change after a suspended game is completed making it a hard bug to reproduce.
We’re going to do our best to squash this bug going forward but if there is a slight error in our numbers, this is the reason why.
When we updated our innings pitched number to follow a more conventional format of listing 2/3rds of an innings as “.2″ instead of “.7″ a predictable bug cropped up that any ratio stat that used innings pitched no longer came out quite right because we would divide by “15.2″ instead of “15.6666667″. ERA and similar stats should now show up correctly again.
It’s mostly true that pitchers have little control over their balls in play, so if a pitcher is too far from the league average BABIP his performance is likely going to regress to the mean. However, one reliable way a pitcher has to control his BABIP is by inducing popups. Since popups have the same result as a strikeout 99% of the time (batter is out, runners can’t advance) BABIP-IFB treats popups the same as a strikeout, and removes them from the denominator of BABIP.
By removing popups low BABIP pitchers that looked “lucky” who induce a lot of popups no longer appear to be candidates for regression. For example Jose Quintana has a .266 BABIP this year but he’s second in the AL in infield flies allowed. Normal BABIP would say he’s likely to have more hits start falling in, but by taking away his popups, we see his BABIP is actually close to the league average. If he continues this infield fly rate all year, he’s likely to keep up his .266 BABIP.
BABIP-IFB is listed next to BABIP in the “Balls In Play” table.
Three true outcomes percentage. The percentage of time a plate appearance results in a strike out, walk, or home run. Not incredibly useful for analysis, but it is interesting to see how little a guy like Adam Dunn puts the ball in play. Added to the “balls not in play” table.
Fractional innings pitched are shown in the more common “.0, .1 and .2″ form. Before they were shown as “.0, .3, and .7″.
Zone swing percentage has been fixed to read as “ZSwng%” it was mistakenly shown as “ZSwng” before.
The report page no longer loads the Diamondbacks by default when you click the “Report” link at the top of the page. This will improve load times for people looking to get a report for non D’Back teams.
wOBA previously showed as .000 on yearly splits. Now fixed.
What is wOBA?
wOBA is an offensive stat that seeks to fix some of the flaws in OPS. OPS is a good measure of offense but it has some flaws such as treating a point of on base percentage as equal to a point of slugging, and saying a double is twice as valuable as a single. wOBA looks to correct these flaws and display them in a league neutral context.
What is a good wOBA?
wOBA is scaled with on base percentage. .330 is average, .300 is below average, .360 is above average and .400 is elite.
What numbers factor into wOBA?
There are seven numbers that affect wOBA: singles, doubles, triples, home runs, reaching via error, walks, and hit by pitches. Each of these numbers is multiplied by a coefficient for each league based on the overall league numbers then adjusted to make .330 average.
Where do these coefficients come from?
The wOBA coefficients are calculated with this spreadsheet by JT Jordan. They are based off the total offensive production of the league.
If it’s based on the league numbers do coefficients change everyday?
No. Until the season is over we use wOBA coefficients from the previous season. This means there will be some adjustments during the post season.
Why are these numbers slightly different than wOBA elsewhere?
- Some versions of wOBA scale average to the league average on base percentage, instead of a flat .330 like we do.
- Some versions of wOBA include stolen base numbers
- Some versions of wOBA do not include pitcher numbers. Ours does, and this is a flaw we’ll look to fix in the future.
Is wOBA league adjusted?
Yes. No matter what the run environment is .330 is always average.
Is wOBA park adjusted?
No. In the future we’ll add park adjusted wOBA.
Why doesn’t wOBA show up in players career numbers or across leagues?
The way our stats are calculated currently doesn’t support combining wOBA coefficients from multiple years or leagues. We’re working to fix this.
2013 updates aren’t quite in the state we’d like them to be in yet, and it looks like some games got double counted yesterday. We’ll work to have this resolved before we update tomorrow.