26 August 2015

What WPA Teaches About The 2015 Orioles

Just five games ago, it seemed like the Orioles had finally turned the corner and were set to make a run to at least secure a wild-card spot. The Orioles lost each of those next five games and are now at .500. Orioles’ fans are feeling despondent while Manny Machado apparently boiled over in the locker room after the latest Orioles heartbreaker. Why have the Orioles struggled this year?

In 2014, the Orioles scored 705 runs while hitting 211 home runs and are on pace to score 715 runs and hit 212 home runs this year which means that their offense hasn't gotten worse. The pitching hasn’t performed as well as the Orioles allowed 505 runs through August (3.74 R/G) in 2014 and have so far allowed 498 runs this year (4.01 R/G).  Allowing an extra quarter of a run per game shouldn’t be the difference between winning 96 games and flirting with .500.

Looking at the Orioles’ Win Probability Added (WPA) and Clutch statistics from 2012-2015 can help explain why the Orioles have struggled this year. For those unfamiliar, WPA determines the change in the likelihood that a team will win from one plate appearance to the next and credits or debits the player based on how much their action increased their team’s odds of winning. The Clutch statistic measures how well a player has performed in high leverage situations.

The first chart, which measures the Orioles’ offense from 2012-2015, shows that the Orioles have scored a standard amount of runs. The problem is that they haven’t been in the right situations and is why the Orioles are projected to have a negative clutch score for the first time in the sample. In contrast, the Orioles’ offense was very clutch in 2013 and reasonably clutch in 2012 and 2014. This is also why the Orioles’ offensive WPA is on pace to be worth 2 wins while it was worth over 6 wins in 2013 and 2014.

While the Orioles are on pace to score 715 runs in 2015, the Orioles’ wRC suggests that they should only be scoring 687. This is lower than what the Orioles scored from 2012 through 2014 and could indicate that Orioles’ fans feel like their offense isn’t as good as it has been in past years because they've scored more runs than expected. 

The next chart shows that Orioles’ starters have a better ERA in 2015 than they did in 2012 or 2013 but have a considerably worse win-loss record and have a lower WPA. The data also suggests that the Orioles’ starters have been only average in the clutch as opposed to 2014 where they came up big in crucial situations. Unlike in previous years where the offense was able to offset the starting pitching weaknesses, it appears that the Orioles have been unable to recover from their starting pitching performances.

The third chart shows that the Orioles 2015 bullpen has had the best ERA out of the four bullpens when using either standard ERA or leveraged ERA (a metric that I use which grades relievers based on their innings pitched and game leverage to ensure that a closer is given more credit than a middle reliever for bullpen performance). Unsurprisingly, the solid 2015 bullpen has a high 59% win rate (equivalent to a 96-66 team) which is the best of any Orioles club over that period with the sole exception of the extremely clutch 2012 bullpen that had a 74.4% win rate (equivalent to a 121-41 team). Yet while this bullpen has been extremely effective, it hasn’t resulted in producing wins according to WPA compared to those other bullpens.
The reason for this can be explained partly by game leverage and partly by comparing saves, holds and blown saves to previous years. In 2015, the Orioles bullpen average game leverage has been 1.05. Their average game leverage ranged from 1.15 to 1.30 from 2012 to 2014. This means that they haven’t been able to pitch in as crucial of situations as they have in the past. This explains why the Orioles’ bullpen is on pace to have only 42 saves and 48 holds in 2015 despite having mid-50 saves from 2012-2014 and between 67 and 97 holds from 2012 to 2014. It would appear that the Orioles bullpen hasn’t had as many opportunities to pitch with a lead as they have previously and consequently saves, holds and blown saves have all decreased.

It should come as no surprise that the Orioles are 17-22 in one run games and 10-15 in two run games because those are the ones most dependent on luck or clutch performance. If the Orioles were to have won as many one and two run games as they’ve won games decided by three or more runs, then they would have another ten wins and would likely be in a fight with the Blue Jays to determine the AL East winner.  Their problem is that they simply haven’t been clutch.

In previous years, the Orioles have been successful by getting a lead to their bullpen and letting their bullpen save the day. In 2013, the Orioles ultimately fell short because the bullpen couldn’t hold onto enough leads and win enough games. This year the bullpen has done well when given a lead (except for this past week) but it just hasn’t happened often enough. The story that the data seems to be telling is that the bullpen has been placed in a situation where they’re trailing by a small amount instead of leading by a small amount and the offense just can’t quite get that hit that would give the bullpen a lead. There’s only so much a bullpen can do in those situations.

The Orioles are still .500 and have over a month to play. Their team is really built to win in September and if the Os do go on a run and take the second wild card spot then no one will care what happened earlier this year. But if the Orioles don’t end up making the playoffs, then people will look at these close games and wonder what could have been with just one more solid offensive player or slightly better pitching from the starting rotation. 

No comments: