Posted in MLB, Philadelphia Phillies, Sabermetrics | Print | 100 Comments »
Today was a good sports day in Philadelphia with the Eagles winning by a slim margin over the Detroit Lions and the Phillies needing all nine innings to overcome the Washington Nationals thanks to a Jayson Werth walk-off home run. Aside from it being an exciting way to finish out a series sweep before welcoming the division rival Atlanta Braves, it was also a learning experience. Well, maybe not learning so much as reinforcement, but still — here’s what I noticed.
Charlie Manuel’s in-game strategy is awful.
Back on September 15, the Phillies led by a score of 9-4 entering the eighth inning against the Florida Marlins. FanGraphs had the Phillies as 97 percent favorites. At that point, it may be wise to use some of the less-popular relievers like Danys Baez. Who does Manuel call upon? Ryan Madson. Of course, Madson did as he was told and got three outs rather quickly and the Phillies became 99 percent favorites.
Sure, Madson may be fresher than he normally would be at this time of the year given his two-month vacation after he broke his toe, but if you get a chance to rest your best arms, you do it.
Fast forward to today. Seventh inning, Phillies trailing by one run, still with a 31 percent chance to win. Who does Manuel call upon? Danys Baez. And, of course, Baez did what he does best: he gave up a home run, putting the Phillies behind 5-3. When the inning was over, the Phillies were given a 22 percent chance to win.
It got worse. In the eighth inning, Manuel called on J.C. Romero, who has arguably pitched as badly as Baez. After a double, a walk, and a single, the Phillies were down 6-3 and only nine percent to win.
- Ryan Madson, 2.42 SIERA, used with 0.11 leverage index.
- Danys Baez, 4.78 SIERA, used with 1.02 leverage index.
- J.C. Romero, 4.82 SIERA, used with 0.53 leverage index (became as high as 1.01 by his own doing).
It wasn’t the first time Manuel unnecessarily used Madson. On August 29, he came in with another 5-0 lead in the ninth inning against the San Diego Padres. On August 22, he pitched the eighth inning against the Nationals with a six-run lead. That’s just going back one month. Overall, he’s come into the game:
- With tie game: 8 times (16%)
- With team ahead by one to three runs: 23 times (47%)
- With team ahead by four runs or more: 10 times (20%)
- With team trailing by one to three runs: 5 times (10%)
- With team trailing by four runs or more: 3 times (6%)
So, in 26.5 percent of Madson’s appearances, the team has either been ahead or behind by four or more runs. That is poor leveraging of arguably the most important reliever (and unarguably the best) in the bullpen.
|Blanton, by inning|
Additionally, Manuel chose to use Greg Dobbs as a pinch-hitter in the sixth inning with the Phillies trailing 4-3. They were only 2-to-1 underdogs at the time and the leverage index was 1.16. Dobbs had an awful .260 wOBA going into today’s game. Manuel chose to use him instead of Ross Gload and his .359 wOBA on the season. Manuel eventually did use Gload but it was in the eighth when the Phillies trailed 6-3 and the leverage index was just 0.79. Although Domonic Brown is still nursing an injury, he was available in “emergency” situations. Call me crazy, but I’ll take Brown at 40 percent health over Dobbs at 100 percent any day, especially if all I’m looking for is a home run.
Finally, Charlie Manuel yet again left Joe Blanton in too long. It has been a feature all year long: Blanton tires around the sixth inning. Some examples of Manuel not realizing this:
- May 3 (first start since being activated from disabled list): Allowed one run through six innings; allowed three in the seventh. To be fair, two of the runners were inherited by Nelson Figueroa and were allowed to score.
- May 8: Allowed no runs through five innings; allowed three in the sixth.
- May 15: Allowed two runs through six innings; allowed three in the seventh.
- May 20: Allowed one run through six innings; allowed six in the seventh.
- May 26: Allowed one run through five innings; allowed three in the sixth.
- July 4: Allowed two runs through six innings; allowed three in the seventh.
- July 16: Allowed one run through five innings; allowed two in the sixth and was brought back out for the seventh.
- July 21: Allowed one run through six innings; allowed one run in the seventh.
- July 31: Allowed three runs through five innings; allowed one run in the sixth.
- August 18: Allowed one run through five innings; allowed one run in the sixth and was brought back out for the seventh.
- September 2: Allowed four runs through four innings; allowed two runs in the fifth.
- September 7: Allowed three runs through six innings; allowed one run in the seventh.
- September 13: Allowed no runs through five innings; allowed one run in the sixth.
- September 19: Allowed one run through five innings; allowed three runs in the sixth.
Blanton appears to simply not be conditioned to pitch deep into games this year. The oblique injury that caused him to miss the first month of the season likely has a lot to do with that. That this hasn’t occurred to Manuel or anyone else involved with the Phillies after five months of watching Blanton pitch is baffling.
The other thing I learned today:
The batting average with runners in scoring position stat is the new batting average.
It’s the latest craze that statistically-oriented people are going to have to swat down. All season long, fans and media types alike have been pointing out that Werth’s production with runners in scoring position has been lackluster. And it has — his triple-slash line was .172/.339/.281 coming into today’s game.
However, rather than use it as intended — as a descriptive statistic — people have been using it to cast aspersions on Werth, portraying him as an “unclutch” player or saying that he is pressing because of his looming venture into free agency. The problem with that is RISP varies wildly from season-to-season so you can’t make any strong inferences about a player’s ability from it. Last year, Werth’s triple-slash line with RISP was .279/.407/.510.
RISP will end up hovering around a player’s overall production. Derek Jeter, known worldwide as “Captain Clutch” and “Mr. November”, has a career triple-slash line of .314/.384/.453. With RISP, it’s .303/.398/.428. In terms of OPS, that’s a difference of 11 points, a negligible difference.
Even Ryan Howard, widely regarded as a “run producer”, hits exactly the same over his career (.279 AVG, .575 SLG) as he does with RISP.
Those are my thoughts from Sunday’s action. Looking forward to an important three-game series against the Atlanta Braves. Check back tomorrow for a series preview with Peter Hjort of Capitol Avenue Club.