Tuesday, May 08, 2007

April 2007 Reds Review: Pitching & Defense

Reds pitching was so good in April that it led the National League in FIP at 3.29! Unfortunately, that success was almost entirely confined to the starting rotation. While I'd probably rather have a good rotation and a bad bullpen than the other way around, it will probably take both to make a serious playoff run this year. Let's take a look at Reds pitching and see if the starters can keep it up, and if the bullpen can regain the form it showed during the first two weeks of the season.

How they got hit - OBP against vs. HR/9
An analog to the OBP v. ISO graph I use for hitters, this shows how different pitchers on the Reds staff got hit. Vertical and horizontal lines represent league averages in April--note the league-wide depressed HR/9 rates (typically 1.0-1.2 hr/9)!
  • The best place to be is in the bottom-left, and the Reds happily have 7 of their 12 pitchers in that quadrant. Not coincidentally, those have been our top pitchers.
  • Todd Coffey had a really rough month. As we'll see below, however, that might not all have been his fault.
  • Coutlangus gave up 2 homers in 7 1/3 innings in April. Prior to that, he'd given up only 4 in ~155 minor league innings in his entire career. He generally keeps the ball on the ground, so I wouldn't expect this of him every month.
  • Milton's right where he usually is. Stanton is something of a fly ball pitcher, but historically hasn't given up a lot of home runs. Otherwise, those two had a nice month.
Strike Zone Management - BB/9 vs. K/9
Walks and strikeouts are two of the most important peripherals underlying pitcher performance. National League April averages are noted by the horizontal and vertical lines.
  • The place to be in this figure is the lower-right quadrant (high K, low BB), and again, the Reds are looking pretty darn good.
    • Here we start to see some of the weirdness of Todd Coffey's month. His strikeout and walk rates were phenomenal--and not completely out of line with what he showed he could do last season. He did have some problem with hit batsmen in key situations, but that seems likely to be a temporary issue as long as his walk rate remains solid.
    • Stanton, Belisle, Lohse, and Weathers, all came in at under 1.5 bb/9. All are likely to regress a tad, but I'll say a word about the latter three in particular:
      • Belisle's walk numbers are very encouraging, mainly because the show a completely reversal of his terrible 4.3 bb/9 last season. Prior to 2006, Belisle's control had always been excellent (2.73 bb/9 in 85.7 IP with the Reds in 2005, 2.24 bb/9 over his minor league career).
      • Lohse has been less consistent over his career, but the years that he's been good (2003 and 2005), he's had walk rates in the 2-2.2 bb/9 range. This would seem to be the key to his success, and I hope he can keep something like this up all season.
      • Weathers had his worst walk rate in the past 6 years in 2006, and it historically has been his weakest peripheral (4.0 bb/9 career). He's likely to regress substantially from this number.
  • I'm mildly concerned to see Arroyo on the left side of the strikeout line, as his success seems most tightly coupled to that rate. But then again, his k-rate varied all over the place from month-to-month last season, so I won't be concerned until this starts to become common...and furthermore, it's nothing like his miserable 3.4 k/9 in the second half of 2005.
  • Milton's k-rate was lower than I'd like as well. And the results speak for themselves.
  • Saarloos's walk rates were out of control, but surprisingly, his k-rates were extremely high--especially for a guy who entered the season with a 4.3 k/9 career rate. I'm more encouraged by the k's than discouraged by the walks.
  • Finally, you have to be a bit concerned that Santos is about to fall apart. The guy was amazing in the spring, but has never shown himself to be more than a mid-high 4's ERA kind of guy. I'm hoping that his move to the bullpen will bring his ERA down to the ~4.00 range on the season, but it's more of a hope than anything.
Performance When It Counts -- VORP vs. WPA
VORP is an indication of the total run value above a replacement player that an individual contributed to his team. Different replacement-level standards are used for starting pitchers and relievers. Win Probability added refers to the cumulative affect on win probability of each batter the pitcher faced over the course of the year. Good or bad performance in high-leverage situations will have a larger effect on WPA than in less crucial situations. Individuals with higher WPA than their overall performance would predict are identified as "clutch" in this figure. Regression line is solely a comparison between Reds WPA and VORP in April, nothing more.
  • Despite his relatively high ERA (4.91), Mike Stanton shows up as the most "clutch" pitcher on the club in April relative to his overall performance--he was at his best when the game was on the line.
  • Coffey also shows up on the good side of the line, even though his net WPA was negative. This identifies that there were times in the month when he did come in a shut down the opposition. Unfortunately, there were times like this game...
  • David Weathers maintained a 2.70 ERA on the month and converted 5 of 6 save opportunities, yet had a negative WPA on the month. What gives?
    • His blown save was a doozie, and yet didn't involve giving up any earned runs of his own.
    • He also lost a game when he was entrusted with a tie ballgame.
    • Most of his successful saves haven't been high enough leverage situations to net him much WPA.
    • Overall, while his basic numbers don't look too bad, he was not a very good closer. Unfortunately, given everyone else's struggles, there's no obvious choice to try in his stead.
  • Saarloos was absolutely terrible when it mattered most. I still like him as a bullpen guy who can come in and get a ground ball out, but he didn't get the job done in April.
Bad performance, or just bad luck? -- FIP vs. ERA
FIP, or Fielding-Independent Pitching, is an estimate of what a pitchers ERA should be given his peripherals: his strikeout, walk, hit by pitch, and home run rates. The diagonal line indicate a perfect match between ERA and FIP. Individuals above the line underachieved, meaning that their peripherals indicate that they pitched better than their ERA shows. Overachievers may have gotten lucky.
  • First, the Overachievers:
    • The biggest over-achiever, by far, was Victor Santos. His strikeout (5.6 k/9) and walk (4.8 bb/9) rates were both substantially worse than the league average, and only a rather lucky 0.250 BABIP saved him from what FIP estimates should be a 4.70 ERA. He will be one to watch for signs of collapse as we move forward this season.
    • Coutlangus also shows up here, but that is largely due to his alarmingly high HR/9 rate. However, that is probably an aberration given his propensity to be a ground ball pitcher--his HR/F ratio was an unusually high 28%, indicating some bad luck. So I'm less worried about him than this figure indicates we otherwise might be.
  • And the Underachievers:
    • Weathers' excellent k/bb ratio this month puts him above the line, but I worry that is unlikely to continue. His BABIP (0.281) actually indicates he experienced some luck.
    • Belisle is similarly rated where he is due to his superb walk rate in April. I think he has a better chance of keeping that up--or, at least, something close to it--than Weathers.
    • FIP indicates that Harang should have had an ERA closer to 3.04, which is a nice improvement from the 4.23 ER/9 he actually recorded. A 0.336 BABIP might be to blame...
    • Finally, Todd Coffey comes in as the biggest underachiever of them all. Yet, even his FIP (4.74) is substantially higher than we'd expect from him. Also notable in his April performance was that his home run per fly rate was 24%, which is unusually high. xFIP, which attempts to correct for varying HR/F rates as well as the other peripherals, indicates that his ERA should have been 3.22. Finally, his April GB% was the best on the club at 66%, which is a great sign that he was doing what he could to try to succeed. I really think that he'll be ok, and, in fact, that he wasn't throwing the ball all that badly in April.
Home Runs Allowed - True v. Standardized Distance
This graph shows data HitTracker in April, comparing actual distance the home runs allowed by Reds pitching traveled versus a standardized distance correcting for wind, temperature, and altitude. Diagonal line indicates a perfect match between actual and corrected home run distance. Strangely, two of Aaron Harang's home runs allowed in April have not been plotted in HitTracker.

Of those recorded by HitTracker, only one home run allowed by Reds pitching seemed particularly helped by the elements: Kyle Lohse's first allowed, hit by Mark DeRosa at Great American Ballpark on April 5th, which HitTracker believes would have only traveled 392 feet instead of its actual 423 feet. It might still have cleared the fence, but would have been a much closer play than it otherwise was.

Outlook

There are two basic questions about Reds' pitching over the rest of the season:

1. Can the starting pitching keep it up?
Harang and Arroyo seem to be performing just as well as we'd expect them to, with Harang perhaps even getting a bit unlucky. Lohse and Belisle have been surprises, thanks primarily to amazing walk rates. I'd expect both of them to decline a bit based on what they've started, but if they can maintain their control over the rest of the season in the 2-2.5 bb/9 range, they should continue to be effective.

2. Is there any hope for the bullpen?
I think so. Coffey looks to be coming off a really unlucky month, and I think we'll see him return to grace as a quality set up man over the rest of the season. Having him in good form would be a huge boost. We can also look forward to the imminent returns of Gary Majewski (recovering from a weak shoulder, apparently just about ready at the time I write this) and Bill Bray (threw some today, out with a broken finger)...and perhaps even Eddie Guardado, who threw batting practice today.

However, there is cause for concern. Santos's peripherals were scary in April, and I worry about whether he can put it back together. And Weathers, while he had good stats in April, did so with uncharacteristically-low walk rates. I'm skeptical about whether he'll continue to maintain that level of control. Finally, Saarloos is another key guy in the pen who hasn't pitched very well. His strikeout rate is encouraging, but k's are not his game--he needs to get the ball over the plate to induce ground balls in order to be successful.

Defense


This section will be abbreviated compared to what I hope to do in future months. The reason is that the Hardball Times isn't yet updating their ZR stats for the 2007 season (in an e-mail they told me that they think it will be operational sometime in June), so we'll have to settle for their overall team ratings on ground balls (mostly the infield) and fly balls (mostly the outfield). Here's how the Reds stack up against other teams in the National League:
The Reds are located in that cloud of points at the center of the graph. Given how terrible they were last season in both the infield and the outfield, being squarely league-average has to be considered a huge improvement. Unfortunately, until the ZR data are released, we won't know the degree to which any of the various defensive changes have caused this improvement. But moving Griffey to right, splitting Hamilton and Freel in CF, starting Gonzalez at short, and perhaps even improvements from Encarnacion at 3B(?), might all have contributed to this trend. Now, if they can just keep improving...

Relevant pitching stats for April:
Name IP K/9 BB/9 HR/9 HR/F %GB BABIP ERA FIP xFIP VORP WPA OBP SLG wOBA
Harang 38.3 8.5 2.1 0.7 7% 43% 0.336 4.23 3.04 3.21 4.0 0.13 0.301 0.395 0.311
Arroyo 34.7 6.0 2.3 0.3 3% 38% 0.271 2.86 3.03 3.79 5.9 0.59 0.265 0.282 0.274
Lohse 34.3 6.8 1.0 0.8 7% 34% 0.315 2.88 3.17 3.55 8.6 0.71 0.283 0.397 0.301
Belisle 31.3 6.3 1.4 0.0 0% 45% 0.289 3.45 2.27 3.08 6.9 0.24 0.260 0.304 0.251
Milton 21.0 5.6 2.1 1.3 8% 26% 0.301 4.71 4.53 4.70 -0.2 -0.34 0.316 0.456 0.352
Coffey 12.3 10.2 2.2 1.5 24% 66% 0.441 6.57 4.74 3.22 -2.7 -0.12 0.388 0.419 0.379
Saarloos 11.3 7.2 4.8 0.8 11% 58% 0.250 4.76 4.62 3.96 -1.3 -0.41 0.321 0.349 0.340
Santos 11.3 5.6 4.8 0.8 10% 52% 0.250 2.38 4.70 3.99 3.6 0.14 0.307 0.373 0.303
Weathers 10.0 7.2 0.9 0.0 0% 34% 0.281 2.70 1.90 2.91 0.9 -0.26 0.238 0.317 0.286
Coutlangus 7.3 11.1 2.5 2.5 28% 39% 0.222 4.91 5.94 3.50 -0.6 0.09 0.304 0.415 0.345
Stanton 7.3 7.4 1.2 1.2 11% 29% 0.238 4.91 4.57 3.98 0.3 0.48 0.291 0.358 0.292
Cormier 3.0 3.0 3.0 3.0 32% 44% 0.300 9.00 8.87 5.54 -1.2 -0.34 0.400 0.615 0.431
Burton 0.3 0.0 90.0 0.0 N/A N/A --- 0.00 33.20 30.02 0.2 -0.01 0.769 0.000 0.554

5 comments:

  1. I just finished listening to yesterday's (May 7) game, and heard Coffey once again give it up in long ball fashion. What a weird season thus far for him. 9.5 k/g, 2.1 bb/g, and 64% GB% are all outstanding rates. But he's given up 2.1 hr/g, and 32% of all his fly balls this season have been home runs. That's absurd. So is his 0.381 BABIP.

    It almost has to be a small sample size issue at this point, and there's nothing for it but to keep sending him back out there. But that's not making this tight feeling in my gut any better. Ugh. -J

    ReplyDelete
  2. Agreed - Coffey has to get better, right? In the game you mentioned he looked good against Lee but made one mistake pitch to Scott, and paid for it.

    BTW, you probably saw this but THT has a review of IF defense on grounders only. The graph of the Reds is depressing and interesting, especially EdE's zones. I expected him to be worse on balls hit directly to him but better on balls to his sides, not the opposite.

    ReplyDelete
  3. I hadn't seen that yet, and it is pretty interesting. A few comments:

    * Rich Aurilia also played a lot of 3B, and according to both ZR and PMR (see my '06 fielding review) he was absolutely HORRIBLE, even compared to EdE. So his performance--particularly at the extremes of the third basemen's area of responsibility on that graph--might well be having a big effect on that figure. ... 'Course, Eddie wasn't exactly good there either.

    * The Reds' problems at SS were far worse than anything happening at 3B.

    * The Reds look darn good at 1B. Hatteberg, by all accounts, has excellent range at that position.

    * Phillips' performance at 2B isn't looking real hot...the metrics I've seen rate him as average at 2B, so I'm surprised to see those numbers so low. It's not matching up real well to David Pinto's PMR charts...
    -j

    ReplyDelete
  4. By the way, Encarnacion's PMR chart looks like you expected it to--not so good on balls hit right at him, but just fine on balls hit to either side of him.

    Aurilia was across-the-boards bad, except down the line, where he seemed fine. -j

    ReplyDelete
  5. Right, forgot about the 50 games Richie played at third. Now it makes more sense.

    You're certainly right about SS, and I fully expected to see the defecit there. Sea Bass's 2006 chart is encouraging, and from what I've seen so far it looks on the money. Though I'm not sure if there's any place to see zone-type stats for '07.

    ReplyDelete