« Draft day 2 recap | Main | Lefties leave O's out to pasture »

Will the real, slimmer Sidney please stand up?

Head over to Michael Wolverton's Support-Neutral starter rankings at the Baseball Prospectus site and at the top of the "Flakiest ML Starters" rankings, you will find a name familiar to Orioles fans:

Flakiest 10 ML Starters (ranked by variance of SNVA):
(7 starts minimum)

Pitcher       Team   APW  SNVA   SNVA Var.
Ponson,S       BAL  -2.1  -1.6    0.088
Hampton,M      ATL  -0.8  -0.4    0.081
Valdez,I       SDP  -0.5  -0.6    0.079
Estes,S        COL  -0.5  -0.4    0.079
Willis,D       FLA   0.6   0.8    0.077
Dickey,R       TEX  -1.1  -0.8    0.077
Pineiro,J      SEA  -0.4  -0.4    0.076
Ainsworth,K    BAL  -1.7  -1.2    0.075
Gobble,J       KCR  -0.5  -0.2    0.075
Lackey,J       ANA  -0.8  -0.4    0.074

(as of June 14, 2004)

Less mathematically versed readers may want to know what "variance of SNVA (Support-Neutral Value Added)" is. Essentially, a high SNVA variance means that a pitcher has had wide swings in the quality of his starts. And sure enough, a look at Sidney Ponson's game log indicates that among his fourteen starts this season, he has had four quality starts (at least six innings pitched with at most three earned runs allowed) but also seven starts in which he has allowed six earned runs or more.

Contrast that with last season, in which he had quality starts in 58% of his outings and not once allowed as many as six earned runs in a start. That kind of consistent quality attracted the interest of the San Francisco Giants going into the stretch run and made the Orioles feel confident enough in January to offer him $22.5 million in guaranteed money for 2004-2006.

A royal riddle

Ponson's 2004 failures have largely mystified him. After some of his poor starts this year, he has told the press essentially that he didn't think he pitched all that badly, but that opposing batters got hits on a lot of good pitches. This kind of remark can be interpreted in two ways. It could come across as a charitable act of sportsmanship that gives credit to a worthy opponent. Or it could look like a denial of personal responsibility that blames his lack of success on bad luck or even poor defense. As former Oriole Mike Mussina has learned in New York, fans and journalists will usually take the latter view and label the commenter a whiny complainer.

Either way, Ponson has refused to accept complete responsibility for his frequently gruesome results, and in particular he has never admitted that his weight (initially 265 pounds, now reportedly in the 250's) may have played a role in his struggles. This combination of unsatisfactory performance and a defensive posture with the press has irked everyone in the Orioles community, from the badgering media and demanding fans to Baltimore's coaches and management. The following articles indicate the scope of the unrest:

Ponson thinks that his command of his pitches has gone awry just enough to make him ineffective. As quoted on MLB.com:

“I am throwing the ball one or two inches the wrong way right now,” he said. “Sometimes in this game, that's all you have to be off is by a few inches. I feel good when I am pitching but I am not throwing the ball where I want to throw it.”

More than his command, Ponson's weight has provided a convenient target for his detractors. Jim Palmer, who at 58 years old is still trim enough to credibly pose in Jockey briefs, has been an especially shrill critic of Ponson's lack of fitness. In the Sheinin story, he delivered this scathing rebuke:

“Part of the responsibility of being a professional athlete is you show up in shape,” Palmer said. “The tragic irony of Steve Bechler's death is that the one guy came in overweight. The other guy came in overweight. One guy makes a minor league salary; the other makes $4 million. The one guy dies tragically, and the other guy keeps carrying around extra weight and gets a huge contract. That's irresponsible.”

Ponson's overweight is certainly a concern, but it is hardly a new phenomenon. He carried about 250 pounds last year, too, and it did not appear to hamper his pitching all that much. So there are probably other, less obvious factors that have led to his problems this season. What are these factors (if they exist)? Are they likely to change in the near future?

Peeling away the onion's layers

Sidney Ponson has made fourteen starts so far in 2004 (through June 13) and has gone 3-8 with a 6.60 ERA—atrocious numbers by any measure, but particularly so for a man who went 17-12 with a 3.75 ERA over 31 starts in 2003. But wins, losses, and earned run average do not tell the whole story. Let's dig a little deeper into the component parts of Ponson's 2004 performance, using his successful 2003 season as a frame of comparison.

Strikeouts, walks, home runs, and ground/fly ratio: these are statistics over which a pitcher exercises some control, and over which his defense has no control (except for the rare homer-pilfering grab). Some people refer to them (the first three, anyway) as Defense-Independent Pitching Statistics, or DIPS.

Ponson's Key DIPS Ratios
Year K/9 BB/9 HR/9 G/F
2003 5.6 2.5 0.67 1.69
2004 5.0 3.1 0.83 1.55
Change -.63 +.56 +.16 -.14
%Diff -11% +22% +24% -8%
2004 AL 6.3 3.5 1.1 1.21

(League averages as of June 10, 2004, from The Hardball Times via Baseball Info Solutions.)

Ponson has fared slightly worse than he did in 2003 in all of these fielding-independent categories. His strikeouts are down a shade, he's giving up about half a walk extra per nine innings, he's yielding homers a bit more often, and his ground/fly ratio is down a tick. But these are hardly earth-shattering differences—his 2003 and 2004 numbers are quite consistent overall. It's easy to believe that they were put up by the same pitcher in adjacent seasons.

And in fairness to Ponson, these rate stats remain respectable compared to the rest of the league. Although his 2004 strikeout rate is subpar, his walk rate is about league average, and his homer and groundball rates are significantly better than most of his peers'. (At last check Ponson was #11 among AL starters in ground/fly ratio as measured by STATS, Inc.) For the most part, Ponson is still keeping the ball down in the strike zone.

Batted ball outcomes: A great deal of a pitcher's statistical line stems from events over which he has very little control: the outcome of balls put in play by the batter.

Batting average yielded on balls in play
  2003 2004
Ponson .286 .352
Balt. .302 .307
AL .290 .297

(Team and league averages as of June 16, 2004, from the Baseball Prospectus. Batting average on balls in play is calculated as (H-HR)/(TBF-HR-HB-BB-SO).)

This year, Ponson's pitching has not worked in harmony with his defense, which has allowed a painfully high .352 average behind him on balls in play. This far exceeds the .307 team average and .297 league average on balls in play in 2004. In 2003, Ponson fared much better in this statistic, allowing a more typical .286 batting average on balls in play.

The Fielding-Independent Pitching (FIP) ERA as explained at The Hardball Times estimates that with a league-average defense behind him, Ponson's 2004 ERA would be a surprisingly decent 4.33. (Last year, his FIP ERA was 3.77, essentially identical to his actual ERA.) So take away the shoddy defensive support, and Ponson has been only about a half-run per game worse than last year.

The Oriole defense apparently is no better this year than last. The outfield of Larry Bigbie, Luis Matos, and Jay Gibbons is essentially unchanged from the 2003 version, so there is no reason to expect a large dropoff there. In the infield, Brian Roberts is still holding the fort at second, and Rafael Palmeiro is at least as good at first base as Jeff Conine was last year. At shortstop, Miguel Tejada has a fine arm, but he has also made several misplays (officially recorded or not) and has not been the great improvement over Deivi Cruz defensively that some thought he would be. However, the biggest problem has been third base, where Melvin Mora so far has been much less reliable than Tony Batista was last year. If Mora improves with the leather in the second half of the season, the Orioles' defense will draw closer to league-average efficiency, helping the pitchers in the process. Even without Mora's defense improving, the law of averages would predict that Ponson's high in-play batting average will regress to the mean sooner or later.

Clutch performance: Many learned scholars have studied clutch situations in baseball, and most have determined that raising one's performance in the clutch is not a repeatable skill, at least at the major-league level. However, no one denies that within a season a few well-timed successes or failures can make a noticeable difference in a player's overall statistics. How has Ponson fared in pivotal situations?

Ponson's OPS allowed by situation

2003 2004
None on .634 .754
Runners on .795 .935
RISP .782 .876
RISP, 2 out
.607 1.117

These numbers go a long way toward explaining Ponson's astronomical ERA. Ponson has utterly failed in clutch situations this season. With runners on, he has allowed an OPS of .935. With runners in scoring position, hitters have been nearly as lethal at .876.

But where Ponson has really been hurt is with runners in scoring position and two out. Hitters have clobbered him like a collective Manny Ramírez in those situations, driving home an amazing 27 runs in just 54 at-bats. By contrast, last year opposing batters drove in just 23 runs against Ponson with RISP and two outs—and that was over an entire season comprising 97 at-bats.

Fielding and holding runners: Could Ponson's weight be causing him to have trouble fielding his position? Let's take a look.

Ponson's fielding stats
Year TC PO A DP E FPct. RF
2003 50 21 28 1 5 .980 2.04
2004 19 11 8 0 1 1.000 1.96

Apparently not. Ponson's range factor shows no appreciable decline compared to last year.

What about holding runners? Is Ponson's bulk causing him to be slower to the plate?

Stolen bases off Ponson
2003 12 4 75%
2004 2 2 50%

Again, the answer appears to be no. If anything, Ponson is doing a better job of controlling the running game in 2004.

Relief support: Another minor contributor to a starter's statistics is the performance of relievers who inherit his baserunners when he leaves the game. Although he has now dropped out of the top 20, prior to his last start Ponson ranked among the 20 ML Starters most hurt by their relievers (ranked by Bullpen Support) as determined by the Baseball Prospectus's Wolverton. Overall, the Baltimore bullpen, keyed by Mike DeJean, has been one of the worst in the league this year. So to compound his problems, Ponson has not been helped by his bullpen.

Under the surface of Flake Superior

None of the preceding analysis directly addresses the matter of Ponson's flaky performance. Wolverton was asked about flaky starters in a recent online chat. The exchange went as follows.

Bill Johnson (in the boonies of New Mexico): I've always been intrigued by your "flakiest starters" segment in the SNWL, not least because Matt Morris, basically one of my favorite pitchers, is such a regular there. Any idea *why* pitchers qualify for the flaky label with regularity the way Morris does?

Michael Wolverton: I haven't ever looked into it in detail. Two possibilities:

  • Pitchers who are just variable in the amount of stuff they bring to a game. Especially pitchers who are prone to bouts of wildness.
  • Pitchers who are much worse with runners on base than with bases empty (and so are prone to collapse in a game when more than a handful of guys reach).

Those are two logical-sounding reasons, although their real-world validity is yet uncertain. Let's examine them in Ponson's case.

  1. Stuff and wildness: If Ponson's stuff is truly fickle, we should see significant swings in his strikeout rates from start to start. If he has wildness issues, his walk rates should show significant inconsistency as well.

    I don't have access to league statistics for per-start stats, so I'll compare the strikeout and walk rates for Ponson's starts in the 2003 and 2004 seasons. To improve the quality of the data, I'll use strikeouts and walks per batter faced (K/BF, BB/BF) rather than per nine innings (K/9, BB/9). Also, each start will be weighted equally in the calculations, no matter how many batters were faced in the start.

    2003 K/BF BB/BF
    Ave. .15 .07
    StDev .07 .06
    SD/Ave. .48 .88

    2004 K/BF BB/BF
    Ave. .12 .08
    StDev .04 .04
    SD/Ave. .37 .52

    The SD/Ave. ratios indicate that Ponson's strikeout and walk rates, start by start, have actually varied less this year than they did last year. So at least in this case, the "variable stuff" explanation does not seem to apply.

  2. Runners on base v. bases empty: These splits were explored earlier in the section on Ponson's clutch performance, and they revealed that Ponson has fared significantly worse this year with runners on base than with the bases empty. Bingo!

    What this means on a more microscopic level is harder to discern. Is Ponson having trouble with his mechanics or command pitching from the stretch? Has he modified his pitch selection with runners on base? How has the change of catchers (from Brook Fordyce to Javy López) affected Ponson's decision-making on the mound?

Since the questioner wanted to know a systematic reason for starters' flakiness, Wolverton did not mention a significant, though less logically satisfying, determinant: plain luck. Pitchers in general are notoriously unpredictable from start to start and from year to year. Much of this variability comes from the high degree of chance involved in the outcome of batted balls (or non-batted balls, for that matter). There also can be significant differences from game to game in the quality of the opposing team's offense, the ballpark's friendliness to hitters, the weather, and the umpire's strike zone. And the starting pitcher, who has no choice in any of these matters, must adjust to all of these conditions to succeed.

In a short time frame—a start, or a series of starts, or even a full season—assuming a normal distribution of luck, a few pitchers will experience much wider swings of fortune than others. As a result, their performances will look extraordinarily flaky. Luck alone doesn't explain why a pitcher like Morris would regularly appear among the flakiness leaders. But it plays a much greater role than people once thought.

Answers forthcoming

In summary, Ponson has not fallen apart as drastically as his gargantuan ERA would suggest, and there's a good chance that his statistics—particularly his ERA—will naturally get better as the season continues. His defense-independent stats are only slightly worse than they were last year, and he is fielding and holding runners about as well as he did before.

Where Ponson has fallen down or been extremely unlucky is in clutch situations, most notably with runners in scoring position and two outs. Unless Ponson is doing something unusual that makes him more hittable with runners on base, this trend cannot be expected to continue for an entire season. It has not helped that his defense has been inefficient at converting batted balls into outs behind him, and that the bullpen has not kept his baserunners from scoring after he has been knocked out of games.

I believe that the Orioles originally wanted Pat Hentgen to be the veteran anchor of their rotation last offseason, but were forced to reconsider Ponson after Hentgen signed with Toronto. Ponson, though possessing more raw talent than Hentgen, has had minor labrum issues and has never performed appreciably better than a league-average pitcher. Hentgen would have come more cheaply and probably would have contributed a level of performance comparable (if slightly inferior) to Ponson while being a better mentor to the young pitchers. But it looks like the Orioles are saddled with Ponson for a while.

There were comparisons between Ponson and the Diamondbacks' Randy Johnson before and after their head-to-head matchup last week (Arizona won, 8-1). Such comparisons are outrageous and irresponsible. Johnson is a unique creature, a 6'10", left-handed strikeout machine who peaked after age 30 and is still going strong at 40. His kind comes along once in a lifetime. As for Ponson, hard-throwing right-handers who experience runs of middling success in their mid-twenties come along every year. To think that Ponson could follow in Johnson's footsteps is unrealistic.

Certainly, Ponson can learn something from Johnson's drive and work ethic. As role models go, Johnson is unquestionably better than, say, Boomer Wells or Scott Erickson. But more than likely, what we have seen from Ponson in the last few years is what we should expect, and that is a relatively durable starter who can elicit grounders and post an ERA in the low to middle 4's. That may not be ace material, and it is certainly not Hall of Fame material, but it is a player who would be welcome on any major-league club. And that is why, despite his horrible results this season, the Orioles are not throwing Ponson off the ship just yet.


This page contains a single entry from the blog posted on June 16, 2004 2:09 PM.

The previous post in this blog was Draft day 2 recap.

The next post in this blog is Lefties leave O's out to pasture.

Many more can be found on the main index page or by looking through the archives.

Powered by
Movable Type 3.33