Roy Halladay is good. I mean, really, really good. Without question, Halladay is one of those guys so consistently great that he gets overlooked due to flashes in the pan and pitchers that cobble together unexpectedly great seasons. Even more impressive than his raw figures and rates is the fact that these statistics have been accrued while facing the beasts of the AL East with regularity over the last several seasons.

This past Thursday, Doc turned in the type of performance we have come to expect, but the suddenly anemic Blue Jays could not muster any offense whatsoever. Despite seven scoreless innings from their ace, the Jays fell 1-0 to “rookie” Kenshin Kawakami and the Atlanta Braves:

Halladay’s performance was matched by Kawakami, as the latter scattered a mere three hits over eight frames, issuing nary a free pass while fanning seven. Kawakami started off a tad slow but has looked extremely solid over his last few outings. Both starting pitchers held the opposition scoreless, ultimately leading to a 1-0 win for the Braves on the heels of a Casey Kotchman sacrifice fly in the eighth inning.

Jesse Carlson, who had relieved Halladay after the seventh, surrendered a ground-rule double to Matt Diaz to begin the inning. Diaz advanced to third on a Yunel Escobar groundout, before scoring thanks to Kotchman’s flyball in a plate appearance with a 3.29 leverage index. Mike Gonzalez closed out the game for Kawakami, extending the Blue Jays losing streak and preventing Halladay from notching career win 140.

Regardless, Halladay has been absolutely phenomenal in the win category that actually means something: wins above replacement. Since 2002, including a couple seasons with lesser playing time due to injuries, here are Roy’s win values: 7.8, 8.0, 3.2, 4.4, 5.7, 5.8, 7.6. Following this particular outing, Halladay’s present win value shot up to 2.7. And if you consider that, in 2001, his 2.34 FIP in 105.1 innings bests his output in 2004–3.79 FIP in 133.0 innings–then Halladay had to have been in the range of 3.2-3.5 wins back then as well.

Since becoming a full-time starting pitcher, Roy Halladay has never seen his overall contribution fall lower than at least a win above an average pitcher, and those numbers were produced in partial seasons. His lowest win value in a full season is an astonishing 5.7 wins. In other words, Roy’s “worst” full season exceeded the production level of Johan Santana last season.

In last week’s top game, Halladay provided but another valid reason as to why he should not be left out of best active pitcher discussions. At 31 years old, he might not have another decade in the tank, but another few healthy seasons at his current pace and those Hall of Fame discussions are going to pick up steam almost exponentially.

I’ll be honest, I have not really been following Alex Rodriguez since his return from injury, so his consecutive games with a home run streak led me to believe that he had been performing just fine. A quick glance at his OBP/SLG and wOBA entering last night’s action confirmed that, at .412/.595, and a wOBA north of .400, Rodriguez had certainly been helping the Yankees throughout their impressive winning streak. Then my attention quickly shifted to the first component of the slash line, BA, since his seemed ridiculously low. Coming into last night, Rodriguez had been slashing .189/.412/.595. Despite a sub-.200 BA, A-Rod was OPSing over 1.000!

After an 0-4 showing, his numbers look like this: .171/.382/.537, .398 wOBA. Can you recall an odder looking line of statistics? Rodriguez’s BABIP currently rests at a microscopic .069. He has seven hits in 41 at bats, five of which are jimmy-jacks. Some more oddities: the league average hitter is swinging at 24% of his pitches seen out of the zone, though A-Rod has swung at such pitches just 12% of the time. On these outside pitches, the league average makes contact 63% of the time; just 33% for Rodriguez.

Curious, I looked through my database for the leaderboards in two different areas: the highest OPS for players with batting averages of .200 or lower, and the lowest BA for players with OPS marks of 1.000 or higher.

Highest OPS, BA < = .200
#  NAME           YEAR      SLASH         OPS
1) Mark McGwire   2001  .187/.316/.492   .808
2) Elijah Dukes   2007  .190/.318/.391   .709
3) Roger Repoz    1971  .199/.333/.374   .707
4) Ruben Rivera   1999  .195/.295/.406   .701
5) Rob Deer       1991  .179/.314/.386   .700

You knew McGwire’s name had to be on here, but this definitely shows that A-Rod is currently performing in uncharted territories. The highest single-season OPS for a player with a sub-.200 BA is .808, over 100 points lower than Rodriguez’s current mark. There is no way Rodriguez continues to get hits on balls in play just 6.9 percent of the time, but don’t you dare ruin my fun! Here are the lowest batting averages for players with an OPS >= 1.000:

Lowest BA, OPS >= 1.000
#  NAME           YEAR      SLASH         OPS
1) Barry Bonds    1999  .262/.389/.617   1.006
2) Mark McGwire   1997  .274/.393/.646   1.039
3) Mark McGwire   1995  .274/.441/.685   1.126
4) Jim Edmonds    2003  .275/.384/.617   1.001
5) Reggie Jackson 1969  .275/.409/.608   1.017

So the lowest BA with correspondingly awesome OPS marks is .262, which is in no way what we consider to be a poor percentage of hits in at bats. The league averages, for hitters with at least 200 plate appearances:

OPS >=  1.000: .324 BA
BA  < =  .200: .542 OPS

A-Rod isn't going to sustain this low of a batting average if he continues to take walks and put the ball in play the way he usually does, but a slash line like his at the present certainly merits a few different double-takes. When those balls inevitably do fall in, watch out. Rodriguez is still an absolute beast and his slash line right now serves as the perfect example for why batting average tells us next to nothing on its own.

One of the most difficult tasks a responsible baseball analyst must take on involves avoiding small samples of data to make definitive claims about a player. If Victor Martinez goes 4-10, it does not automatically make him a .400 hitter. We have enough information about Martinez from previous seasons to know that his actual abilities fall well short of that mark. Not everything, however, should merit a house call from the small sample size police because there are some stats that stabilize more quickly than others. Additionally, a lot of the small sample size criticisms stem from the actual usage of the information, not the information itself. If Pat Burrell struggled mightily after the all star break last season and started this season with similarly poor numbers, we can infer that his skills may be eroding. Isolating these two stretches can prove to be inaccurate, but taking them together offers some valuable information.

The question asked most often with regards to small sample sizes is essentially – When are the samples not small anymore? As in, at what juncture does the data become meaningful? Martinez at 4-10 is meaningless. Martinez at 66-165, like he is right now, tells us much, much more, but still is not enough playing time. What are the benchmarks for plate appearances where certain statistics become reliable? Before giving the actual numbers, let me point out that the results are from this article from a friend of mine, Pizza Cutter over at Statistically Speaking. Warning: that article is very research-heavy so you must put on your 3D-Nerd Goggles before journeying into the land of reliability and validity. Also, Cutter mentioned that he would be able to answer any methodological questions here, so ask away. Half of my statistics background is from school or independent study and the other half is from Pizza Cutter, so do not be shy.

Cutter basically searched for the point at which split-half reliability tests produced a 0.70 correlation or higher. A split-half reliability test involves finding the correlations between partitions of one dataset. For instance, taking all of Burrell’s evenly numbered plate appearances and separating them from the odd ones, and then running correlations on both. When both are very similar, the data becomes more reliable. Though a 1.0 correlation indicated a perfect relationship, 0.70 is usually the ultimate benchmark in statistical studies, especially relative to baseball, when DIPS theory was derived from correlations of lesser strength. Without further delay, here are the results of his article as far as when certain statistics stabilize for individual hitters:

 50 PA: Swing %
100 PA: Contact Rate
150 PA: Strikeout Rate, Line Drive Rate, Pitches/PA
200 PA: Walk Rate, Groundball Rate, GB/FB
250 PA: Flyball Rate
300 PA: Home Run Rate, HR/FB
500 PA: OBP, SLG, OPS, 1B Rate, Popup Rate
550 PA: ISO

Cutter went to 650 PA as his max, meaning that the exclusion of statistics like BA, BABIP, WPA, and context-neutral WPA indicates that they did not stabilize. So, here you go, I hope this assuages certain small sample misconceptions and provides some insight into when we can discuss a certain metric from a skills standpoint. There are certain red flags with an analysis like this, primarily that playing time is not assigned randomly and by using 650 PA, a chance exists that a selection bias may shine through in that the players given this many plate appearances are the more consistent players. Cutter avoids the brunt of this by comparing players to themselves. Even so, these benchmarks are tremendous estimates at the very least.

While discussing the virtues of Mark Buehrle earlier this week, I made sure to point out that the White Sox lefty is only 30 years old despite having the veteran reputation of being older. In the comments of that thread, somebody mentioned that Wandy Rodriguez could be the next pitcher to not only match Buehrle in terms of consistent durability but also with regards to consistently high win values. I hate to break it to those on the Wandy-Wagon, but this is not going to happen.

Don’t get me wrong, Rodriguez has pitched quite well since 2007, but he is currently in the opposite situation of Buehrle. Though Rodriguez may have the reputation of a younger prospect-type pitcher who has finally turned a corner, raise your hand if you knew, without looking it up, that he, too, turned 30 years old this year. Buehrle jumped onto the scene at the tender age of 21 and has been pretty fantastic ever since. Wandy debuted at the age of 26 and did not really become effective until he turned 28.

At 30 years old, Rodriguez likely still has several effective seasons in the tank, but there is absolutely no comparison between he and Buehrle. Through nine starts this season, however, Rodriguez has been tremendous. In 59 innings, he has surrendered just 43 hits with a 53/16 K/BB ratio. Put together, Wandy has already amassed +2.1 wins on the heels of a 1.83 ERA and 2.52 FIP.

Since his 2005 call-up, Wandy has essentially improved in FIP, UBB/9, K/9, HR/9, and WAR each season. He relies more on acumen than raw stuff since, well, 88-89 mph with average-ish movement isn’t all that impressive.

The two big red flags relative to his current production level are the strand rate and percentage of home runs per fly ball. The league averages for both metrics are 72% and 11%, respectively; Wandy currently sits at 80% and 1.6%. The 80% strand rate becomes more significant given his individual context; Wandy has been at 72% or lower in all four of his previous seasons. When the home run rate and strand rate both normalize, his numbers are going to venture into the wrong direction.

Even so, ZiPS sees Rodriguez capable of a 4.32 ERA and 3.83 FIP over the remainder of the season, meaning his bottom line marks of 3.51 and 3.37 would still be career bests. After two consecutive +2.7 win seasons, he is currently on pace to produce right around +5 wins.

Wandy is certainly deserving of attention and multiple looks, but do not get him confused with someone like Johnny Cueto or Ubaldo Jimenez, young flamethrowers who seem poised to turn a corner before turning 26. Rodriguez’s future may include a few more +3-4 win seasons, but he does not belong in the category of young studs finally reaching their potential.

Though more of a Strat-o-Matic fan than a fantasy baseball enthusiast, I aid some friends in their decision making processes for the latter from time to time. Last night, while hanging out for the Phillies-Reds game, one of these friends posed a question I have been asked countless times this season: What do you think of Shairon Martis? While my response involved nothing more than a gesture, encapsulated in the body language was my lack of confidence in the pitcher.

For fantasy purposes, I can see where Martis has value since he sports a 5-0 record, but he isn’t really offering much outside of that meaningless metric, for both fantasy and reality.

Through eight starts, Martis is averaging right around 6 IP/GS with 47.2 total innings. His 4.53 ERA and 4.58 FIP match up rather well and scream average. Now, an average pitcher is certainly an asset for a team, but I am quite skeptical that Martis can continue to perform at a league average level with his current peripherals and skillset.

For starters, his 4.34 K/9 and 3.59 BB/9 do not exactly imply dominance. Martis is also very unlikely to sustain a .269 BABIP over the course of the season, an inevitable regression that will taint his current numbers regardless of a below average strand rate bound to improve.

Further, it also isn’t as if Martis throws like Johnny Cueto, a young stud with a blazing fastball that needs to work out a few kinks before graduating from Throwing College into Pitching University. Martis throws a 90.3 mph fastball with about average movement and does not exactly resemble Derek Lowe in the groundball department. On top of that, none of his pitches, per the fantastic linear weights addition to the site, are disgusting or supremely dominant in any fashion.

Listing comparables for a pitcher proves to be a rather tall order given the numerous facets of performance that must be taken into account, but the way Martis is currently pitching seems to closely match the output of Jason Marquis last season, at least in the ERA, FIP, K/9, BB/9 and HR/9 areas. Martis gets hype because of the 5-0 record and the fact that nobody has seen him in action before, but deep down, his performance to date is akin to 2008 Jason Marquis.

ZiPS projects Martis to end the season with a 4.89 FIP in about 162 innings, which, when coupled with the projected rates of controllable skills, resembles the 2008 campaign of Tim Redding. Martis is pitching like Marquis right now with the projection of Redding for the remainder of the season.

Martis is only 22 years old, which makes him very attractive in the world of keeper leagues, be it for fantasy or Strat purposes, but he has not done anything right now to suggest that he will be anything more than a marginal fourth starter. Bear in mind that my opinion is in no way concrete after some starts this year and a few more last season, and that these feelings are derived more from questioning his selection, movement and velocity as opposed to the actual results.

To walk hitters at his current rate and remain successful, a pitcher either needs to have a flukily good year preventing the longball or strike hitters out at a rate far superior to 4.34 per nine innings. Color me skeptical that Martis’ combination of velocity, movement, and pitch selection can definitely get that job done.

Usually when I begin my posts with discussions centering on those with very solid controllable skillsets and the consistent durability to log 200+ frames each season, the topic of interets is Javier Vazquez. Not this time. No, instead of Vazquez, I want to give some props to another pitcher who might not fit this bill in the same exact manner but who has put together a pretty great career that gets overlooked far too often. That pitcher happens to be South Side southpaw Mark Buehrle, who just recently turned 30 years old, much younger than most realize.

Though 30 years of age does not necessarily constitute a spring chicken, Buehrle’s veteran savvy and resume dating back to the year 2000 make him seem older, somewhere in the 34-36 years old range. Since 2001, and including this season to weed out the retired, only four pitchers have thrown 1750+ innings: Vazquez, Buehrle, Barry Zito, and Livan Hernandez.

Hernandez is likely on his last legs, we all know the modus operandi of Vazquez, and Zito, despite a strong showing early on this season, has come nowhere near replicating his performance in Oakland. Buehrle has a 123 ERA+ in this span that leads the bunch. He is not exactly a strikeout machine, hovering around the league average in that department, but Buehrle walks next to nobody and breaks even on the home run front. All told, in exactly 300 major league outings, Buehrle has a 3.78 ERA and 4.14 FIP to his name.

These are not the kinds of numbers that beg for end of season award voting or spots on the all star team; however, the consistency with which these numbers were accrued speaks volumes for what he can bring to a team. Outside of the 2006 season–which was clearly an outlier in which his K/9 dropped to depths never before seen and his HR/9 rose almost exponentially–Buehrle has never posted an ERA greater than 4.21 (which actually came in his rookie season) or an FIP north of 4.26.

Here are his win values from 2002-08: +4.8, +4.4, +4.9, +6.3, +1.9, +3.8, +4.6. It is also fairly safe to assume that his numbers in 2001, which were very similar to those produced in 2002, would have resulted in right around +4.6 wins. Keeping in mind that Buehrle debuted as a rookie at the tender age of 21 years old, it is pretty incredible the kind of impact he made right out of the gate. At +0.9 wins already this season, he is right on pace for another season besting four wins.

Certain types of pitchers garner plenty of attention in the media: flashy ones like Carlos Zambrano, downright awesome ones like Johan Santana and Roy Halladay, the young ones reaching their potential like Tim Lincecum and Zack Greinke, and the ones that come from nowhere like Cliff Lee. Buehrle does not fit into any of these, and tends to fall into the same category as Roy Oswalt and Aaron Harang, very durable pitchers with quality numbers that, for one reason or another, get overshadowed.

Buehrle is not even close to being done, especially when you consider that Derek Lowe has a similar skillset and just signed a contract that will expire close to when he turns 40. The wear and tear on Buehrle’s arm may come into play but he doesn’t throw exceptionally hard and grades decently with mechanics. It’s about time we start appreciating what guys like Buehrle bring to the table.

When an otherwise average player suddenly puts up one of the best seasons of the decade, fans are not only going to take notice, but they will also be very skeptical that the performance can continue. The kneejerk reaction is to suggest regression to the mean in the coming season(s), especially if the solid numbers were more contingent upon luck based indicators than the controllable skills that tend to stabilize quickly. In the case of Cliff Lee, the bandwagon of fans suggesting he would stink this year grew as vast as the group that swore last season was “for real.”

Even though Lee’s strikeout, walk, and groundball rates were more responsible for his turnaround, the majority of fans felt he would come nowhere near a repeat performance this season. When Lee’s first two starts ended with 17 hits and 11 earned runs in just ten innings of action, confirmation biases reigned supreme.

A confirmation bias is essentially what happens when someone actively seeks evidence to confirm what they think is true. With regards to Lee, someone who felt he would significantly regress would view these two game lines and proclaim that the Cliffer had finally gone back to his old ways.

Well, eight starts into the season, Lee’s line looks like this: 8 GS, 54 IP, 3.00 ERA, 3.17 FIP, 7.0 K/9, 2.2 BB/9. His 1.41 WHIP is not likely to stay as high nor is the 81% strand rate that has virtually canceled out the walks and hits ratio. Lee has already produced +1.7 wins in spite of two poor outings to start the season.

Since his second start, Lee has gone 44 innings while surrendering just seven earned runs. He has a tidy 4.0 K/BB thanks to 32 punchouts and eight free passes. Even though Lee got off to a rough start, he has rebounded very nicely to the tune of a 1.43 ERA and 2.78 FIP over his last six starts. With the dominating season turned in last year and his stellar performance so far, ZiPS sees Lee finishing this campaign with a 3.43 FIP in 208 innings. Numbers like that resemble John Danks‘ line from 2008, which produced +5.2 wins.

Cliff Lee might never again surpass the +7 win plateau but his current pace will still likely merit some award votes and a spot on the all star team. He does not have the pedigree or reputation of a Halladay that would instill supreme confidence into the minds of fans, but Lee has, through eight starts, done plenty to show that last season was not a fluke.

Quantifying the contributions of a manager is a very difficult task in the world of baseball evaluations. Regardless, even those who feel that managers do very little can agree that one responsibility involves accurately filling out the lineup card prior to a game. Rays manager Joe Maddon seemingly could not handle this responsibility yesterday, leading to a ruling that places Sunday’s Rays-Indians game atop all others this week.

Though Clayton Kershaw and his seven no-hit innings certainly made a solid argument and the three consecutive walk-off wins for the Yankees stated an impressive case, Maddon’s blunder and the results surpassed all others.

For those who have not yet heard, Maddon penciled in both Evan Longoria and Ben Zobrist at third base. The ‘5’ on the lineup card next to Longoria’s name was circled, suggesting he would actually be the designated hitter. After Zobrist played the field in the top of the first, Indians skipper Eric Wedge pointed out the flaw. Thirteen grueling minutes later, the umpires stripped the Rays of their right to use a designated hitter, meaning that Andy Sonnanstine would be their third hitter.

B.J. Upton led off with a walk and promptly stole second base. Carl Crawford then singled him in, bringing Sonnanstine to the plate. Sonny tried to sacrifice Crawford along but ended up reaching first on a fielder’s choice force out.

In the top of the second, Ben Francisco hit a three-run homer to put the Tribe ahead, 3-1. A half-inning later, Jason Bartlett tripled and crossed home plate courtesy of a Gabe Kapler groundout. The game remained 3-2 in favor of the Indians until the bottom of the fourth inning, when the Rays erupted for five runs. Prior to that inning, Sonnanstine had batted again and struck out.

The five-run rally kicked off when Ben Zobrist started the frame with a triple. Bartlett then singled him home to knot the game at three runs apiece. Kapler walked and Akinori Iwamura singled to load the bases with nary an out. Michel Hernandez then delivered a bases clearing double to put the Rays ahead, 6-3. After Upton and Crawford were both retired, Sonnanstine launched a double over the head of the quite shallow Ryan Garko, scoring Hernandez.

Sonnanstine lasted until the sixth inning and left with a 7-5 lead. Though he got the win, he did not necessarily pitch that well. Maybe he took his poor at bats back to the mound with him. The last time this odd ruling came into play was on July 23, 1999, when Mike Hargrove messed up the fielding positions of Manny Ramirez and Alex Ramirez, meaning Charles Nagy had to bat. It isn’t uncommon to see a pitcher record a win while going 1-3 with an RBI at the dish… but it is very odd to see this occur in a league specifically designed to avoid pitchers hitting.

The Dodgers entered spring training with guaranteed rotation spots assigned to Chad Billingsley, Hiroki Kuroda, Randy Wolf, and Clayton Kershaw. The fifth spot was largely up for grabs between the likes of Eric Stults, James McDonald, Jason Schmidt, Jeff Weaver, and Eric Milton. The latter three had battled injuries and bouts of ineffectiveness in recent years but had a previously established track record that merited a low risk opportunity.

Well, Stults seemingly won out, but an injury to Kuroda opened up another spot, affording McDonald the chance to build upon his relief success last season and add starts to his resume. In his last three starts, McDonald surrendered eight runs in 11 innings with a 4/11 K/BB ratio. Joe Torre was not impressed and opted to move McDonald back to the bullpen.

Jeff Weaver got the call to replace McDonald after an impressive four-inning relief outing in which he held the Padres scoreless. After two solid starts, Weaver’s seasonal line sits at 14 innings with a 2.57 ERA and 3.01 FIP. He isn’t the same guy from the Tigers days or the first Dodgers stint but Weaver at least deserved to stay in the rotation.

Instead, Joe Torre has called upon Eric Milton to get at least one start in favor of Weaver. I’m not sure if Jeff is hurt, but I would hope he is for the sake of this decision. Granted, Weaver’s sample of performance is small, but when evaluating pitchers from start to start, small samples and scouting info are the only bits of info available.

The Dodgers certainly have depth at their fifth starter spot with the aforementioned quintet pitching in either the show or at the farm, but Weaver was pitching pretty well. Did he get hurt, or is Torre simply trying to see what he has with Milton?

Well, our site continues to expand, adding more and more features at an almost exponential rate. The most recent addition incorporated Dan Szymborski’s in-season ZiPS projections into a sortable page. We now have the capability to not only find out who looks the best or worst from this moment on but also to see how prior stats and the updated true talent expectations combine to form the new end of season projection.

Naturally, once the data was implemented, I raced over to the leaderboards and jumped to the very last page on the pitchers section. Yeah, I wanted to see which starters projected the worst from here on out. Now, starters fitting this bill are not automatically the guys that were terrible from the get-go, who have pitched very poorly thus far after abysmal pre-season projections. The list could also include some hurlers with awful projections who have pitched well and are bound to regress.

Over the rest of the season, here are the worst updated projections for starting pitchers that will realistically toe the rubber:

Adam Eaton, Orioles:       19 GS,  99 IP, 1.53 K/BB, 6.09 ERA
Franklin Morales, Rockies: 19 GS,  97 IP, 0.89 K/BB, 6.03 ERA
Rick Porcello, Tigers:     20 GS,  95 IP, 1.19 K/BB, 5.97 ERA
Dustin Moseley, Angels:    20 GS, 111 IP, 1.71 K/BB, 5.92 ERA

Assuming that these all come to fruition, the worst starters by season’s end would be:

Adam Eaton, Orioles:       25 GS, 130.1 IP, 1.55 K/BB, 6.35 ERA
Carlos Silva, Mariners:    27 GS, 149.2 IP, 1.94 K/BB, 6.31 ERA
Sidney Ponson, Royals:     21 GS, 125.0 IP, 1.28 K/BB, 5.83 ERA
Franklin Morales, Rockies: 21 GS, 101.0 IP, 1.01 K/BB, 5.83 ERA

These in-season projections alter whenever new data enters the sample but they provide a solid snapshot of what has occurred to date and how that affects our knowledge of a player’s true talent level. If Radhames Liz (not shown here due to not making any starts yet but clearly the worst in both areas) and Eaton live down to their projections, they are not likely to garner 19-20 more starts, skewing the playing time results. Still, this is an incredibly useful tool when it comes to determining the extent to which performance up to a certain point is meaningful.