The Hardball Times -- Glenn DuPaul

The Hardball Times -- Glenn DuPaul http://www.hardballtimes.com/main Baseball. Insight. Daily. en studes@hardballtimes.com Copyright 2013 2013-05-21T08:09:15+00:00 Five questions: San Diego Padres http://www.hardballtimes.com/main/article/five-questions-san-diego-padres6/ http://www.hardballtimes.com/main/article/five-questions-san-diego-padres6/#When:07:09:15
Almost no one is talking about the San Diego Padres. It is tough to fault anyone for not talking about the Padres, though; they made no high-profile acquisitions (either through trade or free agency), and they're coming off back-to-back seasons in which they finished near the bottom of the division with fewer than 80 wins.

Do the 2013 Padres have enough talent to compete with the likes of the Giants and Dodgers or will this season be more of the same?

How will moving the fences in at Pecto affect its run environment?

Almost since the day Petco Park opened in 2004, San Diego's home field has been considered the "anti-Coors Field." Right field at Petco was where home runs went to die.

Since 2004, the run environment (specifically in terms of home runs) at Petco has been one of the most pitcher-friendly in baseball. There were rumors that Petco's dimensions were causing hitters to become frustrated, which resulted in some clubhouse issues.

This offseason, the Padres organization decided to drastically alter Petco's dimensions. This move could be purely psychological, in hopes of improving the confidence of the team's hitters. The idea could also have been to increase the number of runs scored at Petco, although, the jury is still out on whether the fences in significantly changes the run environment of a park.

Recently, on MLB Network's Clubhouse Confidential, FanGraphs' Jeff Sullivan discussed this issue:

If you look at the data, sure enough home runs went up relative to games on the road... but you might suspect intuitively that with more home runs there will be a lot more runs scored. And what actually happened in those ballparks (that moved in their fences) is that runs did go up by a very slight amount, but it was a pretty insignificant gain.

To summarize Sullivan's point, more home runs do not necessarily equate to more runs. Moving the fences in at Petco probably will increase the number of home runs hit there, but at the same time it will also cut down on the number of other extra base hits (especially triples) that were a result of the outfield gaps.

The Padres organization seems to agree with Sullivan. The team's president and CEO Tom Garfinkel told reporters that Petco would still very much be a pitcher's park:

Petco Park will still be a pitcher's ballpark, but the changes in the outfield dimensions will eliminate some of the extreme bias. When a ball is crushed, it should be a home run. That didn't happen at Petco Park, particularly on balls hit toward right-center and left-center.

Expect to see more home runs hit in San Diego in 2013, but I would caution those who think substantially more runs will be scored.

What should we expect from the Padres' rotation?

In 2012, the Padres' starting rotation was unquestionably bad. Their starters ranked second-last in the National League in adjusted ERA, 122, and last adjusted fielding independent pitching (FIP), 121.

It is extremely difficult to be a successful major league team without good starting pitching.

According to the early spring training depth chart on the Padres' official team website, seven pitchers were vying for the five spots in the rotation. Added was veteran righty Freddy Garcia, who signed a minor league deal with San Diego and will get the chance to compete for a roster spot.

Below are each of those pitchers' 2012 innings, strikeout rate, walk rate and ERA, as well as their predictive FIP (pFIP) for 2013:

Pitcher	IP	K%	BB%	ERA	pFIP
Anthony Bass	97	19.5%	9.5%	4.73	4.04
Edinson Volquez	182.2	21.7%	13.1%	4.14	4.13
Freddy Garcia	107.1	19.3%	7.6%	5.20	4.23
Casey Kelly	29	19.1%	7.4%	6.21	4.23
Eric Stults	99	13.3%	6.5%	2.91	4.31
Jason Marquis	127.2	16.2%	7.5%	5.22	4.56
Clayton Richard	218.2	11.8%	4.6%	3.99	4.59
Tyson Ross	73.1	13.5%	10.8%	6.50	4.80

In 2012, the league average ERA for National League starters was 4.04. Stults and Richard are the only two pitchers pegged for possible rotation spots who pitched to an ERA-figure below that number. Once Richard's ERA is adjusted to account for the fact that he made half of his starts at Petco, Stults ends up with the only above-average ERA in this group.

Predictive FIP ignores balls in play and uses the three true outcome components (K, BB, HR) to project a pitcher's future performance. According to pFIP, only Bass is projected to have an un-adjusted league average ERA in 2013.

So the outlook is not too promising for the Padres' starters and 2013 could look far too similar to the 2012 season. But there may be some hope for San Diego in the form of three young arms.

For some time, Kelly has been ranked fairly high among the pitchers in the Padres' farm system. In a brief stint in the majors last season, his strikeout numbers were impressive. However, Kelly's strikeout rates throughout his minor league career have never been spectacular, so we'll see if he'll be able to generate whiffs and success when he is given a shot to be a part of the rotation next season.

Two others who show promise are left-hander Cory Luebke and Joe Wieland.

Luebke had a strong 2011 season, splitting time between the bullpen and rotation, with high strikeout numbers and a 3.29 ERA to go along with a shiny 2.93 FIP. He looked poised for a breakout in 2012, but after five starts he was shut down and underwent Tommy John surgery.

Wieland was a highly thought of prospect coming into 2012 and was given a shot to start the season in the Padres' rotation. Unfortunately his season, like Luebke's, ended with Tommy John.

Both will return sometime in 2013, possibly providing the shot in the arm this rotation will almost assuredly need.

How much will Chase Headley regress?

In 2012, Headley's fourth full major league season, he played like a bona fide superstar. The Padres third baseman set career highs in home runs (31), on-base percentage (.376), slugging percentage (.498) and wins above replacement (7.5). If Buster Posey had not been so spectacular last season, Headley could have won the National League's Most Valuable Player award.

I'm sure the Padres' fan base and organization have high hopes for Headley in 2013. But will he be able to repeat his 2012 production?

The rational answer to that question is quite simply, no. Any time a player has a career year, we would expect his performance to regress back toward his career norms. Even before the thumb fracture that will cost him the first month or so of the season, Headley was undoubtedly due for some regression in 2013, but how much?

When studying Headley's statistical line, the first metric that I looked at was adjusted weighted runs created (wRC+). This statistic gives a park and league adjusted measure of total offensive (hitting) value. In 2012, Headley's wRC+ was 145, which meant he was 45 percent better than the league average player. This number was more than 20 points higher than his still very solid 2011 wRC+ (121).

Intuitively, I would assume that if a player's wRC+ increased significantly after a season in which it was already fairly high, then that player would have a wRC+ somewhere between those numbers in the third season; i.e. regression. Thus, I tested this hypothesis.

I found a sample of 172 players, since 1970, who had at least 400 plate appearances in three consecutive seasons, a wRC+ between 110 and 130 in year one and a jump of at least 20 points in wRC+ in year two. I then examined those players' wRC+ in year three, and tabulated the results below:

Year 3 wRC+	% of Sample
Higher than Year 2	18.6%
Between Year 1 and Year 2	46.6%
Below Year 2	34.8%

The largest percentage of this sample had wRC+ in year three that was somewhere between their wRC+ in year one and year two (when the jump occurred), which backed my hypothesis. Based on this sample of hitters similar to Headley, it seems that his overall production will likely be somewhere between his 2011 and 2012 seasons.

One other drastic change that pops out when examining Headley's statistics is the jump from four home runs in 2011, to 31 in 2012. This home run boom in all likelihood was fueled by Headley's home run to fly ball rate (HR/FB) increasing from 4.3 percent in 2011 to 21.4 percent in 2012.

To test how much possible regression we should expect from Headley, in terms of home runs per fly balls, I ran a similar study to the one with wRC+. The result was surprising.

Headley's 17.1 percent increase in HR/FB is the largest since 2002 (when batted ball data became available), by far, for a player who had a HR/FB rate below five percent in the previous season and 400 plate appearances in each season.

The next largest jump for a player whose HR/FB was below five percent was by Chad Tracy from 2004-2005 when his HR/FB went from 4.2 percent to 14.8 percent.

Headley's home run increase is unprecedented; thus, my best guess for what his 2013 HR/FB will be is his career average rate, 10.2. A 10 percentage point reduction in HR/FB sounds like it would reduce production significantly, but 10.2 is not a horrible percentage and the reduction would also likely lead to more doubles for Headley.

Expect a very good season from Headley in 2013, but not nearly as great as 2012.

Are the Padres winning last offseason's big trades?

Before the 2012 season, the Padres made two fairly large trades involving young players. In the first, they sent starter Mat Latos to the Cincinnati Reds and in the second they moved Anthony Rizzo to the Chicago Cubs.

Looking at just the season after a trade is never the best way to evaluate which team won or lost the deal, especially when it involves young players. At the same time, we can ask whether, after one season, it looks like the Padres made the right moves last offseason.

Players Received	2012 WAR	Players Sent	2012 WAR
Yonder Alonso (Reds)	2.0	Mat Latos (Reds)	3.1
Yasmani Grandal (Reds)	2.7	Anthony Rizzo (Cubs)	1.8
Brad Boxberger (Reds)	-0.1	Zach Cates (Cubs)	--
Edinson Volquez (Reds)	1.3
Andrew Cashner (Cubs)	0.3
Kyung-Min Na (Cubs)	--
Total	6.2	Total	4.9

Seven of the players involved in those trades played in the majors last season; five for the Padres.

Those five players were more productive, in terms of wins above replacement, than the two players they gave up. Grandal was one of the Padres' most productive hitters last season and although he's suspended to start the 2013 season, the young catcher could end up being the key acquisition out of these two trades.

The Padres may have gotten the better end of the stick in 2012, but how will they win these deals overall?

Some would point to whether Latos continues to develop into an ace, or whether Cashner can becomes a valuable starter or stays a reliever.

I think the deciding factor will be whether Alonso becomes a better first baseman than Rizzo. The acquisition of Alonso originally made Rizzo expendable, and even if their return, quite specifically Cashner, for Rizzo turns out to be questionable, the Padres will still have won if Alonso becomes the better player.

Is the future, at least, bright?

The Padres do not have nearly enough pitching, especially starting, to compete for a playoff spot in 2013. Baseball Prospectus' PECOTA system currently projects the Padres to finish with 77 wins, just one more than last season. The 2013 Padres seem to be in store for another below .500 season near the bottom of the NL West. However, their organization is not without young talent.

ESPN's prospect expert, Keith Law, ranked the Padres' farm system as the best in baseball before the 2012 season. The system was flush with prospects on the cusp and good talent at the lower levels.

Coming into this season, Law and Minor League Ball's John Sickels both ranked the San Diego system as the sixth best, which is still very solid.

The fact that former prospects like Grandal, Alonso, Cashner and Kelly are now with the club is the major reason why the Padres' system has fallen a bit.

San Diego could have a strong young core in the coming seasons if some of its pitching prospects work out. Padres fans should be looking past 2013 optimistically, because better days may be on the horizon.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2013-03-20T07:09:15+00:00 Searching for home field advantage with PITCHf/x http://www.hardballtimes.com/main/article/searching-for-home-field-advantage-with-pitchf-x/ http://www.hardballtimes.com/main/article/searching-for-home-field-advantage-with-pitchf-x/#When:07:02:15 Derek Jeter steps into the batter's box in the bottom of the eight inning at Yankee Stadium. The count is full with two outs.

Jon Lester delivers a cut fastball that just crosses over the outside corner of the plate for strike three. However, the umpire rules ball four and Jeter takes his base. Lester is halfway to the dugout when he realizes that the umpire did not call the pitch a strike, and he cannot believe it.

Would the pitch have been called a strike at Fenway Park? Was the umpire biased when making the call because Jeter was playing in his home stadium?

This scenario is hypothetical, but I'm sure that as baseball fans we've come across situations like it countless times. Maybe the umpire simply missed the call, but fans of Boston would almost assuredly argue that the umpire was biased.

Luckily, with the availability of PITCHf/x data we can measure possible umpire home field bias more accurately.

Are umpires biased toward the home team on balls just off the plate?

My goal is to test whether umpires will actually rule in favor of the home team more often than the away team when calling balls and strikes on pitches that are close. The first step was to define a "close" pitch.

Based on PITCHf/x, the strike zone is typically defined as thus:

For right-handed batters: -1.03 < px (horizontal location) < 1.00 and (0.92 + batter_height*0.136) < pz (vertical location) < (2.60 + batter_height*0.136)‬

For left-handed batters: -1.20 < px < 0.81 and (0.35 + batter_height*0.229) < pz < (2.00 + batter_height*0.229)‬

I defined "close" pitches as those either three inches outside of the zone off either corner while within the strike zone vertically, or three inches above or below the zone vertically while within the strike zone horizontally.

All these pitches should be called balls, but umpires are obviously not perfect and thus will call these pitches strikes fairly often.

Next came the study. First, I looked at every pitch in the 2012 regular season that was either a called strike or ball located within the parameters that I defined as close, and found the number of these pitches that were called a strike.

To test for any sign of home field advantage, I found the percentage of strikes called on these types of pitches at each home field and separated them by the strike percentage when the home pitcher was on the mound and the strike percentage when the away pitcher was throwing. Below are the results:

Team	HomePerc	Away Perc	Gap at Home
Yankees	31.1%	23.2%	7.9%
Rays	23.0%	16.1%	6.9%
Padres	27.8%	21.3%	6.5%
Braves	27.2%	20.7%	6.5%
Brewers	26.1%	21.3%	4.7%
Giants	25.0%	21.1%	3.8%
Nationals	24.0%	20.2%	3.8%
Cardinals	21.7%	18.4%	3.3%
Astros	24.8%	21.9%	2.8%
Orioles	18.2%	16.7%	1.6%
White Sox	23.3%	22.1%	1.2%
Angels	22.3%	21.5%	0.9%
Diamondbacks	22.0%	21.4%	0.6%
Rockies	19.8%	19.2%	0.6%
Phillies	22.3%	21.8%	0.5%
Twins	16.0%	15.7%	0.4%
Indians	18.1%	18.0%	0.1%
Mets	22.3%	22.2%	0.0%
Red Sox	25.5%	26.1%	-0.6%
Rangers	18.1%	18.9%	-0.7%
Blue Jays	20.1%	20.8%	-0.8%
Athletics	18.1%	18.9%	-0.8%
Tigers	19.5%	20.4%	-0.8%
Dodgers	21.6%	22.5%	-0.9%
Reds	24.8%	25.8%	-1.0%
Pirates	18.4%	19.7%	-1.4%
Marlins	22.0%	23.5%	-1.5%
Cubs	20.8%	22.9%	-2.1%
Royals	14.9%	20.0%	-5.0%
Mariners	15.4%	24.9%	-9.6%

When these results are taken at face value it seems that the New York Yankees got the benefit of almost 8 percent more strikes called on close pitches than their opposing team when playing at Yankee Stadium. Teams such as the Atlanta Braves and Tampa Bay Rays also seemed to have benefitted from some type of home field advantage.

The only issue with interpreting the results in this way is the fact that these results can in no way be taken at face value.

According to research done by Max Marchi at Baseball Prospectus, Brian McCann (Braves), Jose Molina (Rays), Russell Martin (Yankees) and David Ross (Braves) ranked as four of the top six catchers in terms of catch framing over the 2008-11 seasons.

Would one be crazy to assume that much of what these results were picking up on was, in fact, each catcher's ability to frame pitches that were close rather than umpires actually favoring the home team?

If people were not convinced, they would need to look no further than the bottom of this table and see that the Seattle Mariners had almost 10 percent fewer strikes called with their pitchers on the mound at home than their opponents. Last year, the Mariners' catching corps was Jesus Montero, Miguel Olivo and John Jaso, none of whom are known for their ability as catchers.

Also, it is possible that certain teams have pitchers who consistently throw pitches that are in general "closer" to the plate, within this close definition, than other pitchers.

Thus, I adjusted these results by attempting to control for catcher framing, looking at the strike percentage on close pitches for each team when it was playing on the road. I list the results below:

Team	Road Strike Perc
Braves	26.8%
Brewers	26.1%
Mets	25.0%
Yankees	24.9%
Cardinals	24.3%
Giants	24.3%
Rays	24.1%
Reds	24.1%
Astros	23.1%
Blue Jays	22.5%
White Sox	22.5%
Red Sox	22.1%
Nationals	21.7%
Padres	20.9%
Phillies	20.7%
Angels	19.7%
Orioles	19.2%
Diamondbacks	19.2%
Marlins	18.9%
Cubs	18.4%
Rangers	17.8%
Dodgers	17.7%
Royals	17.2%
Rockies	17.0%
Athletics	17.0%
Pirates	16.7%
Tigers	15.9%
Twins	15.5%
Mariners	15.5%
Indians	14.0%

Not surprisingly, these numbers jibe pretty well with past studies on which catchers are adept at framing pitches.

I found the average strike percentage on close pitches for each team on the road (20.4 percent). This average was then subtracted from each team's road strike percentage to create a quasi-expected home field strike percentage gap.

Expected home field strike gap = road strike percentage - Lg. Avg. road strike percentage

In theory, this expected home field strike gap should be close to the gap found in the first piece of this study (the gap between a team's strike percentage on close pitches at home and that of its opponents.)

Below, I compared each team's actual gap in strike percentage at home and its expected home field gap:

Team	Actual Home Field Gap	Expected Home Field Gap	Adjusted Home Field Gap
Indians	0.1%	-6.5%	6.6%
Padres	6.5%	0.4%	6.0%
Twins	0.4%	-4.9%	5.3%
Rockies	0.6%	-3.4%	4.0%
Tigers	-0.8%	-4.5%	3.7%
Yankees	7.9%	4.5%	3.4%
Rays	6.9%	3.7%	3.1%
Orioles	1.6%	-1.2%	2.8%
Athletics	-0.8%	-3.5%	2.7%
Nationals	3.8%	1.3%	2.5%
Pirates	-1.4%	-3.7%	2.3%
Rangers	-0.7%	-2.7%	2.0%
Diamondbacks	0.6%	-1.2%	1.9%
Dodgers	-0.9%	-2.8%	1.9%
Angels	0.9%	-0.7%	1.6%
Phillies	0.5%	0.3%	0.2%
Astros	2.8%	2.7%	0.2%
Braves	6.5%	6.3%	0.1%
Marlins	-1.5%	-1.5%	0.0%
Cubs	-2.1%	-2.1%	0.0%
Giants	3.8%	3.9%	0.0%
Cardinals	3.3%	3.9%	-0.6%
Brewers	4.7%	5.6%	-0.9%
White Sox	1.2%	2.1%	-0.9%
Royals	-5.0%	-3.2%	-1.8%
Red Sox	-0.6%	1.7%	-2.3%
Blue Jays	-0.8%	2.1%	-2.9%
Mets	0.0%	4.6%	-4.5%
Mariners	-9.6%	-4.9%	-4.6%
Reds	-1.0%	3.7%	-4.7%

The final column (Adjusted Home Field Gap) is my best attempt at controlling for a home field advantage on balls and strikes for last season. The average strike percentage for home teams on close pitches was 0.9 percent higher than expected for 2012.

This possible home field advantage effect sounds small, but is it?

The average number of close pitches thrown by a home team in 2012 was 1,237. Thus, the ~one percent home field advantage resulted in only 11 more strikes per team. According to a rough estimate that comes from Tom Tango, the run value of a called strike over a ball is 0.16 runs.

Thus, based on this study the average home field advantage for 2012 on close pitches was less than 2 runs per team. The effect of home field advantage seems to be very small, even at the extreme.

According to this study, the Cleveland Indians were the worst team in terms of strike percentage on close pitches on the road. Yet at home, they received about the same amount of strike calls as their opponent.

These facts resulted in 82 more strikes being called on close pitches than expected for the Indians at home; which is roughly equivalent to 13 runs. This number is larger in comparison to what was found at the league level, but 13 runs is worth just over one win over the course of the entire season and of course, Cleveland's case is at the extreme.

My conclusion, based on this study, is that there seems to be some effect of home field bias on umpires calling balls and strikes, but the effect is not large enough to cause any real significant difference over what is expected during one season of play.

A few brief shortcomings

Although this study came back to show that teams did not seem to be receiving any really significant home field advantage on balls and strikes, that doesn't mean home field advantage does not exist in baseball or that umpires are not subconsciously biased towards the home team.

I say this because this study was in no way perfect.

For one thing, the sample may have not been the best. I tested only one season in this sample and I also looked at only pitches that were outside the strike zone. There could be some home field bias on pitches in the strike zone that are called balls. Also, I did not control for pitch type or for umpire, as each umpire's strike zone is different.

Two critical assumptions when calculating the expected home field gap also may have been flawed.

First it likely was unsafe to assume that pitchers' and catchers' playing times for each team were evenly distributed between home and road games. Also, teams do not have an even proportion of road games across all of the stadiums in baseball. This is an issue because it has been shown that PITCHf/x cameras have biases that vary by stadium. The flaws in both of these assumptions could have resulted in some bias within this study.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2013-03-06T07:02:15+00:00 More on standard deviation and ERA estimators http://www.hardballtimes.com/main/article/more-on-standard-deviation-and-era-estimators/ http://www.hardballtimes.com/main/article/more-on-standard-deviation-and-era-estimators/#When:07:10:15 pFIP). The statistic was a modified version of the original Fielding Independent Pitching statistic(FIP) that was meant to predict future performance rather than describe performance.

Predictive FIP was highly correlated with future ERA (or runs). Unfortunately though, as I pointed out earlier this month, the spread of the projections for individual pitchers was much smaller than any other commonly accepted ERA estimator or projection system.

The pFIP equation, as it stood, led to ERA projections that were tightly centered on the mean ERA. This leads to a really high overall correlation, but fairly useless individual projections that are not close to a real reflection of a pitcher's true talent level.

Instead of scrapping the metric completely, I decided to manually force the spread of projections to be larger.

This meant putting more weight on the measures of a pitcher's skill (homers, walks, strikeouts) and in turn taking away from the component that regresses the projection to the mean, the constant term.

In the comments of my latest piece, MGL gave some great insight into the size of spread I should be looking for:

Honestly, the best thing to do is to do a rough estimate. For example, my experience in doing projections for 20-some odd years and in closely watching and analyzing baseball for almost 30 years is that true talent is right around league average plus or minus 1.5 runs, which happens to be a SD of true talent of around .5 runs! So you did actually get around the right answer—sort of by semi-accident though!

Thanks to MGL, I set out with the goal of creating a pFIP equation that would result in something similar to a true talent of the population spread of ERA projections, with a standard deviation of ~0.5.

I began with the original pFIP equation (ERA version):

pFIP = (18.5*HR + 6*BB - 8*K)/TBF + 4.75

At first, I simply guessed and checked by multiplying the different weights by random constants, to see what combination of weights would lead to a larger spread, yet still be fairly similar to pFIP's original weights.

Quite interestingly, this is the new pFIP equation that I came up with:

pFIP = (20*HR + 10*BB - 10*K)/TBF + 4.60

This new equation is fairly similar to the previous one; however, the skill components get slightly more weight. What made this equation both interesting and ironic was an exchange that I had with Tom Tango, the inventor of the original FIP statistic.

When I first came up with the idea of pFIP, I emailed Tango to get his thoughts on the pursuit. Here was Tango's original response:

(FIP's) 13,3,2 (weights) are "descriptive," if by that you mean it correlates Year T BB, SO, HR to Year T Runs.

If you wanted BB, SO, Hr to be "predictive" (correlating those components in year T to year T+1 runs), they would have different weights. My guess is that it would be something like 2,1,1 or maybe 3,1,1 for HR, SO, BB, respectively. This is because you would regress HR the most and K the least.

Tango essentially came up with the strongest version of the pFIP equation off the top of his head, months ago. It is amazing to me, after months working with the statistic, that he was understood what pFIP should look like almost instantaneously.

Another interesting note about this version of the metric is how similar it is to another powerful ERA estimator, kwERA:

pFIP = 4.60 + 10*(2*HR +BB - SO)/BF

kwERA = 5.40 - 12*(SO-BB)/BF

kwERA is elegant in that it is both simple and a very powerful predictor of future ERA. pFIP was originally modeled after kwERA, but was a more powerful predictor, because it included home runs within the equation.

This new pFIP equation is right in line with kwERA in that it is extremely simple, and more importantly now has a much wider spread.

Does this new equation still have the same strong predictive power?

In the piece that originally introduced pFIP, I showed that over the years 2004-2012, for pitchers who threw at least 120 innings in Year X and at least 100 innings in Year X+1 that pFIP was more highly correlated with future ERA than other established estimators, such as kwERA, FIP, xFIP and SIERA.

My goal in this article is to attempt to duplicate those results and show that pFIP is still the strongest predictor, even with the wider spread of projections.

I modified the test slightly, though, by removing kwERA and adding in ERA as a baseline, as well as, also lowering the minimum number of innings in Year X from 120 innings to 100 innings.

Below, I display the results:

Predictor	Correlation (r)	STDEV
pFIP	0.447	0.516
SIERA	0.424	0.583
FIP	0.423	0.690
xFIP	0.418	0.572
ERA	0.367	0.858

pFIP very clearly was still the strongest predictor (most highly correlated with) of next season ERA.

I listed the standard deviation of each estimator to give an idea of the spread of each metrics estimations. pFIP's standard deviation is still the smallest of the predictors, which benefits its correlation, but at the same time it is now much closer to the others in terms of spread, and right around the 0.5 mark that I was shooting for.

I was pleased to see these results, because after my last piece, I feared that pFIP's strength may have rested entirely upon the fact that the spread of projections was extremely tight. However, these results seem to indicate that the methodology behind pFIP is both powerful and possibly a very good indicator of true talent level.

Is pFIP better than a projection system?

I've asked this question before and have changed my mind a few times on the subject, but I hope this will serve as a final-ish answer to that question.

In an earlier piece, I showed that for pitchers who threw at least 100 innings in 2011 and pitched in at least five games in 2012, pFIP had a higher correlation with 2012 ERA than the 2012 ZiPS projections.

That test was very obviously a small sample, it used the old pFIP equation and I compared pFIP to only one projection system. Thus, I decided to put this new formula to the test with a broader sample.

I found a sample of pitchers (n=354), who threw at least 100 innings in the year prior and started at least five games in the next season, for the years 2010-12. I tested pFIP against three projection systems: ZiPS, Bill James and Marcel.

The results were as follows:

Predictor	Correlation (r)	STDEV
ZIPS	0.386	0.630
Bill James	0.324	0.484
pFIP	0.314	0.517
Marcel	0.277	0.496

A good place to start when looking at these results is the standard deviation of each projection.

Bill James' and Marcel's spreads are similar to that of pFIP and right around our estimation of the "true talent" spread. It is surprising, at least to me, that ZiPS had the strongest correlation with ERA while also having the largest spread of projections. For all intents and purposes, that makes ZiPS the clear winner.

pFIP was able to pass above the simple Marcel system's baseline, but was not as strong a predictor as the other two systems tested.

Based on these results, if one were to ask which ERA estimator projects future ERA the most successfully I'd feel fairly safe to say pFIP.

However, if one really wanted to project future ERA with the most success at this point my answer would be to use a projection system.

I created a Google Doc with pFIP's 2013 ERA projections alongside the projections of four other systems (Marcel, Oliver, ZiPS and Steamer) for any pitcher who threw at least 100 innings in 2012.

For what it is worth, I listed the spread of each systems's projections below:

Predictor	STDEV
pFIP	0.463
Marcel	0.495
Steamer	0.370
Oliver	0.473
ZiPS	0.623

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2013-02-20T07:10:15+00:00 Standard deviation and ERA estimators http://www.hardballtimes.com/main/article/standard-deviation-and-era-estimators/ http://www.hardballtimes.com/main/article/standard-deviation-and-era-estimators/#When:06:01:15
One of the main reasons I became interested in studying baseball statistics was fielding independent pitching (FIP) and ERA estimators. Over the past few months, my interest in ERA estimators turned into an obsession. During that time, I developed predictive FIP (pFIP), an ERA estimator of my own.

In almost every test that I ran, pFIP came out ahead (higher correlation) of other more established ERA estimators. This lead me to believe that pFIP was the best ERA estimator currently available, but that in no way meant that the metric was without its own flaws.

Below I listed the standard deviation for pFIP, FIP, ERA and two other ERA estimators (SIERA and xFIP), for pitchers who threw at least 100 innings in a season for the years 2010-12:

Metric	STDEV
ERA	0.862
FIP	0.652
SIERA	0.510
xFiP	0.502
pFIP	0.387

Unsurprisingly, on a single season basis, ERA has the widest distribution, while pFIP has the tightest.

Quite honestly, the fact that pFIP's standard deviation is significantly smaller than xFIP and SIERA (which are known for typically having small standard deviations) was a cause for concern.

In Colin Wyers' piece on SIERA and other ERA estimators this issue is discussed in great detail:

In a real sense, that’s what we do whenever we use a skill-based metric like xFIP or SIERA. We are using a proxy for regression to the mean that doesn’t explicitly account for the amount of playing time a pitcher has had. We are, in essence, trusting in the formula to do the right amount of regression for us. And like using fly balls to predict home runs, the regression to the mean we see is a side effect, not anything intentional.

Simply producing a lower standard deviation doesn't make a measure better at predicting future performance in any real sense; it simply makes it less able to measure the distance between good pitching and bad pitching. And having a lower RMSE based upon that lower standard deviation doesn't provide evidence that skill is being measured. In short, the gains claimed for SIERA are about as imaginary as they can get, and we feel quite comfortable in moving on.

My understanding of Colin's argument is that metrics like xFIP and SIERA "crudely" regress each pitcher to the mean which would lead to a higher correlation (lower RMSE), but at the same time may not be an accurate measure of a pitcher's true talent level.

It is evident that of the four ERA estimators discussed in this piece, pFIP has the largest regression to the mean for each individual. This fact brings me to a question whose answer I've found myself switching sides on countless times.

What is the point of an ERA estimator?

There are two answers to that question that could hold serious weight in an argument:
{exp:list_maker} To be the best at predicting (highest correlation with) future ERA
To be the best representation of a pitcher's true talent level{/exp:list_maker}
It would be nice if an ERA estimator came along that could fulfill both of those requirements, but I would argue that that is not the case.

For the first (high correlation with future or next season ERA), the estimation should be seriously regressed to mean. But when one is attempting to estimate a pitcher's true talent level, should that regression be as harsh? At lower innings pitched totals there should be some, but not nearly as strong as when the goal is to simply predict future ERA.

My main issue with the true talent level idea for ERA estimators is how difficult it actually is to calculate that number. An ERA estimator that reflected a pitcher's skill should be able to account for all of the possible factors within the pitcher's control and weed out all of the other factors around the pitcher correctly. The problem is that it is nearly impossible to do.

In the extreme, relievers throw so few innings on a per-season basis that by the time they throw enough innings for us to get a fair idea of their ERA talent, years will have passed. And in all likelihood, their true talent will have changed. Even for starters who throw more innings, their true talent level is tough to decipher out of all the different factors that go into run prevention.

The consensus at this point is that the estimator with a standard deviation as wide as one's true talent ERA and a high correlation with future ERA is the best at measuring true talent. However, there are issues with this approach too.

Pinning down an exact number for the standard deviation of a pitcher's true talent ERA is difficult. This issue was raised in a FanGraphs Community Post by Steve Staude. He showed that from 200 to 1,000 innings pitched, the standard deviation of ERAs range from 0.8 down to 0.5, as the innings increase.

I think most would agree that true talent does not reveal itself at the 200-inning mark, but then where? 500? 750? 1,000?

Most pitchers never reach 1,000 career innings; many do not reach 500. It also takes most pitchers at least three seasons to reach 500 innings, and it seems reasonable that an individual's talent level could change significantly over the course of those seasons.

For a moment though, let's ignore that and look to Wyers' original piece to see that ERA true talent seems to be revealed somewhere between 400 and 500 innings. According to Staude's study, the standard deviation of ERA between 400 and 500 innings ranges from about 0.65 to 0.6; thus, it would make logical sense that an ERA estimator with a high correlation with future ERA and a standard deviation of around 0.6 or 0.65 would be the best true talent estimator.

Interestingly, if we look at the standard deviation that I found for FIP in this article, it falls right in line with that logic. FIP has a higher correlation (in small to medium samples) with future ERA than ERA and it has a standard deviation that is similar to "true talent" ERA. This assumption also falls in line with a trend we often see: A pitcher's career FIP lines up fairly closely to his career ERA.

The fact that most logic would lead one to conclude that FIP is best true talent ERA estimator we have available fascinates me.

Why? Because the structure of FIP is in no way meant to predict future ERA.

FIP is commonly used in that fashion because it does a fairly good job of predicting future ERA, but that is not the statistic's purpose. FIP is meant to be a describer of a pitcher's performance that is scaled to look like ERA. It's best described as a what a pitcher's ERA should have been. That type of description may make FIP sound similar to a true talent evaluator, but it is in no way correlated or meant to describe future performance.

This idea brought about the birth of pFIP.

pFIP regresses the components of FIP (strikeouts, walks, home runs) to predict future performance rather than describe of past performance. In plain English, that idea sounds great and interestingly the math works out, too.

FIP's more volatile components (walks and homers) receive a fair amount of regression, while strikeouts (the least volatile) receive little or no regression and these regressions result in a stronger correlation with next season ERA than when simply using FIP.

But is pFIP really saying anything about a pitcher's true talent level?

I would argue that it may give one some of indication of a pitcher's talent, but it is not a true talent evaluator. If you look at either the pFIP equation, the standard deviation of pFIP or an individual's numerical pFIP, what the statistic actually does becomes very clear.

Essentially, pFIP starts each individual's ERA projection at the same point (the mean ERA) and then moves each number slightly away from the mean based on the player's individual peripherals. This strategy works great when your goal is to predict with the highest rate of success, but it does not give you a great idea of a pitcher's actual true ERA or skill.

Thus, when one decides to evaluate pFIP as a statistic one must return to our original question: What is the purpose of an ERA estimator?

pFiP is essentially useless if you'd like to evaluate a pitcher's talent level, but if your goal is to predict next season's ERA then pFIP will serve you well. However, if predicting future ERA is the only real purpose of pFIP then is there any real reason for the statistic?

Projection systems are a very real thing, and their goal (at least from what I understand) is to do exactly what pFIP does; project future performance. I've shown before that pFIP is fairly comparable to projection systems when looking at overall correlation with next season runs (or ERA). Although, I'm fairly certain that simple correlation with the next season is notthe best way to test how well a projection system works, let's say for a moment that it is.

Is pFIP really better or equivalent to a projection system?

The short answer is quite obviously no, but the evidence behind that assessment is fairly educational.

I'm not saying this is true, but consider a fantasy scenario where pFIP has exactly the same correlation with future ERA as an average projection system. How would we test which one was actually doing a better job? A good starting point would be to consider the standard deviations of the projected ERAs.

I looked at the standard deviations of ERA projections for three projection systems (Marcel, Bill James and ZIPS) for the years 2010-2012 for pitchers who were projected to have at least 100 innings in that season:

Projection System	STDEV
Bill James	0.514
Marcel	0.520
ZIPS*	0.657

*ZIPS does not project playing time, so Marcel's playing time projections were used for the pitchers in the sample.

Under the assumption that pFIP and the projection systems have similar or equivalent correlations, it would seem that projection systems do a better job at really projecting future performance/skill as their distributions are wider.

This should not really be too surprising, as projection systems take a great deal more information into account than pFIP. Also projection systems are even more useful as they project playing time and counting stats as well as the rate stats (like ERA, FIP, etc.)

This all brings me back again to my original question: What is the point of an ERA estimator? Or to be more clear, if we have projection systems, then what is the point of ERA estimators?

I can think of only two answers to that question.

The first is that some people don't trust or find utility in projection systems and thus find sanctuary in using much simpler ERA estimators, which are still fairly predictive and easier to understand. The second is that ERA estimators should be a reflection of a pitcher's true talent level, which, of course, is almost impossible to define.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2013-02-06T06:01:15+00:00 2012 minor league leaders in predictive FIP (Part 2) http://www.hardballtimes.com/main/article/2012-minor-league-leaders-in-predictive-fip-part-2/ http://www.hardballtimes.com/main/article/2012-minor-league-leaders-in-predictive-fip-part-2/#When:05:23:15 first part of a series of 2012 minor league leader boards. In that piece, I ranked pitchers who threw at least 80 innings in Triple-A and Double-A in 2012, based on my predictive FIP (pFIP) statistic.

My original thought was to do the same for the lower levels of the minor leagues, but then I ran across an interesting interview that changed my mind.

About a month ago, former Hardball Times writer Mike Fast, currently an analyst for the Houston Astros, gave an interview to the "What the Heck, Bobby?" blog. You can read the full interview here. But for the purposes of this article, I'd like to focus on a specific quote that hit home with me and is the most relevant to this piece:

One thing that I would be careful with, with those advanced pitching metrics … just looking at batted ball outcomes and whether it fell in for a hit or not is sort of a poor way to get an idea over the course of a season of a pitcher’s talent. That’s pretty true at the major league level. It’s less true in the minor leagues.

The ability to prevent hits is a pretty good indicator of talent in the minor leagues. The farther you get away from the major leagues the more important that is. One way to think of it is that, at the major league level, pitchers who get there have sort of been screened for their ability to prevent good contact, so if you’ve got a guy who gets hit hard, he’s never going to make the majors. The guys who come to the majors are all at least pretty good at that. And there may be some differences between them but they’re all pretty good at preventing good contact because their stuff moves or they throw hard or they mix their pitches well, know how to locate, whatever. They’ve got some skills in that. And that’s less true particularly in the lower minors.

If you have a pitcher in the low minors that just gave up a ton of hits, I don’t know that I’d be so quick to ascribe that to 'Oh, he’s got bad luck or he had bad fielders behind him.' That might be true, but it’s also quite likely that the batters are just saying this guy’s stuff is not that hard to deal with.

There's a lot in this quote to digest. I think Fast makes a really interesting and convincing point that I had never heard before.

I typically work with major league statistics andtalent evaluation, so I typically consider hit prevention to have more to do with luck than defense or actual skill. However, Fast's argument is that all major league pitchers have some hit prevention skills, which means in the minors (especially in the lower levels) having the ability to suppress hits is important.

What Fast said made me think that pFIP leader boards for the levels below Double-A may not be relevant or informative at all for readers.

In the next paragraph of the interview, Fast suggests that using a combination of strikeout rate and WHIP (walks + hits/IP) would be a good place to start when using statistics to evaluate lower level pitchers.

Using K-rate and WHIP in tandem made me think of an ERA estimator that is based solely on strikeouts and walks known as kwERA . Carson Cistulli's minor league kwERA leader board in the THT Annual was the inspiration for this series; kwERA has been shown to be a powerful predictor at the major league level. However, kwERA, considering only strikeouts and walks, ignores a pitcher's hit prevention skills.

Based on Fast's suggestion, I decided a version of kwERA that included the number of hits a pitcher gave up could be interesting for lower level minor league leader boards.

Below is the typical kwERA equation and the modified version used in this exercise that includes hits allowed:

kwERA = 5.40 - (12*(K-BB)/BF))

kwhERA (with hits) = 5.40 - (12*(K-BB-H)/BF))

For this piece, I listed the top five pitchers in terms of kwhERA at each lower level (minimum 80 innings pitched for High-A/Single-A and minimum 50 innings pitched for Short Season-A/Rookie), in 2012. I also listed each pitcher's pFIP and where it ranked in the league for a comparison in the differences between the two metrics.

Note: kwhERA was made up specifically for this exercise and has not been shown to have any predictive ability on a player's ability in the majors or chance of reaching the big league level. The same can be said for pFIP. I'd also like to mention that many very smart baseball people do not use statistics at all at or below the Double-A level.

High-A

Pitcher	2013 ORG	Age	ERA	kwhERA	pFIP (rank)
Adam Morgan	Phillies	22	3.29	2.69	3.06 (1st)
Wilmer Font	Rangers	22	4.11	2.89	3.36 (3rd)
Tyler Wilson	Orioles	22	3.49	2.99	3.51 (9th)
Jerry Sullivan	Padres	24	4.2	3.04	3.50 (8th)
Burch Smith	Padres	22	3.78	3.07	3.43 (5th)

All five of these pitchers also ranked in the top 10 in pFIP; which was a theme across the four leader boards in this piece, as only one pitcher failed to reach the top five in kwhERA and top 10 in pFIP.

Phillies left-hander Adam Morgan ranked first in both kwhERA and pFIP. He was also recently listed as the Phillies' third-best prospect by Baseball Prospectus. In 2012, Morgan reached Double-A, where he performed well, and he should begin the 2013 season there.

Rangers right-hander Wilmer Font was dominant in High-A ball in 2012. which led to a promotion to Double-A and eventually a major league debut. John Sickels of Minor League Ball wrote a glowing prospect review of Font's ability late in the 2012 season.

Tyler Wilson, a righty in the Orioles system, is not known as much of a prospect. That idea may change during the coming years; as Wilson struck out 114 batters while walking just 19 in 111 innings last season.

The next two players on this list are members of the Padres organization and, like Wilson, have not been known as prospects.

Jerry Sullivan struggled as a starter in 2011, but he was very effective coming mainly out of the bullpen in 2012. Sullivan is a little old for never having reached the Double-A level, but he'll be a non-roster invitee to spring training in 2013.

Last year was Burch Smith's first full professional season and he was very impressive. The righty struck out 137 batters and walked just 27 in over 125 innings. Sickels suggested that Smith could end up as a mid-rotation guy someday.

Single-A

Pitcher	2013 ORG	Age	ERA	kwhERA	pFIP (rank)
Clayton Blackburn	Giants	19	2.54	2.76	2.97 (1st)
A.J. Cole	Nationals*	20	2.07	2.89	3.27 (7th)
Noah Syndergaard	Mets*	19	2.60	2.89	3.02 (2nd)
Jose Mavare	Rangers	22	3.57	2.92	3.21 (5th)
Mason Radeke	Indians	22	3.29	3.02	3.47 (11th)

Again for this list, the pitcher who led in kwhERA also led in pFIP: FanGraphs recently ranked 19-year-old righty Clayton Blackburn as the number two prospect in the Giants' system. Blackburn struck out 143 batters and walked only 18 while giving up just three home runs in over 130 innings on the hill.

Quite interestingly, the next two pitchers on this list were both traded this offseason.

A.J. Cole was originally drafted by Nationals and was one of the highest-ranked prospects in their system before he was included in the package Washington sent to Oakland in exchange for Gio Gonzalez. Cole put up impressive numbers for Oakland's Single-A affiliate in 2012. He was recently traded back to Washington as part of a three-team deal that included Seattle, and looks to again be one of the top prospects in the Nationals system for 2013.

Noah Syndergaard was involved in the highest-profile trade this offseason, part of the prospect haul the Mets received from the Blue Jays in the R.A. Dickey swap. Syndergaard was originally ranked as Toronto's number two prospect for 2013, but is now slotted as New York's number three prospect in a loaded system.

Jose Mavare is the second member of Texas' organization to find his way into this article. Mavare spent this entire season in the bullpen and projects to pitch out of the 'pen going forward. He has only two career professional starts.

Mason Radeke split time between starting and relief roles in the Indians system last season. He struck out more than a batter an inning at Single-A before his eventual promotion to Double-A. Radeke was recently ranked just outside the top 20 Indians prospects.

Short-Season A

Pitcher	2013 ORG	Age	ERA	kwhERA	pFIP (rank)
Luis Mateo	Mets	22	2.45	2.49	2.81 (1st)
Rainy Lara	Mets	21	2.91	2.58	3.15 (5th)
Javier Avendano	Blue Jays	21	1.27	2.82	3.02 (2nd)
William Cuevas	Red Sox	21	1.40	3.04	3.30 (8th)
Aaron West	Astros	22	2.04	3.09	3.31 (9th)

Luis Mateo is probably not brought up enough in discussions of the Mets system, which is a testament to wealth of pitchers they boast. Mateo ranked first in both pFIP and kwhERA on this list and was ranked as the number one prospect on the rise in the Mets' system going into 2013. Over 70 innings of work in 2013, he struck out 85 batters while walking nine and giving up only two home runs.

Rainy Lara is also a member of the Mets system. He is not thought of nearly as high as Mateo though; ranking as their 35th best prospect. Lara, like Mateo, struck out a ton of batters in 2012 and it'll be interesting to see if he can continue to do so as he moves up their system.

Javier Avendano, mainly as a starter, had a shiny ERA (1.27) to go along with strong peripherals at Low-A last season and was named the MVP of the Vancouver Canadians. Toronto then promoted Avendano to Single-A, where he continued to put up strong numbers out of the bullpen in 30.1 innings of work (1.78 ERA, 3.60 kwhERA, 3.23 pFIP). I have yet to find Avendano show up on any list of Toronto's top prospects, which makes me guess that his stuff may not project at the big league level.

William Cuevas split time between the bullpen and rotation for the Red Sox' affiliate, Lowell. Cuevas kept walks and home runs low, while striking out almost a batter per inning last season.

The Houston Astros drafted Aaron West in the 17th round of the 2012 draft. The righty did not disappoint. West started 12 games (61.2 innings pitched) and struck out 59 batters while walking just six and giving up only three home runs. West was recently ranked as the Astros' 22nd best prospect going into 2013.

Rookie

Pitcher	2013 ORG	Age	ERA	kwhERA	pFIP (rank)
Sam Selman	Royals	21	2.09	2.13	2.44 (1st)
Spencer Patton	Royals	24	6.32	2.72	3.02 (2nd)
Miguel Sulbaran	Dodgers	18	2.51	2.81	3.07 (3rd)
Thomas Lee	Cardinals	22	4.03	3.00	3.26 (5th)
Yorfrank Lopez	Tigers	21	2.28	3.02	3.26 (6th)

Two members of the Royals organization led Rookie ball in both kwhERA and pFIP in 2012. The first, Sam Selman, was drafted in the second round of 2012 and was dominant in his first taste of professional baseball; striking out 89 batters in 60.1 innings. Baseball Prospectus ranked Selman as the number one prospect on the rise in the Royals' system and FanGraphs ranked him as their eight best prospect for 2013.

Spencer Patton spent his second straight season in Rookie ball and will turn 25 next month without ever playing above that level. He struck out a ton of hitters last year (84 in 57 innings), but his .435 BABIP and 6.32 ERA are serious signs of concern.

The Dodgers' lefty Miguel Sulbaran was extremely good in Rookie ball before his eventual promotion to Single-A last season. He'll turn 19 before the start of the season and you could see his name fly up the charts on any prospect list for Los Angeles.

Thomas Lee was signed by the St. Louis Cardinals in 2012 as an undrafted free agent. There is almost no information on him, other than that he went to Sonoma State University and that he struck out 57 batters in 51.1 innings last season.

Yorfrank Lopez spent his third straight season in Rookie ball in 2012. The right-hander split time between relief and starting appearances and put up some impressive strikeout numbers.

Other top prospects and Google Docs

Two names, Dylan Bundy and Jose Fernandez, were nowhere to be found on these leader boards.

Bundy, quite famously, is the top prospect in the Orioles' system and Fernandez is currently the top prospect in the Marlins' system, both among the top rated prospects in baseball.

Bundy did not throw enough innings at any particular level to qualify for these boards as he flew through the Orioles' system and made his major league debut in 2012. I would still like to note that across a total of 103.2 minor league innings Bundy posted a 2.75 kwhERA and a 3.08 pFIP.

Fernandez missed qualifying for the Single-A leader board by just one inning because of his promotion to High-A. Fernandez pitched a total of 134 innings across those two levels. His kwhERA was 2.65 and his pFIP was 2.81.

I expect both of these pitchers to become household names in future years. Bundy may have already reached that status.

Below I linked Google Docs with a list for each level. I hope the readers found this at least half as interesting and informative as it was for me:

pFIP and kwhERA numbers for High-A
pFIP and kwhERA numbers for Single-A
pFIP and kwhERA numbers for Short Season A
pFIP and kwhERA numbers for Rookie

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2013-01-23T05:23:15+00:00 2012 minor league leaders in predictive FIP (Part 1) http://www.hardballtimes.com/main/article/2012-minor-league-leaders-in-predictive-fip-part-1/ http://www.hardballtimes.com/main/article/2012-minor-league-leaders-in-predictive-fip-part-1/#When:06:55:15
I find the study of scouting, projecting and evaluating prospects to be absolutely fascinating. It's just not my area of expertise.

However just recently, I found myself inspired to delve into some uncharted territory.

My inspiration came from the fantastic 2013 Hardball Times Baseball Annual. Within the book's pages are a series of interesting (different) leader boards provided by FanGraphs' Carson Cistulli.

One of these is a 2012 minor league kwERA leader board. kwERA (or strikeout and walk ERA) is an ERA estimator based solely on strikeouts and walks; which I've been fascinated with for some time.

I've done a lot of work with the predictive strength of kwERA, and that work led to the development of my own ERA estimator, predictive FIP (pFIP).

In the THT Annual, Cistulli grouped all minor league levels together into an all-encompassing top-10 (minimum 550 batters faced). Cistulli was confined to just two pages in the annual, which left some room for improvement/expansion off of his idea.

Thus, I've decided to make some leader boards for each level of the minors (Triple-A through Rookie ball) for the 2012 season, with two little tweaks on Cistulli's original methodology:

First, I used pFiP rather than kwERA as my metric of choice. Second, I used a minimum of 80 innings pitched (~320 batters faced) rather than 550 batters faced, to get a broader scope.

Disclaimer: Predictive FIP works really well at projecting ERA at the major league level. I have yet to test how well it works when translating minor league numbers into major league equivalents. It's something I have been working on, but it is a daunting task. These leader boards are meant solely to be interesting and to possibly shed light on prospects that could be underrated.

Also, pFIP is an ERA estimator that ignores balls in play; which works great at the major league level, but the further away you go from that the more valuable batted ball information becomes. But I'll delve into that issue in much greater detail in Part 2 of this series. Please take all of the leader boards with a grain of salt, but at the same time, enjoy!

Triple-A

The minor leagues' highest level is broken up into just two leagues, the Pacific Coast League (PCL) and the International League (IL). It would be an understatement to say the run environments in these two leagues are distinct. The PCL boasts some parks that really promote offense, such as Colorado Springs, Albuquerque, Salt Lake, Tucson and others. Given the differing environments, I broke the Triple-A top 10 into two separate top-fives for each league.

Notes: The individual home run numbers for each pitcher were adjusted using multipliers posted at Baseball Think Factory for the years 2009-2011. Also, for reference: For pitchers who threw at least 80 innings at Triple-A in 2012, the average pFIP was 4.31.

PCL

Pitcher	2013 ORG	Age	ERA	pFIP
1. Trevor Bauer	Indians	21	2.85	3.49
2. John Ely	Astros	26	3.20	3.52
3. Tyler Lyons	Cardinals	24	4.28	3.60
4. Yusmeiro Petit	Giants	27	3.46	3.63
5. Wade LeBlanc	Marlins	27	3.74	3.69

Interestingly, the PCL's two leaders in pFIP were both traded this offseason.

Mega-prospect Trevor Bauer posted the top pFIP in the PCL, while also spending time in Double-A and the majors. Bauer posted a fantastic pFIP at Double-A (3.27) before being promoted to Reno.

In four major league starts, he was hampered by control issues, but Bauer, 21, isn't the first top prospect to have trouble finding the strike zone in the bigs at such a young age. Surprisingly, Arizona moved the right-hander to Cleveland this offseason in a three team swap that included Cincinnati.

Also this offseason, John Ely was traded by the Los Angeles Dodgers to the Houston Astros for Rob Rasmussen. Despite pitching in a hitter-friendly Albuquerque, Ely was able to post really good peripherals and had great results last season. It will be interesting to see if Ely gets a chance to have success in the Astros' rotation next season. He has 115.1 career major league innings with a 5.70 ERA and 4.46 FIP

Tyler Lyons is a left-handed starter in the St. Louis Cardinals system who has moved quickly up the ladder after his 2011 pro debut. In 2012, Lyons not only posted a 3.60 pFIP in 88.1 Triple-A innings, but he also had an above-average pFIP (4.00) in 64.1 Double-A innings. Viva El Birdos ranked Lyons as the Cardinals' eighth-best prospect for 2013.

Yusmeiro Petit debuted in the majors all the way back in 2006 with the Marlins. He's spent time in the minors/majors with Arizona and in the Mariners system, as well as in Mexico. Petit's career major league numbers are uninspiring, with a 5.54/5.45 ERA/FIP in 234 career innings. However, Petit was dominant in his first season with the Giants organization in 2012: He posted a 3.63 pFIP across 166.2 innings. He's currently on the Giants' 40-man roster for the 2013 season.

Wade LeBlanc has spent a good deal of time in the majors. The left-hander debuted with the Padres, the team that drafted him, in 2008. He was a serviceable starter for San Diego across four seasons before being traded to the Marlins prior to the 2012 season.

LeBlanc spent the first three months of 2012 in Tiple-A, where he was very good, posting a 3.69 pFIP in 96.2 innings. He was called up at the beginning of July and pieced together a solid 3.64/4.04 ERA/FIP in 68.2 innings between the rotation and bullpen. LeBlanc should start 2013 in the Marlins' rotation.

IL

Pitcher	2013 ORG	Age	ERA	pFIP
1. Chris Tillman	Orioles	24	3.63	3.58
2. Chris Archer	Rays	23	3.66	3.61
3. Bryan Morris	Pirates	25	2.56	3.63
4. Corey Kluber	Indians	26	3.59	3.65
5. Dylan Axelrod	White Sox	26	2.88	3.69

Chris Tillman was a Mariners prospect who came to Baltimore in the Erik Bedard/Adam Jones swap. The right-handed starter began 2012 in Triple-A after three rough stints in the Orioles' rotation across 2009-2011.

Tillman may have figured it out in 2012 though, as his 89.1 Triple-A innings were the best in the International League, according to pFIP (3.58) and he posted a 2.93 ERA in 15 starts at the major league level for Baltimore. Tillman is expected to begin 2013 in the Orioles rotation.

Chris Archer is one of baseball's top prospects. Archer was ranked in Baseball America's top 100 coming into 2012. The Rays right-hander built off that ranking with a great season in Durham; which led to his major league debut.

Archer posted a 4.60/3.40 ERA/FIP in 29.1 big league innings last season. He will have a shot to make the Rays' rotation in 2013, but even if he ends up back in Triple-A, the righty seems to have a bright future.

Bryan Morris is the only reliever on this list. Morris pitched in 46 games at the Triple-A level for the Pittsburgh Pirates. His pFIP could benefit from coming out of the bullpen, as relievers tend to have higher strikeout rates than starters.

Morris made five major league appearances in 2012, striking out six batters and allowingjust two runs (one earned). Morris is currently on the Pirates' 40-man roster.

Corey Kluber started 21 games (125.1 innings) for the Indians' Triple-A affiliate, Columbus, posting a very good 3.65 pFIP. He also started 12 games at the major league level. Kluber's xFIP (3.99) and SIERA (3.87) show that he may performed much better than his 5.14 ERA would indicate. It's possible that Kluber could join Bauer in the Indians' rotation in 2013.

White Sox right-hander Dylan Axelrod, has been superb in back-to-back seasons in the International League. His brief stint in the majors in 2011 indicated that he could translate his Triple-A success in the majors, but he struggled in 51 major league innings in 2012 (5.47/5.04 ERA/FIP). Axelrod is currently on the White Sox 40-man roster, but most signs point to him not starting the season with the big league ball club.

Double-A

The Double-A leader board is, again, based on predictive FIP. All the numbers are adjusted for park using the same multipliers that were used for the Triple-A leade rboards. However, I did not separate the three leagues into different sets of top pitchers, as previous research has shown that the Double-A leagues have similar run environments (although the Texas League is slightly more hitter-friendly).

Below are the top 10 pitchers based on pFIP with at least 80 innings at Double-A in 2012 (for reference, the average pFIP for this sample was 4.21):

Pitcher	2013 ORG	Age	ERA	pFIP
1. Daniel Straily	Athletics	23	3.38	3.06
2. Justin Grimm	Rangers	23	1.72	3.36
3. Chris Heston	Giants	24	2.24	3.41
4. Matt Magill	Dodgers	22	3.75	3.42
5. Tony Cingrani	Reds	22	2.12	3.48
6. Zack Wheeler	Mets	22	3.26	3.51
7. Hiram Burgos	Brewers	24	1.94	3.58
8. Jose Cisnero	Astros	23	3.56	3.60
9. Eddie Gamboa	Orioles	27	3.33	3.64
10. Daryl Maday	Giants	26	3.43	3.68

The 2012 season was nothing short of incredible for Oakland's Dan Straily. His pFIP was by far the best on this list, thanks to his incredible 108 strikeouts (to go with only 22 walks) in just 85.1 Double-A innings.

Straily's success continued in his 66.2 innings at Triple-A, with 2.85 pFIP that would've led the leaderboards there had Straily gotten just 10j more batters out. Straily was given the chance to pitch at the big league level, where he had some struggles, in 39.1 innings, with control and the home run ball. I'd argue that Straily was the best pitcher in the minors in 2012, and that he deserves a spot in the Athletics' 2013 rotation.

Justin Grimm, like his AL West counterpart Straily, reached the majors in 2012. Grimm pitched extremely well with the Rangers' Double-A affiliate, Frisco. His numbers at Triple-A (4.35 pFIP in 51 IP) weren't nearly as good, though.

His 9.00 ERA in 14 innings with the Rangers does not look so great, but that sample size is minuscule and his 2.81 FIP was anything but horrible. Grimm is on the Rangers' 40-man roster and could start the year in their bullpen.

The Giants' Chris Heston struck out a ton of batters (135) at Double-A in 2012 while giving up just two home runs across almost 150 innings. The word on the righty is that he does not have great stuff, despite his shiny numbers. It'll be interesting to see if he can continue to be successful with below-average stuff and eventually reach the majors.

Coming into 2012, right-hander Matt Magill was ranked as just the Dodgers' 45th best prospect. I've got to think that his 2012 age-22 campaign had to have helped that ranking; he led all of Double-A in strikeouts (168). Los Angeles also seems to be impressed with the numbers he put up Chattanooga, as Magill is currently listed on the Dodgers' 40-man roster.

All Tony Cingrani has done since his 2011 professional debut is strike batters out, and 2012 was no different.

He spent some time dominating, with the Reds' high-A affiliate, where he posted a 1.11 ERA and a 32.3 percent strikeout percentage before being moved up to Double-A. Cingrani struck out over a batter an inning there and actually made his major league debut last season. The lefty struck out nine batters in five innings with Cincinnati. Cingrani is on the Reds 40-man roster and could start the year in their bullpen.

Most baseball fans have heard of Zack Wheeler, as he was famously traded to the Mets for Carlos Beltran just two seasons ago. Wheeler was very good in High-A in 2011 and he picked up right where he left off in 2012.

Wheeler was promoted to Triple-A near the season's end and posted an above average pFIP (3.90) in 33.1 innings. The Mets expect big things from their top prospect, who will most likely begin the year back in Triple-A. His major league debut should come sometime in 2013.

The Brewers' Hiram Burgos was nominated for Minor League Pitcher of the Year by MiLB.com, for good reason. The righty was very good at High-A, Double-A, and Tfriple-A in 2012. Like many others on these leader boards, Burgos is on his team's 40-man roster.

Jose Cisnero was not ranked as one of Astros' top- 0 prospects in a recent listing at Baseball Prospectus. Cisnero reached Triple-A as a 23-year-old last season, and posted a 4.12 pFIP in almost 40 innings there. Cisnero is on the Astros 40-man roster.

A 27-year-old who is only at Double-A is not usually anything to write home about. A 27-year-old in High-A is even worse. But for whatever reason, Baltimore's Eddie Gamboa pitched two games in High-A last season.

His pFIP at Double-A was one of the best of the season, mainly because he gave up very few home runs while keeping walks low. This success led the righty to reach Triple-A for the first time in his career. Gamboa has career minor leaguer written all over his profile.

Daryl Maday joins Heston as the second member of the Giants' organization on this top 10. Maday came primarily out of the bullpen in 2012, starting just three games. He, like Gamboa, probably will never make the majors, but found his way onto this list because he was able to keep walks and home runs low, while striking out a few guys at Double-A in 2012.

Part 2 and Google Docs

In the coming weeks, I'll post the second of this series, where I hope the discussion will become more interesting, as pFIP may be less relevant at the lower levels of the minors. My main focus will be how I attempted to account for this fact in those leaderboards.

For a full list of 2012 pFIP numbers for Triple-A click here.

For a full list of 2012 pFIP numbers for Double-A click here.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2013-01-09T06:55:15+00:00 Why an increase in payroll may not lead to sustained success http://www.hardballtimes.com/main/blog_article/why-an-increase-in-payroll-may-not-lead-to-sustained-success/ http://www.hardballtimes.com/main/blog_article/why-an-increase-in-payroll-may-not-lead-to-sustained-success/#When:14:16:15
Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2012-12-31T14:16:15+00:00 The myth of going for broke http://www.hardballtimes.com/main/article/the-myth-of-going-for-broke/ http://www.hardballtimes.com/main/article/the-myth-of-going-for-broke/#When:07:37:15
There are times when teams build new stadiums or sign new TV deals and use these cash inflows to invest in premium talent that could put their organization over the top. Some teams develop a deep farm system for years, investing in young talent that they hope to some day cash in on. Other teams get a taste at success by reaching the playoffs or coming close and spend big the next offseason in hopes of reaching that next step.

This new investment in talent or increase in spending is commonly referred to as "going for broke."

Before the 2012 season, the Miami Marlins supposedly cashed in on their new stadium and became one of the biggest spenders of that offseason.

They signed major free agents Jose Reyes, Mark Buehrle and Heath Bell.

The Marlins organization believed it had a shot at taking NL East supremacy and increased its payroll from $57.7 million in 2011 to $101.6 million in 2012, an incredible 76 percent increase.

Unfortunately for Miami, adding a wealth of talent to an already established core did not lead to success. In 2012, the Marlins won three fewer games than in 2011 (72 wins in 2011).

The Marlins aren't the only team in recent memory who "went for broke" and came out worse in the next season.

The 2008 Detroit Tigers may be the most famous case. They increased their payroll from $98 million in 2007 to $138 million (40.8 percent increase). Detroit traded a slew of prospects for Dontrelle Willis and Miguel Cabrera, and then signed both to big extensions. They also traded prospects for Edgar Renteria, signed Carlos Guillen to an extension and re-signed free agent pitchers Todd Jones and Kenny Rogers.

Everyone expected big things from the Tigers in 2008, but their big additions did not lead to a World Series championship. Instead the Tigers won just 74 games, 14 fewer than they won 2007.

This offseason it seems that both the Kansas City Royals and Toronto Blue Jays are going for it in 2013. Both teams have cashed in their farm systems to acquire quality major league talent that they hope will put them over the top and into the playoffs.

Toronto hasn't made the playoffs since 1993, and it's been even longer for Kansas City, which hasn't been to the postseason since 1985. Both teams have increased their payrolls from last season (the Blue Jays substantially) and hope their big offseason moves will end these long playoff droughts.

When a team goes for it and uses extra revenue or prospects to attempt to win now, fans of the team and writers become excited about the upcoming season. The 2012 Marlins and 2008 Tigers are cautionary tales, though, that success may not be as automatic as one would expect. A lot of baseball teams look great on paper, but the games aren't played on paper, or on spreadsheets for that matter.

I decided to test whether a significant increase in payroll actually leads to success or if it doesn't mean much at all.

The study

I looked at each team since 2002 whose overall payroll has jumped up by at least 20 percent during the offseason (payrolls have been increasing across baseball at a fairly steady rate of about four percent over this time).

There have been 79 teams with payroll increases that qualified for this sample, and I looked at each team's change in win total from the previous season to see whether the payroll increase would coincide with an increase in wins. On average, the teams in the sample saw a change of -0.96 wins in the season in which they increased payroll. That's right, not only did these teams not improve, they actually, on average, lost one more game than in the previous season.

I wouldn't consider this result as evidence that a payroll increase actually made these teams worse, but instead that the effect was negligible. However, just because there was no real effect overall does not mean some teams did not improve. I looked at which teams improved the most after their increase in payroll, as well as those teams that performed much worse than in the previous season.

Top five win improvements

Team	Payroll X+1	Payroll X	Change in payroll	% Change	Change wins	Wins X+1	Wins X
2008 Tampa Bay Rays	$47,124,500	$25,790,800	$21,333,700	82.70%	31	97	66
2002 Anaheim Angels	$61,414,167	$46,945,167	$14,469,000	30.80%	24	99	75
2006 Detroit Tigers	$85,198,456	$67,868,500	$17,329,956	25.50%	24	95	71
2004 San Diego Padres	$59,172,333	$43,565,000	$15,607,333	35.80%	23	87	64
2012 Washington Nationals	$94,568,929	$70,794,429	$23,774,500	33.60%	18	98	80

The 2012 Nationals and 2008 Rays had suffered through numerous losing seasons while they built from within before they both picked their spot and went for it in 2012 and 2008, respectively. After signing Jayson Werth the year before, the Nationals added Edwin Jackson and Gio Gonzalez, called up Bryce Harper and welcomed Stephen Strasburg back from injury. They went on to win the NL East.

For Tampa Bay, many young stars were coming into their prime and thus received raises through arbitration, which lifted the payroll. The Rays also may have realized how talented their core was as they added veteran free agents Cliff Floyd and Troy Percival to a team that would win the AL pennant.

The 2002 Anaheim Angels also had a slew of arbitration raises, but they also traded for Brad Fullmer and signed free agent Aaron Sele on their way to a World Series championship.

The 2006 Detroit Tigers were the AL Champions after their major payroll increase from 2005. Star outfielder Magglio Ordonez got a big raise, Kenny Rogers was their main free agent signing, and many other core players were given raises.

The 2004 San Diego Padres moved into brand-new Petco Park, won 23 more games than they had in the previous season, and reached the playoffs. Their payroll had risen from the season before (possibly because of their new digs) through the additions of Brian Giles (late in the '03 season), Ramon Hernandez, Terrence Long, David Wells and Jeff Cirillo.

Top five drops in win totals

Team	Payroll X+1	Payroll X	Change in payroll	% Change	Delta wins	Wins X+1	Wins X
2008 San Diego Padres	$74,010,117	$58,571,067	$15,439,050	26.40%	-26	63	89
2003 Anaheim Angels	$77,001,667	$61,414,167	$15,587,500	25.40%	-22	77	99
2011 San Diego Padres	$45,869,140	$37,799,300	$8,069,840	21.30%	-19	71	90
2007 Oakland Athletics	$80,777,050	$64,615,625	$16,161,425	25.00%	-17	76	93
2008 Colorado Rockies	$70,706,500	$57,062,000	$13,644,500	23.90%	-16	74	90

The 2008 and 2011 Padres have similar stories. Both teams were coming off successful seasons and wanted to build off that in the next season. The '08 Padres brought back Greg Maddux, signed free agent Randy Wolf and traded for Jim Edmonds but failed to duplicate their success. The '11 Padres kept much of their roster intact from the previous season, with many of their key players becoming more expensive through arbitration, but like the '08 Padres, they incurred a serious drop in wins.

The 2003 Angels were coming off a World Series victory and, like the '11 Padres, thought that keeping the same (more expensive) roster together would be result in more success, but injuries and regression to the mean kept that from happening.

The 2007 Oakland A's are an interesting case. Despite playoff success in 2006, many would have expected Billy Beane to trade his costly veterans—Jason Kendall, Eric Chavez and Esteban Loaiza—exchange for cheaper prospects in hopes of keeping the payroll low while constantly restocking the team with talent.

Instead, Beane chose to keep his veterans and signed another veteran free agent, Mike Piazza, to a big contract. The one time that Beane decided to hold his cards did not work out: Oakland won 17 fewer games than in the previous season.

The 2008 Rockies were coming off their first World Series appearance, and like the other teams on this list, wanted to build off this appearance with added revenues and a more confident ownership. They signed star outfielder Matt Holliday to an extension, brought in free agent reliever Luiz Vizcaino, and a large portion of the rest of their core received raises, but Colorado regressed a great deal, winning only 74 games in 2008.

I thought that looking at the sample as a whole and simply looking at teams that improved or regressed was not enough, so I considered a few other factors.

For one, an increase of 20 percent in payroll is a bigger jump for a team in a big market that already had a large payroll than for a small market team that began with a low payroll. Also, if a team was coming off a successful playoff season, the ownership's taste of some success, as well as possible added revenues, could cause a team to increase payroll in hopes of continuing that success in the subsequent season.

Small market vs. big market

I had trouble deciding whether small market teams that increase payroll should expect more success in the next season, given the number of reasons why they might decide to increase. The most obvious reason would be young players beginning to receive arbitration raises.

Small market teams tend to have low payrolls and the majority of their players have less than three years of major league service and thus are eligible to make the league minimum salary. As these players accrue more service time, if the team elects not to trade them, its payroll will inevitably rise.

It's difficult to say what effect this would have their team's overall performance. Typically, arbitration years coincide with a player's prime; thus, you'd expect that if a small market team has a bunch of players moving into their arbitration seasons at the same time (large increase in overall payroll, you'd see an improvement in overall team performance. However, every player does not age the same; thus, you can't assume that prime years will always coincide with a player's prime.

Small market teams also increase payroll when their front office believes their young core is good enough that the addition of one or two free agents or veterans through trade will be enough to put their team over the top; again, the effect on overall success of this type of thinking is uncertain.

So I controlled for small market teams to see whether their payroll increases tended to lead to more wins, as compared to the rest of the sample. I defined small market as a team with a payroll that was below 75 percent of the major league median payroll in the year before it increased spending.

Of the 79 teams that increased payroll by 20 percent from 2002-12, 47 had a payroll below 75 percent of the median in the year before the increase. On average these teams lost 1.7 more games in the season after the increase.

It seems that whatever the cause of the increase is (free agents, trades or arbitration raises), on average, small market teams do not win more in the year of the increase.

For teams in big (or above average) markets a 20 percent increase in payroll is much larger nominal increase. The ways in which payroll could increase for teams with higher payrolls are the same as small market teams: free agency, trades and arbitration raises.

The assumption, by most, would be that typically for teams that already have a high payroll, major increases would come from free agent contracts rather than arbitration raises.

If that assumption were in fact true (it's obviously true some of the time, but is not always the reason for an increase) do free agents make a team much better the next season?

I apologize for sounding like a broken record, but the average increase in wins was -0.26 (essentially negligible) for the 23 teams that had a higher than median payroll in the year before they increased payroll by 20 percent or more.

Playoff teams

My final idea for why an organization would want to increase payroll significantly would be to maintain or build off success in the previous season.

Sometimes, a team uses added revenues from reaching the postseason or winning the World Series to invest in more expensive talent, or a team comes so close to reaching the postseason that its ownership can taste success and is willing to expand the budget for the next season.

Of the 79 teams in my sample, 28 won at least 88 games in the season before they increased payroll. My hypothesis would be that these 28 teams will have increased their payroll either to get back to the playoffs or get in after being in thick of playoff contention the year before.

What was the effect of the increase in payroll on these 28 teams?

Of those teams, 14 won 88 or more games again after increasing their payroll; the other 14 failed to repeat their success> Again, it seemed that the increase in payroll had no overall effect on a team's ability to repeat its success.

Conclusion

I'm obligated to point out that this entire study was based on a fairly small sample. Also most of the discussion was on the aggregate.

Obviously, as I showed, some teams did improve a good deal after "going for it" and breaking the bank the next season. At the same time, many teams did not improve but instead regressed despite a much larger overall budget.

I wouldn't conclude based on these results that "any large increase in payroll, even if it means adding impact talent, does not change anything on the field."

Instead, my real point is that an increase in payroll does not guarantee results.

Even after adding Albert Pujols, C.J. Wilson and Zack Greinke (in mid-year) last season, the Angels failed to make the playoffs. The Tigers added a ton of talent in 2008, but all was for naught.

No matter how great a team's offseason may seem, it really means nothing until games are played on the actual field.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2012-12-26T07:37:15+00:00 Reinforcing the power of predictive FIP http://www.hardballtimes.com/main/article/reinforcing-the-power-of-predictive-fip/ http://www.hardballtimes.com/main/article/reinforcing-the-power-of-predictive-fip/#When:07:25:15 Predictive FIP (or pFIP for short). This metric is a slightly modified version of Tom Tango's commonly used fielding independent pitching (FIP) statistic.

Tango's version of FIP is meant to describe a pitcher's performance in terms of the three true outcomes (walks, strikeouts and home runs). The FIP equation weights each of those three outcomes in a descriptive manner:

FIP = (13*HR + 3*BB – 2*K)/IP + Constant (typically ~3.20)

FIP works fairly well as a predictor of future ERA or runs allowed (RA9); thus, many use the statistic to predict, despite the fact that it is not meant to do so. A good way to think about FIP is what a pitcher's ERA should have been, or better yet, what his ERA would be based solely on Ks, BBs and HRs. FIP is not meant to tell us what a pitcher's ERA is going to be in the future.

I set out to convert FIP from its descriptive form into a predictive metric.

After a few tests and some advice, I changed some of the methodology behind FIP. First, the FIP weights and constant are meant to describe ERA; I decided make pFIP a predictor of runs allowed per nine innings rather than ERA. Second, I made plate appearances (or batters faced) the denominator of the statistic rather than innings pitched.

The result was this equation:

pFIP = (17.5*HR + 7*BB – 9*K)/PA + Constant (typically ~5.18)

The major differences between FIP and pFIP come in the weighting of strikeouts and home runs. Strikeouts become more important when predicting future runs, while home runs become less important.

pFIP held up very well against other more commonly accepted "ERA estimators" (including descriptive FIP). That being said, just because something works fairly well does not mean one should not at least attempt to improve it.

A while back, I attempted to reform pFIP by regressing each of its components (Ks, BBs, HRs), to the mean. Strikeouts and walks are less volatile over one to two year samples; thus, their regression was not nearly as significant as the regression for home runs. Interestingly, regressing the components to the mean, did not improve the metric.

My next idea to improve pFIP was to focus only on the home run component of the statistic.

Dave Studeman, the leader of the Hardball Times, converted Tango's FIP into a version known as expected fielding independent pitching (xFIP).

According to the THT Glossary, xFIP is:

An experimental stat that adjusts FIP and "normalizes" the home run component. Research has shown that home runs allowed are pretty much a function of fly balls allowed and home park, so xFIP is based on the average number of home runs allowed per outfield fly. Theoretically, this should be a better predictor of a pitcher's future ERA.

The FanGraphs Sabermetrics Library explains how xFIP is calculated:

(xFIP) is calculated in the same way as FIP, except it replaces a pitcher’s home run total with an estimate of how many home runs he should have allowed. This estimate is calculated by taking the league-average home run to fly ball rate (~9-10 percent depending on the year) and multiplying it by a pitcher’s fly ball rate.

Over most small-to-medium samples xFIP is a better predictor of future than FIP; thus, I decided to apply this concept to pFIP.

xFIP simply inserts the expected number of home runs directly into the FIP equation:

xFIP = ((13*(FB% * League-average HR/FB rate))+(3*(BB+HBP))-(2*K))/IP + constant

I decided against inserting the expected number of home runs into the pFIP equation with its current weights.

An attempt to contrive an xpFIP

I took a sample of starting pitchers who had at least 100 innings in Year X and at least 100 innings in Year X+1 for the years 2007-12 (n = 479).

Then, I ran a multiple regression with strikeouts, walks and flyball percentage times the league average HR/FB in Year X against RA9 for each starter in Year X+1. This regression resulted in this regressed or xpFIP equation:

xpFIP = ((5*FB%*League-average HR/FB rate))+ (9*BB) + (9*SO) )/PA + constant**

**In this case the constant was 5.23**

By estimating the home run total, the home run coefficient of pFIP is only about half of the weights of Ks and BBs, as opposed to being weighted twice as much as those two coefficients in the original equation.

Then, this xpFIP equation was tested against these other ERA estimators:
{exp:list_maker} pFIP
FIP
xFIP
kwERA
SIERA{/exp:list_maker}I ran a linear regression, on the same sample, between each starter's ERA estimator in Year X and his RA9 in Year X+1.

I used r-squared as the measure of the predictive value of each estimator, and found these results:

Predictor	r^2
pFIP	18.50%
xpFIP	17.78%
kwERA	17.73%
SIERA	15.63%
FiP	15.33%
xFIP	14.82%

This new xpFIP equation did fairly well, beating almost all of the other estimators tested. However, regressing the home run component hurt predictive ability of the original pFIP; which was the strongest predictor.

Before scrapping the idea of regressed home runs in pFIP completely, I tested the equation on a different sample. I used the same minimum requirements (100 IP) and the same estimators and ran the same linear regression for the years 2002-07 and found these results:

Predictor	r^2
pFIP	19.19%
SIERA	16.56%
FiP	16.33%
xpFIP	16.03%
kwERA	15.79%
xFIP	15.29%

The xpFIP equation did not predict future RA9 nearly as well for this sample. My original pFIP equation did significantly better than the other ERA estimators at predicting future RA9.

Why does the pFIP with a regressed home run component do worse than the non-regressed pFIP?

It's interesting that the statistic that uses actual home runs is more predictive than the regressed version, despite the random variation that affects home run numbers.

My best guess for the reason behind this finding has to do with survivor bias. It has been shown that some pitchers have the ability to suppress home runs and consistently have lower than average HR/FB rates. I think it is entirely possible that a fair number of pitchers who are allowed to throw 200+ innings over the course of two seasons have some ability to control their home run rates.

Also there is the issue of park factors. The majority of these players did not change teams during the span of two seasons. It makes abstract sense that a pitcher who made half of his starts in a park that suppressed home runs would have a lower than average home run rate over those two seasons, and vice versa for a pitcher in a home run-friendly park.

I think it's well within the realm of possibility that regressing the home run component of pFIP would benefit the statistic when looking at pitchers who change teams between Year X and Year X+1.

pFIP vs. ZIPS

At this point, I'm pretty confident in the strength of pFIP as a predictor.

However, I had always simply assumed that projection systems were more useful, as they consider many more factors other than just the three true outcomes, when attempting to project future runs for pitchers. (Although, this Matt Swartz article caused me to be a little uncertain about that opinion.)

So, mainly for fun, I compared pFIP's RA9 projections for last year (2012) to the RA9 projections of the popular ZIPS projection system.

First, I looked at a sample of every pitcher who threw at least 100 innings in 2011 and at least one inning, in 2012 (n=137) and compared how well each system (or metric did) at projecting future RA9:

Predictor	r^2
pFIP	17.72%
ZIPS	14.65%

Much to my surprise, pFIP explained over three percent more of the variation in RA9 than ZIPS. However, my minimum inning threshold for 2012 (one!!) was admittedly silly.

Thus, to eliminate some outliers and converted relievers, I set the minimum threshold in 2012 to be at least five games started in the season (n=118). I found these results:

Predictor	r^2
pFIP	19.84%
ZIPS	17.20%

This change improved the predictive ability of both systems, and closed the gap slightly between pFIP and ZIPS. Interestingly though, pFIP still came out ahead of the much more sophisticated system.

This is very obviously a small sample. I looked at starting pitchers in only one season; thus, it could have been pure luck that pFIP was a better predictor of future runs than ZIPS. Also (and more importantly) ZIPS and other projection systems are built to predict many more factors (IP, GS, Ks, BBs, etc.) than just runs.

At the same time, I think these two short studies (regressing home runs and comparing to ZIPS), do a fair job at reinforcing the strength of this simple predictive re-weighting of the FIP equation.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2012-12-12T07:25:15+00:00 Relievers who could improve in 2013 http://www.hardballtimes.com/main/article/relievers-who-could-improve-in-2013/ http://www.hardballtimes.com/main/article/relievers-who-could-improve-in-2013/#When:05:43:15
For many pitching statistics, one season for a full-time starter (more than 160 innings), is actually a small sample size. Because of this, evaluating individual performances by relievers is rather difficult.

Last year, Josh Roenicke, of the Colorado Rockies, led all pitchers in innings out of the bullpen, with 88.2, a sample size only about half what a full-time starter would throw.

If a reliever throws 60 innings in back-to-back seasons, which is a fair amount out of the bullpen, the overall sample (120 IP) would still be very small.

Only a few statistics stabilize, or become reliable, over a ~60 IP sample:
{exp:list_maker}Strikeouts
Walks
Groundballs/Flyballs{/exp:list_maker}
I've found that the small sample that a reliever's numbers are subject to make future runs allowed extremely difficult to predict.

In that study, I found that many of the more advanced ERA estimators did not do a good job of projecting future runs for relievers.

For me, this was not the most important takeaway from that study. That honor rested with how a reliever's strikeout percentage beat more established estimators (FIP, xFIP and SIERA) at predicting future runs.

This fact brought me to the idea for this piece.

The random variation that affects ERA, especially for a small sample like one season for a reliever, causes it to not be a very reliable number to help look into the future. Thus, I decided to use strikeout percentage's predictive ability to look at a few relievers who had higher than average strikeout percentages (K%) and higher than average ERAs in 2012.

My theory is that these pitchers will end up with better run prevention results in 2013.

Here are the 2012 average strikeout percentage, ERA and adjusted ERA (FanGraphs' ERA-) for relievers:

Statistic	Average
K%	21.9%
ERA	3.67
ERA-	91

Using a minimum of 30 innings pitched, I found that 32 relievers had both a higher than average strikeout percentage and a higher than average non-adjusted ERA. Also, 32 relievers had both higher than average strikeout percentage and higher than average adjusted ERA, but the two lists were not identical.

I don't plan on highlighting each pitcher, but for those who are interested in seeing all of the names, I've attached a google doc.

Instead, I'd like to highlight five pitchers who could see a big boost in results out of the 'pen next season.

Antonio Bastardo

2012 numbers: 52 IP, 224 BF, 36.2 K%, 4.33 ERA, 109 ERA-

A Philadelphia Phillies left-hander, Bastardo, posted a superb 2.63 ERA (69 ERA-), in 2011. Bastardo's ERA was supported by by a .179 batting average on balls in play (BABIP), and a 81.1 percent left-on-base rate.

Both of those numbers regressed in 2012 (.306 BABIP, 72.5 LOB%) and his ERA ballooned.

Interestingly, Bastardo's strikeout percentage improved by over five percent and his ground ball percentage improved by over two percent, while his walk rate (11.6 percent) remained constant.

Among relievers (minimum 30 IP), Bastardo ranked sixth in strikeout percentage, ahead of the likes of David Robertson and teammate Jonathan Papelbon, yet his adjusted ERA was 18 percentage points above league average.

If Bastardo's strikeout percentage is any indication of what the future holds, then the Phillies could receive some serious value from the lefty, who is projected to make only $1.1 million next season.

John Axford

2012 numbers: 69.1 IP, 310 BF, 30.0 K%, 4.67 ERA, 118 ERA-

The Brewers' closer, Axford, was one of the best relievers in baseball in 2010 and 2011. Over those two seasons, he racked up 70 saves, while posting a 2.18 ERA and 2.29 FIP. Axford saved 35 games in 2012, but the results weren't nearly has good.

Axford blew nine saves (he had only five career blown saves coming in), and his adjusted ERA and FIP (104) were significantly lower than league average.

His strikeout numbers were very good and right around his career average (29.8 percent). Axford did struggle with his control (12.6 walk percentage, but that wasn't his main issue in 2012.

In 139.1 career innings coming into 2012, Axford had given up only five home runs. In 2012, his home run to fly ball rate (HR/FB) ballooned to 19.2 percent (career 8.9 percent, 2012 average: 10.3 percent); he gave up 10 home runs in 2012. Also, Axford could not strand anyone (which could partly be due to the home runs), as his LOB rate dipped to 68.2 percent, much worse than his career or league average.

I think there's a good chance both of those statistics rebound, and if he can keep the strikeouts up, a bounce-back year could be in store.

Jason Frasor

2012 numbers: 43.2 IP, 191 BF, 27.8 K%, 4.12 ERA, 100 ERA-

The former Blue Jays reliever's adjusted ERA was nine percent worse than the average major league reliever in 2012 and the second worst adjusted ERA of his career. Despite these sub-par results and his first stint on the disabled list, Frasor posted the highest strikeout rate of his career. He was especially successful against righties, striking out 35 of the 109 he faced.

Some may look at Frasor's 2012 season see a guy who is on the wrong side of 30, coming off a career-worst year. But I see a 35-year-old reliever who has had a solid and durable career, and could end up as a free agent bargain, because of his ability to strike out right-handed hitters.

Brad Brach

2012 numbers: 66.2 IP, 280 BF, 26.8 K%, 3.78 ERA, 104 ERA-

This was Brach's first full season in the majors and his unadjusted ERA was close to league average for relievers. But when it was adjusted to take his home park (Petco) into account, it looks much less pretty.

Brach had a surprisingly low BABIP (.245), but I think that was a product of his high walks and high home run totals, which I expect to come down in 2013. He is the kind of hard-throwing, splitter-yielding, high-K guy I think translates to a successful reliever. His services will come very cheap to San Diego next season.

Fernando Rodriguez

2012 numbers: 70.1 IP, 309 BF, 25.2 K%, 5.37 ERA, 138 ERA-

The Houston Astros right-hander ate innings last year, but his results were just terrible: His adjusted ERA ranked as the 12th worst among relievers (minimum 30 IP). However, Rodriguez throws hard and he had a well above-average K-rate over those innings. Rodriguez struggled slightly with his control, walking 11 percent of batters (league average was just over nine), but I can't overstate how more important strikeouts are.

Despite his struggles with walks and home runs, Rodriguez's peripherals (FIP, xFIP, SIERA, etc.) indicate that he was much better last season than his ERA would reflect.

Rodriguez was a solid reliever in 2011 with similar peripherals, but his 2012 strand rate (65.2 percent) caused his ERA to inflate. I think his velocity, strikeouts and peripherals are a better indicator for his future.

Houston could see Rodriguez return to being a very good reliever next year for very little cost: Rodriguez will make the league minimum.

Conclusion

Sample size matters.

The amount of innings that these five pitchers and any other reliever will throw next season will be a small sample. Thus, it's possibley that all five of these pitchers end up with higher ERAs in 2013.

Strikeouts aren't everything, but I think given the small sample that relievers' results are subject to, they give us a much better idea of who could improve, or perform more capably within that sample. Strikeouts hit an all-time high last season around baseball, and relievers like Aroldis Chapman and Craig Kimbrel put up some unheard of numbers.

Given baseball's current landscape, when it comes to the guys who come out of the 'pen, it is all about the Ks.

Click here to learn about THT's download subscriptions.]]> Glenn DuPaul 2012-12-05T05:43:15+00:00