Scouting Report: Hiroyuki Kawahara

In 2011, a young Japanese pitcher named Hiroyuki Kawahara pitched in the Puerto Rican winter league. The 21 year old lefty is not by any means "crafty" or "soft-tossing". According to scouting reports, he throws about 91 but has hit 96 before, and often.He was originally drafted in the NPB but there wasn't room in the rotation for him so he had been pitching in the minor league equivalent of the NPB. However, he is expected to take a leap and join the rotation of the Hawks in 2012. In Puerto Rico he had a great groundball ratio of 2.44 GB/FB. He gave up just 1 homer in 26.1 innings (7 starts, less than 4 innings a start). On the other hand, his ERA was 4.78. He somehow walked more batters than he struck out (control problems was originally part of his scouting report when he was just 18), walking 27, striking out 17 (5.86 K/9IP). He had big time splits, with a 6.19 ERA against righties and 2.61 ERA against lefties (even though his BAA was higher against lefties, he didn't walk many or give up many extra base hits to lefties). His numbers also would have been better if he didn't struggle so much with runners in scoring position (he gave up a BAA of .321 in those situations compared to .233). In most baseball circles LOB% is considered more luck than skill. One would expect this number to get better with a larger sample size, which would make his overall numbers better. He has had some should and elbow problems, and he was drilled in the head by a line drive earlier this year, but none of this seems too serious. Kawahara is an exciting prospect, and it should be fun to see him pitch in the NPB for the next few years, and at the very least could be a dominant left-handed specialist if made available to the U.S.

MLB Draft Talk: High Schoolers or College Players?

College and High School baseball is about to start. So even though we are several months away from the baseball draft, teams are already sending out scouts across the nation to look at prospects. So yeah, its time to start thinking about the draft. Here, I wanted to see which players are usually better choices, high school players or college players.
Since 1965:
1st overall picks (rated according to WAR):
The top 6 Players are out of high school. Only 3 out of top 10 were out of college. 19.5 WAR was "average" for each player selected first overall. Excluding 2008, 2010-2011, the 3 players drafted first overall that did not make the majors were all out of high school. Out of the 8 with negative WAR, 4 were out of college, 4 were out of high school. Out of the 27 (that's excluding the 2008, and 2010-2011 picks) "under-average" 1st overall picks, 14 were out of high school, so about split.

2nd overall
4 out of the best 5 went to college. Excluding the 2010-2011 picks, 13 players either had negative WARs or didn't play in the Majors. 8 of them were drafted straight out of high school. The 2nd overall pick averaged just 12.4 WAR. That is a big drop off from the 1st overall pick, so in 2012 there will be a big expected difference from the player the Astros get opposed to the one the Twins get at number 2 overall.

In the first round of the "Moneyball" draft of 2002, the top 4 players all came out of high school. Only 3 of the top 10 came out of college, and 2 of them were drafted by the Athletics. Only 5 out of the 14 that never made the MLB were drafted straight out of high school. Every single player, except 1, that reached the Majors and had a 0 or negative WAR was out of college. In the 2003 draft, 3 of the top 5 were out of college, while 11 of the 18 players that didn't make the Majors or had negative WARs were out of high school. However, in 2004-2006, 4 of the top 5 were college players each year.

When we look at all the pitchers drafted in the first round since '65, 51% of high school pitchers made the Majors. Those players averaged a 5.9 WAR. When you add Junior College players into this list, the same amount make the Majors, but the average WAR goes up to 6.1. When College players are added, we go up to 58% of players making the Majors. However, the average WAR for those players is 5.8. In round 2, 42% of all draftees make the Majors, and those average a 5.2 WAR. High schoolers in the 2nd round made the Majors 36% of the time, but had an average WAR of 6. Junior College players reached the Majors just 28% of the time and those players had a WAR of just 4.6 on average. So it seems that high school players have much higher ceilings, while college players are less risky, but provide less reward (and Junior College players are hit and miss).

Scouting Report: Angel Castro

Angel Castro is a pitcher that has been pitching in the Dominican winter league. He throws 95-96 MPH (he went up to 97 according to one radar gun in the Dominican Republic). In the Dominican he had a 1.86 GB/FB rate with a 2.53 ERA. Overall, he has a pretty impressive -2.73 PE with a .253 OBP against. This projects to a .287 OBP against in the Majors, which is pretty dominant. 

He spent 2011 in the Mexican League and gave up an ERA of 5.45 despite having a 3.00 GB/FB and 7.7 K/9IP. He gave up an amazingly bad 12 H/9IP and a WHIP over 1.8. He suffered from a .389 BABIP that one can deduce is from a bad defense, since he gave up 6 unearned runs as well, and his FIP was 3.88. He did keep the ball in the ballpark with a .5 HR/9IP rate.

AA is where Castro had problems, as he had a 4.48 ERA in 64.1 innings, with a 1.477 WHIP and 6.2 K/9IP. Most of the problems stemmed from his 18 inning stint with the Phillies organization, where he had a 49.7 LOB% and .369 BABIP. His stint with the Tigers in AA wasn't that bad with a 3.96 FIP. This still isn't very good and projects to a ERA of around 5.40.

After watching Castro throw a few times in the Dominican this off-season, I don't have much doubt that he can pitch in the big leagues. It would be a real shame if he has to go back to the Mexican League to pitch.

Phillies and Reds swap Valdez for Horst

The Phillies traded Wilson Valdez for Jeremy Horst from the Reds.

Valdez has a 1.8 WAR over the last 3 seasons or .6 WAR per year. He is scheduled to make just under 1 million dollars in 2012, or a solid 1667 projected WASP for 2012. Offensively, he has just a .290 OBP and 67 OPS + for his career. He has improved some offensively, as he has a 76 OPS + over the last 3 years. This still is pretty bad, and assuming Valdez plays 100 games, he will "create" 13 runs less than an average player (remember to distiquish between average and "replacement" player, average is .500 winning percentage, while replacement is .320). Pretty much every offensive metric shows that Valdez is a really poor hitter, I won't bore you with the details. However, does his defense wash this out? According to Baseball Info Solutions, the only position he is good at is shortstop. He has cost his team 6 runs more than average in 105 games at 2nd, and 3 runs in 41 games at third base (he has played just 5 games in the outfield). At shortstop however, he has saved his team 13 runs in 190 games. Adjusted for 100 games, like we did with hitting, he saves about 7 runs, meaning he would cost his team 6 runs more than an average player overall (with offense and defense combined) if he only plays shortstop and plays 100 games. However, he makes more than 2 million dollars less than an average player, so it would seem Valdez is still worth his salary.

Horst made his MLB debut in 2011, pitching 12 games in relief and posting a .1 WAR. He has appeared in 42 AAA games (65.2 innings), a much bigger sample, and the one we will be using here. There, he had a -1.104 PE, or -.911 Adjusted PE. According to our AAA metric, this projects to a 1.426 PE or 1.619 Adjusted PE. Neither of those numbers are very good at all for a reliever. If you assume he throws 60 innings (like most "full time" relievers) in 2012, his projected CWAR would be 1.05 (an average reliever with a 4.2 FIP and 0 PE that pitched 60 innings would have a CWAR of about 1.06, despite Horst's bad PE, he is helped by a pretty good projected FIP in the 3.5 range), which is better than Valdez's WAR of .06 average in the last 3 years. At least in his time in the Majors, he threw an under 89 MPH on his fastball. The velocity is certainly concerning, but other than that he projects to be about an average reliever, while Valdez is an under average player. This is why it looks like the Phillies have won this trade, even if by only 6 or so runs a season.

Projecting Leonys Martin

The Rangers signed Leonys Martin to a 5 years 15.5 million contract back in May of 2011. He played 29 games in AA and had a .435 OBP, but that dropped significantly in AAA, where he had a .316 OBP (and .630 OPS in the hitter friendly PCL). Neither of these sample sizes are particularly helpful.

In his last 3 years in Cuba (his 3 seasons with more than 200 AB) Martin batted .345/.477/.518. Clay Davenport projects 263/.339/.399 and a 2.1 WAR for Martin in 2012. Davenport also has a Yeonis Cespedes projection, where he turns his Cuban numbers of .333/.434/.667 into .251/.309/.451. This is a .082/.125/.216 regression. This turns Martin into a .263/.352/.392, so really close.
Davenport also says that Cuban baseball can be graded like "low A ball." According to this (and our A metric): Martin projects to be a .428 OBP and .451 SLG (.879 OPS). This is nowhere close to Davenport's own projections, which means our A-ball projection is not helpful for Cuban statistics.

So Martin seems to be a guy with an average batting average, solid OBP, and under average slugger with .744 OPS. This is a 1.52 Simple WAR. According to Runs Created formula, Martin is projected to create about 82 runs in 140 games, or 17 runs more than an average hitter. Defensively, Martin had a 2.45 Range Factor per Game in the minors. 2.59 Range Factor per Game is about league average for center fielders. This means that Martin will get to almost 20 less balls than an average center fielder. If we assume that each time Martin doesn't get to a ball that an average fielder would get, he cost his team one base, or about 5 runs overall if he plays 140 games. This still leaves him responsible for 12 positive runs more than an average player would. 

To be worth his contract, he really needs to be about league average, or provide 5 wins over replacement in his 5 years. According to Davenport's WAR projection, he will be worth over the about 3 million dollars a year. So it seems according to the projections and all metrics, Martin will be worth the contract the Rangers gave him in 2011. Now to see if this actually becomes the case. 

Marlins sign Gary Glover

The Miami Marlins have signed pitcher Gary Glover to a minor league contract. Glover has been in the KBO since the middle of 2009. In Korea, he had a -.22 PE as a starter. This projects to a 2.518 PE ceiling and 4.248 PE floor in the Majors. Assuming Glover throws 100 innings for the Marlins in 2012 (just for theoretical purposes), his ceiling would be a .58 cWAR and under replacement would be the floor. Glover also spent 2006 in Japan in the NPB and really struggled with a 4.97 ERA. He has also had a long MLB career, pitching over 500 innings, with a .3 WAR in that time. Using our hypothetical 100 inning season, he would be about replacement (technically .058 WAR). His career PE in the MLB is 3.598 with an Adjusted PE of 3.428, meaning he performed between the ceiling and the floor of what his KBO projection looks like. He gave up a .338 OBP and .793 OPS.  His FIP - is 110, and he threw about 93 MPH in 2008. One expects that his velocity has fallen a bit. Add to this that he gives up more contact than an average pitcher, and this gives us a pretty good idea of how Glover will pitch in 2012. The idea we get is not very good. Glover being a 35 year old replacement type pitcher isn't very attractive, and its hard to believe that Glover will provide any kind of impact for the Marlins in 2012.

Tigers make big signing

The Tigers made big news on Tuesday, signing a huge big ticket free agent to a 9 year 214 million dollar deal. However, this post isn't about Prince. The Tigers also signed Warwick Saupold, an Australian Pitcher who was pitching for the Perth Heat in the Australian Baseball League. In 70 innings this season, he had an ERA of 1.41 and WHIP of .93. This projects to about a 3.53 ERA in the MLB according to our minor league and ABL metrics. However, the previous year for Saupold was much less pleasant. In the 2010-2011 season, he had an awful 5.52 ERA in 44 innings. Is this just a case of things evening out with larger sample sizes, or was there another explanation. Just on the surface, while he still got more groundballs than flyballs, he had less groundballs last season than this season. He gave up 5 homers last season, while he only gave up 2 this year. Strangely, he had a better strikeout in his bad season than his good season. This would suggest a bit of BABIP luck, but also the factoring of the difference in flyballs and homers. So which Saupold will the Tigers be getting? That is very hard to predict, but one would figure that the more recent season should hold more weight, and that his improved groundball and home run rate is probably not flukey. That is to say, there seems to be a genuine improvement from the previous year to this season. So it would make sense to call his 2011 season that projects to a 3.53 ERA his ceiling. However, we could create a "floor" for Saupold, using his average from his 2 seasons in the ABL. Overall, he has a 3.00 ERA in 114 innings, which is not quite a full season. This projects to an ERA over 5, meaning a below replacement pitcher. This is a wide range of projections, and keep in mind its a relatively small sample size. While I like his home run rate (.55 HR/9IP overall), his lack of strikeouts are concerning. His best hope is to make it as a groundball pitcher who pitches to contact. While there are some decent pitchers in the MLB, they are highly volatile (for the most part) and not in the elite category. Even though the ERA projection projects a ceiling of middle rotation, I would say the ceiling is likely back of rotation or long reliever. However, even if that happens and he has some success down the road, it would be a nice pickup for the Tigers, and may make some MLB franchises take the ABL more seriously.

How we did: Championship Weekend

Big D in bold, I in italics

Patriots hold off Ravens

Giants sneak past 49ers
Excuse: I definitely thought this would be a close game, and it was. Special teams are hard to predict and highly volatile, but this week showed that it is tremendously important. 2 bad plays by Kyle Williams gave the Giants the win.

Big D: 0-2, 3-7 in playoffs, 60-44 overall
I: 1-1, 4-6 in playoffs, 63-41 overall

Athletics sign Jonny Gomes

The Athletics signed Jonny Gomes to a one year 1.1 million dollar deal. He split 2011 between the Nationals and Reds and posted a .7 WAR, worth 2.1 million dollars according to the Halladay Standard. Over the past 3 years, Gomes has averaged a .3 WAR (worth 900,000 dollars) with a .297 BABIP in that time. Offensively, he has a .329 OBP, 106 OPS +, and .330 Secondary Average. He is an above average home run hitter, with an ISO of .209, walks slightly above average, and has a 3.84 Pit/PA. So basically in every offensive category Gomes is slightly above average. Defensively, its a whole other story. His fielding percentage and range factor are both way under average. Depending on which metric you trust more, Gomes has cost his teams 46-40 runs more than an average fielder would in his career. Using runs created, we can measure (we did this in the Jack Cust article) how many runs and wins Gomes will be worth for the Athletics. According to Runs Created per Game metric, Gomes is worth .59 runs per game (technically he has a 5.3 Runs Created per game metric, but that assumes an offense of 9 Gomes', here I am isolating it to 1 Gomes'). Since he has played 428 games in the outfield, he will cost his team a run more than average about every 10 games. An average offensive player would create 65.3 runs in 140 games, while Gomes would create about 82.6 runs, or 17.3 runs more than an average player. Defensively in this time (assuming he never plays DH), he will cost his team 14 runs more than average. This is all a long way of saying that Gomes is worth 3.3 more runs a year than an "average player". There are two ways to look at this, I believe. One way makes Billy Beane and the Swinging A's look like geniuses, and one way makes Gomes' agent look like a genius. First, Beane might point out that the average MLB player salary is 3.1 million. Gomes is slightly better than average, and yet will be playing for a third of that salary. Gomes needs a new agent. Gomes' agent might try to save his job by pointing out that according to the Halladay Standard (a standard we have relied on heavily in this blog), Gomes is worth just $900,000, and so he is making 200 thousand extra dollars. It all depends on which metric you rely on more.

NFL Picks: Championship Weekend

Big D in bold, I in italics.
                                                                  49ers versus Giants
Patriots versus Ravens

Projecting Gerardo Concepcion

18 year old Cuban pitcher Gerardo Concepcion has moved to Mexico and will become a free agent. So I found Cuban pitchers who pitched both professionally in Cuba and the MLB. I could find just 3 with reliable statistics.

Aroldis Chapman
Cuba: 10.04 K/9IP, 3.75 ERA
MLB: 12.8 K/9IP  ,3.27 ERA

Yuneisky Maya
Cuba:  7.6 K/9IP, 2.45 ERA
MLB: 4.14 K/9IP, 5.52 ERA

Jose Contreras
Cuba: 14.97 K/9IP, 4 ERA
MLB: 6.8 K/9IP, 4.54 ERA

Contreras and Maya saw significant decline in the Majors, while Chapman saw improvement. The biggest difference between Chapman and Contreras/Maya is that Chapman throws an average 99 MPH fastball, hitting 100 frequently, and neither Contreras or Maya hit 95. As I had shown in a previous article, pitchers that threw 95 MPH were much more likely to have success than pitchers that didn't. According to reports I have seen, Concepcion can barely hit 90 MPH, which is about what Maya throws. So here we will look at Concepcion's statistics through Contreras and Maya's difference in the two countries.

Gerardo Concepcion
Cuba: 4.7 K/9IP, 3.36 ERA
MLB Projection: (negative strikeouts somehow), 5.165 ERA

Just for fun, a projection including Chapman:  1.74 K/9IP, 4.4 ERA

As you can see, the strikeout projections are particularly unhelpful, but a 4.7 K/9IP at any level is terrible. His ERA is also very pedestrian and unimpressive. He is only 18, and could develop more (I would be surprised if he didn't), but throwing 90ish with no apparent strikeout abilities isn't exactly something teams should be clamoring over.

A's Acquire Seth Smith

The Rockies traded Seth Smith to the A's for Guillermo Moscoso and Josh Outman.

Smith agreed to a 2.145 million dollar deal the day before being traded. Smith had a 0 WAR in 2011, and has averaged a 1 WAR over the past 3 years, worth 3 million dollars. Terrible on defense (-1 WAR in his career, and a -1.4 WAR in 2011), he has a below average fielding percentage and range factor in the outfield. He has a career OBP of .348 and 110 OPS +. He has a very good 6.3 Runs Created per Game metric, with a .636 Offensive Winning percentage. He has good power at .209 ISO, and is reasonably efficient with a .330 Sec Average. The above average walker sees 3.95 pitches per plate appearance, giving him a PPS of 95.84.

Moscoso had a 2.2 WAR in his first season with significant pitching time. As mostly a starter, he has a PE of 1.227 with an Adjusted PE of 1.977. His FIP - is 105, with a SIERA of 4.96. He had a horrible groundball rate of .50, and gives up just under 1 HR/9IP. These statistics, combined with Coors (the horrible groundball rate really makes you scratch your head), makes me believe he won't repeat his 2011 success.

Outman, besides having an awesome pitching name, has a career 2.7 WAR in 3 seasons (although it is just 151.1 innings, mostly as a starter, so it is about a season). His FIP - is 97, but his SIERA is a 4.51. He also gives up more flyballs than groundballs.

I really don't understand this deal from the Rockies side. It really shows a lack of understanding by the organization of their own ballpark and advanced pitching statistics. For the A's, they get a much needed bat, although it comes with defense baggage, and trade 2 pitchers that appear to be products of the ballpark. 

How we did: Divisional Round

Big D in bold. I in italics.

49ers dunk over Saints
Excuse: Obviously didn't see the Saints turning over the ball 5 times. Alex Smith looks like a totally new guy.
Patriots destroy Broncos
Ravens knock off Texans
Giants shock Packers
Excuse: Wow. This was my second most sure pick this week. Horrible secondary play, fumbles, and bad line play made the Packers look like a very pedestrian team. We probably should have saw this coming, but 15-1 made it hard to take those flaws seriously.

Big D: 2-2, 3-5 in Playoffs, 60-42 overall
I: 2-2, 3-5 in Playoffs, 62-40 overall

Australian Baseball and Minor League Pitching Metric

As the Australian Baseball League 2011 season wraps up, I thought it made sense to come up with a pitching metric to evaluate pitchers pitching in the league. Since we didn't come up with a reliable A and AA pitching metric, I filled out the rest of the spreadsheet with minor league pitchers (from the White Sox system) who had pitched at all 3 levels.

The ABL players that have pitched in A-ball had an average ERA of 3.39 in the ABL. Those same pitchers have a 3.51 ERA in A-ball, a difference of .12. Those pitchers averaged a 1.19 WHIP in A-ball and 1.16 in the ABL, a .03 difference. So there is a slight upgrade from the ABL to A-ball, but the difference is very small, so no one will judge you if you just use the A-ball metric we are about to create and use it on ABL players.
The pitchers on the list that pitched in AA had a 3.16 ERA and a 1.21 WHIP in A, and a 3.7 ERA and 1.32 WHIP in AA (a .54 ERA difference and .11 WHIP difference). The pitchers that pitched in AAA had a 3.43 ERA and 1.233 WHIP in AA, and 4.32 ERA and 1.37 WHIP (a .83 ERA difference and .137 WHIP difference).

The Howie Kendrick extension

Howie Kendrick has signed a 4 year 33.5 million dollar extension with the Angels. Over the past 4 years, Kendrick has posted a 10.9 WAR, or 2.7 WAR per season. If Kendrick performs at the same level for the next 4 years, he will has a WASP of 3073, right around the Halladay Standard. However, in the last year of the deal, he will be 31, which is usually past the prime for a baseball player. It is not that Kendrick can't perform at a high level, or even the level he has been performing at, it is just unlikely he will be as good at age 31 as he is at 27. This makes a contract that is seemingly border-line like this one, much less attractive. Defense usually regresses quicker than offense, and Kendrick relies much more on offense than defense (just 1.7 D-WAR in his career). He is a solid defender, with a positive D-WAR every year in his career except 1. Offensively, he has a career .329 OBP and 105 OPS +. His offensive winning percentage is .495, with 4.8 Runs created per game. As those statistics point to an average offensive player, his secondary average and ISO point to something a little worse than average. He is an under average home run hitter, well below average walker, and slightly above average extra base hitter. He is an extreme groundball hitter, and sees just 3.63 Pit/PA. To me, it appears that Kendrick is really just a slightly above average player with a lot of flaws. I think the Angels overpaid a bit for him, but it is a good long term deal for Kendrick himself. If I was an Angel fan, I would really start worrying about how my team is overpaying a bunch of guys, and wonder how they are going to manage salary in a couple of years.

NFL Picks: Week 19 Divisional Round

Big D in Bold. I in Italics

Saints versus 49ers
Patriots versus Broncos
Ravens versus Texans
Packers versus Giants

How much does QB success depend on the team?: Part 1

Unlike baseball, success for a player in football depends a lot on the success of the players around the player. For example, Justin Verlander's strikeout totals have nothing to do with how his teammate Brad Penny pitched. Ichiro Suzuki's terrible 2011 defense had nothing to do with the Mariner's anemic offense or problems on the infield. I wanted to see if I could measure how much a quarterback's success had to do with him or the team he was on. To do this, you have to some how isolate the quarterback from the team. The best way to do this, that I can think of, is to use quarterbacks who switched teams during their careers. So I included the last year's passer rating with the previous team, and then either the first year with the new team or his total numbers with the first team. I used 200 Attempts as a rough requirement (I did make some exceptions). I only wanted to use consecutive years so people like Vick and Grossman are thrown out. I did this because a lot can change in two years for a player (same reason I didn't do baseball write-ups on people like Jamie Moyer and Manny Ramirez for the 2012 season).I including Russell Wilson in the spreadsheet because I found the difference between his 2010 season with NC State (where he was a solid quarterback) and his 2011 season with Wisconsin (where he was one of the best in the nation). I then included some notable running backs that have played with multiple teams and compared their average yards per carry.

Average difference in passer rating (excluding Russell Wilson) is 6.59. Just to give you an idea, that is about the difference between Tony Romo and Matt Schuab in 2011 (as well as the difference between Ryan Fitzpatrick and Tim Tebow). The biggest difference was McNabb's 15.8, which is about the difference between Phillip Rivers and Kevin Kolb. So it would seem, at least in the sample we have, that there is a difference, but for the most part not a huge difference for quarterbacks depending on what quality of a team they play on.

Runningbacks had an average difference of .65 yards per carry. This is a huge difference. So, at least it would seem according to these samples, that things like offensive line affect a running back much more than a quarterback. In Part 2, I will use college statistics for quarterback and try to see which statistics best translate from NCAA division 1 to the NFL.

What offensive statistics are most important?

I wanted to see what base statistic most correlates to runs scored. This way we know which statistics to value when looking at offensive players. I used Fangraphs customizable reports, and looked at all 30 teams from 2000-2011. I selected a few different statistics to see the correlations.

The Yankees have scored the most runs during this time period, and also have the most home runs and walks. They are 27th in Sacrifice hits and have less strikeouts than average. They have also grounded into the 2nd most double plays and are 5th in stolen bases. They are second in batting average and first in OBP. They are tied for first in slugging and tied for second in ISO. As far as BABIP goes, they are tied for 5th. The worst team in this era is the Expos/Nationals, scoring 8177 runs or 77.4% of the runs the Yankees scored (according to Sabermetric formula of 10 runs equaling a win, this is 238.3 more losses than the Yankees or almost 20 more losses a season). The Expos/Nationals were in the bottom half in baseball in homers and walks. They were 12th in baseball in strikeouts and had the second most sacrifice hits. They grounded into the 13th most double plays, and were about mid-pack in stolen bases. They were second to last in both batting average and OBP. They were 5th to last in ISO and 3rd to last in SLG with the 7th worst BABIP.

3 of the top 5 home run teams were top 5 run teams. The same ratio applied to walks. 3 of the worst home run teams were part of the bottom 5 in runs scored. 2 of the worst 5 walk teams were among the worst 5 offenses, while none of the 5 best walk teams were part of the worst 5 offenses. Only 1 of the top 5 strikeout teams were 1 of the worst 5 offensive teams, and 1 of those were actually part of the top 5 offenses. 1 of the teams with the 5 lowest strikeout totals were one of the worst 5 offenses and none of them were in the top 5. As far as sacrifice hits go, they seem to be more helpful than I originally thought, as 2 of the top 5 offenses were also in the top 5 in sacrifice hits (and of course the Expos were in the bottom 5 of offenses and top 5 of sacrifice hits). Only 1 team that were in the top 5 in grounding into double plays were 1 of the worst 5 offenses (of course the best offensive team had the second most GIDP). None of the teams that hit into the least double plays were in the top 5 offenses. The Yankees were the only team in the top 5 in stolen bases and the top 5 in runs scored. The Red Sox were among the worst teams as far as stolen bases, but also among the best offenses. Batting average had a huge correlation, all 5 teams with the most runs scored were in the top 6 in BA. 3 of the worst 5 teams in batting averages were among the worst 5 in runs. OBP also had a high correlation as 4 of the best 5 run scoring teams were among the best 5 in OBP. 4 of the worst offensive teams were also in the bottom 5 in OBP. All 5 of the top teams in runs scored were in the top 5 in Slugging, and all 5 of the bottom in runs scored were in the bottom 6 in SLG. The top 5 in Isolated Slugging contained 4 of the top 5 run scoring teams. All five of the worst offenses were in the bottom of 7 in Isolated Slugging. BABIP doesn't have quite as strong of a correlation as 2 of the top 5 offenses were also the top 5 offenses. None of the bottom 5 offenses suffered from being one of the bottom BABIPs in the league.

So it seems that offensive statistics can be ranked by importance in this order: Great correlation: OBP, BA, SLG, ISO, BB, HR,
Small correlation: SH, BABIP
No correlation: SB, GIDP, SO

John Bowker to go to Japan

The Phillies have released John Bowker so he can pursue a career in the NPB. In 622 PA (about a full season), Bowker had a .283 OBP, .664 OPS, 3.31 PAPP, and .61 PPG. Through the NPB metric, he is projected to have a .316 OBP and a .06 Simple WAR. In his Major League career, he had a 74 OPS +, with 3.6 Runs Created per a game. His Secondary Average was .216, and an Isolated Slugging of .150. He was an average home run hitter and good flyball hitter, but a below average walker. He saw an average amount of pitches per plate appearances. This all added up to a Offensive Winning Percentage of .374. Defensively, he was horrible at first base, and had an above average fielding percentage in the outfield, but got to a lot less balls than an average outfielder according to range factor. In total, it is estimated that he saved 5 runs above average in his career, mainly because of his outfield play. Assuming he plays above average in the outfield in the NPB, and his projected just above replacement offense, he would certainly be worth the equivalent of 1 million American dollars (because most NPB teams can pay about 1 million dollars per 1 WAR as we saw in the Lastings Milledge article).

How much does velocity matter?

I wanted to see the correlation between velocity and effectiveness in a pitcher. This has huge ramifications for scouting, as a pitcher's velocity is one of the first things looked at by scouts. So here I used any pitcher who threw at least 20 innings in 2011 in the NL West. I used Fangraphs customizable reports, and used FIP - as the standard of effectiveness, and used their velocity charts. My assumption before hand was that top velocity matters very little and speed differential matters much more, so I also subtracted the highest velocity from the lowest velocity for each pitcher (click on the images to enlarge).

Interesting to note that the Phillies, the best pitching team in baseball, had only one qualifying player throw 94 MPH, and no one threw harder. All 5 pitchers that threw over 95 were above average (above average is having a FIP - below 100, below average is anything above 100), so it does seem there is a positive correlation between big time velocity and success. 22 pitchers didn't throw pitches at 90MPH or more, and 9 were above average and 13 were below average.
There were only two pitchers who had differentials more than 20 MPH, and they both were above average. 20 pitchers had differentials of 15 or more MPH, 15 were above average and 5 were below average. However, 16 players had differentials of less than 10 MPH and 12 of those pitchers were above average. So perhaps there isn't much to speed differential. At the very least, we have to say the results are inconclusive. It doesn't even seem to help to have a slow pitch necessarily, as 3 of the 4 pitchers with pitches less than 70 MPH were below average.
So it does seem that having a 95+ MPH pitch correlates to success, while not being able to throw 90 correlates to failure.

How we did: Week 18 picks

Big D in Bold, I in Italics

Texans pick apart Bengals
Excuse: The two things I counted on in Cincinnati: Andy Dalton and the defense, failed miserably. Marvin Lewis also showed complete incompetence in managing the game, especially with his two silly challenges. How he still has a job every year is almost as amazing as Norv Turner.
Saints pass all over Lions
Giants bash Falcons
Excuse: Really 2 4th and inches plays turned this thing around. Mike Smith didn't look very impressive either.
Broncos shock Steelers
Excuse: I thought an injured Big Ben and Tebow was as bad as quarterback play gets. Then I watched LSU and Alabama play.

Big D: 1-3, 58-40 overall
I: 1-3, 60-38 overall

Cubs aquire Rizzo

The Cubs sent Andrew Cashner and Kyung-Min Na to the Padres for Anthony Rizzo and and Zach Cates.

Cashner has 65 career MLB innings (basically all in relief), and has earned a .3 WAR in that time. His PE is .115, with an adjusted PE of .665. His FIP - is 117, but his SIERA is 3.94. He has a nice groundball ratio (1.52), but gave up 1.25 HR/9IP. He has a league average Pit/PA. A bigger sample size is his AA numbers, where he has pitched 97 innings and had a PE of -1.95, and gave up just .1 HR/9IP. His FIP was about 2.8, meaning he had a CWAR of about 1.26. This would be a negative WAR in the MLB according to our AA metric.

Na is a below average outfielder according to both fielding percentage and range factor. He has played 149 games in the minors, almost exclusively in different levels of A-ball (just 2 games in AA). There he has an OBP of .335 and OPS of just .619. This projects to a terrible .286 OBP and .453 OPS.

Rizzo has a -.7 WAR in 49 career MLB games. He has played 93 games in AAA, in the PCL. There he has a .404 OBP, 1.056 OPS, 2.2 PAPP, 2.7 PPG, for a great 6.44 Simple WAR. This projects to a 4.95 Simple WAR with an OBP of .358. Defensively, he isn't bad, with a 9.51 Range Factor and .989 Fielding Percentage at 1st base.

Cates has spent just one year in A ball, and had an even 0 PE as a starter, with an adjusted PE of -1.4. He has an okay walk rate, and gave up just .31 HR/9IP. His CWAR was 2.11, and if we assume that pitcher regression would be the same as hitter regression (that is the ratio of CWAR lost to Simple WAR lost is 1:1), he would have a -1.45 CWAR in the MLB. A different way might be to use the 22% regression that hitters see in their OPS from A ball-MLB and apply that to his FIP. This would project his FIP at 3.94, which would obviously lead to a positive WAR.

The Cubs get a first baseman that projects to be very good in the Majors, and a pitcher with a good strikeout rate and doesn't give up homers while the Padres get a player that can't play along with a reliever that doesn't project very well either and struggles with homers. This is a very lopsided trade.

The Mariners sign Muneno Kawasaki

The Mariners have reportedly signed Muneno Kawasaki to a minor league contract. At age 30, in his past 5 years in the NPB, he has a .344 OBP, .721 OPS, 1.1 PPG, and 2.76 PAPP. According to the Japanese metric, this is projected to be a .311 OBP, .61 PPG, 3.5 PAPP, and a -1.15 Simple WAR. In 2011, Kawasaki was worse, with just a .310 OBP, which is just a .277 OBP once adjusted for the MLB. He is also a below average shortstop according to range factor, although it is worth noting that he is above average according to fielding percentage. He homers less than 1% of the time, and walks just 6% of the time. He did have a decent stealing rate, at about 72% success rate. You usually want a 75-80% rate for stealing to be worth it. One would expect some kind of regression in the MLB, meaning that his one decent weapon would be taken away from him. It is just a minor league contract, so its not like the Mariners lose anything, but I don't see Kawasaki having success in the Big Leagues.

NFL Picks: Week 18 Wild Card Round

Big D in Bold, I in italics


Falcons versus Giants
Steelers versus Broncos
Bengals versus Texans
Saints versus Lions

Another College Metric: NJCAA

175 players from NJCAA (Colleges that are considered Junior Colleges in baseball) were drafted in the 2010 draft. Here I tried to judge the difference between A-ball (only Bryce Harper out of these have made it to AA) and the NJCAA. Of course not all 175 are included for different reasons (perhaps they didn't sign, don't have reliable statistics, or changed from a pitcher to a batter). But here it is:

Hitters:

 Pitchers:

 See you when the College season starts.

Red Sox sign Doug Mathis

The Red Sox signed about 11 minor league free agents this past week, including some former familiar names such as Brandon Duckworth, Carlos Silva, and Tony Pena Jr. However, the one I wanted to focus on former Ranger pitcher Doug Mathis. He pitched for the Samsung Lions in the KBO in 2011 and had a 2 PE (with almost as many walks as strikeouts). When this is converted to the MLB, this a horrendous 4.54 PE. So it doesn't seem that he is ready for MLB success according to his KBO statistics. In his Major League career, his PE is 5.015 with a -.1 WAR in 87.1 innings. So he actually performed slightly better in the KBO than expected. In his MLB career, he has a FIP - of 121, SIERA of 4.79, and 5.625 Adjusted PE. Obviously these numbers are horrible. His AAA numbers are pretty bad too, with a 2.88 PE in the PCL, which translates to a 4.31 PE in MLB. So according to his AAA numbers, he has underperformed in the Majors, but nothing could really be expected of him anyway. I am not sure what the Red Sox see in Mathis, there isn't much there, and several different metrics show this.

Cubs resign Reed Johnson

The Cubs have resigned Reed Johnson to a 1 year, 1.15 million dollar deal. In 2011, he had a 1.3 WAR, and has a .63 WAR average over the last 3 years, which is worth about 1.9 million dollars. Defensively he is not very good, at a -.6 career D-WAR, -.1 in 2011. While his fielding percentage is above league average, he doesn't make many errors, he doesn't get to many balls either, with a well below average range factor. BaseballProjection says that he cost his team 4 runs more than an average fielder. Offensively, Johnson had a .348 OBP, 122 OPS +, and .130 ISO with a gigantic .394 BABIP. Obviously those statistics are pretty useless, as his 1.3 WAR is not something you can expect to repeat. That certainly taints the .63 WAR average over the last 3 years, as his 2009 and 2010 WARs were both .3. Over those two years, Johnson has an OPS + of 84, and OBP of just .309 with a BABIP of .310. These are pretty abysmal offensive numbers as evidenced by his 4 runs created per game and .426 offensive winning percentage in those two years. In his career, he has walked just 4.8% of the time, and is an under average extra base and home run hitter. Even at such a small low risk contract, the Johnson signing doesn't make a whole lot of sense. It would have been much better if they had been able to bring him back on a minor league deal.

Dominican Winter League Metric

I picked 50 batters that played at least 20 games in the Dominican Winter League in 2011, and compared the OBP and OPS of their Winter League season and compared it to their MLB or AA numbers (sometimes both). A few didn't have either, so I put their A numbers and then put up just about everyone's A numbers afterwards so we could go ahead and create a A metric. Click the picture to enlarge of course (the best bet is to right click and open in new tab and zoom)

 The average OBP in the DWL by the players surveyed was .313, while the OPS was .662. Of the 24 that have MLB experience (at least enough to qualify, I set the cutoff at about 50 games), they averaged an OBP of .299 and OPS of .664. However, we do have a AA metric, so we will project the players that have AA statistics but no MLB statistics and then we will be able to calculate both a A metric and a DWL metric. Projections are in bold (-.058 OBP from AA to MLB, and -.136 OPS from AA to MLB)

So the 48 Projected/Actual MLB players averaged a .279 OBP and .630 OPS. This means that from DWL to MLB, there is a -.034 OBP and -.032 OPS (which means that the Slugging actually improves a little bit in the MLB, suggesting that the DWL is not exactly a friendly slugging environment). For some perspective on Venezuelan League Statistics, I would direct you here: http://mlb.sbnation.com/2012/1/3/2679279/venezuelan-league-stats-carlos-zambrano-bryan-lahair

Now for the A-ball projection. The 34 players we have A-ball and MLB numbers (or projections) for averaged a .342 OBP and .775 OPS in A ball. Those same players averaged a .293 OBP and .659 OPS in the MLB (or MLB projections). This means there is an average regression of -.049 OBP and -.116 OPS.

The Orioles pull off a couple of trades

The Orioles traded cash for Jai Miller from Oakland, and moved Brandon Snyder to Texas for cash. Since we don't (and won't) know the actual cash considerations, we will just compare Jai Miller to Brandon Snyder and see if the player they acquired is better than the one they got rid of (i.e. is Miller better than Snyder).

Miller has played in just 28 MLB games, so it doesn't seem wise to take anything from that. He has played in 423 AAA games, so it does make a lot of sense to look at those statistics. Defensively, he is under average, as his outfield range factor shows. Offensively, he has a .351 OBP, 867 OPS, 1.5 PPG, 2.36 PAPP, which is a 2.42 Simple WAR. This projects to be a .93 Simple WAR with an OBP of .305.

Snyder has just 16 games played in the MLB, so like Miller, we will rely on his minor league numbers. Defensively, he is under average at first base. Offensively at AAA (285 games), he has a .317 OBP, .711 OBP, 1.11 PPG, and 3.05 PAPP. This adds up to a -.43 Simple WAR. So Snyder is a below replacement AAA player.

This makes it quite obvious that the Orioles acquired the better player and dumped the worse one. Assuming the cash considerations were about the same, then it appears that the Orioles made a pretty decent roster move.

Chiba Lotte and CWAR

In Japan, Chiba Lotte has resigned both Hayden Penn and Carlos Rosa, both will make the equivalent of 750,000 American dollars in 2012 (thanks again to Yusuke Ueno @inter_ueno on Twitter). In 2011, Chiba Lotte had a payroll was 2,237 million yen, or 29 million dollars. This means Chiba Lotte has exactly 1 million dollars to spend per Win Above Replacement, using the formula we used in the Lastings Milledge article. This means they need about a .67 WAR from each Rosa and Penn. However, since they are both pitchers, so we have no simple WAR. Since no Simple WAR calculators for pitchers exist (at least as far as I could find), it made a lot of sense to design one.

So I decided to use a formula that contains of FIP (ERA when FIP is not available), PE, Innings, and Fangraphs WAR. What we will do to determine a WAR ratio is add a pitchers PE and FIP, add 100 to that number (so we don't have to deal with negative numbers) divide by innings pitched, and divide that number by Fangraph's WAR. The average ratio we get will be the number we use for CWAR (since BWAR is Baseball Reference WAR and FWAR is for Fangraphs, CWAR will be Clint WAR). For example, Justin Verlander in 2011 had a PE + FIP of -.74, get 99.26 when you add 100, and divide by 251 innings pitched to get about a .395 number. We divide that number from his Fangraphs WAR of 7, to get a 17.7. We will use the average of Verlander's and other's to determine a ratio for WAR.


I used the 2011 Brewers, which you can view by clicking on the image:

The Second picture here is some notable Japanese pitchers whose WARs were calculated here: http://img.ly/brmL thanks to @yonada on Twitter.

As you can see, the Brewers yielded about a 1.75 average ratio, while the top NPB pitchers registered a 1.93 average ratio. We will take the middle ground and use 1.84. This means that when calculation CWAR, take the initial number, which is FIP+PE divided by Innings Pitched, and have it divided by the ratio like this: 1.84/Initial Number.

So back to the pitchers in question. Carlos Rosa had a PE+FIP of 2.8 (you can find the FIPs for most NPB pitchers here. He threw 73.2 innings, giving him (after you add 100 to the PE+FIP) a 1.40 initial number, which translates to a 1.31 CWAR. This means he is certainly worth the 750,000 dollar contract. Rosa also had 31 appearances in the MLB, and earned a 3.912 PE (.4 WAR), or expected 1.866 PE in the NPB. This means that Rosa has just pitched better than expected in Japan, or that the 2011 success was fluky. Either way, I still like this deal for Chiba Lotte.

Hayden Penn has had a 5.84 PE+FIP in 2 years in Japan, earning a 1.25 CWAR over his 2 years, or .625 CWAR per year. This means he is probably slightly overpaid according to his new deal. In the MLB, he had a PE of 4.926 (-4.1 WAR), or a 2.88 Expected PE in NPB. So he is pitching as about expected in the NPB.

How we did: Week 17 Picks

Big D in Bold, I in Italics

Titans edge Texans
Excuse: I knew some starters would be held out, but I didn't think they would hold Jonathan Joseph out, between that and the injury of Yates forcing Delhomme to play, you can see why they lost. They were just a good snap away from winning anyway.
Saints blow out Panthers
Falcons destroy Buccaneers
Chiefs score 1 touchdown against Broncos
Giants rock Cowboys

Big D: 2-3, 57-37 overall, 60.6% success rate
I: 4-1, 59-35 overall, 62.8% success rate. I win the regular season.

White Sox trade Jason Frasor

The White Sox traded Jason Frasor to the Blue Jays for Myles Jaye and Daniel Webb.

Frasor will make 3.75 million in 2012, and had a 1 WAR in 2011. Over the last 3 years, he has a WAR average of 1.13, which would be a WASP of 3319, slightly above the Halladay Standard. He had a -1.03 PE (-.54 Adjusted PE) in 2011, with a career PE of -.978 (-.975 Adjusted PE). His FIP - was 100, which is about league average in 2011. SIERA liked him quite a bit more, at a 3.56, but for just the second time in his career, he gave up more flyballs than groundballs. This could explain why his BABIP was slightly higher than usual (.304 against a .288). He also gave up the most HR/9IP since 2006, a further sign of some kind of regression. It is worth noting, however, that he pitched well if the BlueJays in the first half, but pitched poorly (a 0 WAR) with the White Sox after being traded at the deadline. For whatever its worth, the Bill James projections have him at a 3.69 FIP for 2012.

Jaye has spent 1 short season in rookie ball, making 9 starts in 13 appearances after being drafted in the 17th round. There he had a -1.63 PE, had a decent walk rate but gave up 1.2 HR/9IP.

Daniel Webb has only pitched in A-ball, and has pitched mostly as a starter. He has been pretty bad, with a PE of 3.37. He has kept the ball in the ballpark and hasn't walked too many batters, but still has a BAA of over .280, despite having a reasonable BABIP. He was a hitter in college at Illinois, and had a .903 OPS, which would translate to a .653 OPS in the MLB according to our college metric. So it seems it was the right choice to change him to a pitcher, but it is not like that experiment is going real well either.

So the White Sox get two pretty inconsequential minor leaguers without too much promise (Jaye is certainly better than Webb though), while the Blue Jays get a slightly overpaid apparently regressing reliever. I guess the White Sox have a small advantage in this trade because they cut payroll, but I don't like this trade for either team really (although if Frasor returns to form, it could be advantage Blue Jays).