Saturday, October 31, 2015

Wall Street Journal Book Review Seems To Have Error About 1919 White Sox

See Was It So, Shoeless Joe? Here is the letter I sent them:

"David M. Shribman's review of "The Betrayal" by Charles Fountain ("Was It So, Shoeless Joe?" Oct. 31) seems to have a major statistical error. It reads "In the Series they (the White Sox) batted .255, significantly higher than their .215 season average." But Baseball Reference shows the White Sox to have a team batting average of .287 during the 1919 season while batting just .224 during the World Series.
http://www.baseball-reference.com/postseason/1919_WS.shtml"

Thursday, October 22, 2015

Did Chris Sale Have Bad Luck This Year?

It seems like he pitched worse in high leverage situations than otherwise this year and that has not been the case before. The table below shows his OPS allowed for High, Medium and Low leverage situations from 2012-15. All data from Baseball Reference and the BR Play Index.


Year High  Medium  Low 
2012 0.639 0.626 0.715
2013 0.556 0.618 0.693
2014 0.612 0.522 0.607
2015 0.709 0.658 0.616

So he allowed a much higher OPS in High leverage cases than otherwise. But it looks like he did better in High leverage cases in 2012 & 2013. So maybe he had some good luck then. My guess is that most pitchers have this kind of pattern, where they have bad luck one year and good luck the next or vice-versa (I hope I can take a look at that).

But I did look at how pitchers in general did over the years 2012-15 (with stats for each pitcher being cumulative). I only looked at pitchers who had 100+ IP in all three leverage situations, so I think that means only starters got in (or mostly starters-the only guy with under 500 total IP was Wade Davis and I took him out).

Then I calculated a weighted average (by batters faced) of the OPS allowed for all three leverage situations over the years 2012-15 for this group. There were 72 pitchers (Chris Sale was removed as well). Here are the results:


High  0.706
Medium  0.706
Low 0.692

So pitchers generally do just a bit worse in High and Medium cases than Low cases (in the long run, as I combined years). So Sale might have had bad luck this year, after some good luck in the past.

Thursday, October 15, 2015

Dizzy Dean's 1934 Season, When He Went 30-6

The Cardinals came in first place, just 2 games ahead of the Giants. The Giants were only one game out going into the final day of the season. Dean won a complete game on that day.

Dean was outstanding the last two months of the season. In August, his ERA was 1.41 in 51 IP (46H, 33K, 10BB, 2 HR).

In September, his ERA was 1.54 in 64.1 IP (45H, 41K, 16BB, 3HR). He started 7 games, completing 6 AND he pitched 5 games in relief (for the whole season, he started 33 games and pitched 17 in relief).

In the last 6 days of the season, he pitched 3 complete games, allowing just a total of 2 earned runs (one game was on 2 days rest and the other was on 1 day of rest!!). In those three games, he had 20H, 19K, 3BB, 1HR. He won all three games.

He saved 2 games in September (he had a total of 7 saves that year with a 4-2 record as a reliever). On September 11, he had a 2.1 IP save the day after pitching a complete game. He also pitched one-third of an inning on September 12.

He did have a gap of 12 days when he did not pitch at all, from August 12-August 24. He had a 10 day suspension. For details, see A Closer Look: Dizzy Dean Wins 30 in 1934 by Gabriel Schechter. There was also a victory awarded to him in a questionable decision by the official scorer. Otherwise, he wins only 29.

Twice he pitched 3 innings in relief the day after pitching 9 innings. Another time it was 2.1 IP the day after pitching 9 innings. Still another time it was 2 IP the day after pitching 8 innings. In all, he pitched in relief six times the day after starting and going at least 8 innings.

He also pitched in relief in both ends of a double header on September 23rd, after a complete game on Sept. 21. Then he pitched a complete game on Sept. 25 followed by CGs on Sept. 28 & Sept. 30!

The he started game 1 of the World Series on October 3, on just 2 days rest and got a complete game victory. His next start was Oct. 7. He lost in 8 IP, but it was still a complete game. He then started game 7 on October 9, on only one day of rest. He pitched a complete game shut out against a lineup that included the following Hall of Famers:

Mickey Cochrane
Charlie Gehringer
Goose Goslin
Hank Greenberg 

So he had 6 starts in a 15 day stretch and completed all of them. He allowed only 5 earned runs in 26 IP in the World Series.

From Baseball Reference:

Dean's 1934 splits

Dean's 1934 game logs
 
Dean's post season game logs

Wednesday, October 14, 2015

Post Season Game Scores Of 90+

They are from the Baseball Reference Play Index. If you click on the link you may need to be a subscriber to see it. You might try this one, too.

The list is posted below, although not without all the stats for each pitcher. If you click on the date, you can see the box score.

Roger Clemens has the highest ever, 98, in the 2000 ALCS. In 9 IP.

Whose record did he break? (2 guys, each had 97)

Those two guys each pitched more than 9 innings. So whose 9 IP record did Clemens break? (2 guys, each had 94)

Who is the only guy to have 2 90+ Game Scores in post season history?

Who had the only no-decision with a 90+? (his team also lost the game) There are 21 such games (going back to 1903) and 20 of them gave the pitcher a win.

Who are the only teammates to have 90+ Game Scores in the same year?

Who are the only pitchers to reach 90 in less than 9 IP? (2 guys)

This was all easy to find with the Baseball Reference Play Index. I was wondering about something and then when I saw the list, I thought it would make for some interesting questions.  

Player Date Tm Series App,Dec IP GSc
Roger Clemens 10/14/2000 NYY ALCS SHO9, W 9 98
Dave McNally 10/5/1969 BAL ALCS SHO11, W 11 97
Babe Ruth 10/9/1916 BOS WS CG 14, W 14 97
Tim Lincecum 10/7/2010 SFG NLDS SHO9, W 9 96
Roy Halladay 10/6/2010 PHI NLDS SHO9, W 9 94
Don Larsen 10/8/1956 NYY WS SHO9, W 9 94
Ed Walsh 10/11/1906CHW WS SHO9, W 9 94
Josh Beckett 10/12/2003 FLA NLCS SHO9, W 9 93
Ken Holtzman 10/9/1973 OAK ALCS CG 11, W 11 93
Bob Gibson 10/2/1968 STL WS SHO9, W 9 93
Kevin Brown 9/29/1998 SDP NLDS GS-8, W 8 92
Randy Johnson 10/28/2001 ARI WS SHO9, W 9 91
Randy Johnson 10/16/2001 ARI NLCS SHO9, W 9 91
Cliff Lee 10/18/2010 TEX ALCS GS-8, W 8 90
Livan Hernandez 10/12/1997 FLA NLCS CG 9, W 9 90
Nolan Ryan 10/14/1986 HOU NLCS GS-9 9 90
Mike Scott 10/8/1986 HOU NLCS SHO9, W 9 90
Vida Blue 10/8/1974 OAK ALCS SHO9, W 9 90
Monte Pearson 10/5/1939 NYY WS SHO9, W 9 90
George Earnshaw 10/6/1931 PHA WS SHO9, W 9 90
Bill Dinneen 10/2/1903 BOS WS SHO9, W 9 90

Thursday, October 8, 2015

Interesting Splits For The Playoff Teams

To see the splits for MLB over the years 2010-14, go to Major League Situational Stats, 2010-2014. Here they are to provide reference for the numbers I give below:

Compiled using the Baseball Reference Play Index.


Split PA BA OBP SLG OPS
Total 923779 0.254 0.319 0.398 0.717
None On 520158 0.249 0.310 0.393 0.702
Men On 403621 0.261 0.332 0.405 0.737
RISP 238074 0.255 0.339 0.394 0.733
Late & Close 153559 0.240 0.316 0.365 0.681

Here is what I have for the years 1991-2000. The relative differences are not too much different than they used to be.

At the end of this post I provide link for the batting and pitching splits pages of each team.

Rangers have an OPS differential of -.008 (.739-.747). But somehow they won 88 games. It could be their bullpen that made the difference. They allowed an OPS of just .667 in Late & Close situations. That is .080 below their overall OPS allowed, a much bigger than normal differential.

The Blue Jays had OPS of .878 with men on. With None on it was .735. 4 times normal gap. With RISP, it was .839. Again, abnormally high when compared to the None On OPS. Their batters had just a .716 OPS in Late & Close situations. Overall, it was .797. So they were abnormally low in Late & Close situations. Maybe that partly explains their 15-28 record in 1-run games.

Their pitchers allowed an OPS of .769 with RISP while with None On it was .688. That is much bigger than normal.

The Royals had an OPS differential of just .024 (.734 - .710). Using my estimate of

Winning Pct = .5 +1.3*OPSDIFF

That gives them a .531 pct., lower than their actual pct. of .586. So how did they do so well? They had a .778 OPS with runners on and .701 with none one.  With RISP it was .772. Their pitchers allowed an OPS of just .668 with RISP while it was .700 with none on. That is impressive given that teams usually hit better with RISP.

The Royals pitchers allowed an OPS in Late & Close situations of just .629, far below their overall OPS allowed of .710. That beats the normal differential by quite a bit.

The Astros had an OPS differential of .071 (.752 - .681).  Astros estimate to .592  but they actually had only .531. Their pitchers allowed a .660 OPS with None On but it was .713 with Men On and .707 with RISP. They allowed .684 Late & Close situations. So that is a weakness, given that normally it is lower than overall OPS allowed.

The Cardinals had an OPS differential of .040 (.716 - .676). Their hitters had a .713 OPS with None On but it only goes up to .719 with Men On and actually falls to .683 with RISP. In Late  Close situations it is .713, so it drops off alot less than normal.

But the Cardinal pitchers have been great with Men On and with RISP. The allow a .711 OPS with None On but in the other two it is .618 & .630, respectively. Those are very large differentials considering that they usually go up.

The Cubs had an OPS differential of .057 (.719 - .662). So larger than the Cardinals'. Nothing unusual about their hitters but their pitchers allowed a .626 OPS with None On but it shoots way up with Men On and with RISP where it is .712 and .718. So the reverse of the Cardinals.They allowed only .608 in Late  Close situations, a bigger drop off than normal.

The Dodgers had an OPS differential of .070 (.739 - .669).  Nothing unusual about their hitters but their pitchers allowed a .633 OPS with None On but it shoots way up with Men On and with RISP where it is .724 and .721. They also allow .694 in Late & Close situations, well above their normal level, the reverse of the MLB average.

The Mets had an OPS differential of .040 (.712 - .672). The Mets don't appear to to have any interesting splits.

Dodger batting splits
Dodger pitching splits

Mets batting splits
Mets pitching splits
 
Cardinals batting splits
Cardinals pitching splits
 
Cubs hitting splits
Cubs pitching splits
 
Royals batting splits
Royals pitching splits
 
Astros batting splits
Astros pitching splits
 
Blue Jays batting splits
Blue Jays pitching splits
 
Rangers batting splits
Rangers pitching splits

Sunday, October 4, 2015

Andrelton Simmons Will Likely Join Elite Who Have Led Their League In Defensive WAR Three Straight Years Or More

He leads the NL with 3.5 and Brandon Crawford has 2.9 going into today, the last day of the season. He would be the 12th guy to do it. Joe Tinker led the NL every year from 1905-09 except 1907, when his double play mate at 2B, Johnny Evers, led the league. Evers was 1st with 3.3 and Tinker was 2nd with 2.8.

Here are the players who have done it in the AL


Year American League dWAR
2001   Rey Sanchez   (KCR) 2.4
2000   Rey Sanchez   (KCR) 2.8
1999   Rey Sanchez   (KCR) 2.7
1991   Cal Ripken+   (BAL) 3.4
1990   Cal Ripken+   (BAL) 3.4
1989   Cal Ripken+   (BAL) 3.4
1988   Ozzie Guillen   (CHW) 3.4
1987   Ozzie Guillen   (CHW) 3.7
1986   Ozzie Guillen   (CHW) 3.1
1978   Mark Belanger   (BAL) 3.7
1977   Mark Belanger   (BAL) 3.4
1976   Mark Belanger   (BAL) 3.9
1975   Mark Belanger   (BAL) 4.9
1974   Mark Belanger   (BAL) 3.5
1973   Mark Belanger   (BAL) 4
1936   Billy Rogell   (DET) 1.5
1935   Billy Rogell   (DET) 2.7
1934   Billy Rogell   (DET) 2.2
1915    George McBride   (WSH) 2.2
1914    George McBride   (WSH) 3
1913   George McBride   (WSH) 3.6
1912   George McBride   (WSH) 2.7

Now the NL


Year National League dWAR
2015   Andrelton Simmons   (ATL) 3.5
2014   Andrelton Simmons   (ATL) 3.9
2013   Andrelton Simmons   (ATL) 5.4
2002   Andruw Jones   (ATL) 2.3
2001   Andruw Jones   (ATL) 2.9
2000   Andruw Jones   (ATL) 2.6
1989   Ozzie Smith+   (STL) 4.7
1988   Ozzie Smith+   (STL) 3.1
1987   Ozzie Smith+   (STL) 2.4
1949   Pee Wee Reese+   (BRO) 2.3
1948   Pee Wee Reese+   (BRO) 1.7
1947   Pee Wee Reese+   (BRO) 2.1
1922   Dave Bancroft+   (NYG) 2.7
1921   Dave Bancroft+   (NYG) 3.2
1920   Dave Bancroft+   (2TM) 4
1919   Art Fletcher   (NYG) 3.9
1918   Art Fletcher   (NYG) 3.4
1917   Art Fletcher   (NYG) 5.1

Thursday, October 1, 2015

Is Manny Machado The Best Young 3B Man Since Eddie Mathews?

Maybe. Here are all the seasons of 6+ WAR for 3B men through age 22 using the Baseball Reference Play Index:

Rk Player WAR Year Age
1Dick Allen 8.8 1964 22
2 Eddie Mathews 8.3 1953 21
3 Eddie Mathews 7.8 1954 22
4 Jimmy Williams 6.9 1899 22
5 Manny Machado 6.5 2015 22
6 Manny Machado 6.4 2013 20
7 Hank Blalock 6.4 2003 22
8 Aurelio Rodriguez 6.4 1970 22
9 Freddie Lindstrom 6.3 1928 22

The only two guys to have two such seasons are Machado and Mathews. If the age is raised to 23, here is the list, Mathews has three and then Machado still has two. Only Dick Allen joins the group.

Rk  Player WAR Year Age
1 Dick Allen 8.8 1964 22
2 Eddie Mathews 8.3 1953 21
3 Troy Glaus 7.8 2000 23
4 Eddie Mathews 7.8 1954 22
5 George Brett 7.5 1976 23
6 Eddie Mathews 7.3 1955 23
7 Evan Longoria 7.0 2009 23
8 Frankie Frisch 6.9 1921 23
9 Jimmy Williams 6.9 1899 22
10 Ron Santo 6.7 1963 23
11 Rogers Hornsby 6.7 1919 23
12 Scott Rolen 6.6 1998 23
13 Manny Machado 6.5 2015 22
14 Manny Machado 6.4 2013 20
15 Hank Blalock 6.4 2003 22
16 Aurelio Rodriguez 6.4 1970 22
17 Dick Allen 6.4 1965 23
18Freddie Lindstrom 6.3 1928 22
19 Edgardo Alfonzo 6.2 1997 23
20 Gary Sheffield 6.2 1992 23
21Eric Chavez 6.0 2001 23

Here is the top 10 in WAR through age 22:


Rk Player WAR From To Age
1 Eddie Mathews 18.4 1952 1954 20-22
2 Manny Machado 16.9 2012 2015 19-22
3 Buddy Lewis 12.6 1935 1939 18-22
4 Freddie Lindstrom 12.5 1924 1928 18-22
5 Buddy Bell 9.6 1972 1974 20-22
6 Bill Dahlen 9.0 1891 1892 21-22
7Bob Horner 8.9 1978 1980 20-22
8 Dick Allen 8.8 1963 1964 21-22
9 Denny Lyons 8.3 1885 1888 19-22
10 Adrian Beltre 8.2 1998 2001 19-22

If we raise the age to 23 we have


Rk Player WAR From To Age
1Eddie Mathews 25.7 1952 1955 20-23
2 George Davis 18.2 1890 1894 19-23
3 Manny Machado 16.9 2012 2015 19-22
4 Buddy Lewis 16.2 1935 1940 18-23
5Dick Allen 15.2 1963 1965 21-23
6 Freddie Lindstrom 15.0 1924 1929 18-23
7George Brett 14.5 1973 1976 20-23
8 Denny Lyons 13.8 1885 1889 19-23
9 Buddy Bell 12.5 1972 1975 20-23
10 Billy Nash 12.1 1884 1888 19-23