Strikeout Quality: Actual vs. Expected Strikeout Ratio – 2010

Traditional measures of a pitcher’s strikeout ability focus on the actual number of strikeouts or the number of strikeouts per nine innings. One of the misleading parts of these measures, though, is that they do not reflect the strikeout frequency of the batters faced. If a pitcher faces batters that strikeout often, that pitcher would be expected to have a higher number of strikeouts and a higher rate of strikeouts per 9 innings. What would be helpful in determining the quality of a strikeout is an adjusted strikeout measure that reflects the strikeout frequency of the batter faced.

To do this, you would need to look at the pitchers actual strikeouts in relation to the number of strikeouts the pitcher would be expected to have given the batters they face. For each batter, I calculated the strikeout frequency for the season, and took this as the frequency with which the average pitcher would strikeout this batter. I then compared the number of times a pitcher actually struck out the batter to the number of times it is expected that an average pitcher would strike out this batter. If the number of actual strikeouts is higher than the number of expected strikeouts, then the pitcher is striking out batters at a higher rate than average. If it is lower, the pitcher is striking out fewer batters than average. I have labeled this ratio the AESR (Average vs. Expected Strikeout Ratio), and it is simply calculated as the actual number of strikeouts divided by the expected number of strikeouts.

2009 Results

The top 10 pitchers in terms of AESR from 2009 were:

Of the pitchers with the top 10 AESR’s in 2009, only three of them made the top 10 in 2010.

2010

For 2010, the pitchers with the top 10 AESR’s (at least 100 innings pitched) are:

Of the pitchers that were in the top 10 in 2009, the pitchers that stayed in the top 10 in 2010 are Jon Lester, Justin Verlander, and Tim Lincecum. Lester remained at 4, while Verlander dropped from 1 to 7 and Lincecum dropped from 3 to 10.

There are also some interesting insights when comparing the AESR rank to the strikeouts per 9 innings rank. Brandon Morrow had the highest AESR and the highest number of strikeouts per nine innings, and was significantly better than the second ranked pitchers in both categories. However, there were significant differences between the AESR and strikeouts per nine innings rank for many of the other pitchers. Of the pitchers in the top 10, Lincecum had the largest negative difference between the AESR rank and the strikeouts per nine innings rank. He was ranked second in strikeouts per nine innings but 10th in AESR. This was due to the fact that although he had 231 strikeouts, the batters he faced struck out more often than average, so his AESR suffered as a result. Conversely, American League Cy Young winner Felix Hernandez had an AESR rank that was 17 places higher than his strikeouts per nine innings rank. His rank in strikeouts per innings pitched was significantly lower because he pitched significantly more innings. He pitched more than 25 innings more than any other pitcher in the top 10, and was second in the major leagues only to National League Cy Young winner Roy Halladay who pitched 250 and 2/3 innings.

Career

The top 10 pitcher AESR’s for pitchers that have pitched more than 2000 innings since 1974 is shown below.

On this list, there are no pitchers that were active in 2010. Interestingly, as dominant a performance as Morrow had in 2010, there are three pitchers who, for their career, have a higher AESR. Nolan Ryan sits significantly above the rest of the field.

Of the pitchers that were active in 2010, the top 10 AESR’s are shown for pitchers with at least 1,000 innings pitched.

Of the pitchers on this list, five are starters and barring any unforeseen circumstances would get to 2,000 innings pitched. Santana is the closest to 2,000 innings pitched, and he would rank 6th on the all time list. At their current AESR, Beckett and Kazmir would make the top 10 in 10th place. It looks as if the top 5 spots are solidified for at least the near future.

Over the next few weeks, in preparation for the start of the 2011 season, I will look at the strikeout performance of the top 5 pitchers in more detail.

Actual vs. Expected Strikeout Rate – June 9 Update

In a blog post on May 24, I introduced a measure for the determination of strikeout quality called the Actual vs. Expected Strikeout Ratio (AESR). You can see that post for a description of the calculation of the AESR. I have updated the AESR ratios for games through June 9, 2010. The updated top 10 in AESR ratio are shown below.

Tim Lincecum

Tim Lincecum still leads the list, even though he ranks just fourth in the major leagues in number of strikeouts per 9 innings. This is because of the 112 different batters that Lincecum has faced, he has struck out 61% of them more than the average pitcher based on the expected strikeout ratio of the batters. Of the batters he has struck out, he strikes them out an average of almost 2.5 times more than average. Of the batters that Lincecum has faced more than once, there are several batters that Lincecum had significant success against. Lincecum has faced Jorge Cantu three times this year and has struck him out all three times, although Cantu normally strikes out only once every six times at the plate. He has also faced Raul Ibanez three times and struck him out twice, even though Ibanez on average once every seven plate appearances. He has also been successful against Shane Victorino, striking him out twice in four plate appearances, although Victorino strikes out in just one out of every eight plate appearances.

Francisco Liriano

Francisco Liriano is a close second to Lincecum, and he ranks high for similar reasons as Lincecum ranks high. Liriano has faced 77 different batters this season, and has struck out 62% of them more than expected based on their average strikeouts per plate appearance. Of the batters he has struck out, he has struck them out on average about 2.3 times more than expected. Of the batters that Liriano has faced more than once, he has had significant success against a number of batters. Gabe Gross has struck out three times in three plate appearances against Liriano, even though on average Gross on strikes out in 18% of his plate appearances. Ramon Santiago and Mike Sweeny have both struck out twice in three plate appearances against Liriano, even though they both strike out on average in just 13% of their plate appearances.

Other Top 10 Pitchers

Of the other pitchers in the top 10 in AESR, Phil Hughes remains in the top 10 even though he does not rank in the top 10 in strikeouts per 9 innings. Adam Wainwright also makes the top 10 even though he ranks 18th in strikeouts per 9 innings.

Bud Norris

The interesting omission from the top 10 in AESR in the pitcher that ranks number one in strikeouts per 9 innings. The current leader is Bud Norris, who is striking out just over 11 batters per 9 innings, however he ranks 11th in AESR. The reason Norris ranks lower in this statistic is because of the 77 different batters he has faced, he has only struck out 49% of them greater than the expected number of times. Of the batters he has struck out, he strikes them out more than 2.5 times than average.

Stephen Strasburg

Stephen Strasburg made his major league debut on June 8th, and the expectations for Strasburg were greater for him than for any other major league debut. By most accounts, Strasburg did not disappoint in his debut. While it is only one game, I could not resist seeing how Strasburg’s 14 strikeouts ranked in terms of the AESR.














As can be seen in the table above, based on the average strikeouts per plate appearance, the average pitcher would have struck out just under five batters. Strasburg’s 14 strikeouts yield an AESR of just under 3, significantly higher than any other pitcher that has pitched at least 40 innings. He struck out all of the batter he faced at least once, and even struck out Lastings Milledge twice even though he strikes out normally only once every seven at bats. It will be interesting to see how his numbers develop as he has more starts and batters see more of Strasburg.

Pujols & Cabrera: Three Home Run Games

On Friday, May 28th, Miguel Cabrera hit three home runs in a 5 – 4 loss to Oakland. Two days later, on Sunday, Albert Pujols hit three home runs in a 9 – 1 win over the Cubs. This was the first time Cabrera hit three home runs in a game, but the fourth time Pujols hit three home runs in a game. Pujols is now tied for sixth on the all time list for the number of games with three or more home runs. At the top of the list with six is Sammy Sosa.

Three home runs in a game is a rare occurrence, happening only an average of just over nine times per year since 1987. Interestingly, two of the previous three times that Pujols hit three home runs in a game, he did it on the same day as another player. On September 3, 2006, Pujols hit three home runs against the Pirates, and on the same day Ryan Howard hit three home runs against the Braves. On July 20, 2004, Pujols hit three home runs against the Cubs on the same day that Travis Hafner hit three home runs against the Angels. On April 16th, 2006, Pujols hit three home runs against the Reds just five days before Alfonso Soriano hit three home runs against the Braves. And when Pujols hit three home runs on Sunday, he did it just two days after Cabrera.

As rare as the occurrence, Pujols tends to hit three home runs in a game within days of others accomplishing the same feat.

Data Source: http://www.baseball-reference.com/

Do More Foul Balls Really Unnerve a Pitcher – Part 2

After some comments on the last blog post (“Do Foul Balls Really Unnerve a Pitcher”), I went a little deeper and looked at the number of foul balls for a fixed number of balls and a two strike count. The conclusion from just looking at the number of foul balls was that it appears that an increased number of foul balls unnerves both the pitcher and the batter. Looking at the number of foul balls based on the count reinforces this conclusion.

Batting Average

Based on seasons from 1988 – 2009, the batting average by number of foul balls and the place in the count is shown below.

As the number of foul balls increases for a fixed ball/two strike count, the batting average increases consistently. The batting average increases between 17 and 50 points over the zero foul ball batting average. It is possible that pitchers actually can become frustrated because they are unable to put the batter away early. The longer the batter stays alive, the more likely it is that they will actually get a hit.

Even though the batting average increases as the number of foul balls increases, the highest it reaches is about .250. The batting average is significantly lower for two strike counts with a fewer number of balls. This points to the unnerving of the batter with the overall lower batting average, and also with the decreasing batting average as the number of balls decreases.

There appear to be a number of factors influencing the result of the at bat, including whether the pitcher is behind or ahead in the count, as well as how long the plate appearance continues. The closer either the pitcher or batter is to losing the at bat, the more the pressure appears to increase.

On Base Percentage
The on-base percentage shows the same trend as the batting average except for the 3/2 count.

For two strike counts with two balls or less, the on base percentage increases as the number of foul balls increases. The amount that the OBP increases over the 0 foul ball count is about 17 to 50 points as well.

For the 3/2 count, however, the on base percentage increases by about 23 points for one foul ball and then stays fairly level as the number of foul balls increases. This is due to the fact that the walk rate for a 3/2 count is about 30% and remains consistent regardless of the number of foul balls. For the 3/2 count, batters end up on base in nearly one out of every two plate appearances.


Strikeout Rate

The strikeout rate by count and number of foul balls tells a similar story of the pressure increasing on both the pitcher and batter.

As the number of foul balls increases for a given count, the strikeout rate decreases. The longer a batter is allowed to hang around, the more difficult it is for the pitcher to put the batter away. This is also shown by the fact that as the number of balls increases for a two strike count, the strikeout rate decreases as well.

Data Source: http://www.retrosheet.org/

Do More Foul Balls Really Unnerve a Pitcher?

I am traveling in San Diego this week, and I have the chance to go see the Padres play the St. Louis Cardinals this evening. Jon Garland gets the start for the Padres. In an interview with Corey Brock from mlb.com on Monday, Garland was asked about his reaction to Phil Hughes enduring 41 foul balls on last Saturday against the Mets, which was the most since Garland’s 42 foul balls in 2005. Garland was partially quoted as saying that “allowing so many foul balls can be unnerving for a pitcher.”

I took a quick look at some data on the number of foul balls in an at-bat and analyzed the ultimate result of the at bat to determine if batters have more success if they are able to keep the at-bat alive longer by fouling off more pitches. The answer is that it seems to be unnerving to both the pitcher and the batter.

Batting Average

I looked at the batting average based on the number of foul balls in an at-bat using the 2005 – 2009 seasons. The batting average is shown below.

Number of Foul Balls - Batting Average
0 - .295
1 - .230
2 - .208
3 - .217
4 - .228
5 - .228
6 - .229

The batting average drops by between 65 and 87 points after just one foul ball. This suggests that as a batter fouls off more pitches, he is less likely to successfully get a hit. This could perhaps be because the batter becomes a bit “unnerved” or frustrated after fouling off a few pitches. This unnerving also be seen in the strikeout rate increasing as the number of foul balls increases.

Number of Foul Balls - Strikeout Rate
0 - 9.5%
1 - 26.5%
2 - 33.0%
3 - 30.3%
4 - 27.7%
5 - 27.2%
6 - 26.4%

The strikeout rate triples as the number of foul balls increases from 0 to 2 or more. This could actually increase the confidence of the pitcher, knowing it is easier to strike the batter out.

On Base Percentage

The on base percentage of a batter as the number of foul balls increases is shown below.

Number of Foul Balls - On Base Percentage
0 - .359
1 - .296
2 - .286
3 - .316
4 - .337
5 - .350
6 - .343

The on base percentage decreases as a batter fouls off one or two balls, but then begins to increase with the third foul and ends up near the on base percentage for the 0 foul on base percentage level. This suggests that as a player fouls off more pitches they are more likely to end up on base. The obvious reason for this is that as there are more pitches thrown, it is more likely that the pitcher walks the batter. The rate of batters either walked or hit by a pitch by the number of foul balls is shown below.


Number of Foul Balls - Walk Rate
0 - 9.9%
1 - 8.7%
2 - 10.0%
3 - 12.7%
4 - 14.2%
5 - 16.0%
6 - 14.8%

The walk rate increases from 10% to as high as 16% as more pitches that are fouled off. This could be some evidence of the “unnerving” of the pitcher.

It seems that as the stress of any sports situation increases, nerves can possibly come into play. As a batter and a pitcher get deeper into an at bat, nerves can play a part on both sides.

Strikeout Quality: Actual vs. Expected Strikeout Ratio

As of May 23rd, Brandon Morrow of the Toronto Blue Jays was leading the major leagues in strikeouts per 9 innings, with 11.8. Tim Lincecum of the San Francisco Giants was leading the majors in the number of strikeouts, with 75. These measures have long been used as a measure of the ability of a pitcher to strike out batters, with strikeouts per 9 innings being especially useful. One of the misleading parts of these measures, though, is that they do not reflect the strikeout frequency of the batters faced. If a pitcher faces batters that strikeout often, that pitcher would be expected to have a higher number of strikeouts and a higher rate of strikeouts per 9 innings. What would be helpful in determining the quality of a strikeout is an adjusted strikeout measure that reflects the strikeout frequency of the batter faced.

Actual vs. Expected Strikeout Ratio

To do this, you would need to look at the pitchers actual strikeouts in relation to the number of strikeouts the pitcher would be expected to have given the batters they face. For each batter, I calculated the strikeout frequency for the season, and took this as the frequency with which the average pitcher would strikeout this batter. I then compared the number of times a pitcher actually struck out the batter to the number of times it is expected that an average pitcher would strike out this batter. If the number of actual strikeouts is higher than the number of expected strikeouts, then the pitcher is striking out batters at a higher rate than average. If it is lower, the pitcher is striking out fewer batters than average. I have labeled this ratio the AESR (Average vs. Expected Strikeout Ratio), and it is simply calculated as the actual number of strikeouts divided by the expected number of strikeouts.

For example, in 2010, Mark Reynolds has struck out in nearly 40% of his plate appearances. If a pitcher faces Reynolds 10 times, given his propensity to strikeout, the pitcher would be expected to strike him out four times. If the pitcher struck out Reynolds six times, he would be doing better than average (AESR = 6/4 = 1.5). If he struck Reynolds out twice, he would be doing worse than average. Conversely, consider David Eckstein, who has struck out only four times in 158 plate appearances. Striking out Eckstein even once in ten tries would be four times better than average.

2010 Season

For 2010, as of May 23rd, the top 10 pitchers in strikeouts per 9 innings are (for pitchers with at least 30 innings pitched):

1 - Brandon Morrow (TOR) - 11.80
2 - Bud Norris (HOU) - 11.22
3 - Tim Lincecum (SFG) - 11.07
4 - Hisanori Takahas (NYM) - 10.69
5 - Carlos Zambrano (CHC) - 10.27
6 - Tommy Hanson (ATL) - 10.10
7 - Yovani Gallardo (MIL) - 10.07
8 - Clayton Kershaw (LAD) - 10.02
9 - Jered Weaver (LAA) - 9.92
10 - James Shields (TBR) - 9.68

For pitchers in the top 10 in strikeouts per 9 innings in 2010, most of them remain relatively high in the AESR rank. In this case, Morrow, the top pitcher in strikeouts per 9 innings, has ann AESR of 1.60, which means hestrikes out about 60% more batters than the average pitcher. Lincecum, who ranks at the top in AESR, strikes out 65% more batters than the average pitcher. Lincecum performs well at striking out both batters that strike out a lot as well as those that do not strike out as often. Lincecum has faced 83 different batters so far this year, of those 83 he has struck out two-thirds of them at least once.

One of the most interesting names on this list is Carlos Zambrano. Although he is fifth on the list of strikeouts per 9 innings, he ranks 28th in AESR (1.20), only striking out 20% more batters than average. Before he was placed in the bullpen, Zambrano was performing impressively in the strikeout category based on strikeouts per 9 innings, and his AESR of 1.20 is still better than average. However, relative to the other pitchers in the top 10, he has not performed as well in striking out more batters than expected.

The top 10 pitchers in AESR for 2010 are:
1 - Tim Lincecum (SFG) - 1.646
2 - Hisanori Takahas (NYM) - 1.606
3 - Brandon Morrow (TOR) - 1.604
4 - Jon Lester (BOS) - 1.553
5 - Clayton Kershaw (LAD) - 1.441
6 - Phil Hughes (NYY) - 1.404
7 - Justin Masterson (CLE) - 1.391
8 - Jered Weaver (LAA) - 1.390
9 - Yovani Gallardo (MIL) - 1.387
10 - Ricky Romero (TOR) - 1.381

On this list, the biggest jump from rank in strikeouts per 9 innings to actual vs. expected rank in by Phil Hughes, who rose 18 spots. One of the main reasons is that although Hughes does not strike out as many batters as others, of the 65 different batters he has faced, he has struck out 60% of them at least once.

While the comparison of actual to expected strikeouts does track with strikeouts per nine innings for many pitchers, there are some exceptions with which it does not track.

2009 Season

For the 2009 season, the top 10 list of pitchers with the best actual to expected strikeout ratio is shown below.

1 - Justin Verlander - 1.522
2 - Rich Harden - 1.512
3 - Tim Lincecum - 1.506
4 - Jon Lester - 1.486
5 - Zack Greinke - 1.468
6 - Javier Vazquez - 1.415
7 - Jake Peavy - 1.376
8 - Yovani Gallardo - 1.368
9 - Clayton Kershaw - 1.349
10 - Ricky Nolasco - 1.347

All ten of these pitchers were in the top ten in strikeouts per 9 innings in 2009 except for Ricky Nolasco, who was 11th. The pitcher that was in the top 10 in strikeouts per 9 innings that did not make the top 10 in AESR was Jonathan Sanchez. Last year, Sanchez only struck out about 30% more batters than the average pitcher, placing him 12th on the list.

The top pitcher on the list in 2009 was Justin Verlander. Verlander was effective at striking out a lot of batters. Of the 191 different batters he faced throughout the season, the struck out 121 of them at least once, but no batters more than 6 times throughout the year (Nick Punto and Mike Jacobs). Of the batters that Verlander struck out in 2009, he struck them out an average of 1.83 times more than expected.

Career

To place this in historical perspective, I have calculated the actual vs. expected ratio for pitchers from 1974 to 2009 that have pitched at least 2,000 innings. This will help place in context the greater significance of some of the more recent numbers shown above. The top 10 AESR’s for this time frame are:

1. Nolan Ryan - 1.744
2. Randy Johnson - 1.626
3. Pedro Martinez - 1.622
4. Roger Clemmons - 1.445
5. Curt Schilling - 1.398
6. David Cone - 1.349
7. Ron Guidry - 1.329
8. Tom Gordon - 1.328
9. Steve Carlton - 1.295
10. John Smoltz - 1.277

The top three pitchers in 2010 are off to an AESR start that rank in the top three, but what makes them seem less impressive is that the numbers shown above were accomplished over long careers. Nolan Ryan, for his career, struck out just under 75% more batters than the average pitcher during his career. The next closest pitcher, Randy Johnson, is almost 12 points behind.

There are several active pitchers on the list above, but none with long enough left in their careers to catch Nolan Ryan. Among active pitchers with significant time left in their careers, the closest is Kerry Wood at 48%. Given his current role as a closer, he could improve his numbers over the remainder of his career, but it would be next to impossible to even get close to Ryan. Rich Hardin is the next closest at about 46% over his career, but even he would have a long way to go to even challenge Ryan for the top spot in recent history. This is yet another reminder of just how special Ryan’s career really was.

Data Sources: www.retrosheet.org and www.baseball-reference.com

Welcome to the Baseball Data Analytics Blog

Welcome to the Baseball Data Analytics Blog. My name is Roosevelt Mosley, and I have worked in the statistics and data analysis field for over 15 years. Over the past few years I have been analyzing baseball statistics, and I created this blog to share some of my thoughts and to hopefully get some feedback and additional contributions. Hopefully you will find the posts interesting.

I welcome all comments, thoughts, and suggestions for additional analysis.

Enjoy.

Roosevelt