This is just messing around with the data, having fun with it.

                When I was a kid, I used to think of 41 starts as a full season by a starting pitcher.  In the 1960s you could lead your league in Games Started with 41—or 40, or 39, but somewhere in that neighborhood.  The most games anyone started in the 1960s was 42, but that only happened five times.   41 was a full season.

                I started looking, then, for 41-start strings by pitchers.  I would like to stress that these are actual and legitimate records; these are things that actually happened. They’re not projections or simulations or anything like that; these are records of actual events, just as much as single-season totals are. We’re just looking at the records in a slightly different form than we usually look at them; that’s all.  Checking 41 consecutive starts for a pitcher is like checking 162 consecutive games played for a hitter. 

                OK, I’d better budget myself to 25 points here, or I’ll go on with this stuff forever:

                1)  Dwight Gooden, in 41 consecutive starts beginning August 11, 1984 was 29-5, 352 strikeouts, 74 walks, 1.49 ERA.   This is the fourth-greatest stretch of 41 starts in my data:

                2)  Sandy Koufax, in 41 consecutive starts beginning June 4, 1964, was 31-4 (the same record as Lefty Grove in 1931), with 368 strikeouts and a 1.70 ERA.   This is the third-greatest stretch of 41 starts in my data. 

                3)  Pedro Martinez, in a stretch of 41 starts beginning August 24, 1999 and running into 2001, struck out 434 batters and walked only 48!:

                The 434 strikeouts in a stretch of 41 starts is actually not a record.    But it’s pretty good. 

                4)  Bob Gibson, in a stretch of 41 consecutive starts beginning April 26, 1968 pitched 16 shutouts and posted a 1.19 ERA:

                You can rate them however you want to rate them, but the formula I established ranks this as the greatest 41-start streak in my data.

                Gibson’s stretch of games is famous, but I think, in a sense, that people don’t quite process it because it happened in 1968, and a lot of other great pitching stuff happened in 1968—Drysdale’s scoreless inning streak, McLain winning 31 games, Luis Tiant’s phenomenal season, etc.  I think people kind of write 1968 off as a fluke season all around, and diminish what Gibson did as a part of that. 

                But even if you adjust Gibson’s 1968 campaign to a more normal context, it’s still a phenomenal season.    The National League ERA in 1968 was 2.98.  Gibson’s ERA was 1.12.   That’s 38% of the league.   A normal league ERA, over time, is 3.90, 4.00.  Adjusted to a normal league, Gibson’s ERA is still about 1.50.   How often do we see a pitcher with a 1.50 ERA?

                And the ERA is really only half the story.  The other half is, Gibson was pitching 9 innings a start.   Gibson, during this exact same stretch of games, was also pitching more innings per start than any other pitcher in the data.   He had the lowest ERAs ever and the highest innings per game ever, working at the same time.  That’s a potent combination. 

                5)  Surprisingly, Nolan Ryan does not hold the record for the most strikeouts in a 41-game stretch; we’re so used to Nolan Ryan holding any and every strikeout record you can find that it’s surprising to find him third on the list here.   The most strikeouts by a pitcher in a stretch of 41 starts is 436, by Randy Johnson, beginning either June 19, 1998 or May 15, 1999.    Pedro had 435 in a stretch, and Ryan is third on the list with 406, except that Randy actually has several non-overlapping stretches with more than 406:  May 20, 1994 to September 13, 1995 (407); April 28, 1998, to May 30, 1999 (408); June 4, 1999 to July 4, 2000 (431); and July 9, 2000 to August 8, 2001 (413).   Johnson actually has more than 150 stretches that total up to more than 400 strikeouts, although, of course, most of those are formed by overlapping periods.   There are four pitchers who struck out 400 batters in a stretch of 41 starts:  Randy, Pedro, Nolan and Curt Schilling.  

                6)  Ten pitchers in the data won 30 starts out of a string of 41.   I’ll give you the years in which the stretch started and ended:  Pedro Martinez, 1998 to 2000, (32-6); Denny McLain, 1968 to 1969, (32-6); Sandy Koufax, 1964 to 1965 (31-4); Tom Seaver, 1969 to 1970, (31-7); Ron Guidry, 1977 to 1978 (30-3); Roger Clemens, 1989 to 1991 (30-7); Camilo Pascual, 1963 to 1964, (30-9); Barry Zito, 2001 to 2002, (30-5); Robin Roberts, 1952 to 1953, (30-9); and Bob Welch, 1989 to 1991, (30-7).  

                7)  Bob Gibson holds the 41-start records for Innings Pitched (369.2), ERA (1.11), fewest Earned Runs Allowed (45), Complete Games (38) and Shutouts (16), and that’s all basically one stretch, just shifting the boundaries back and forth a few starts.

                8)  Highest Winning Percentages:   1. Randy Johnson, 1995-1997, (27-2, .931).   2.  Dave McNally, 1968-1969, (26-2, .929).   3.  Ron Guidry, 1978-1979, (30-3, .909).  4.  Matt Latos, 2012-2013, (19-2, .905).    5.  Max Scherzer, 2012-2013, (27-3, .900).   6.  Whitey Ford, 1961-1962, (26-3, .897).  

                Mat Latos is the surprise there; somehow I didn’t realize that that was happening. 

                9.  Worst Winning Percentages:  1. Jose DeLeon, 1984-1985, (3-28, .094).   2.  Matt Keough, 1978-1979 (3-27, .100).   3.  Jerry Koosman, 1977-1978 (3-25, .107).  

                10.  Most Hits Allowed:  Rick Langford, 1980 to 1982, 359, and Bill Lee, 1974-1975, also 359.

                11.   Most Runs Allowed and Earned Runs Allowed:  Mike Hampton, 2001-2002, 209 and 191.  

                12.  Fewest Strikeouts:  Sandy Consuegra, 1952 to 1956, 48 strikeouts.

                13.   Fewest Walks, Carlos Silva, 2004 to 2006; most Walks, Bob Turley, 1954 to 1955:

                14.   The only pitcher in the data to lose 30 out of 41 starts was Roger Craig, 1962-1963:

                Kip Wells lost 29, 2005 to 2007.

                15.  Worst ERA, Russ Ortiz, 2005-2009:

                16.   Highest Run Support:  Jamie Moyer, 1996 to 1997, 307 runs.   7.49 per start.    Lowest Run Support:  Hal (Skinny) Brown, 1962 to 1964, 82 runs.   2.00 per start.   Moyer went 25-7 in his run; Brown went 7-25 in his.

                17.   Longest time span required to make 41 starts:  Lindy McDaniel, August 16, 1957 to June 23, 1974.    McDaniel made 9 starts in the closing weeks of 1957, then made 17 starts in 1958, 7 in 1959, 2 in 1960, 2 in 1962, 3 in 1967, and 3 in 1973.   That totals up to 43, but the stretch reached 41 with the first start in 1973.   Lindy was 10-19 with a 5.34 ERA in his 17-year "season". 

                18.  Shortest time span:  Wilbur Wood, May 5 to September 20, 1972.    138 days.

                19.  Best stretch of 41 games with a losing record:  Nolan Ryan, 1972-1973.    Ryan in a stretch of 41 starts struck out 402 batters, pitched 7 shutouts, and posted a 2.26 ERA—but lost 20 games:

                Part of what makes that stretch absolutely phenomenal is that Ryan did this in less than one calendar year:   July 27, 1972 to July 19, 1973.    Here are some other really good stretches of games in which a pitcher was stuck with a losing log:

                20.   I figured "trend lines" for each pitcher, measuring how fast his 41-game performance line was improving or declining.    The fastest "uptrend" in the data was by Frank Tanana in late August/early September, 1975.    In a stretch of 9 games starting July 19, Tanana went 7-1 with a 1.10 ERA, striking out 83 batters in 82 innings.    This changed Tanana’s running 41-game record as below:

                21,  The fastest DOWNTURN that any pitcher ever took was Gaylord Perry in 1975. Pitching tremendously well in 1974 and early 1975, Perry had a 16-game stretch beginning May 4, 1975, in which he went 3-12 with a 5.07 ERA.  

                Perry actually pitched some good games in that 16-game stretch; what was peculiar about him was that his managers would let him keep pitching even after he got knocked around pretty good, so he also had several games in there where he gave up 7 or 8 runs.  

                22.   Roger Clemens had 417 "strings" of 41 games in which he won 20 games or more.  Clemens had 709 career starts, which means that there are 669 41-game strings for him.  He had 20 or more wins in 417 of those strings, the most of any pitcher.   He had 25 wins in 104 of those strings, and had 200 or more strikeouts in 663 of the 669 strings.  

                Warren Spahn had 20 wins in 91% of the 41-game strings in the data we have for him, which is only a little over half of his career.   Adam Wainwright is at 88%, Roy Halladay 81%.   Chien-Ming Wang was 77%, Bob Gibson 76%, Verlander 75%.    These are the highest percentages for pitchers for whom we have reasonably complete data.

                Nolan Ryan had 200 strikeouts in every 41-start string of his career, Randy Johnson (the Walnut Creek Horror) in 562 out of 563.  

                23.   The 25 best pitchers in the data in their FIRST 41-start stretches:

                Of the 25 best pitchers in the data, in their first 41 starts, only two are now in the Hall of Fame, Ferguson Jenkins and Phil Niekro.    About ten others had truly outstanding careers, however, while only six of the 25 faded or burned out quickly (Hardin, Guzman, Fidrych, Nash, Score and Prior.)

                24.  The best pitcher in the LAST 41 starts of his career was, of course, Koufax, and his record in his last 41 games is simply his last-season record, since Koufax made exactly 41 starts in 1966.     J. R. Richard in his last 41 starts was 22-13, 338 strikeouts, 2.20 ERA.  (The 338 strikeouts for Richard is the record.)     Hoyt Wilhelm, in his last 41 starts (1959-1963), was 18-14, 2.69 ERA.     Don Drysdale—paired with Koufax for almost all of his career—was the fourth-best pitcher in his last 41 games, posting a 2.48 ERA, followed by Roger Nelson (??), 17-12, 2.86, Dave Giusti (15-15, 2.78), and Brandon Webb (26-8, 3.53 ERA).  

                Richard, Nelson and Webb, of course, were finished by sudden injuries while pitching well.   Giusti is interesting because he was moved to the bullpen, by a team that needed starting pitching, at a time when he was pitching well.   He had almost ten good years in the bullpen.  

25.  The actual reason I started this research bit. .. .I was curious about this question.    If pitchers now still worked on a four-man rotation, would we have as many 20-game winners per team now as we did in the 1970s?   In other words, is all of the change in the number of 20-game winners now as opposed to 1965 explained by the change from a four-man to a five-man rotation, or is some of it due to the bullpens, etc.?

I realized that I could get the answer to that question by figuring pitcher records in 41-game strings, and. ..well, I like doing that kind of stuff.  In my data I have 8,974 41-start strings which end in the 1950s.  Of those 8,974 strings, 2,423 include 20 or more wins by the starting pitcher.   This is 27%, and the chart below gives the data for each decade:

                So clearly, the decline in the number of 20-game winners does NOT result entirely from the switch from a 4-man to a 5-man rotation.   Some of it is attributable to increasing use of the bullpens, or to other causes.