Username:	Password:

Remember me

Forgot your username/password?

Print Email

Home>Articles

Lady Luck and the 20-Game Winner

By Bill James

August 14, 2017

2017-39

Lady Luck and the 20-Game Winner

Most pitchers who win 20 games benefit from at least a little bit of good fortune in reaching that number.

Since 1950 there have been 353 pitchers who won 20 games. Of those 353, in my est bestimates, 123 won 20 or more games purely on merit, but 230 were pitchers who "should" have won 19 or less, but benefitted from some good fortune. 65% of twenty-game winners are luck-assisted. On the other side, 77 pitchers since 1950 should have won 20 or more games, but fell short of that standard due to bad luck.

It’s not pure luck that makes 20 game winners; it is mostly skill mixed with a little bit of good luck. Of the 240 who won 20 games but should have had 19 wins or less:

38 should have been 19-game winners,

48 should have been 18-game winners,

57 should have been 17-game winners,

38 should have been 16-game winners,

25 should have been 15-game winners,

17 should have been 14-game winners,

6 should have been 13-game winners, and

One should have been a 12-game winner.

I pointed out in a previous article that Red Ruffing, who won 20 games every year from 1936 to 1939, should have won 19 every year. Adam Wainwright is like that. He has won 20 games twice. He was a little bit lucky, but not all that lucky; he should have been 19-8 both years. Randy Johnson in 1997 was 20-4; he was lucky. He should have been 19-5. OK, he was a little bit lucky; he wasn’t really lucky. The really lucky 20-game winners were:

						Deserved
First	Last	Team	Year	Won	Lost	Wins	Losses
Johnny	Sain	Boston Braves	1950	20	13	14	18
Billy	Hoeft	Detroit Tigers	1956	20	14	14	14
Denny	McLain	Detroit Tigers	1966	20	14	14	16
Jim	Merritt	Cincinnati Reds	1970	20	12	14	12
Dave	McNally	Baltimore Orioles	1971	21	5	14	11
Stan	Bahnsen	Chicago White Sox	1972	21	16	14	15
Vida	Blue	Oakland A's	1973	20	9	14	16
Joe	Niekro	Houston Astros	1980	20	12	14	15
Bob	Welch	Oakland A's	1990	27	6	14	13
Scott	Erickson	Minnesota Twins	1991	20	8	14	9
Bill	Gullickson	Detroit Tigers	1991	20	9	14	12
Jack	Morris	Toronto Blue Jays	1992	21	6	14	13
John	Burkett	San Francisco Giants	1993	22	7	14	12
Rick	Helling	Texas Rangers	1998	20	7	14	10
Tim	Hudson	Oakland A's	2000	20	6	14	10
Jamie	Moyer	Seattle Mariners	2001	20	6	14	10

Bob	Grim	New York Yankees	1954	20	6	13	10
Ray	Sadecki	St. Louis Cardinals	1964	20	11	13	12
Catfish	Hunter	Oakland A's	1973	21	5	13	16
John	Smiley	Pittsburgh Pirates	1991	20	8	13	10
Andy	Pettitte	New York Yankees	2003	21	8	13	11
Russ	Ortiz	Atlanta Braves	2003	21	7	13	12
Jered	Weaver	LAA Angels	2012	20	5	13	9

Bob	Forsch	St. Louis Cardinals	1977	20	7	12	13

How did those pitchers do in the following seasons? This chart adds their won-lost records in the next seasons:

						Deserved		Next Season
First	Last	Team	Year	Won	Lost	Wins	Losses	Won	Lost
Johnny	Sain	Boston Braves	1950	20	13	14	18	7	14
Billy	Hoeft	Detroit Tigers	1956	20	14	14	14	9	11
Denny	McLain	Detroit Tigers	1966	20	14	14	16	17	16
Jim	Merritt	Cincinnati Reds	1970	20	12	14	12	1	11
Dave	McNally	Baltimore Orioles	1971	21	5	14	11	13	17
Stan	Bahnsen	Chicago White Sox	1972	21	16	14	15	18	21
Vida	Blue	Oakland A's	1973	20	9	14	16	17	15
Joe	Niekro	Houston Astros	1980	20	12	14	15	9	9
Bob	Welch	Oakland A's	1990	27	6	14	13	12	13
Scott	Erickson	Minnesota Twins	1991	20	8	14	9	13	12
Bill	Gullickson	Detroit Tigers	1991	20	9	14	12	14	13
Jack	Morris	Toronto Blue Jays	1992	21	6	14	13	7	12
John	Burkett	San Francisco Giants	1993	22	7	14	12	6	8
Rick	Helling	Texas Rangers	1998	20	7	14	10	13	11
Tim	Hudson	Oakland A's	2000	20	6	14	10	18	9
Jamie	Moyer	Seattle Mariners	2001	20	6	14	10	13	8

Bob	Grim	New York Yankees	1954	20	6	13	10	7	5
Ray	Sadecki	St. Louis Cardinals	1964	20	11	13	12	6	15
Catfish	Hunter	Oakland A's	1973	21	5	13	16	25	12
John	Smiley	Pittsburgh Pirates	1991	20	8	13	10	16	9
Andy	Pettitte	New York Yankees	2003	21	8	13	11	6	4
Russ	Ortiz	Atlanta Braves	2003	21	7	13	12	15	9
Jered	Weaver	LAA Angels	2012	20	5	13	9	11	8

Bob	Forsch	St. Louis Cardinals	1977	20	7	12	13	11	17
Average				21	9	14	13	12	12
Winning Percentage				.705		.523		.504

Catfish Hunter, after an atypical season in 1973, had deserved won-lost records of 23-13 and 23-14 the next two seasons, and exactly matched his deserved record in 1975 (23-14). But many of these pitchers had crash-and-burn seasons the next year. It is what I refer to as the Reality Bites syndrome.

When we get lucky, we inevitably start to believe that we weren’t lucky; we are just really good. All of us do. When a movie director makes a hit movie, he doesn’t think, "I was sure lucky to get that script, and to get that actor to work with me." He thinks he’s a genius, and he can produce great movie after great movie from now on. It’s inevitable. If you invest in a few stocks that do well, you think you can pick winners in the stock market.

When a pitcher who has been lucky one year starts to contend with ordinary luck or some tough luck the next year, he thinks he’s lost it. "What am I doing wrong now?" he asks. "I was so good last year; what have I lost?"

Based on that assumption, he starts to make adjustments, trying to reclaim a magic which in reality was external to him, but which he believes was a part of him. Sometimes those adjustments backfire on him, and the next thing you know, he’s completely messed up.

There are also 77 pitchers since 1950 who deserved to win 20 games, but did not. These are those 77 pitchers. I will bold-face those pitchers who got cheated out of a 20-win season, and never did have one:

					Deserved
First	Last	Year	W	L	Wins	Losses
Joaquin	Andujar	1982	15	10	20	10
Kevin	Appier	1993	18	8	20	7
Vida	Blue	1976	18	13	22	12

Bert	Blyleven	1971	16	15	20	12
Bert	Blyleven	1972	17	17	20	13
Bert	Blyleven	1974	17	17	20	12
Bert	Blyleven	1975	15	10	20	11
Bert	Blyleven	1976	13	16	20	14
Bert	Blyleven	1985	17	16	20	13

Kevin	Brown	1996	17	11	20	7
Kevin	Brown	1998	18	7	20	9

Jim	Bunning	1965	19	9	20	13
Jim	Bunning	1966	19	14	24	12
Jim	Bunning	1967	17	15	22	12

Dean	Chance	1968	16	16	20	13

Roger	Clemens	1988	18	12	22	9
Roger	Clemens	1991	18	10	22	9
Roger	Clemens	1992	18	11	20	8

Don	Drysdale	1964	18	16	23	13
Don	Drysdale	1960	15	14	21	10

Bob	Friend	1960	18	12	20	12

Bob	Gibson	1964	19	12	21	12
Bob	Gibson	1972	19	11	21	10

Zack	Greinke	2009	16	8	21	6
Zack	Greinke	2015	19	3	21	6

Roy	Halladay	2011	19	6	20	8
Ken	Holtzman	1970	17	11	21	12

Randy	Johnson	1999	17	9	23	8
Randy	Johnson	2000	19	7	22	8
Randy	Johnson	2004	16	14	21	8

Clayton	Kershaw	2013	16	9	20	7
Clayton	Kershaw	2015	16	7	20	7

Jimmy	Key	1987	17	8	20	10
Bill	Lee	1973	17	11	20	12

Tim	Lincecum	2008	18	5	20	7
Tim	Lincecum	2009	15	7	20	7

Greg	Maddux	1995	19	2	21	3
Greg	Maddux	1996	15	11	20	8
Greg	Maddux	1997	19	4	21	6
Greg	Maddux	1998	18	9	21	8
Greg	Maddux	2000	19	9	20	9


					Deserved
First	Last	Year	W	L	Wins	Losses
Pedro	Martinez	1997	17	8	22	6
Pedro	Martinez	2000	18	6	22	3

Jon	Matlack	1974	13	15	20	10
Sam	McDowell	1965	17	11	21	10
Andy	Messersmith	1975	19	14	23	14
Joe	Niekro	1982	17	12	20	11

Phil	Niekro	1977	16	20	21	17
Phil	Niekro	1978	19	18	24	14

Gaylord	Perry	1967	15	17	20	14
Gaylord	Perry	1969	19	14	22	15
Gaylord	Perry	1973	19	19	21	18
Gaylord	Perry	1975	18	17	20	15

Billy	Pierce	1953	18	12	20	11
J.R.	Richard	1979	18	13	21	13
Jose	Rijo	1993	14	9	21	9

Steve	Rogers	1977	17	16	20	14
Steve	Rogers	1982	19	8	22	10

Nolan	Ryan	1977	19	16	21	13
CC	Sabathia	2008	17	10	21	9

Johan	Santana	2005	16	7	20	8
Johan	Santana	2006	19	6	20	8

Curt	Schilling	1998	15	14	21	9
Mike	Scott	1986	18	10	23	8

Tom	Seaver	1968	16	12	20	11
Tom	Seaver	1970	18	12	22	11
Tom	Seaver	1973	19	10	23	10

Chris	Short	1965	18	11	20	14

Mario	Soto	1983	17	13	20	12

Dave	Stieb	1982	17	14	21	12
Dave	Stieb	1983	17	12	21	11
Dave	Stieb	1984	16	8	20	10
Dave	Stieb	1985	14	13	20	11

Don	Sutton	1972	19	9	21	10
Frank	Tanana	1976	19	10	20	13
Bob	Veale	1964	18	12	20	12
Mike	Witt	1986	18	10	20	11

There have been twenty pitchers since 1950 who have won 25 games or more—but there have been only 14 pitchers since 1950 who have won 25 games based on merit, without luck. Those 14 are as follows:

					Deserved
First	Last	Year	W	L	Wins	Losses
Steve	Carlton	1972	27	10	28	11
Gaylord	Perry	1972	24	16	27	12
Sandy	Koufax	1966	27	9	27	10
Bob	Gibson	1968	22	9	26	8
Robin	Roberts	1953	23	16	26	13
Vida	Blue	1971	24	8	26	10
Sandy	Koufax	1965	26	8	26	13
Sandy	Koufax	1963	25	5	25	10
Dwight	Gooden	1985	24	4	25	6
Denny	McLain	1968	31	6	25	13
Wilbur	Wood	1972	24	17	25	18
Bert	Blyleven	1973	20	17	25	12
Bob	Gibson	1969	20	13	25	11
Steve	Carlton	1980	24	9	25	10

All fourteen pitchers who deserved to win 25 or more games did in fact win at least 20, and they averaged between 24 and 25 wins.

And finally (for today) there have been 19 pitchers since 1950 who either did win 20 games or should have won 20 games, and who had a deserved winning percentage of .750 or higher:

					Deserved
							Winning
First	Last	Year	W	L	Wins	Losses	Percentage
Pedro	Martinez	2000	18	6	22	3	.882
Greg	Maddux	1995	19	2	21	3	.856
Pedro	Martinez	1999	23	4	21	4	.856
Roger	Clemens	1997	21	7	24	6	.809
Dwight	Gooden	1985	24	4	25	6	.795
Zack	Greinke	2009	16	8	21	6	.795
Randy	Johnson	2001	21	6	23	6	.792
Zack	Greinke	2015	19	3	21	6	.788
Clayton	Kershaw	2014	21	3	18	5	.786
Pedro	Martinez	1997	17	8	22	6	.783
Jake	Arrieta	2015	22	6	21	6	.782
Randy	Johnson	1997	20	4	19	5	.781
Greg	Maddux	1997	19	4	21	6	.778
Roger	Clemens	1990	21	6	20	6	.772
Johan	Santana	2004	20	6	20	6	.763
Bob	Gibson	1968	22	9	26	8	.762
Ron	Guidry	1978	25	3	24	7	.760
Randy	Johnson	2002	24	5	23	7	.759
Pedro	Martinez	2002	20	4	18	6	.757

16 of these 19 pitchers did in fact have winning percentages over .750. Also Kevin Appier, 1993, had a deserved winning percentage of .749. Most of those pitchers (16 of the 19) have been from the post-1990 period. I think that’s due to the five-man (vs. four-man) rotation. When a pitcher is worked hard, that pushes him back toward the center. When he has more space, he is more likely to perform at a dominant level, perhaps.

COMMENTS (15 Comments, most recent shown first)

FrankD
Interesting question DaveNJNews: how many that were "inside baseball" looked at W/L? Stengel didn't, he picked up people like Don Larsen and others who only won with the Yankees. I dunno about if they used ERA but I think the management back in the day saw what the pitchers could throw and judged accordingly. I think the fans fixate on W/L - I know I did before I got into thinking about the game. Like Bill said: teams win games not just pitchers .......
7:52 PM Aug 16th

DaveNJnews
Looking at the first list in this article, one wonders how many bad trades were made in the 1960s or 1970s because someone decided to go out and get "a proven 20-game winner."
4:38 PM Aug 16th

OldBackstop
FWIW, and this may be a total justification of the methods here, nine out of the top ten luckiest guys had more cheap wins than tough losses, and ten of the top ten unluckiest guys had more tough losses than cheap wins.

Sometimes the number of cheap wins equal Bill's to a win or two, and sometimes the net comes close, but the trend is almost absolute.
1:57 PM Aug 16th

evanecurb
The first two major league games I attended in person were Angels-Yankees and A's-Orioles, both in the summer of 1970. All four starting pitchers were 20 game winners that year. Clyde Wright, Mel Stottlemyre, Jim Palmer, and Catfish Hunter. This is my sole contribution to the comments section of a very entertaining series. Thanks, Bill.
10:20 AM Aug 16th

smbakeresq
evanecurb
On the list of guys who had at least three seasons where they deserved to win 20, I see a bunch of Hall of Famers, Clemens, and Dave Stieb. Elite company for a guy who didn't garner much HOF support.

If those 20 win seasons were on his resume he would have.
10:29 AM Aug 15th

KaiserD2
Just a couple of overview comments.

I reviewed the post about BIll's method again. He has tried to go a step further than baseball-reference.com does, because he evidently compiled extremely complicated data (how he got it, I'm not sure) on how many runs pitchers should have given up, based on their walks, hits, etc, rather than simply look, as b-r does (and therefore, I did), at how many runs they did give up, and compare that to what would have been expected. To put this another way, Bill was trying to eliminate what I call "run luck" from the equation--that is, the fewer or more runs that occur when hits and walks are spread further apart, on the one hand, or bunched closer together, on the other. I think this was a sensible thing to do and indeed I'd be very curious to see a list of pitchers whose runs allowed (actual) differed the most from what Bill's formula predicted. It would be an attempt to identify "clutch pitchers," if such exist.

Despite this difference, it seems that Bill's method and mine (adapted from b-r) reach very similar conclusions about what a great season was and what it wasn't. I'm delighted, for instance, to see that Bill also picked out Bill Lee's 1973 season, which was genuinely great.

I do, however, see one possible problem in Bill's method. It is how he calculates expected decisions for pitchers. He bases it on the average number of innings per decision in the particular era the season under consideration took place. What I noticed is that this seems to lead to a systematic overestimate of the number of decisions the pitcher was likely to get. If you look for instance at the table of seasons where the pitcher didn't win 20 but in Bill's opinion should have, you will see his estimate for decisions is consistently higher than the actual number.

I have an idea as to why this might be so. Every game generates two decisions, a win or a loss. For obvious reasons, the losing pitcher, on the average, will have pitched substantially fewer innings than the winning one, in most cases. But in that table Bill is looking at pitchers who were very good, who won most of their games. Only a .500 pitcher, it seems to me, could be genuinely expected to have the same number of decisions that league average of D/IP would predict.

I still think W-L is too flawed to be a preferred basis for judging pitchers but Bill did find a way around a lot of the problems with it.

David K
7:49 AM Aug 15th

OldBackstop
Correction below: Sain had 6 cheap wins.
11:31 PM Aug 14th

OldBackstop
Bill, I'd be interested to see how your model comes up as compared to "Chead Wins" and "Tough Losses".

Just poking through, but I looked at Johnny Sain in 1950. He is listed as leading the league in "cheap wins" with 5 (defined as surrendering more the 3 earned runs or going less than 6 innings or losing a quality start.

But he also had four tough losses....losses in quality starts. So he was plus two. The other five guys with 19 or more wins were at leat minus 6 except for Don Newcombe, That year Don Newcombe, who was plus four at 19-11. Newcombe also had a full half run better in run support.I wondered if he had 20 wins would he topped Sain on the lucky list?
11:30 PM Aug 14th

shthar
Good question, how many years in a row does it take for luck become 'the way it is'?
10:48 PM Aug 14th

Gfletch
Good to see Dave Stieb getting some love here in these articles. I saw a bunch of his games back in the day and I can tell you he is one of the best pitchers I ever saw.

Question: are teams telling pitchers to ignore their W/L records nowadays? At least to the extent of emphasizing their overall stats to counteract any tendency to destructive adjustments?
6:33 PM Aug 14th

FrankD
Love the article. Looking at Blyleven's numbers. Even a stronger argument for HOF for him. But, a question: If you are unlucky for 5 years, is it really luck or just statement of reality? Also - I haven't gone through all the data but is there a difference in 'luck' based on type of pitcher, i.e., left vs right, extreme fastball vs control, etc? Or am I just wanting to push the data too far?
5:07 PM Aug 14th

shthar
I just hear Rodney Dangerfield saying, "luck, I don't get no luck"
2:31 PM Aug 14th

OldBackstop
rofl turtle line. Too funny.
11:49 AM Aug 14th

Davidg32
A little surprising to me to see Greg Maddux missing out on 20 wins so many times from 1995 - 2000. The Braves were a good team, and Maddux was the ace of the staff. (Although the pitching staff was awfully strong.) You would think that being on a good team would tend to help a pitcher win a couple of games more than he might have deserved.
8:47 AM Aug 14th

evanecurb
On the list of guys who had at least three seasons where they deserved to win 20, I see a bunch of Hall of Famers, Clemens, and Dave Stieb. Elite company for a guy who didn't garner much HOF support.

8:29 AM Aug 14th

Lady Luck and the 20-Game Winner

COMMENTS (15 Comments, most recent shown first)

Leave a comment

Report inappropriate comment


Type of Abuse:
Comments: