Bob Rush in 1956 and Don Cardwell in 1965 both finished the season with won-lost records of 13-10, a not particularly notable fact in that there are 67 pitchers in baseball history (through 2006) who finished their season 13-10. Both Rush and Cardwell, however, pitched 240 innings—240.0. Each pitcher allowed 101 runs, which in each case included 85 earned runs, giving each of them a 3.19 ERA. Rush struck out 104 and walked 59; Cardwell struck out 107 and walked 59.
Let us say for the sake of argument that these are the two most-similar seasons by pitchers in baseball history, which I don’t know that that’s true, but we’ll debate that later. When you look closely at Rush and Cardwell, you find any number of secondary or incidental similarities. Both Cardwell and Rush, for example, had relatively long careers for pitchers who had losing records. Rush pitched in the majors for 13 years, winning 127 games but losing 152. Cardwell pitched in the majors for 14 years, winning 102 games but losing 138. There are only 35 pitchers in major league history who won 100 or more games in their careers but had winning percentages under .460. Rush and Cardwell are two of them.
Both Rush and Cardwell spent significant portions of their careers with the Cubs, and both pitched for the Braves at the end of their careers (Rush winning two games for the Braves in 1960, his last season, and Cardwell winning two games for them in 1970, his last season). Rush pitched 417 games in his career and struck out 1,244 batters; Cardwell pitched 410 games and struck out 1,211 batters. Rush in his career finished 44 games that he did not start (GF) and saved 8 of them; Cardwell had 44 GF and 7 saves. Cardwell’s career high in strikeouts was 156; Rush’s was 157. Cardwell’s career high in innings pitched was 259; Rush’s was 255.
Both men were big men and right-handed pitchers. Cardwell is listed in the books a 6’4”, 210 pounds; Rush was 6’4” and 205, in a time when there were not nearly as many pitchers of that size. (Only 10% of major league pitchers in 1955 were 6’4” or taller, as opposed to 27% in 2005.) Both men were regarded as having excellent fastballs, and both men basically lived off the fastball.
Both men had blue eyes. Both listed “Golf” as their number one hobby. Rush, who was born in December, 1925, was 30 years old when he had his 13-10 season; Cardwell, born in December, 1935, was 29 years old when he had his. Two years earlier, both pitchers had gone 13-15. The next season, each pitcher won only 6 games.
Both men were very slightly above-average hitters for a pitcher, Rush with a career OPS of .426 against a norm for his era of .419, Cardwell at .387 against a norm of .374, although both men struck out 40% more often than an average pitcher and walked even less often than an average pitcher. Both men were used as pinch hitters a few times in their careers—Cardwell six times, Rush three times.
The similarities between two pitchers. . ..or two hitters, or two trees, or two automobiles, or two office managers, but we are talking for the moment about pitchers. . .the similarities between two pitchers can be put under several headings: essential similarities, defining similarities, accidental and random similarities, apparent similarities, etc. No one could doubt that Rush and Cardwell were essentially similar pitchers—but when you think about it, these are the not the defining similarities in the comparison above. I started with the fact that both men had seasons in which they were 13-10, pitched 240.0 innings and gave up 101 runs, 85 of them earned. This is not essential information about the men, as pitchers; it’s just kind of a random match. The defining similarity there is a fairly arbitrary happenstance. But when two players happen to share a range of random characteristics, it is likely that they will share essential characteristics as well. There are a wide variety of mechanisms that cause this to happen, so much so that it becomes difficult to say for sure that any shared characteristic is truly an accident.
If you look up the most-similar season to any season in history, most often you find another season that occurred within three or four years of the focus year, and almost always within 20 years. Remarkably often, the most-similar season to a line from 1973 will be another line from 1973 (or 1916, 1952. . .whatever.) The most-similar season for a right-handed pitcher will usually be another right-hander; the most-similar season for a lefty is usually another lefty. The most-similar pitcher to a tall pitcher is usually another tall pitcher; the most-similar season to a short pitcher is usually another shorty. The most-similar pitcher to a pitcher who walks 100 batters will usually be another pitcher who walks 100 batters, even if you eliminate strikeouts and walks from the similarity routine. The most-similar season to a season by a 35-year-old will usually be another season by a pitcher in his mid-thirties; the most-similar season to a pitcher by a 22-year-old is often another season by a 22- or 23-year-old. The most-similar season to a season by a rookie is often a season by another rookie. The most-similar season to a season by a Hall of Famer is often another season by a Hall of Famer—even if you focus on seasons which are not indicative of Hall of Fame performance.
The most-similar season to a season by a knuckleball pitcher is often another season by a knuckleball pitcher. If you look up the pitcher’s repertoire in the Neyer/James Guide, the most-similar season is often by a pitcher who has a very similar repertoire.
Astonishingly often, when you identify the most-similar pitcher to someone who went to jail shortly after his career, or died of prostate cancer at age 57, or went into politics after his career, you find another pitcher who went to jail shortly after his career, or died of prostate cancer at age 56, or went into politics after he got out of jail. One never knows what to make of these similarities. .whether they are truly accidents, or whether one has accidentally tapped into some profound underlying similarity the root of which is not within our view.
Back to the time when I first became a baseball fan in 1961, I have always been fascinated by players who had very similar statistics for a season or a career. In the early 1980s I introduced a method to “score” the similarity between any two seasons or two careers. The method isn’t important and I don’t really remember what it was; what was significant was not the method but the realization that, even though “similarity” is essentially a subjective concept, we could nonetheless define it and measure it, so long as we were willing to accept that our definition and our measurement were somewhat arbitrary.
This is now so much a standard practice for me. . .and I think for many others, but I won’t speak for others. ..this is so much a standard practice for me that I could hardly go through a day without it. I think that the only use of the concept of similarities in baseball analysis before the early 1980s was in Hall of Fame arguments, and there only in a very narrow and one-dimensional way. People would argue that George Kelly was in the Hall of Fame and Dixie Walker had very similar numbers to George Kelly, therefore Dixie Walker should be in the Hall of Fame, too. This argument was extremely common, although the obvious other side of the argument—that Dixie Walker had very similar numbers to Carl Furillo and Carl Furillo was not in the Hall of Fame, therefore Dixie Walker should not be in there either—never came up. Occasionally this sort of narrow and one-sided “similarities” argument would also surface in MVP discussions, but not much, since there is an obvious problem with it in that context.
Anyway, I use similarity scores, in my work with the Red Sox, on almost a daily basis. I would guess that I might use similarity scores to study an issue more often than I use runs created, secondary average or Win Shares. Let us say that the Red Sox are considering trading for. . .oh, I don’t know, Joe Rudi or Drungo Larue Hazewood. One of the first things I would do, to evaluate the target player, is to try to identify the ten or twenty-five other players who looked most similar to him at the same age.
Suppose that the Red Sox were considering offering a three-year, $28 million contract extension to, I don’t know, Sandy Consuegra or David Cone. One of the first things I would do, to evaluate that investment, would be to identify all of the pitchers who looked just like Sandy Consuegra or David Cone up until this point in his career.
Most of the time this analysis is not dispositive—but occasionally it is. Occasionally you look up the 25 most-comparable players to Joe Rudi at age 32 and realize that, of the entire group, not a single one ever came back after a bad year to re-establish his value. This doesn’t dispose of the issue from the standpoint of the Red Sox, because that’s not my decision—but it does close the book for me. At that point I know what my recommendation will be: stay away from Joe Rudi. Nothing else that I could find would convince to recommend otherwise.
Most often, of course, I get a mixed bag of comparable players, and then I move on to other types of analysis. Comparability or similarity has worked its way into analysis of almost any question related to baseball. Suppose you are wondering what the chances are that Jason Bay, now 28 years old, will one day be in the Hall of Fame. One way to study this is to identify 20 other players who had very similar numbers to Jason Bay at age 28. If 15 of those 20 other players are now in the Hall of Fame, it is likely that Bay will also reach that standard. If only 2 of the 20 went on to be Hall of Famers, it is unlikely.
Suppose that you are wondering whether Jason Schmidt is likely to get 200 career wins. Identify the 20 pitchers most similar to Jason Schmidt at the same age, and look at their career win totals. Suppose that you are wondering whether Team X, which has four regular infielders under age 27, is likely to emerge as a perennial contender. Identify 20 similar teams with young infields. Suppose that you are wondering whether intentional walks have a measurable effect in reducing a pitcher’s home runs allowed. Identify matched sets of pitchers with very different intentional walk totals, and look at their home runs allowed. It’s a ubiquitous method.
Well, no. It’s a ubiquitous approach. The statistical method used to implement the approach varies with the issue being studied. What makes it so ubiquitous is exactly that: that there is no standard and inflexible method for evaluating similarity—therefore, we can adapt it to the needs of the study. If we need to find players coming off of injury seasons, we can change the system so that it emphasizes coming off a bad season. If we need to compare pitchers who had winning records but bad strikeout/walk ratios who are then switched to the bullpen, we can change the system so that it emphasizes won-lost records and strikeout/walk ratios. It’s arbitrary; therefore, we can change it to suit our needs. Being arbitrary doesn’t make the system LESS useful; it makes it MORE useful. Who would have thunk it?
I am now balanced precariously between two articles here. When I sat down to write this article it was just a fun little riff on the most-similar pitcher’s seasons that I could find. Then I started thinking about the history and applications of similarity scores, and I wandered into a quasi-philosophical reflection on the nature and usefulness of similarity. I don’t know which article to finish. I guess I’ll tackle the rumination first.
Similarities between two pitchers (or office managers) can be put under several headings: essential similarities, defining similarities, accidental and random similarities, apparent similarities, etc.
Accidental or random similarities are similarities which could just as easily not exist—for example, the fact that two pitchers are both 13-10. If you take the same pitchers the next season, one of them could just as easily go 17-8 and the other one 8-13, totally dissimilar records.
Defining similarities are those similarities which form the basis of the comparison. Defining similarities can be accidental, or they can be essential similarities. The defining characteristics are those that we change, within the similarity formulas, to focus the study on whatever it is that we are trying to learn more about.
Essential similarities are those similarities which focus on the central characteristics of the player. For example, the essential characteristics of Stan Musial are that he was thin, durable, swift, hit for a very high average with excellent power, and batted left-handed. If you match him with a player who happens to have very similar batting stats in some season but who was a right-handed hitting steroid monster from the 1990s who only had two good seasons, that’s a arbitrary or non-essential similarity. A group of pitchers who share several essential characteristics can be referred to as a “family” of pitchers.
Profound similarities could be defined as multiple essential similarities including similarities on a non-obvious level—for example, Roger Maris’ career is profoundly similar to Hack Wilson’s, although Maris and Wilson never had seasons with similar batting stats. Maris and Wilson had careers of extremely similar shape, duration and value. Both men held very important single-season slugging records, and both men died young. They are profoundly similar although their basic hitting numbers do not really match.
Apparent similarities are random or accidental similarities between players who are profoundly dissimilar. For example, Jeff Pfeffer in 1923 and Mark Williamson in 1987 both finished the season 8-9, and both had ERAs of 4.03. However, there are profound differences between them. Pfeffer was 35; Williams was 27. Pfeffer was a starting pitcher in 1923; Williamson was a reliever in 1987. Pfeffer struck out 32 batters in 152 innings; Williamson struck out easily more than twice as many hitters in 125 innings. They really have nothing significant in common, despite having apparently similar records.
Serious studies on which actual decisions might be premised require that we identify players who are profoundly similar or at least essentially similar. You can write similarity formulas so that they find players who are essentially similar. You cannot write formulas to find players who are profoundly similar, I don’t think; that you have to recognize on your own. You can use formulas to help you.
This article is not about players who are essentially similar or profoundly similar, except by accident. This is just about random matches between pitcher’s seasons which look a lot alike:
| Tommy Bond, 1877, and Tommy Bond, 1878 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| 1877 |
58 |
521.0 |
40-17 |
.702 |
530 |
170 |
36 |
2.11 |
| 1878 |
59 |
532.2 |
40-19 |
.678 |
571 |
182 |
33 |
2.06 |
My similarity score for these two seasons is only 932, as opposed to 980 for Rush and Cardwell, but the differences magnify with the innings, and Bond’s consistency is extraordinary for a pitcher pitching 500+ innings.
| Bob Caruthers, 1886, and Pete Conway, 1888 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Caruthers |
44 |
387.1 |
30-14 |
.682 |
323 |
166 |
86 |
2.32 |
| Conway |
45 |
391.0 |
30-14 |
.682 |
315 |
176 |
57 |
2.26 |
Similarity: 950. Caruthers was 22 years old and the time; Conway, 21. Caruthers also played the outfield that year, and hit .334. Conway didn’t play a position that year—he had played some in the outfield earlier in his career—but hit .275 with 3 homers, 23 RBI.
| Frank Kitson, 1903, and Red Donahue, 1903 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Kitson |
31 |
257.2 |
15-16 |
.484 |
277 |
102 |
38 |
2.58 |
| Donahue |
32 |
267.2 |
15-16 |
.484 |
287 |
96 |
34 |
2.59 |
Similarity: 972.
| Deacon Phillippe, 1903, and Jesse Barnes, 1919 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Phillippe |
36 |
289.0 |
25-9 |
.735 |
269 |
123 |
29 |
2.43 |
| Barnes |
38 |
296.0 |
25-9 |
.735 |
296 |
92 |
35 |
2.40 |
Similarity: 944. Phillippe pitched five complete games in the 1903 World Series, winning three of them, although his team lost the series. Barnes didn’t become a World Series hero until two years later, 1921.
| Sam Leever, 1906, with Waite Hoyt, 1927 and Mort Cooper, 1944 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Leever |
34 |
260.0 |
22-7 |
.759 |
232 |
76 |
48 |
2.32 |
| Hoyt |
36 |
256.0 |
22-7 |
.759 |
242 |
86 |
54 |
2.64 |
| Cooper |
34 |
252.0 |
22-7 |
.759 |
227 |
97 |
60 |
2.46 |
Similarity: Hoyt to Cooper, 960; Hoyt to Leever, 959, Leever to Cooper, 944.
| Doc White, 1910, and Dennis Martinez, 1988 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| White |
36 |
236.2 |
15-13 |
.536 |
219 |
111 |
50 |
2.66 |
| Martinez |
34 |
235.1 |
15-13 |
.536 |
215 |
120 |
55 |
2.72 |
Similarity: 973. This one is included because it is so unusual to have very-similar seasons separated by 78 years. A couple of others of those are Gregg and Barber (below) and Joe McGinnity, 1905 with Robin Roberts, 1951 (not included.)
| Vean Gregg, 1912, and Steve Barber, 1963 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Gregg |
37 |
271.1 |
20-13 |
.606 |
242 |
184 |
90 |
2.59 |
| Barber |
39 |
259.0 |
20-13 |
.606 |
253 |
180 |
92 |
2.75 |
Similarity: 960
| Lefty Williams, 1919, and Burleigh Grimes, 1920 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Williams |
41 |
297.0 |
23-11 |
.676 |
265 |
125 |
58 |
2.64 |
| Grimes |
40 |
304.0 |
23-11 |
.676 |
271 |
131 |
67 |
2.22 |
Similarity: 954. Grimes threw the spitball. Williams threw the series.
| Dazzy Vance, 1928, with Whitlow Wyatt, 1941 and Dick Ellsworth, 1963 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Vance |
38 |
280.1 |
22-10 |
.688 |
226 |
200 |
72 |
2.09 |
| Wyatt |
38 |
288.0 |
22-10 |
.688 |
223 |
176 |
82 |
2.34 |
| Ellsworth |
37 |
291.0 |
22-10 |
.688 |
223 |
185 |
75 |
2.10 |
Similarity: Wyatt to Ellsworth, 965, Ellsworth to Vance, 962, Wyatt to Vance, 946. Ellsworth and Vance both pitched for relatively poor teams that finished just over .500, well below .500 with other pitchers on the mound.
| Tiny Bonham, 1940, and Mel Stottlemyre, 1964 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Bonham |
12 |
99.0 |
9-3 |
.750 |
83 |
37 |
13 |
1.91 |
| Stottlemyre |
13 |
96.0 |
9-3 |
.750 |
77 |
49 |
35 |
2.06 |
Similarity: 953. Both rookies, both Yankees. Mel Stottlemyre, called up when the Yankees were floundering in August, 1964, arrived just in time to save the Yankees’ bacon in 1964. Bonham, called up in August, 1940, got there too late to save the Yankees in 1940.
| Bill Lee, 1940, with Neal Heaton, 1985 and Jimmy Anderson, 2001 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Lee |
37 |
211.0 |
9-17 |
.346 |
246 |
70 |
70 |
5.03 |
| Heaton |
36 |
207.2 |
9-17 |
.346 |
244 |
82 |
80 |
4.90 |
| Anderson |
34 |
206.1 |
9-17 |
.346 |
232 |
89 |
83 |
5.10 |
Similarity: Anderson to Lee, 969, Lee to Heaton, 967, Anderson to Lee, 932.
| Dizzy Trout, 1942, and Jim Wilson, 1955 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Trout |
35 |
223.0 |
12-18 |
.400 |
214 |
91 |
89 |
3.43 |
| Wilson |
34 |
235.0 |
12-18 |
.400 |
200 |
96 |
87 |
3.45 |
Similarity: 976.
| Joe Dobson, 1947, and Sal Maglie, 1952 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Dobson |
33 |
229.0 |
18-8 |
.692 |
203 |
110 |
73 |
2.95 |
| Maglie |
35 |
216.0 |
18-8 |
.692 |
199 |
112 |
75 |
2.92 |
Similarity: 980. Dobson’s nickname was “Burrhead”. Maglie’s was “The Barber”.
| Johnny Sain, 1947, and Joaquin Andujar, 1985 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Sain |
38 |
266.0 |
21-12 |
.636 |
265 |
132 |
79 |
3.52 |
| Andujar |
38 |
269.2 |
21-12 |
.636 |
265 |
112 |
82 |
3.40 |
Similarity: 965. Somehow, it just struck me as funny to pair Joaquin Andujar with anyone named “Sain”. You had to know Joaquin to get the joke.
| Luis Arroyo, 1955, and Alex Kellner, 1955 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Arroyo |
35 |
159.0 |
11-8 |
.579 |
162 |
68 |
63 |
4.19 |
| Kellner |
30 |
163.0 |
11-8 |
.579 |
164 |
75 |
60 |
4.20 |
Similarity: 972. Kellner and Arroyo were both lefties who had one shining moment at one end of very modest careers. However, although the two men were about the same age (Kellner two years older), Kellner’s shining moment had come in the previous decade, when he won 20 games as a rookie for the 1949 Philadelphia A’s. Arroyo’s shining moment came in the next decade, in 1961, when he had a brilliant year as a reliever for the 1961 Yankees.
| Jack Sanford, 1957, and Mark Langston, 1991 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Sanford |
33 |
237.0 |
19-8 |
.704 |
194 |
188 |
94 |
3.08 |
| Langston |
34 |
246.1 |
19-8 |
.704 |
190 |
183 |
96 |
3.00 |
Similarity: 980.
| Johnny Podres, 1961, and Cliff Lee, 2005 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Podres |
32 |
183.0 |
18-5 |
.783 |
192 |
124 |
51 |
3.74 |
| Lee |
32 |
202.0 |
18-5 |
.783 |
194 |
143 |
52 |
3.79 |
Similarity: 968.
| Whitey Ford, 1963, and Frank Viola, 1987 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Ford |
38 |
269.0 |
24-7 |
.774 |
240 |
189 |
56 |
2.74 |
| Viola |
35 |
255.1 |
24-7 |
.774 |
236 |
193 |
54 |
2.64 |
Similarity: 967. Both veteran left-handers of exceptional quality. Matches this good are common among pitchers who win 12-15 games, but not common among pitchers of Cy Young quality.
| Bill Monboquette, 1963, and Brad Radke, 1997 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Monboquette |
37 |
267.0 |
20-10 |
.667 |
258 |
174 |
42 |
3.81 |
| Radke |
35 |
239.2 |
20-10 |
.667 |
238 |
174 |
48 |
3.87 |
Similarity: 962.
| Robin Roberts, 1964, and John Tudor, 1986 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Roberts |
31 |
204.0 |
13-7 |
.650 |
203 |
109 |
52 |
2.91 |
| Tudor |
30 |
219.0 |
13-7 |
.650 |
197 |
107 |
53 |
2.92 |
Similarity: 983.
| Dave Giusti, 1966 with Brian Lawrence, 2004 and Greg Maddux, 2006 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Giusti |
34 |
210.0 |
15-14 |
.517 |
215 |
131 |
54 |
4.20 |
| Lawrence |
34 |
203.0 |
15-14 |
.517 |
226 |
121 |
55 |
4.12 |
| Maddux |
34 |
210.0 |
15-14 |
.517 |
219 |
117 |
37 |
4.20 |
Similarity: Giusti to Lawrence, 977, Giusti to Maddux, 968, Lawrence to Maddux, 967.
| Nelson Briles, 1969, and Charlie Leibrandt, 1991 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Briles |
36 |
228.0 |
15-13 |
.536 |
218 |
126 |
63 |
3.51 |
| Leibrandt |
34 |
229.2 |
15-13 |
.536 |
212 |
128 |
56 |
3.49 |
Similarity: 986.
| Sonny Siebert, 1970, and Pascual Perez, 1983 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Siebert |
33 |
222.2 |
15-8 |
.652 |
207 |
142 |
60 |
3.44 |
| Perez |
33 |
215.1 |
15-8 |
.652 |
213 |
144 |
51 |
3.43 |
Similarity: 982.
| Paul Splittorff, 1972, and Jack Billingham, 1972 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Splittorff |
35 |
216.0 |
12-12 |
.500 |
189 |
140 |
67 |
3.13 |
| Billingham |
36 |
217.2 |
12-12 |
.500 |
197 |
137 |
64 |
3.18 |
Similarity: 979.
| Ferguson Jenkins, 1973, and Bret Saberhagen, 1988 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Jenkins |
38 |
271.0 |
14-16 |
.467 |
267 |
170 |
57 |
3.89 |
| Saberhagen |
35 |
260.2 |
14-16 |
.467 |
271 |
171 |
59 |
3.80 |
Similarity: 977. Jenkins and Saberhagen were both former Cy Young Award winners, Jenkins two years earlier (1971) and Saberhagen three years earlier (1985). The next year, Saberhagen bounced back to go 23-6 with a 2.16 ERA, claiming his second Cy Young Award. Jenkins bounced back to go 25-12 with a 2.82 ERA, but lost the Cy Young voting to Catfish Hunter (below).
| Ferguson Jenkins, 1974, and Catfish Hunter, 1974 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Jenkins |
41 |
328.0 |
25-12 |
.676 |
286 |
225 |
45 |
2.82 |
| Hunter |
41 |
318.0 |
25-12 |
.676 |
268 |
143 |
46 |
2.49 |
Similarity: Only 890. I just included them because of the Jenkins/Saberhagen matchup above. Jenkins and Hunter were extremely similar except for the large difference in strikeouts.
| Jim Colborn, 1977, and Rudy May, 1977 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Colborn |
36 |
239.0 |
18-14 |
.563 |
233 |
103 |
81 |
3.62 |
| May |
37 |
251.2 |
18-14 |
.563 |
243 |
105 |
78 |
3.61 |
Similarity: 980. They were not similar pitchers, but had very similar records that year. .
| Steve Rogers, 1977, and Bert Blyelven, 1985 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Rogers |
40 |
301.2 |
17-16 |
.515 |
272 |
206 |
81 |
3.10 |
| Blyleven |
37 |
293.2 |
17-16 |
.515 |
264 |
206 |
75 |
3.16 |
Similarity: 975.
| Vern Ruhle, 1980, and John Tudor, 1990 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Ruhle |
28 |
159.1 |
12-4 |
.750 |
148 |
55 |
29 |
2.37 |
| Tudor |
25 |
146.1 |
12-4 |
.750 |
120 |
63 |
30 |
2.40 |
Similarity: 960.
| Tom Henke, 1980, and Akinori Otsuka, 2006 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
Saves |
| Henke |
61 |
74.2 |
2-4 |
.333 |
58 |
75 |
19 |
2.17 |
32 |
| Otsuka |
63 |
59.2 |
2-4 |
.333 |
53 |
47 |
11 |
2.11 |
32 |
| Ron Guidry, 1983, and Tom Glavine, 2000 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Guidry |
31 |
250.1 |
21-9 |
.700 |
232 |
156 |
60 |
3.42 |
| Glavine |
34 |
241.0 |
21-9 |
.700 |
222 |
152 |
65 |
3.40 |
Similarity: 969.
| Buddy Black, 1984, and Bob Knepper, 1986 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Black |
35 |
257.0 |
17-12 |
.586 |
226 |
140 |
64 |
3.12 |
| Knepper |
40 |
258.0 |
17-12 |
.586 |
232 |
143 |
62 |
3.14 |
Similarity: 976. Black and Knepper, though very different personalities, were profoundly similar pitchers.
| Dave Righetti, 1984, and Lee Smith, 1990 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
Saves |
| Righetti |
64 |
96.1 |
5-6 |
.455 |
79 |
90 |
37 |
2.34 |
31 |
| Smith |
64 |
83.0 |
5-6 |
.500 |
71 |
87 |
29 |
2.06 |
31 |
Similarity: 948.
| Bert Blyleven, 1984, and Roy Halladay, 2002 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Blyleven |
33 |
245.0 |
19-7 |
.731 |
204 |
170 |
74 |
2.87 |
| Halladay |
34 |
239.1 |
19-7 |
.731 |
223 |
168 |
62 |
2.93 |
Similarity: 968. Watch out, Roy; they’re going to screw you in the Hall of Fame voting.
| Oil Can Boyd, 1984, and Kris Benson, 2004 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Boyd |
29 |
197.2 |
12-12 |
.500 |
207 |
134 |
53 |
4.37 |
| Benson |
31 |
200.1 |
12-12 |
.500 |
202 |
134 |
61 |
4.31 |
Similarity: 980. Think about it. Boyd was a colorful, eccentric character who had a famous emotional melt-down after being left off the All-Star team in 1986. Benson is more famous for his wife’s hooters than for his pitching.
| Danny Cox, 1985, and Joe Magrane, 1989 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Cox |
35 |
241.0 |
18-9 |
.667 |
226 |
131 |
64 |
2.88 |
| Magrane |
34 |
234.2 |
18-9 |
.667 |
219 |
127 |
72 |
2.91 |
Similarity: 977. Cox and Magrane were very dissimilar pitchers and very dissimilar personalities. But both pitchers faded instantly after giving the Cardinals just a taste of their ability. Magrane was 24 at the time, Cox 25 (both pitching for St. Louis. Sorry. . .I tend to assume that you’re old enough to remember those things.) Both finished under .500 the next season, and both finished under .500 in relatively short careers.
| Bill Gullickson, 1986, and Tom Browning, 1989 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Gullickson |
37 |
244.2 |
15-12 |
.556 |
245 |
121 |
60 |
3.38 |
| Browning |
37 |
249.2 |
15-12 |
.556 |
241 |
118 |
64 |
3.39 |
Similarity: 988. The measured similarity of these two seasons (988) is even higher than that of Rush and Cardwell. Gullickson and Browning both pitched for the Reds at the time of these two very-similar seasons. Gullickson had other seasons in which he was 20-9, 14-13, 14-12 and 10-14. Browning had other seasons of 20-9, 14-13, 14-14 and 10-13.
| Mike Schooler, 1990, and John Franco, 1994 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
Saves |
| Schooler |
49 |
56.0 |
1-4 |
.200 |
47 |
45 |
16 |
2.25 |
30 |
| Franco |
47 |
50.0 |
1-4 |
.200 |
47 |
42 |
19 |
2.70 |
30 |
Similarity: 967.
| Tom Browning, 1991, and Mark Buehrle, 2003 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Browning |
36 |
230.1 |
14-14 |
.500 |
241 |
115 |
56 |
4.18 |
| Buehrle |
35 |
230.1 |
14-14 |
.500 |
250 |
119 |
61 |
4.14 |
Similarity: 982.
| Jack McDowell, 1992, and Pat Hentgen, 1996 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| McDowell |
34 |
260.2 |
20-10 |
.667 |
247 |
178 |
75 |
3.18 |
| Hentgen |
35 |
265.2 |
20-10 |
.667 |
238 |
177 |
94 |
3.22 |
Similarity: 968.
| Jimmy Key, 1992, and Tom Glavine, 2005 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Key |
31 |
216.2 |
13-13 |
.500 |
205 |
117 |
59 |
3.53 |
| Glavine |
39 |
211.1 |
13-13 |
.500 |
227 |
105 |
61 |
3.53 |
Similarity: 972. Key and Glavine are profoundly similar pitchers.
| Rick Aguilera, 1993, and Jason Isringhausen, 2001 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
Saves |
| Aguilera |
65 |
72.1 |
4-3 |
.571 |
60 |
59 |
14 |
3.11 |
34 |
| Isringhausen |
65 |
71.1 |
4-3 |
.571 |
54 |
74 |
23 |
2.65 |
34 |
Similarity: 954.
| Ramon Martinez, 1995, and Kenny Rogers, 1995 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Martinez |
30 |
206.1 |
17-7 |
.708 |
176 |
138 |
81 |
3.66 |
| Rogers |
31 |
208.0 |
17-7 |
.708 |
192 |
140 |
76 |
3.38 |
Similarity: 970. It just struck me that Rogers, still going strong more than ten years later, was actually a year OLDER than Martinez when the two had similar seasons in 1995. And still is, I guess.
| Erik Hanson, 1996, and Darryl Kile, 1998 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Hanson |
35 |
214.2 |
13-17 |
.433 |
243 |
156 |
102 |
5.41 |
| Kile |
36 |
230.1 |
13-17 |
.433 |
257 |
158 |
102 |
5.20 |
Similarity: 966.
| Dave Burba, 1998, and Jimmy Haynes, 2002 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Burba |
32 |
203.2 |
15-10 |
.600 |
210 |
132 |
69 |
4.11 |
| Haynes |
34 |
196.2 |
15-10 |
.600 |
210 |
126 |
81 |
4.12 |
Similarity: 972.
| Roger Clemens, 1998, and Johan Santana, 2004 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Clemens |
33 |
234.2 |
20-6 |
.769 |
169 |
271 |
88 |
2.65 |
| Santana |
34 |
228.0 |
20-6 |
.769 |
156 |
271 |
54 |
2.61 |
Similarity: 946. Perhaps the best-matched Cy Young seasons. (McDowell won the Cy Young Award in 1993, not 1992, and Whitey Ford would have won the American League Cy Young Award in 1963 had there been such an award.) In truth, Clemens and Santana could hardly be LESS similar as pitchers.
| Bob Wickman, 1999, and Robb Nen, 1999 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
Saves |
| Wickman |
71 |
74.1 |
3-8 |
.273 |
75 |
60 |
38 |
3.39 |
37 |
| Nen |
72 |
72.0 |
3-8 |
.273 |
79 |
77 |
27 |
4.00 |
37 |
Similarity: 941.
| Todd Jones, 1999, and Danys Baez, 1999 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
Saves |
| Jones |
65 |
66.1 |
4-4 |
.500 |
64 |
64 |
52 |
3.80 |
30 |
| Baez |
62 |
68.0 |
4-4 |
.500 |
60 |
52 |
29 |
3.57 |
30 |
Similarity: 961.
| Chad Durbin, 2001, and Jeff D'Amico, 2005 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Durbin |
29 |
179.0 |
9-16 |
.360 |
201 |
95 |
58 |
4.93 |
| D'Amico |
29 |
175.1 |
9-16 |
.360 |
204 |
100 |
42 |
4.77 |
Similarity: 969.
| Mark Mulder, 2001, and Bartolo Colon, 2005 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Mulder |
34 |
229.1 |
21-8 |
.724 |
214 |
153 |
51 |
3.45 |
| Colon |
33 |
222.2 |
21-8 |
.724 |
215 |
157 |
43 |
3.48 |
Similarity: 980. These may be the best-matched twenty-win seasons in baseball history. Mulder, however, doesn’t look anything like a hippopotamus with a hand-knit doily on his head.
| Matt Clement, 2002, and Kerry Wood, 2002 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Clement |
32 |
205.0 |
12-11 |
.522 |
162 |
215 |
85 |
3.60 |
| Wood |
33 |
213.2 |
12-11 |
.522 |
169 |
217 |
97 |
3.66 |
Similarity: 973. Teammates. Nolan Ryan, 1988, was also very similar.
| Cory Lidle, 2006, and Claudio Vargas, 2006 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Lidle |
31 |
170.2 |
12-10 |
.545 |
181 |
130 |
58 |
4.85 |
| Vargas |
31 |
167.2 |
12-10 |
.545 |
185 |
123 |
52 |
4.83 |
Similarity: 983.
It occurs to me that all of the similarities I have shown you are exceptional—and thus atypical—matches. I probably should have shown you a few more typical comparisons. If you take any two 200-inning seasons and match them together at random, you get an average similarity score, by the method I used here, of 658. This is 656:
| Murry Dickson, 1952, and Billy Pierce, 1953 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Dickson |
43 |
278.0 |
14-21 |
.400 |
278 |
112 |
76 |
3.56 |
| Pierce |
40 |
271.0 |
18-12 |
.600 |
216 |
186 |
102 |
2.72 |
Dickson and Pierce have about the same workload, but very different levels of effectiveness.
This is 659:
Pud Galvin, 1889, and Gary Peters, 1971 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Galvin |
41 |
341.0 |
23-16 |
.590 |
392 |
77 |
78 |
4.17 |
| Peters |
34 |
214 |
14-11 |
.560 |
241 |
104 |
70 |
4.37 |
Peters and Galvin have about the same level of effectiveness, but radically different workloads.
This match scores at 657:
Clyde Wright, 1973, and Kenny Rogers, 1993 |
| Pitcher |
G |
IP |
W-L |
Pct |
H |
SO |
BB |
ERA |
| Wright |
37 |
257.0 |
11-19 |
.367 |
264 |
64 |
54 |
4.42 |
| Rogers |
35 |
208.1 |
16-10 |
.615 |
210 |
140 |
71 |
4.10 |
Bill James
Ft. Myers, Florida
March 28, 2007 |