Does the quality of starting pitching faced by various teams, over the course of a season, even out?   Does it more or less even out?   To what extent does it even out?

            Are pennant races won and lost because one team or another catches a good break on the starting pitching they face?   Or is it, over the course of a season, a non-factor?

            In this year’s Gold Mine (The Bill James Gold Mine 2010), I authored an article about these issues.   This article is a continuation of that research.    This article examines the same issues, but with much more data and with a better-developed method.   To the extent that these two articles contradict one another—and they do—this is the later article, and this represents the more current state of the research.

            OK, we start here with

            1)  A list of every game played within the Retrosheet data (the full-account data, not the box-scores-only data), and

            2)  A Season Score for each pitcher.

            The Season Scores are explained somewhere on Bill James Online; I’m not going to explain them all again.   For this study, I converted the Season Scores into league-normalized numbers that look like and function like winning percentages, although they’re actually not.   Example:  Felix Hernandez, 2009.   The American League average Season Score in 2009 was 78.3.   Felix Hernandez’ season score was 299.0.   Hernandez was 220.7 points better than the league average.    We add 500 points to that, and divide by 1000.   Hernandez’ Season Score is expressed in this study as .7207, or .721.

            You may know (or you may not). . ..Hernandez’ team won 73.5% of his starts.    We hope these things are in the same range.    It’s nice when they are.    Much later in the article, I will demonstrate that the “normalized percentages” do in fact track very well with the relevant winning percentages—thus, that the measured differences in our study, derived from these normalized percentages, can be treated as Wins or Win Advantages.

            We looked at every team since 1952, every game within the Retrosheet data, to determine which teams had faced the toughest (and easiest) starting pitching.   I’ll give you the basic results up front here, in case you’re in a hurry, or you’re just tired of my bullshit.

            It pretty much evens out.

            It is not clear that there has ever been a pennant race that was decided as it was because one team got lucky in terms of the starting pitchers that they faced.   I looked for such a race.   I didn’t really find anything.   There is probably one out there if you want to work hard enough to find it.

            Let me go back now to the data.   In 2009 the major league team which faced the toughest starting pitching—as was reported in the Gold Mine—was San Diego.    The team which faced the weakest opposition starting pitching was the Cubs.   In the Gold Mine it says the Yankees, but that was without league-normalizing the data.  The National League Season Scores run a little bit higher than the American League (78.3 average in the American League, 84.5 in the National), which means that the National League pitchers APPEAR to be a little bit better, but when you normalize that so that National League pitchers are .500 and American League pitchers are .500, then the team facing the weakest opposing pitchers wasn’t the Yankees, but the Cubs.

            In all of the years 1952 to 2009, the team which faced the toughest opposition starting pitching was. . .well, let’s say it was the 1965 Houston Astros.   The starting pitchers faced by the Astros in 1965 had an average normalized percentage of .5224.    There are two teams in the data which had higher percentages, but they were both 1950s teams for which significant numbers of games are missing.   If we assume that the missing data is league-normal, both teams would be below the Astros, so the Astros are the winners of this contest.

            The lowest opposition pitcher percentage by any team in the data was by the 1975 Dodgers, .4778.   The standard deviation was .0065.   In a 162-game schedule, .0065 is almost exactly one game.  One game on a 162-game schedule is .0062; one game on a 154-game schedule was .0065.   Thus, the standard deviation of the difference in the quality of pitchers faced by different teams is one game.

            The largest difference between any two teams competing in the same league was between the 1965 Houston Astros (.5224) and the 1965 Dodgers (.4806).   This is a difference of 6.77 games.

Let’s look more closely at that comparison.  We start by making a list of the starting pitchers who faced each of these two teams.    Chris Short, for example, started six times that year against the Astros.   He also started six times against the Dodgers.   No difference.    Bob Gibson started 4 times against each team, Bob Veale started 5 times against each team, Bob Buhl started 4 times against each, Bill Faul started one game against each, Nelson Briles started one against each, etc.    All of those games are the same against the two teams, and thus can be thrown out of a study of the differences for the two teams.

            Juan Marichal started 5 times against the Astros, 4 times against the Dodgers.  We can, then, throw four of those games out for both teams, or set 4 of them aside.   Maybe I shouldn’t have used Marichal; your mind goes to the Roseboro’s bleeding head.    Sammy Ellis started 4 times against Houston, 5 times against the Dodgers.   We can set four of those games aside, and record that as one game in which the Dodgers faced Sammy Ellis, one game in which the Astros faced Juan Marichal.   In those 9 games for each team, that is the difference between the starting pitching faced by the two teams.

            Proceeding in this way, we can set aside 101 games of the 162 games played by each team, because 101 of the 162 are the same.   The difference between the starting pitching faced by the Astros and the starting pitching faced by the Dodgers was in the other 61 games.

            Let’s look at who those 61 pitchers were, or, stated more accurately, who the pitchers were who started those 61 games:

            Sandy Koufax started five times against the Astros and no times against the Dodgers, of course, because Sandy Koufax was on the Dodgers, and consequently could not pitch against them.  Of the 61 games that are different between the two teams, 18 games are a product of the fact that the Dodgers and Astros played each other 18 times during the season.

            And, of course, the Dodger’s pitching was much better than the Astros’ pitching, so in these 18 games, the pitching faced by the Astros was much better than the pitching faced by the Dodgers.     Koufax had a normalized percentage—a Season Score percentage—of .836, one of the highest of all time.   No Astros starting pitcher had a normalized percentage higher than .562—and they traded that guy off early in the year.  These are the percentages for the 18 starters faced by each team in those head-to-head games:

            Of the difference between the quality of starting pitchers faced by Houston and Los Angeles in the 1965 season, a little more than one-half is simply a natural effect of the fact that, when Houston faced Los Angeles, Los Angeles usually had a better starting pitcher on the mound.   If you total up those “percentages” above, they total up to 11.42 for LA vs. Houston, and 7.88 for Houston vs. LA. That’s an advantage of 3.54 wins for Los Angeles.  Their total advantage on the season—the largest difference between any two teams in the same league in the major leagues since 1952—was 6.77.

            In other words, even in the most extreme case, the difference between the quality of pitching faced by the teams facing the toughest pitching and the teams facing the poorest pitching is mostly a natural effect of the fact that if you don’t have good pitching on your team, you have to play the teams that do.

            [Excuse me; I have to stop and explain something irrelevant here to keep somebody from being confused.   Early in the 1965 season the Astros had a pitcher named Ken Johnson, who had been with them for several years and was a good pitcher.   Early in the season he pitched once against the Dodgers for Houston, then was traded to Milwaukee.

          With M’waukee, he started once against the Dodgers and four times against the Astros—thus, he actually pitched four times against Houston, only twice against the Dodgers (plus one relief appearance, which is not part of this study.)   Thus, when we “zeroed out” the matching pitchers in the exercise above, we showed Johnson pitching two times against Houston, no times against LA.    But when we go game-by-game as we did in the later example, we find that Johnson actually DID pitch against LA; it was just masked by the greater number of times he started against the Astros.    It’s confusing, but it’s not an error; the math works out, it’s just that the examples are confusing.]

            We have to account for the other 3.23 wins that the Dodgers gained from facing softer starting pitcher than did the Astros.    For the Cincinnati Reds in 1965, veteran pitcher Joe Nuxhall was 11-4 with a 3.44 ERA.   Nuxhall started four times against the Astros.    He never took the mound against the Dodgers. . .well, he did, but in relief.    The Dodgers didn’t have to face Nuxhall; instead, they got three games of Jim O’Toole, and an extra game, let us say, of Joey Jay, who started 3 times against the Astros but 4 against the Dodgers.

            The advantage to the Dodgers in those four games is 0.69 wins—2.08 expected losses for Houston, based on the quality of the starting pitchers, vs. 1.39 for Los Angeles.    Let’s stop to ask:  Is there a reason that Nuxhall started four times against Houston, no times against LA, whereas O’Toole, having a poor year, started three times against Los Angeles but not at all against Houston?

            Well. . .not really.   Of course there is SOME reason for it.   Both Nuxhall and O’Toole were veteran pitchers, both left-handers.   Nuxhall was an old pitcher by then—you will remember that he pitched a game as a 15-year-old during World War II.   O’Toole, only 28 years old, came into the season off a run of four outstanding seasons, 1961-1964, and was expected to be the Reds #2 pitcher.    O’Toole pitched three times against the Dodgers, essentially, because he was in the rotation early in the year when the Reds played the Dodgers.  Nuxhall pitched four times against the Astros, essentially, because he was an old pitcher, what was called a swing man, who had beaten the Astros three times in 1964, and the Reds no doubt figured he was good enough for that kind of work, facing the Astros.   This turned out to be an advantage for the Dodgers, but it was never intended to be that; it just worked out that way.

            A similar thing is that a kind of broken-down old pitcher for the Phillies, Ray Herbert, made five starts against the Dodgers, whereas he made only one against the Astros, and the colorful but ineffective Bo Belinsky made two starts against the Dodgers, one against the Astros.    That’s five games that the Dodgers got to pick on the butt end of the Phillies’ pitching staff, while the Astros had the privilege of tangling with Ray Culp and Jim Bunning:

            Bunning made 5 starts against the Astros, 3 against the Dodgers; Culp made 4 against the Astros, 1 against the Dodgers.   We count that as two extra Bunnings and three extra Culps facing the Astros.  The advantage to the Dodgers in these five games was another 0.92 wins.  Even then, Bunning was opposed to extending welfare for the Astros.

            A third example like that was the Pirates.   For the Pirates, a veteran pitcher named Joe Gibbon made three starts against the Dodgers, none against the Astros.   The Astros had the honor of facing, instead, two extra games against Vern Law, and one against Don Cardwell:

            There’s another 0.64 wins for the Dodgers, vs. the Astros.   Of the 6.77 wins in the difference between the quality of the starting pitching faced by these two teams, we have so far accounted for 5.79--3.54 in the head-to-head games, 0.69 in the Cincinnati games we showed above, 0.92 in the Philadelphia games, 0.64 in the Pittsburgh games.

            Digression.  Joe Gibbon was a college basketball player, like Koufax and Bob Gibson except a better basketball player, and was the same age as Koufax and Gibson.  In 1961, which was Bob Gibson’s first year as a rotation starter, Gibson was 13-12, 3.24 ERA.    Gibbon, pitching for Pittsburgh, was almost the same—13-10, 3.32 ERA.   Gibson stayed healthy and built from that platform.   Gibbon got hurt in ’62, and that was the best year he ever had.   I always think about that, the huge unstated role that simply staying healthy plays in determining who goes on to be great.

            There were two organizations in that era that just really loved college basketball players, the Pirates and the Dodgers.   The Pirates in 1952 signed Dick Groat, a great college basketball player who also played some in the NBA.    After that they signed the O’Brien twins, one of whom I think held the NAIA record for career points scored or something like that, and then later they signed Bob Veale, who had played college basketball at Benedictine College in Atchison, Kansas, Gibbon, who I am certain had been an outstanding college basketball player although Wikipedia doesn’t know about it, Donn Clendenon, who had been a multi-sport star at Morehouse College, and I think some other guys who had been pretty good college basketball players.

            The Dodgers, of course, had Koufax, who had been a college basketball player briefly, and Frank Howard, who was an All-American basketball player at Ohio State, and I believe they also had some other guys.   It was a lot different then; athletes didn’t have to focus on one sport at the age of twelve the way they do now.   It was a popular way to try to find athletes in that era, to go after those great athletes from the other sports.

            Anyway, we have accounted for 5.79 of the 6.77-win difference in the quality of starting pitching faced by these two teams.  For the Braves, we can contrast three extra starts facing Tony Cloninger and two extra starts facing Wade Blasingame versus starts for the Dodgers facing Denny Lemaster (3 extra times) and Hank Fischer (twice):

            There’s another 0.95-win advantage for the Dodgers, in facing much weaker starting pitching.   There was a difference between the Dodgers and the Astros in 61 starting pitchers from 1965.   I have now shown you 35 of those games.   In the other 26, the pitching was different but equal, overall as good for one team as the other.

            Now, is that difference nothing?  Well, no, it’s not nothing.   You get to face Denny Lemaster for three games instead of Tony Cloninger, that’s not nothing; that’s an advantage.   It’s not a negligible advantage.   The Dodgers had an advantage over the Astros that season of 3.24 wins, essentially because of the luck of the draw.

            I’m trying not to overstate it.  We have a tendency, whenever we are studying X, to leap immediately into the argument that X is important. . ..I forget who it was that said that we always greatly overestimate the importance of whatever it is that we are focused on right at the moment.   It’s an advantage for the Dodgers in 1965, vs. the Astros, of something like three games.

            We can say that it was 3.24 games if we can assume that the “percentages” above were winning percentages.   They would have to be winning percentages not for that pitcher; what they would actually need to be, for this to be a true measure of the difference stated in games, is the winning percentage of the team with that pitcher as the starting pitcher.

            How close do we come to that?   Well. . .I think pretty close.   The highest “percentage” for any starting pitcher in the last ten years was .795, by Randy Johnson in 2002.   The Diamondbacks with Johnson on the mound that year were 29-6, which is. . .I don’t know, pretty close to .795.  The second-highest percentage of the last decade was .779, by Pedro Martinez in 2000, and the Red Sox were 21-8 (.724) with Pedro starting in 2000.    The third-highest percentage was .769, by Randy Johnson in 2001, and the Diamondbacks were 24-10 (.706) with Johnson on the mound, while the fourth-highest percentage was .753, by Johnson’s teammate Curt Schilling that same season, and the D’Backs were 27-8 (.771) with Schilling on the mound.    The fifth-highest percentage was .746, by Cliff Lee in 2008, and the Indians went 24-7 (.774) with Lee as the starting pitcher.    Average those five out, and I think you’ll find we’re on target.  Throw in Felix Hernandez, 2009, whom we mentioned earlier, and you can see that a “percentage” in the .700s consistently yields winning percentages for the team in the .700s, although not in every case.

            Let’s go down to .650, and see what we have.   Since 2000 there have been twelve pitchers who had “percentages” by this method between .645 and .655:   Chris Carpenter in 2006, Fausto Carmona in 2007, Ryan Dempster in 2008, Greg Maddux in 2001 and 2002, Pedro Martinez in 2005, Daisuke Matsuzaka in 2008, Mike Mussina in 2001, Roy Oswalt in 2004, Jarrod Washburn in 2002, David Wells in 2000, and Chien-Ming Wang in 2006.    I looked up the actual won-lost records of those pitchers’ teams with those pitchers on the mound.   In the same order as before, the records are 20-12 (Carpenter), 22-10 (Carmona), 22-11 (Dempster), 20-14 and 23-11 (Maddux), 17-14 (Martinez), 23-6 (Dice-K), 20-14 (Mussina), 21-14 (Oswalt), 22-10 (Washburn), 25-9 (Wells), and 22-11 (Wang).    Add that up, it’s 257-136, which is a percentage of .654.    It appears to be generally true that a “percentage” of .650 in this system yields an actual winning percentage for that pitchers’ starts of about .650.

            Of course, it could be asked:  if you wanted a winning percentage, why didn’t you simply state the pitcher’s ERA as a winning percentage, since we know how to do that, and have known how to do that since about 1980?   But there are all kind of problems with that, too.   If you work from ERA, a pitcher named Dennis Musgraves in 1965 made one start for the Mets, four relief appearances, and finished the season with an ERA of 0.56.   If you work from ERA or runs allowed to expected winning percentage, you reach the conclusion that the most difficult starting pitcher to beat in the National League in 1965 was Dennis Musgraves.    I am much, much more comfortable with a system that says that the most difficult starting pitcher to beat in the National League in 1965 was Sandy Koufax, rather than one that says it was Dennis Musgraves.

            We have Dennis Musgraves for that season with a normalized percentage of .379, meaning that we do not treat him as a particularly difficult assignment for the opposing team.   This number is at odds with his ERA for the season, true—but on the other hand, if you look at ALL of the pitchers who made one start during the season and a few relief appearances, you will find that most of them lose.  I’m confident that over 60% of them would lose, as our system says they should.   The Mets actually lost this game, that Musgraves started.  Our operating assumption is that a wide range of information, including the strikeouts and walks by the pitcher  and the number of innings that he pitched, gives us a better line on the quality of his season than does a narrow projection based on his rates of efficiency. 

            Anyway, I believe that the “percentages” used in this method, derived from Season Scores, do in fact track very well with real-life winning percentages for those pitcher’s teams in their starts—thus, that our measures of differences between teams are usable as win differentials.

            If the Astros and the Dodgers had been in the pennant race in 1965, and if the Astros had lost that race by 3 games or less, then we would have to conclude that the Astros lost the race because they had tough luck in the drawing opposing starting pitchers.   In fact, though, the Astros lost by a modest 32 games, which renders the luck element pretty much irrelevant.  

Is there any case in which a team lost a close pennant race because they had tough luck in the pitchers they had to face?

            I can’t find it.   If there is such a race, it has to be a race that was decided by three games or less, since that appears to be the outside margin of the luck differentials in this area.   In 1952 the Yankees beat the Indians by only two games, but, since the Yankees faced tougher starting pitching than the Indians did, that doesn’t seem to be a factor.   In 1955 the Yankees beat Cleveland by only three games, but the quality of the starting pitching faced by the two teams is essentially the same.    In 1957 the Dodgers beat the Braves by only one game and the Reds by only two games, and the Dodgers did face the easiest opposing starting pitching in the league that year.   However,

            a)  the Dodgers had the league’s best pitcher that year, Don Newcombe, and probably faced the weakest opposing pitching because they had the best, and

            b)  the differences are not notably large.

            The “opposing pitcher winning percentage” was .484 for the Dodgers, .503 for the Reds, but

            a) that’s just a 3-game difference, and

            b) the Reds starting pitching was just fair, which probably explains most or all of the difference between the two.

            In 1959 the National League pennant race ended in a tie between Los Angeles and Milwaukee, and the Braves did face more difficult starting pitching (.503 to .489).   You can argue that the Braves lost the race because of tough luck in the starting pitching if you want to, but

            a) I wouldn’t bet on the claim to stand up to a more rigorous analysis, and

            b) when you do as many stupid, stupid things as Fred Haney did that summer, you’re not allowed to bitch about your luck.

            One of the worst managerial performances in the history of baseball.

            I am sure there is some race somewhere in which the random difference in the quality of pitchers faced was greater than the margin of victory in the race, but. . .it’s like a bad break in a basketball game.   Maybe a referee makes a bad call against you, and you wind up losing the game in overtime.   Certainly the referee’s call cost you the game if you want to look at it that way, but it’s not a healthy way to look at it.   If the referees make twenty bad calls that go against you and five that go your way, then you’re entitled to bitch about it if you want to.   Then it’s a real factor that decides the game.   If it’s one, you move on.

            My point here is that it is, almost without exception, a much smaller factor than I would have guessed.   The New York Mets faced the strongest opposition starting pitchers in the National League in 1962, 1963 and 1964.   Why?   Because they didn’t have any pitching themselves.  Because their pitching was weaker than anyone else’s, the opposing pitching they faced was stronger than anyone else’s.   It’s a natural effect.

            The data shows that, year-in and year-out, the teams which face the strongest opposition pitching are the teams that have the weakest, and vice versa.    My point is:  that’s really not a very large effect.    Let us suppose we are dealing with a ten-team league, and that the starting pitching of your team is at .410 (.500 being average.)   The starting pitching you FACE should average about .510.   That’s not a large effect.   We’ll call this the “natural” effect of having poor pitching.

            IF there were large “random” effects here, they would entirely swamp the natural effects, and the natural effects would be almost invisible in the data.   In fact, though, the natural effects DOMINATE the data.   It is almost ALWAYS true that the teams which have the best pitching are the teams which face the weakest pitching, and vice versa.   That means that the other effects, the random effects, have to be very, very small. 

            The standard deviation of the expected wins based on opposition pitching differences is one game per season—INCLUDING the natural effect.   Without the natural effect, it would be even smaller—probably much smaller.   There’s just not much there.

            In the article in the Gold Mine, I included this sentence:

The difference between the Padres and the Yankees in 2009 amounts to about 15 games started by a Cy Young-quality pitcher as opposed to a replacement-level pitcher.  It’s a huge difference. 

            Am I now saying that I was wrong about that?

            Well, yes, sort of.    The average Season Score for a pitcher facing the Padres in 2009 was 97.1; the average facing the Yankees was 71.6.   That’s a difference of 25.5 points, times 162 games is 4,000+ points.    Assuming that the difference between a Cy Young pitcher and a replacement-level pitcher is not more than 300 points, my statement is literally true.

            But first, I wasn’t adjusting there for league differences.   The norm in the American League was 78.3; in the National League, 84.5.   If we take that out, the difference shrinks from 25.5 points per game to 19.3.

            Much of that difference is the natural effect of the fact that the Yankees had better starting pitching than the Padres did, thus tend to face worse starting pitching.   Let’s say that accounts for another 8 points in the difference—it’s just an estimate—that cuts us down to 11.3, rather than 25.5.  That means that it’s not 15 games by a Cy Young starter, but more like 6.

            That seems like a huge thing, and I’m not saying now that it is negligible, but. . ..it’s not the end of the world.    Let’s say that a team wins 40% of the time with a marginal pitcher on the mound, 73% of the time with a Cy Young starting pitcher on the mound.   That’s a difference of .33, times six is essentially 2.00.   The difference between the team facing the toughest starting pitching in baseball and the team facing the weakest, setting aside the natural effect, is two games. 

            OK, I’ve gone on too long here, let me wrap up my notes.   These are the ten teams since 1952 which have faced the most difficult starting pitching:

            Obviously that’s more than ten teams.   I included the teams for which there were significant numbers of missing games, but didn’t count them.  And these are the ten teams since 1952 which have faced the least difficult starting pitching:

            Noting again the influence of the “natural effect” on the data.   Three of the greatest pitching staffs of all time were the 1954 Cleveland Indians, the 1965 Dodgers and the 1971 Orioles.   All three teams are on the list above, of teams that faced very weak starting pitching—because they had a near-monopoly on the league’s best pitchers.

            Where are the Braves?   The Braves with Maddux, Glavine and Smoltz faced relatively weak opposition starting pitching virtually every year, but that was a 14-team or 16-team league.   The natural effect is weaker in a larger league.    The lists above are dominated by teams from the smaller leagues—demonstrating again just how weak the “random” effects on the data must be.

            I sorted all pitchers into groups “A”, “B”, “C” and “D” in this study, with “A” pitchers being those who had percentages over .600 (meaning Season Scores 100 points better than the league average), “B” pitchers being those between .500 and .600, “C” pitchers those between .400 and .500, and “D” pitchers those under .400.

            The 1973 Kansas City Royals faced 52 “A” quality starting pitchers, the most of any team in the study.  The Royals had a good year.   A fifth-year expansion team, they finished 88-74, six games behind the World Champion Oakland A’s, despite running into a historic number of quality starting pitchers.

            But to a large extent, this, again, is a natural effect.   Even though I normalized the Season Scores to .500 for every league, the distribution of pitchers is very different in different leagues.   Some leagues have a number of pitchers having Cy Young-type seasons; some leagues really have none.   Even after you normalize the data.

            The American League in 1973 had twelve twenty-game winners.   Everybody in the league faced a lot of quality pitchers.    The Royals just happened to face the most.    The 52 “A” quality pitchers faced by the 1973 Royals were Jim Palmer (5 times), Bert Blyleven (6 times), Nolan Ryan (5 times), Ken Holtzman (3 times), Catfish Hunter (5 times), Bill Singer (4 times), Bill Lee (3 times), Jim Colborn (twice), Wilbur Wood (5 times), Luis Tiant (twice), Vida Blue (6 times), Joe Coleman (4 times) and Gaylord Perry (3 times).   The Royals actually were 23-29 in these 52 games—not a half-bad record against that quality of competition.   The second-most “A” quality pitchers ever faced by a team was 50, by the 1965 Houston team that we wrote about earlier.

            On the other hand, the 2004 Red Sox—you remember them?—faced only five “A” quality starting pitchers all year.   The Yankees didn’t have an “A” quality starting pitcher that year, Blue Jays didn’t either.   The only “A” quality starting pitchers in the division were Curt Schilling and Pedro Martinez, and the Red Sox had them both.   With the imbalanced schedule, that’s a lot of games out of the way that you don’t have to worry about facing a real good pitcher.

              On the other side of the scale, the most “D” quality starting pitchers ever faced by a team was 46, by the 1965 Dodgers.   Again, we looked at them earlier.   The Dodgers were 35-11 in the 46 games started by “D” quality opposing starters, which makes them 62-51 in other games.

            On the other hand, the 2006 Chicago Cubs faced only one starting pitcher with a “normalized percentage” under .400.   It was kind of a “flat” year for NL pitchers—a lot of guys who looked the same, without standouts on either end.  The Cubs didn’t get the opportunity to face their own Mark Prior (1-6, 7.21 ERA), Glendon Rusch (3-8, 7.46 ERA) or Angel Guzman (0-6, 7.39 ERA), and they didn’t happen to run into Kyle Davies, Russ Ortiz or Jose Lima.   The only really bad starting pitcher they faced was Jason Johnson, who was bouncing from team to team, and was at that moment pitching for Cleveland.   

            The other thing I looked at was the ratios of “good” starting pitchers faced—normalized percentages over .500—to “bad”.   On a first impressions look, these numbers look startling.   These are teams that faced the most “good” pitchers;

            While these were the teams that faced the most “bad” starting pitchers:

            The 1955 Dodgers faced 114 starting pitchers who were less than average?

            Yes, it’s a big number, but even though we normalize the league to .500, there are usually more pitchers under .500 than over (because those who are over .500 are usually further from .500.)    In the National League in 1955, 62% of the starts that we have accounted for were by pitchers under .500.  The “norm” isn’t 77-77; it’s 58-96.

            The Dodgers are still 18 games off the norm, but the Dodgers—by far the best team in the league that year—put very good pitchers on the mound themselves, which causes them to have an above-average number of games against below-average pitchers as a natural effect.

            Hey, the article’s too long, I know; I appreciate your staying with me.   I love trying to figure out a way to know something about a team that they didn’t know themselves at the time, something that nobody knew.   Nobody knew, at the time, that the 1970 Cincinnati Reds had exploded on the league, in part, because they faced a lot of weak pitching.   It’s a small note, but I like to know this stuff.