Baseball stat basics: identical seasons worlds apart

Sam West had a nice career playing for the Washington Senators and St. Louis Browns from the late 1920s to the early 1940s.

Check out his 1935 and 1936 seasons, which perhaps look identical:

1935:  527 AB  .388 OBP  4 triples  10 homers  70 RBI

1936: 533 AB  .386 OBP  4 triples   7 homers 70 RBI

Pretty much identical, right?


Here’s another illustration of why a lot of raw numbers can be deceptive, especially without context.

In 1935, West played in 138 games, getting 615 plate appearances. He walked 75 times (12.2% of the time) and struck out 46 times (7.5%).

In 1936, West played in 152 games, getting 637 plate appearances. He walked 90 times (14.8% of the time) and struck out 70 time (11.0%).

So, he bumped his walk rate up a little, but his K rate shot up by more than 40%.

His homers dropped from 10 to 7, by itself not all that meaningful. But, he also hit 11 fewer doubles (37 to 26) and lost 61 points off his slugging percentage (.442 to .381)

He also lost 22 points off his batting average (.300 to .278).  All together, despite getting more plate appearances in 1936, his total bases dropped from 233 to 203.

To make matters worse, league-wide run scoring was way up in 1936, at 5.19 runs per game, from just 4.90 runs per game in 1935.

Factoring in all of these differences, and it turns out that West’s OPS+ was 111 in 1935 but just 88 in 1936. He went from having a good above-average season to a pretty disappointing below-average one. For 2011 comps, his 1935 was similar to what Yunel Escobar and Alberto Callaspo did while his 1936 was similar to what Austin Jackson and Martin Prado did.

But, you say, he still drove in 70 runs each year, so what does it matter? Well, we don’t have his detailed batting splits but I’m guessing he hit a lot worse with runners on in 1936 than in 1935. We can see that in the earlier year, he batted a lot more in the 1 or 2 position in the lineup, whereas in 1935 he got a lot more time at the 5 and 6 holes. Almost certainly he had more RBI opportunities in 1936 and yet batted in just the same number of runners.



Baseball stat basics: identical seasons worlds apart — 50 Comments

  1. It’s great when you can explain the difference upon closer inspection, but it’s frustrating when you can’t.

    Here’s an example. If anything the season with the lower OPS+ looks a bit better – better BA, fewer Ks, more runs. The only things better in the other season are triples and a few more walks.

    Player Year 5 Tm G AB R H 2B 3B HR RBI BB SO 2 Doug Rader 1972 109 HOU 152 626 553 70 131 24 7 22 90 57 120 .237 .309 .425 .734
    3 Doug Rader 1973 99 HOU 154 632 574 79 146 26 0 21 89 46 97 .254 .310 .409 .719
    Provided by View Play Index Tool Used
    Generated 3/13/2012.

    Sorry. Don’t know why that’s coming out garbled. But, it’s Doug Rader in 1972 and 1973. Same team and ballpark. Very similar stats, but a 10 point difference in OPS+.

    • Rader also grounded into 8 fewer double plays in ’73.

      The answer of course is the league moved and Rader didn’t. In ’72 NL teams scored an average of 3.91 runs per game. In ’73 that figure was 4.15 which is nearly a quarter of a run per game more. To put that in context, in the 1920 offensive “explosion” AL teams scored about two-thirds of a run more per game (0.66, to be exact) than in 1919. Not as dramatic a change to be sure and not even all that uncommon but that 6+% increase in offense accounts for two-thirds of Rader’s change in relative performance. I would guess that the 7 fewer triples and subsequent lower slugging percentage is the rest.

    • Doug….easy explanation. 1972 MLB: 3.69 runs per age, 1973 MLB: 4.21 runs per game. Like my example above, that was one of the year-to-year quantum leaps in offense.

      • And it was even more dramatic in the AL, which jumped from 3.47 RPG to 4.28 RPG. Almost a full run increase in the AL. So the owners got what they wanted – add a DH, increase run scoring.

  2. Thanks guys.

    I should have looked that up.

    Looks like 1972 was the end of the “second dead-ball era”. Seems like a pretty abrupt end, just like the first time.

    • 1972? – try 1968 (3.42 R/G)…
      What we now call the “second dead-ball era” was from 1963-68, after MLB re-defined the strike zone before the 1963 season. In 1968, 21% of all games ended in shutouts.

      They then re-re-defined the strike zone before 1969, to make it more favorable for hitters. They also limited the maximum height of the pitcher’s mound from 15 to 10 inches (it was probably more than 15 inches, as they really didn’t pay much attention before – see Dodger Stadium). Expansion also contributed to the increase in offense.

      In 1969/70 scoring was up substantially, though still not quite up to average historical norms (4.07, 4.34 R/G). But by 1971, scoring declined a lot (3.89), and even more in 1972 (3.69). That is when MLB introduced the DH, to increase scoring. As Andy in #3 noted, it certainly worked.

      Hope this helps.

      • I still like 1972 as the end of the second dead-ball era. The 1969 changes really only worked for 1969 and 1970. By 1972, pitchers had adjusted and resumed their dominance – that year had the lowest MLB R/G (other than 1968) since 1918.

        What’s more curious is how, even without the DH, the NL also went up a quarter-run in 1973, and stayed basically around that level really until 1992, save for upward blips in 1977 and 1987.

        • Doug, you can call 1972 the endpoint, but I’ve always seen the second dead-ball era defined as 1963-68. 1969 was the second full year that I followed and actually understood MLB, and the contrast between 1968 and 1969 was quite dramatic.

        • 1969 was also an expansion year with 4 teams added. I am sure that was also a reason for the 1969-1970 bump in offense.

  3. Good article, but then again any article about the Pride of Longview, Texas is a good one.

    Anyways, we all know that the context of league scoring is important, but the stats used to show the similarity of the seasons seem kind of arbitrary.

  4. In my lifetime, biggest two year jump MIGHT be MLB 1968 – 1970….scoring went from 3.42/g to 4.34 – roughly a 12.5 % increase each year. But, then again, I didn’t look at 1993 – 1995,; or any of the “steroid” years.

    Hey, can I say it? “Steroids work – real well”

    • Of course, as we saw when expanded the strikezone was expanded in the 1960s, which eventually leading to the year of the pitcher, and then when they lowered the pitching mound in 1969, MLB is very good at changing scoring.

      There the 1919 Black Sox scandal and the 1994 strike-shortened season encouraged MLB to continue trends leading to greater offense during those times by juicing the baseballs. Fans love scoring.

    • It wasn’t just steroids. There is almost indisputable evidence that Rawlings changed the baseball between the ’93 and ’94 season. Runs scored went up from 4.60/g to 4.92/g in ’94. AB/HR went down from one every 38.5 AB in ’93 to one every 33.3 AB in ’94.

      • Edit: change that to “some evidence” instead of “almost indisputable evidence”. I was trying to change that when I must have hit the “submit” button.

        • Bstar, this is not a challenge, but — can you point me to that evidence about changes to the baseball before ’94?

          • I’ll jump into this since I attempted in my botched post above to allude to “rumors” that MLB purposely spiked the baseballs heading into the 1994 season, a year in which there would be a lot of negative stories around a potential strike, one which sadly came about. (There was a similar suggestion that MLB did the same after the 1919 Black Sox scandal.)

            For the record, I don’t believe either story, meaning that MLB purposely juiced the baseballs, but I do believe “something” happened in 1994 that goes beyond steroid narrative.

            In 1994, the two leagues witnessed a significant rise in runs per game and HRs per game. In the five years before 1994 (during which time steroids were certainly being taken), the AL’s HRs ranged from 1.52 per game to 1.83 per game. In 1994, it spiked 136 percent higher over the five-year average to 2.23 game. The NL saw a 128 percent increase.

            I think that spike across all of baseball has led some to believe there was a conspiracy to increase HRs/scoring. I’m not part of that camp, although I won’t discount it either. I just don’t think there’s any proof of a conspiracy. Yet that doesn’t mean that there weren’t changes made that increased scoring without intent, with one likely suspect being a change in the baseballs.

            What I don’t believe is the increase was caused by steroids in one single year, especially since we now know steroids have been part of the game going back to the 1980s. I think a number of things happened in the 1990s that caused an overall increase in scoring, with steroids being one of them, but I think something else happened causing the rapid spike in 1994.

          • I’ll try and find the best article I’ve read about it…there certainly are some out there. I don’t necessarily consider it a conspiracy theory, though….meaning if they did it, they were trying to improve the game for fans and I’m fine with that. I don’t think it’s really that different than lowering the mound in the late 60s… saw an imbalance in the game, thought it might be detrimental to the sport, and did something about it. Just because it wasn’t openly known that they changed the ball doesn’t, to me, make it a big conspiracy. They were trying to help the game.

            John, I actually think the huge offensive spike is proof enough, because no such spike occurred from year to year after ’93-’94. That, and major league baseball proved they knew all about juiced balls when they admitted to using them in the (???) All-Star game at Fenway.

            I will try to find some articles about it tonight, but it may be tomorrow before I get back to you. I knew I was in trouble with suggesting this. I should have come with links in hand, but I didn’t. My bad.

          • Ok, here’s some, John:


            The one above points strongly to baseballs being different today than they were in the past. Two scientific studies came to the same conclusion.

            Here’s one from Jay Jaffe at Baseball Prospectus that mentions yet another study:


            Here’s another pretty convincing article; not a scientific one, but a great look at how something definitely happened ’93-’94:


          • I’ve read that the MLB manafacturer (Rawlings?) changed the method of making the baseballs before 1994, and that they were on the high end of the tolerance for “liveliness”. So while they were still within the specs, on average the ball was indeed more lively.

          • MikeD @19 — A couple of numerical points to square away first:

            1. On a year-to-year basis, the biggest spike in R/G and HR/G of the steroids era occurred in 1993, not ’94.
            – In 1993, R/G rose about 12% and HR/G by 23%, compared to ’92.
            – In ’94, they rose by another 7% and 16%, respectively, compared to ’93.

            2. Using your 5-year average basis of comparison, it’s true that the ’94 spike was bigger:
            – In ’93, R/G and HR/G rose by 10% and 17%, respectively, over the previous 5-year average.
            – The ’94 rates were up by 15% and 31% over the previous 5-year avg.

            3. In your 3rd paragraph, your AL figure of “136 percent higher” is incorrect. An increase of 136% means the figure would have to more than double. Perhaps you meant 36% higher? Ditto for your 128% figure for the NL.

            (All my figures are built from B-R’s numbers copied into Excel. I use the average per team per game, rather than the 2-team average.)

          • John, yes, I noticed my percentages were off as I hit sent. An edit button would be great addition to the site to protect to the innocent, or in my case, the guilty!

            Beyond that, my message remains the same.

          • bstar @20-21 — Thanks for the links. I’ll read them as soon as I can.

            But to your statement that “the offensive spike is proof enough, because no such spike occurred from year to year after ’93-’94,” I would reply: That all depends on what size of spike you consider significant.
            – 1996 saw an 8.2% surge in HR/G.
            – In ’99, R/G and HR/G rose by 6.2% and 9.3%.
            – In 2000, R/G reached a local high of 5.14, or 4.5% more than 1994.

          • John, to answer your question you have to consider the approximately 5-year context of what the run enviroment had been in the game. The spikes you mention in ’96, ’99, and ’00 are not as noteworthy as the ones in ’93 and ’94 because by ’96 it had already been established that this was an era where scoring and offense were on the increase. Looking at the 5-year-previous run contexts, the spikes in ’93 and ’94 stand out more since there wasn’t any previous evidence that offense would spike suddenly. The last link I gave you probably explains this better.

      • “It wasn’t just steroids.”

        I’d go so far as to say it wasn’t steroids at all, or very little of it was. The data doesn’t match up with any reasonable steroids-based theory. We know steroid use was rampant starting in the 80s, and offense went all over the map in the 80s, with one big block jump forward in that 1993-94 period. And it didn’t really drop off again in a significant way until a few years *after* the testing program was put in place. They changed the ball, the parks got smaller, random fluctuation happens just like it always has — all these explanations make a lot more sense than the “steroids work” one. I’m not at all sure they do, in baseball.

        • I have no idea how much impact steroids had, but it’s absolutely true that there were many other factors.

          For one thing, Denver came into the league in 1993. That alone accounted for 1 percentage point of the 12% and 23% increases in R/G and HR/G, respectively, for 1993 over 1992.

          • I have a thought re: steroids that I’ve never seen expressed and I want to put it out there to see what people think. We always hear about the hitters. Not so much about the pitchers. But presumably there were lots of pitchers on steroids as well.

            Now, I started watching baseball in the mid-70s. At that time there were basically two guys who could consistently hit 95+. Gossage and Ryan. But in the steroid era there were lots and lots more pitchers hitting that level. And the one thing we know via physics is that the faster the ball is thrown the further it can be hit. So I’ve been wondering if the increase in homeruns wasn’t at least partially the result of pitchers throwing harder (which was probably due to pitchers being on steroids). Presumably this would help all hitters, regardless of whether they were on steroids themselves.

            As a corollary, I wonder if pitchers taking steroids led to a decrease in the quality of pitching. In other words, pitchers who could throw hard because they were on steroids, were getting promoted to the bigs before they really should have been. (cause obviously most teams prefer power pitchers over junkballers).


          • Ed @34 — I think that’s a very interesting theory, is what I think.

            I have no doubt that the percentage of pitchers using steroids was on a par with the percentage of hitters.

            However, I do not take for granted that today’s average fastball velocity is higher than that of 20+ years ago. I’ve read here and there that the change is more traceable to the new generation of radar guns, though I can’t cite any authority.

            Also, I still believe that the biggest factor in the rise in HR rate was the change in the predominant hitters’ approach, which I believe is also the main factor in the rise in K rates.

            Through 1993, The MLB strikeout rate never topped 6.0 K/9. In 1994, it jumped from 5.8 to 6.2, and has gradually risen to the present 7.1.

            There is a pretty strong correlation between HR and K rates. For instance, 1987 saw the highest HR rate to that point in history, and also the tied the highest K rate to that point (which was last reached in 1967).

          • Well yes, obviously it’s only a theory. But that’s all any of this will ever be. A bunch of theorizing. We’ll never conclusively know what caused the jump in offense.

          • P.S. to Ed — It seems that you may be taking my phrase “an interesting theory” as criticism. I meant it as praise. As you noted, most of what we deal in here is theoretical to some extent.

        • Bill, I agree with almost everything you said since that’s been my position for years. There were many factors that intersected to cause a rise in hitting. The spike in the ’93-’94 range is the one that’s most interesting and the one that I doubt had anything to do with steroids.

          I say almost because I can’t go as far as you and say “it wasn’t steroids at all.” I believe it was a contributing factor, but much smaller than most people assume. Yet, I’m open to any and all evidence that it contributed nothing, or contributed more greatly than my assumption.

        • FWIW, here are the largest percentage changes in HR/G since 1901 (positive or negative, measured against the previous year, rounded to whole numbers):

          +72%, 1919
          +50%, 1977
          +49%, 1921
          +44%, 1911
          +38%, 1910
          -32%, 1931
          +31%, 1925
          +29%, 1947
          -28%, 1988
          +28%, 1920
          +27%, 1932
          -27%, 1926
          +26%, 1934
          +26%, 1982
          +24%, 1929
          +23%, 1993

          • What the heck happened in 1977? Well, I know one thing that happened. That was an expansion year as was 1993. Interesing that 1961, 1969 and 1998 did not show up on the list, especially 1969 since that’s the only time in MLB history when the leagues expanded by four teams, and of course MLB was coming off of the Year of the Pitcher in 1968 and changed both the strikezone and went to a uniform pitching mound height. So many ingrediants in place for a massive increase, yet doesn’t make the list.

          • I forgot 1962, another expansion year that brought us Mets and the Colt .45s, now the Astros. So six expansion seasons, two of which make the list. The off-the-charts ’77 season making the list, and ’69 not are the two surprises to me.

          • MikeD, what happened in 1977 is that was the year that Rawlings reportedly first changed the ball.

        • Bill:
          If Tim Montgomery goes from the fifth fastest man in America to the fastest man in the history of the world, then I have to believe steroids work.
          A friend advised me of a professional wrestler weightlifting training partner of his – 420# in the bench clean; 580#’s on the juice.
          One thing I certainly will agree on: If a pitcher constantly attempts to pitch around hitters, hitters will get ahead in the count and get good pitches to hit. If the strike zone goes from the top of the letters down to the bottom of the knees and Selig says “let’s tighten the strike zone” , then pitchers have even bigger problems.
          We know the 1930’s numbers put up by Foxx Gehrig and Greenberg were done with some form of altered baseball, but I’m just not buying the 1990’s version of “tighter seems” or “more elastic core” when second basemen were 6’1″ and 220#’s.
          Just look at Caminiti – he started juicing well into his career and altered his production. He may have been awful high an awful lot, but I don’t recall him mentioning a juiced baseball

          • But Paul, the next question is, for what purposes do steroids work?

            Neither brute strength nor raw speed are primary components of star ability in baseball.

            There’s an interesting piece in ESPN:Magazine’s recent Analytics Issue on Dodgers head trainer Stan Conte (not to be confused with BALCO founder Victor Conte), who is trying to build a sort of sabermetrics of injuries. Buried within is this little nugget:

            “Though most teams had full strength-and-conditioning programs for the first time, injury rates had increased every season from 1989 to 1999, … peaked in 2001, dipped the following year, then plateaued in 2006. They’ve been higher ever since. … [I]ndustry experts have noted that the window of lower rates could mark the peak of [PEDs], which are thought to accelerate recovery.”

            Well, maybe — but that presumes a big surge in PED use between 1999 and 2005. An alternative hypothesis would be that a significant percentage of players were already using PEDs during 1989-99, and that the PED use contributed to the rising injury rate by promoting an excessive focus on strength training.

  5. Those Browns teams, managed by Rogers Hornsby, had typical Browns seasons (bad), but plenty of good seats were available (80,000 total attendance in 1934). They also featured the first “modern” slugging third-baseman in Harlond Clift.

  6. In 1935 the Browns drew 80922 fans. I already mentioned in another post that Clift set a HR record for third-baseman but I forgot the year , will have to look it up.

  7. First, thanks Andy, that’s a great intro to the idea of OPS+. I’m enjoying these occasional ‘basic’ lessons, even though I’ve had a few years following the B-R blog to understand the stats.

    Second, I’ve wondered for awhile, and this seems like the place to ask – is OPS+ and other ‘adjusted to league’ measures adjusted to the whole of the major leagues, or is it AL-NL specific? It’s a bit more murky now with more interleague games, but for example, it seems to me that the change to the DH in the AL would affect scoring in that league, so players in the AL should only be measured against others with regards to OPS+ and WAR, etc. Is that the case?

  8. I have another example of seemingly similar seasons with a much different result in final evaluation. This time the difference is in WAR for a pitcher.

    I happened to be looking at the Cy Young results for the AL in 1974 when I noticed the top 2 players had eerily similar stats. It made me think of this post. The players are Catfish Hunter (won the Cy Young) and Fergie Jenkins (runner-up).

    Catfish stats – 25-12, 41 GS, 23 CG, 6 SHO, 318 IP, 46 BB, 143 K, 0.986 WHIP, 2.49 ERA, 134 ERA+
    Fergie stats – 25-12, 41 GS, 29 CG, 6 SHO, 328 IP, 45 BB, 225 K, 1.008 WHIP, 2.82 ERA, 126 ERA+

    Those stats seem pretty close, except Fergie has a big advantage in K’s and Catfish has an advantage in ERA+ (although neither value is overwhelming).

    I figured their WAR values would be pretty similar, but Jenkins has a pretty big edge over Hunter (7.6 to 6.4). I checked the defense and Hunter had 14 rdef more than Jenkins (8 to -6), but Fergie held a 1.0 to 0.9 lead over Catfish in aLI (although I am not sure how much that plays into the overall calculation). But it seems like the Strikeouts have much more of an impact on a pitcher’s WAR then I thought. NOTE: “WAR” is the baseball-reference WAR value.

  9. I don’t know what to think about steroids. But I do know a few things:

    A substantial number of ballplayers used
    Many used over a prolonged period
    Many still continue to use

    So, my question is, why would ballplayers continue to use something that in no appreciable way improved their ability to hit a baseball. I mean if they didn’t help AT ALL wouldn’t word have gotten out by now…”the “sh*t don’t work!”

Leave a Reply

Your email address will not be published. Required fields are marked *