Remember me
Forgot your username/password?
 
Print
Home>Articles

The 1963 to 1968 Hitting Desert

By Bill James
June 6, 2020
                              The 1963 to 1969 Hitting Desert

 

            There are 169 major league players who had 800 or more at bats in the years 1963 to 1968, the hitting desert of the 1960s, and who also had 800 or  more at bats in the majors outside of those years.  For the sake of convenience, I will refer to the years 1963-1968 as "the desert", and will refer to years within that scope as "in", and to the years outside that scope as "out".  

            Of those 169 players, exactly 100 hit for higher averages outside the desert than inside.  62 hit for higher averages in the desert than out, and 7 hit for the same average in the desert that they did outside of it, although this counts as being the same Orlando Cepeda, who hit .297 inside the desert and .296 outside of it, but who counts as the same because the difference in his average is less than half a point. 

            I am surprised that the differences are not larger and more consistent than they are; I would have guessed that 80% of players would have hit better outside than in, and I would have thought that a good many players would have lost 20 points or more.  Of the 169 players, 23 are Hall of Famers.  Among the Hall of Famers, 11 hit for a better average outside the desert than in, 3 (including Cepeda) hit for the same average, and 9 hit for a better average in than out.   The other two Hall of Famers who hit for the same average in and out are Bill Mazeroski (.260) and Frank Robinson (.294). 

            As a generalization, players who were born in the years 1934 to 1939, maybe 1940, hit for higher averages in the desert than out, while those who were born before 1934 or after 1940 generally hit for better averages outside than in.  Of course, this does not mean that the players born 1934-1940 were not affected by the tough hitting conditions, or effected, whichever one it is supposed to be; it just means that the losses are offset by the fact that the players were in their prime in those years, and were better hitters then than they were before then or after then. 

            Among Hall of Famers, those who hit better inside than out were, and I am going to give you their birth years from memory because I don’t want to look all of these up. . .those who hit better inside than out are led by Ron Santo (1940), who hit .292 in the years 1963-1968, but only .265 in the rest of his career.   Others include Carl Yastrzemki (1939, .304 and .278), Roberto Clemente (1934, .326 and .312), Brooks Robinson (1937, .276 and .263), Harmon Killebrew (1936, .264 and .252), Willie McCovey (1938, .276 and .266), Billy Williams (1938, .292 and .288), and Hank Aaron (1934, .306 and .305).  Al Kaline, also born 1934, hit only two points better outside than in (.298 to .296).  Joe Morgan, born 1942 or 1943, is the only Hall of Famer born outside those years who hit better in the desert than outside of it (.273 to .271), but of course Morgan’s best seasons were outside the desert. 

            Hall of Famers born away from that period lost much more because, of course, they were not in the prime years anyway.   Rod Carew (born 1945, I think) hit 50 points higher after the desert than in it (.333 to .283).   The four superstars born in 1931 all had significant losses during the desert—40 points for Mickey Mantle (.309 and .269), 35 for Eddie Mathews (.282 and .247), 22 points for Ernie Banks (.282 and .260) and 9 points for Willie Mays (.304 and .296, rounding discrepancy.)  Mantle, as you will remember if you are my age, lost a .300 career batting average by playing through the hitting desert, and regretted that.   His career average dropped from .316 through 1957 down to .298, and Mantle said that he always thought of himself as a .300 hitter, and regretted that he had not retired before the .300 average got away from him.  

            Speaking of Mantle,  here’s an interesting thing that popped up.   You remember Floyd Robinson, anybody?  Chicago White Sox outfielder, hit .300 in ’61 and had a big season in 1962.  Mickey Mantle and Floyd Robinson have almost identical hits and at bats in the desert.   Mantle had 594 hits in 2,206 at bats in the desert; Robinson had 595 hits in 2,206 at bats—one more hit, in the exact same number of at bats.   Robinson hit .270 in the desert, Mantle .269.

            But here’s the thing.  Robinson also out-hit Mantle by one point OUTSIDE the desert, .310 to .309.   That means that Floyd Robinson out-hit Mickey Mantle (for average) both INSIDE the desert (.270 to .269) and OUTSIDE the desert (.310 to .309)—but Mantle outhit Robinson by 15 points overall (.298 to .283).  If you’re a high school math teacher who uses baseball as a teaching tool, remember that one. 

            The ten players who lost the most points during the desert were Rod Carew (50 points), Lee Thomas (48 points, .288 to .240), Wes Parker (.289-.245), Paul Schaal (.262 and .218), Art Shamsky (.273 and .231), Reggie Smith (.294 and .253), Tito Francona (.286 and .246), Mickey Mantle (.309 and .269), Floyd Robinson (.310 and .270) and Norm Siebern (.287 and .248). 

            Norm Siebern and Jose (Who the Hell) Azcue were teammates in 1962; Siebern hit .308, and Azcue hit .229.   The hitting desert began the next year, and Azcue out-hit Siebern, .281 to .272.  During the hitting desert as a whole, Siebern’s average went down 39 points, and Azcue’s went up 39 points.   Siebern is 10th on the list of those who lost the most during the desert; Azcue is first on the list of those who gained the most.   That list is Azcue, 39 points (.226 and .265), Don Wert (.213 and .250), Dalton Jones (.213 and .248), Dal Maxvill (.204 and .234), Gene Alley (.237 and .266), Ron Santo (.265 and .292), Roger Repoz (.208 and .235), Carl Yastrzemski (.278 and .304), Curt Flood (.278 and .303), and Paul Casanova (.214 and .238).    36 players gained 20 or more points during the desert; 16 players lost 20 or more points.

            If there are any players who lost a Hall of Fame career due to the hitting desert, the two best candidates are two players who came up about the same time with the Dodgers—Willie Davis (.290 and .262) and Frank Howard (.279 and .267).  Other players who had long, distinguished careers and 20-point losses during the desert include Reggie Smith (41 points, but it was only two seasons for him), Jose Cardenal (.283 and .255, 28 points), Roy White (.276 and .250, 26 points), Ken Boyer (.297 and .272, 25 points; Boyer is part of the 1931 group), and George C. Scott (.273 and .249, 24 points.) 

            On average, the 169 players lost just five points during the desert (.264 to .259).  I expected the loss to be larger, and I assume that it would be larger if measured by some other method.  

            Thanks for reading. 

 

 
 

COMMENTS (26 Comments, most recent shown first)

vandorn
Correct, Oester played half his games on Riverfront Stadium's turf, plus road games at Philadelphia, Pittsburgh, Montreal, St. Louis and Houston, which resulted in about 114 games on turf. Sax played half his games on grass at Dodger Stadium plus the road games in Chicago, Shea Stadium, San Diego, San Francisco, and Atlanta, leaving about 42 games to be played on turf. Both players hit much higher on turf than grass, but since Oester played many more games on turf he had the higher average. It's a classic example of Simpson's paradox.
4:38 PM Jun 10th
 
lidsky
Allen - the reason this is possible is they didn't have the same number of at-bats in Grass vs. turf. Let's say, for example, that they both hit better on turf than on grass (which I'm guessing is likely). If Oester had a greater ratio of turf-at-bats to grass-at-bats than Sax than he could have a higher average than Sax, as the greater number turf-at-bats will pull up the total average.
3:41 PM Jun 10th
 
Allen Schade
OK, I'm slow, how can Sax outhit Oester on both grass and turf and then Oester have a better average???​
2:18 PM Jun 10th
 
vandorn
Bill's comment about Mickey Mantle and Floyd Robinson is an example of something called Simpson's Paradox. I heard of the paradox years before I got a degree in statistics and found out it had a name.

How did I know? I knew it because Bill mentioned an example in the 1986 Baseball Abstract involving Ron Oester and Steve Sax. In 1985 Sax hit for a better average on turf, Sax hit for a better average on grass, but Oester had the higher batting average overall.
11:44 AM Jun 10th
 
KaiserD2
Ooops again, I meant baseball-reference.
8:05 PM Jun 9th
 
KaiserD2
Bill was right, sorry. All dates from retrosheet.
8:04 PM Jun 9th
 
FrankD
Interesting. I too thought there would have been a bigger difference between hitting in the desert vs out of the desert. Now is Bill going to write the corollary article: Pitching in the Oasis ?
10:51 AM Jun 9th
 
bjames
Just checked. He's actually wrong about Carew, too.
10:24 AM Jun 9th
 
bjames
Again, I still have not checked the dates, but Kaiser was also WRONG on all of his corrections except maybe Carew. Clemente was in fact born in 1934, not 1935 as Kaiser "corrected", and Yastrzemski was in fact born in 1939, not 1940. Kaiser is just subtracting their baseball age from the year, but this gets the wrong answer for anyone born after July 1. Clemente and Yaz were just born after July 1--but in the year that I said. I think all of them were, except possibly Carew.
10:22 AM Jun 9th
 
steve161
Perhaps you missed it, but David also included some corrections within the body of the quote. In each case, the date was one year off. So yeah, I have to echo Meris' question: what was the point?

(FWIW I'm older than both David and Bill, and my memory for dates is godawful.)
7:39 AM Jun 9th
 
jrickert
I was curious about what the ratios would be for a six year non-desert set of years so I duplicated the study on the years 1983-1988. There were 240 players with at least 800 at bats inside and outside the years. 136 of them had a higher batting average during the 1983-8 period, 57%.
Thinking about it further a standard deviation for the differences in batting averages should be about 15 points ( roughly sqrt[2*.3*.7/1500)] ) which makes the 59% number for 1963-8 seem a little low but not unreasonable.
7:21 PM Jun 8th
 
MarisFan61
MW: Do you realize, HE WASN'T EVEN RIGHT?

I mean, I'm often (usually?) not even right, but I try to apologize for it. :-)
6:10 PM Jun 8th
 
MWeddell
You couldn't let it slide, David K? ;-)
4:42 PM Jun 8th
 
MarisFan61
Need I say, I wish you would! :-)

Hey, doesn't anyone care about when the desert really was.....

3:02 PM Jun 8th
 
raincheck
I was going to comment, but I didn’t want to interrupt MarisFan61’s discussion with MarisFan61. :)
1:22 PM Jun 8th
 
MarisFan61
OOPS!
That's what I get for not looking it up first.

Now I really don't know what Kaiser was thinking, because:

The birth year given by Bill was RIGHT.
12:39 PM Jun 8th
 
MarisFan61
(dunno if the below post was tongue-in-cheek, but to give it the benefit of the doubt, let's say it was) :-)

Just for anyone's FYI: It was 1 year off, which counts as right.
12:37 PM Jun 8th
 
KaiserD2
The difference between Bill James (born 1950 I believe) and David Kaiser (b. 1947) is that I know better than to rely on my memory for birth years of players. Corrections below:

Among Hall of Famers, those who hit better inside than out were, and I am going to give you their birth years from memory because I don’t want to look all of these up. . .those who hit better inside than out are led by Ron Santo (1940), who hit .292 in the years 1963-1968, but only .265 in the rest of his career. Others include Carl Yastrzemki (1939, .304 and .278), [NO, 1940] Roberto Clemente (1934 [NO, 1935], .326 and .312), Brooks Robinson (1937, .276 and .263), Harmon Killebrew (1936, .264 and .252), Willie McCovey (1938, .276 and .266), Billy Williams (1938, .292 and .288), and Hank Aaron (1934, .306 and .305). Al Kaline, also born 1934 [NO, 1935], hit only two points better outside than in (.298 to .296). Joe Morgan, born 1942 or 1943 {NO, 1944}, is the only Hall of Famer born outside those years who hit better in the desert than outside of it (.273 to .271), but of course Morgan’s best seasons were outside the desert.

Rod Carew was born in 1946.

David K
10:34 AM Jun 8th
 
MarisFan61
Following up further -- although I'm half-feeling like I shouldn't, because maybe there's no further interest, but, that's OK, I'm happy enough putting it on here for myself and for posterity, provided that the site upgrade doesn't make all these things disappear. :-)

Folks, this is going to be interesting, and I think it's significant.

So -- If offensive numbers didn't rise that much in those first few years after the "desert" period, how was it that run-scoring increased so much from "desert levels" after 1968?

Well, it did and it didn't.

The pattern wasn't what I think most of us thought. For sure it wasn't what I thought.

Runs did jump up from '68 to '69, and jumped some more the next year -- but then the next 2 years reverted to the "early-desert-period" level -- totally.

Here's average Runs Per Game, year by year:

58 4.28
59 4.38
60 4.31
61 4.53
62 4.46

63 3.95
64 4.04
65 3.99
66 3.99
67 3.77
68 3.42

69 4.07
70 4.34
71 3.89
72 3.69

=================

Following up further on there being more of a difference between the PRE-desert period and the desert, than between the POST-desert period and the desert:

Bill didn't give a breakdown for the pattern for players whose non-desert career was all or mostly PRE-desert, and those whose non-desert career was all or mostly POST-desert.

From what I found, I would guess that the former had a greater mean difference than the latter, although players in the latter category whose only 'desert' portion was in '67 and '68 would blunt that effect.

I would take it a little further: I'd say that it's not really correct to consider "1963 to 1968" as the desert that we're talking about.
I realize that it is commonly considered that way -- I CERTAINLY AL.WAYS DID -- but it seems more correct to look at the period as being the fuller stretch of 1963 to 1972, with '67 and '68 being a severe downward outlier, the single year of '70 being an upward outlier, and, very surprisingly, the single year of '72 being another downward outlier, fully on the level of 1967 and even a little toward that nadir year of 1968.

In any event, it looks like '69, '71, and '72 were quite like '63-'66.
12:15 AM Jun 8th
 
MarisFan61
Well, I figured out part of why the differential is less than I would have expected; no idea of course if it's a reason for anyone else.

A good basic frame of reference is, how much WERE offensive numbers lower during that period in general, anyway?

I realized I didn't exactly know.
I checked.
The results surprised me. Actually I feel like saying shocked.

Short answer: They weren't as much lower as I had thought.

Slightly longer answer: Except for 1968, the nadir of that "desert," the general/overall batting average for the next few years after the desert wasn't particularly higher than it was in the desert.

Footnote: (I'd put this in small font if I could)
I wondered if the increase in offense after the desert period was mainly in things other than batting average, so I looked at on-base and slugging also.
Answer: No, not particularly there either. On-base became slightly higher, but slugging average didn't. In fact, except for the 1 year of 1970, where it jumped to a higher level (.385), slugging average in the years from 1969-1972 was actually LOWER (a little) than it had been in most of the desert years.
The overall SLG for the desert period was about .368; not counting 1968, it was about .373.
For 1969-1972 it was about .369.

Back to batting average:

From 1958-1962, the general overall batting average was about .257.
In the desert period (including 1968), it was about .245.
From 1969-1972, it was about .250.

BTW, the next year, 1973, there WAS some jump, to .257/.325/.379.
That's why I left it out of the above data.
You could call it cheating :-) but I didn't intend it to fool anything.
I thought that only including those 4 post-desert years was valid to make the point, and that those 4 years cover the bulk of most of the post-desert at-bats of the players in Bill's study.
To equalize it maybe I should have gone back and eliminated the 1st year of the pre-desert period, to make both of those be just 4 years, but hey, I had already done that period, and anyway eliminating that first year wouldn't have changed the numbers.

I have the detailed data if anyone wants more specifics.

=======================

It still leaves questions like what I noted below, because even though the drop in batting average was less than I thought, it's still more than what Bill's results show -- but not much.
Taking the "pre"-desert and "post"-desert periods together as the basis for comparison, the 'deficiency' in batting average during the desert was 8½ points. Bill's study showed a 5 point difference, so it's just 3½ points that we'd need to explain, if we felt a need to explain any of it.

I'm offering that what I said below is a fair stab at that.
But the main thing is, except for 1968 (and we could add 1967), batting average in general wasn't that much lower in the desert period.
4:45 PM Jun 7th
 
jfenimore
1934 was a fine year for ballplayers to be born.
4:02 PM Jun 7th
 
MarisFan61
(Clarification: When I said "Players who dropped off more than might have been expected (for age and any other known considerations)...." the parenthetical part was in relation to what the usual expectation might have been -- i.e. some drop-off might have been expected due to such factors....
But that these were players whose stats dropped off more than might have been expected from those usual considerations.)

The reason I'm clarifying is, I realize that the parenthetical part could be seen as referring to something else.
12:17 AM Jun 7th
 
MarisFan61
....In fact I wonder if the things I mentioned, or other things like it (meaning just generally other kinds of biasing factors due to the desert itself), might account for the difference being so small.

I should mention, it's not that I think those things necessarily do bias it in that direction; they were off-the-top-of-the-head things that I see to be aspects relating to that period and which might have affected things in any which direction.

But here's a thing, somewhat related to those, which I do think worked toward lowering the difference.

One of the things I mentioned down there was about whether teams took meaningful account of the changed environment in how they judged the players who were continuing from before. I think we'd all agree that hardly anyone, teams or us or anyone, took enough account of it, and most people probably took none at all. But anyway, with regard to what would seem to be relevant to this study, can't we surmise these couple of things, if not assume them:

-- Players who dropped off more than might have been expected (for age and any other known considerations) would have tended to get less playing time, maybe to be released, and therefore would tend to be less likely to have made it into this study because they might not have reached 800 AB's in the desert period.
Besides the stars and superstars and near stars, I think the players who dropped off the most and even who dropped off just an 'average' amount might well not have been kept around, because if they weren't starting from at least near-star-level, it might not have taken that much extra of a statistical drop from the usual for teams to downgrade them, and I can easily see that a disproportionate number of such players wouldn't have reached the 800 AB's from 1963 on.
I mean "disproportionate" in relation to how many such players would normally have been kept around more.

AND...... (this part is harder, but I'm not sure we need it as much)

-- I would think that the new players who came up in that period and who did well enough to stick around to get 800 AB's during the desert period and thus to make it into this study may have tended to be (here's the hard part) :-) ....may have tended to be, for whatever varying reasons, players who weren't hurt that much by whatever the anti-hitting factors were, which helped them rise to the top from that crop -- and then, therefore, weren't benefited that much when those factors were countered and the desert period ended.

Can't wait to see how psychotic Bill and y'all think this thing is....

I think it's pretty good. :-)
12:12 AM Jun 7th
 
MarisFan61
That's impressive that you (or anyone) can even take a stab at writing down so many players' birth years without looking them up!
BTW, I didn't check on them, just assume they're about right, and anyway the fact of your knowing them well enough just to do that is more interesting than whether they're right.


I likewise would have guessed that more than just 59% had lower batting averages in the desert, but I wouldn't have easily surmised anything before bogging down on some questions like, how might the fact of "the desert" (and its lower offensive numbers) have affected the ratio of desert players whose non-desert years were mainly before vs. mainly after the desert, which would get into a couple or more sub-questions like how good of a job did the teams do in taking the lower hitting environment into account before getting rid of players whose numbers weren't as good as they used to be, and as for the newer players, was there some selection effect from the desert factor, like maybe that certain types of hitters tended to look less disappointing in that environment and therefore get selected for, and how would that affect this.....

If you feel like saying again that you have no idea what I'm talking about, I wouldn't blame you. :-)

But yeah, if I'd had to give a figure, I'd have guessed than it would have been closer to 80%.
8:20 PM Jun 6th
 
mpiafsky
Why you bagging on the Immortal Azcue?
8:08 PM Jun 6th
 
bjames
Stated one thing backward. 36 players LOST 20 points during the desert; 16 players GAINED 20 points. Not the other way around.
7:46 PM Jun 6th
 
 
©2024 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy