In 2010 there were 2,430 regular-season games (162, times 30, divided by 2).   Of those 2,430 games, there were:

 

            1,629 games in which both starting pitchers got the decision,

               107 games in which one starting pitcher got a win, the other a no-decision,

               124 games in which one starting pitcher got a loss, the other a no-decision,

               570 games in which neither starting pitcher got the decision.

 

            Per 1000 games, that is:

 

670 in which both starting pitchers got the decision,

               44 in which one starting pitcher got a win, the other a no-decision,

               51 in which one starting pitcher got a loss, the other a no-decision,

             235 in which neither starting pitcher got the decision.  

 

           

            There were 4,860 games started by pitchers.  We could sort these starts into eight classes:

 

            Starting pitcher won, opposing starting pitcher lost (1629)

            Starting pitcher lost, opposing starting pitcher won (1629)

            Starting pitcher won, opposing starting pitcher had no decision (107)

            Starting pitcher lost, opposing starting pitcher had no decision (124)

            No decision, opposing starting pitcher had a win (107)

            No decision, opposing starting pitcher had a loss (124)

            No decision for either starting pitcher, team won (570)

            No decision for either starting pitcher, team lost (570)

 

            These were the starting pitcher ERAs for each of those eight classes:

 

            Starting pitcher won, opposing starting pitcher lost                           &nb​sp;    ERA:  1.95

            Starting pitcher lost, opposing starting pitcher won                                ERA:  7.01

            Starting pitcher won, opposing starting pitcher had no decision               ERA:  2.78

            Starting pitcher lost, opposing starting pitcher had no decision                ERA:  7.41

            No decision, opposing starting pitcher had a win                                    ERA:  2.84

            No decision, opposing starting pitcher had a loss                                    ERA:  4.65

            No decision for either starting pitcher, team won                                    ERA:  4.08

            No decision for either starting pitcher, team lost                                    ERA:  3.85

 

           

            There are two real questions here, both of which we can figure out just by thinking about it for a minute:

 

            1)  Why is the ERA of starting pitchers better (3.85) in a no decision/team loss than in a no decision/team win (4.08)?

 

            If there is a no decision/team loss, what that means is that you (the starting pitcher) weren’t the guy who gave up the runs.    If there is a no decision/team win, that means that the game was won late, in the late innings, which would mean that any runs given up were probably given up early in the game.

 

            2)  Why is the ERA of no-decision pitchers in games in which the other pitcher had a WIN so much better (2.84) than the ERA of no-decision pitchers in games in which the other pitcher had a LOSS (4.65)?  Shouldn’t it be the other way around?        

 

            You might think.   But if your opponent has a win and you DIDN’T get the loss, that means that the decisive runs were scored against the bullpen after you left the game—therefore, that you pitched well, more often than not.   If your opponent has a loss and you didn’t get the win, that means that there is a high probability that the game was a slugfest in which both pitchers were out of the game early.

 

            There were, as we mentioned, 1,629 games in which both pitchers had a decision.   In those games, the won-lost record of the pitcher who had the higher Game Score was 1,537-79, a .951 winning percentage.   There were 13 games in which both starting pitchers had the same Game Score.

            In the 79 games in which the pitcher who lost had a higher Game Score than the pitcher who won, the most extreme case was the Nationals against the Cubs, April 28, 2010, Ryan Dempster vs.  Luis Atilano.  These are the pitching lines for the two pitchers:

 

           

            Dempster, giving up only 4 hits, one walk in eight innings, posted a Game Score of 67; Atilano, giving up 6 hits and 3 walks in 6 innings, had a Game Score of 50—nonetheless, Atilano was the winning pitcher, Dempster the losing pitcher.

 

            What about the better pitcher?  In games in which both starting pitchers got the decision, the better pitcher, based on the Season Score, had a won-lost record of 1063-566, a .653 percentage.

 

 

*   *   *   *   *   *   *

 

Publishable Criteria

 

            Could I explain just a little bit what I look for in a letter to "Hey, Bill"?  I’ll try not to lecture.   Three points:

            First, I want to answer as many questions as I can, so I will answer your question if I can.    Well. . .I will respond to your question; I may not answer it, but I’ll say something if I can.  

            Second, I like short questions and short answers.    One sentence is best; two sentences is better than three.

            Obviously, most questions run longer than that; I just wish they wouldn’t.  Most of the questions go on too long, and I cut things out of them before I print them, but. . .they still go on.  The longer you go on, the less chance there is that I will post the question.

            And third, I like questions that have answers.    Not arguments, not theories, not speculation, not questions that call for opinions; answers.   The best questions are questions that ask me something which I do not know, but which I can find out with some research.

            I don’t get very many of them, actually.   People rarely ask me questions that actually have answers.    90% of what I get is requests for opinions.

 

 

*  *  *  *  *  *  *

 

Verizon

 

 

            So I’m buying an I-Pad; don’t ask me why.   It has to do with this Bill James Baseball IQ App; I figure I look like a dork, representing a product that I can’t actually use because I don’t have the equipment.

            I go to the Verizon store, and I take my son with me so that one of us will have some idea what we’re shopping for, and I ask the sales lady, among other questions, "If I buy this today, can I carry it out today?   Do you have it in stock?"

            "Oh, sure," she says.   Her name Is Amanda Ostmeyer, or at least that’s what it says on the business card.  "It will be all loaded and active when you leave the store."   We talk about some more stuff, and eventually I order the I-Pad 2, or IPad2, or I-Pad II, or whatever it is; I’ll figure it out when I get it, which I now suspect will be on the Greek Kalends.  Finally I order it, and Amanda goes in the back room, and comes out very apologetically.

"We don’t have that particular model in stock," says Amanda Ostmeyer.  "We’ll have to send it to you.   It will take a couple days."

"So I should have it Tuesday?" I asked.

"Tuesday, maybe Wednesday," says Amanda Ostmeyer.   "I’ll write my phone number on this card," and she writes this number on the card, 785-764-4197; don’t worry about it, it’s out of service anyway.    "If you have any problems, give me a call, and we can track delivery."

Wednesday, the IPad isn’t there.    Thursday, it isn’t there; I call Amanda’s number, which is out of service.   I’ve been fake-numbered by a business professional. There’s another number on the receipt, which doesn’t work either, for some other reason; I figure it’s a phone company, they can have as many phone numbers as they want.  There’s a third number, with Extension 3.   I call this number, and I sit through a series of about 23 Verizon commercials while trying to work my way through the system.   Finally I get the opportunity to push 3 for extension 3, and this guy comes on the phone.

"Hey," I say as nicely as I can for a phone freak who has been fighting the phone system for about 12 minutes.   "I ordered this IPad from you; it’s supposed to be here yesterday.  Where is it?"

"What’s your number?" he says.

"What number?", thinking maybe he wants to know how many Win Shares I have.

"Your Verizon phone number."    I give it to him.   He types it into his computer.

"Oh," he says.   "It’s an I-Pad 2.  It’ll be four to six weeks."

"What?   You told me I could carry it out of the store."

"You ordered it on April 10," he says in a voice suggesting I might be a Chihuahua demanding a German shepherd’s ration of Kibbles and Bits.  "IPads ordered on April 10 will be delivered in 4 to 6 weeks."

Oh, well.   I’ve been a Verizon customer for ten years; I knew they were a ring of thieves when I walked in the store.   They’ll charge me for service from the day I ordered the thing, although I won’t have it for two months; I know these people.   I am only sorry for Amanda Ostmeyer’s soul.

 

 

*  *  *  *  *  *  *

 

Publicity

 

 

            I have some publicity out. . .I would tweet about this, but Twitter has changed their interface and I no longer know how to use it.

            1)  I did a podcast with Joe Posnanski—Poscast—which I think is available to anyone, perhaps through SI.com.

            2)  There’s an article about me or something current on Wired.com. . .haven’t seen it but I guess it’s OK.

            3)  I will be on Colbert on May 5.

 

*  *  *  *  *  *

 

Colavito and Kuenn

 

On April 17, 1960—which was the day before the American League season opened, April 18—the Cleveland Indians traded Rocky Colavito for Harvey Kuenn.   The trade excited the public’s interest as few trades ever have, for three reasons: 

1)  Colavito was immensely popular in Cleveland,

2)  Colavito was the reigning American League home run champion, with 42 (actually tied with Killebrew), while Kuenn was the reigning American League batting champion at .353,

3)  The timing.

Just before opening day there is an excitement about the baseball season starting, a spark in the air.  The Kuenn/Colavito trade exploded from that spark.  Neither Kuenn nor Colavito was perceived as a failed or failing player who might be traded; nobody was expecting a trade of two big stars on the day before the season started, although there had been talk of the trade six weeks earlier, when both players were holding out.

The trade became the pivot point of analytical debate in that era.   Who do you want:  the batting champion, or the home run champion?   How you felt about that trade was the key to your offensive philosophy.   Do you believe in the long ball, or do you believe in high averages?   The Detroit Free Press headlined the trade:  42 home runs for 135 singles.  

"Joe (Gordon) and I believe that the home run is overrated," said Frank Lane, who had made the trade for Cleveland.  "Look at Washington.   They almost led the league in home runs, yet finished last .  (Rocky) may hit 50-55 homers in Detroit, but we’ve given up 40 homers for 40 doubles.   We’ve added 50 singles and taken away 50 strikeouts."

Well, technically you didn’t, Frank.  Giving your argument its maximum benefit, you traded 33 homers for 18 doubles.    Colavito hit 42 homers in 1959; Kuenn hit 9.   That’s 33 homers.   Kuenn hit 42 doubles; Colavito hit 24.   That’s 18 doubles.    The Indians did gain 50 singles and lose 50 strikeouts, more or less.

This chart compares their batting stats year by year:

 

HARVEY KUENN								ROCKY COLAVITO
Year	Team	HR	RBI	AVG	SLG	OBA	OPS	Team	HR	RBI	AVG	SLG	OBA	OPS
1952	Tigers	0	8	.325	.400	.349	.749
1953	Tigers	2	48	.308	.386	.356	.742
1954	Tigers	5	48	.306	.390	.335	.725
1955	Tigers	8	62	.306	.423	.347	.769	Indians	0	0	.444	.667	.444	1.111
1956	Tigers	12	88	.332	.470	.387	.857	Indians	21	65	.276	.531	.372	.903
1957	Tigers	9	44	.277	.388	.327	.715	Indians	25	84	.252	.471	.348	.819
1958	Tigers	8	54	.319	.442	.373	.815	Indians	41	113	.303	.620	.405	1.024
1959	Tigers	9	71	.353	.501	.402	.903	Indians	42	111	.257	.512	.337	.849
1960	Indians	9	54	.308	.416	.379	.795	Tigers	35	87	.249	.474	.317	.791
1961	Giants	5	46	.265	.361	.329	.690	Tigers	45	140	.290	.580	.402	.982
1962	Giants	10	68	.304	.433	.365	.799	Tigers	37	112	.273	.514	.371	.885
1963	Giants	6	31	.290	.374	.358	.732	Tigers	22	91	.271	.437	.358	.795
1964	Giants	4	22	.262	.353	.329	.682	A's	34	102	.274	.507	.366	.873
1965	Giants	0	6	.237	.237	.352	.589	Indians	26	108	.287	.468	.383	.851
	Cubs	0	6	.217	.258	.336	.594
	TOTALS	0	12	.223	.251	.341	.593
1966	Cubs	0	0	.333	.333	.333	.667	Indians	30	72	.238	.432	.336	.767
	Phillies	0	15	.296	.352	.333	.686
	TOTALS	0	15	.296	.352	.333	.685
1967								Indians	5	21	.241	.366	.329	.695
1967								White Sox	3	29	.221	.300	.306	.606
1967								TOTALS	8	50	.231	.333	.317	.651
1968								Yankees	5	13	.220	.451	.330	.781
1968								Dodgers	3	11	.204	.310	.295	.604
1968								TOTALS	8	24	.211	.373	.311	.683
		87	671	.303	.357	.408	.765		374	1159	.266	.359	.489	.848

 

And this is a comparison of their Win Shares and Loss Shares:

Harvey Kuenn								Rocky Colavito
	Batting		Field		Total				Batting		Field		Total
Year	W	L	W	L	W	L	W Pct	Year	W	L	W	L	W	L	W Pct
1952	2	2	0	1	2	2	.497
1953	14	12	0	8	14	20	.424
1954	13	14	5	3	18	17	.509
1955	16	9	3	5	19	14	.578	1955	0	0	0	0	1	0	1.000
1956	18	5	4	3	22	8	.741	1956	11	3	3	1	14	4	.781
1957	11	15	4	4	15	19	.448	1957	12	9	2	4	14	12	.536
1958	15	8	4	3	19	11	.629	1958	22	-2	3	3	26	1	.960
1959	17	4	3	3	21	7	.751	1959	17	8	5	2	22	10	.683
1960	13	6	3	3	16	8	.665	1960	12	12	4	3	16	15	.511
1961	9	12	2	3	11	16	.409	1961	24	1	3	4	27	5	.835
1962	13	8	3	3	16	11	.587	1962	16	9	3	4	20	13	.597
1963	12	5	1	4	13	9	.578	1963	16	10	3	4	19	14	.572
1964	7	7	2	3	9	10	.479	1964	18	7	2	4	19	11	.633
1965	1	2	0	0	1	2	.394	1965	21	3	4	3	26	6	.806
	2	4	0	1	2	5	.298
1966	0	0	0	0	0	0	1.634	1966	13	11	4	3	18	13	.574
	3	4	0	1	3	5	.402
								1967	4	4	1	1	5	6	.475
									4	4	2	1	6	5	.525
								1968	3	1	0	1	4	2	.705
									2	4	1	1	3	4	.377
	167	116	35	48	203	164	.553		197	83	42	40	239	123	.660

 

            Kuenn played shortstop in 1953 but was 57 Assists below expectation, leading to the 0-8 defensive record for that season.

            In truth, Kuenn in 1959—the season before the trade--was of essentially the same value as Colavito.   There was not a very significant difference between the overall value of the two players in the 1959 season, but there are two critical factors that define the trade:

            1)  Kuenn was 29 years old on the day of the trade; Colavito, 26.   Colavito was 32 months younger—an immense difference in projecting value from that point forward.

            2)  Kuenn was not really a .353 hitter.   Kuenn was a good hitter, but his career average at that point was .314, not .353.    Colavito hit .257 in that season, but his career average at that point was .272.

            Kuenn was coming off of an "up" year; Colavito, off a down year.    Including these two factors, it’s obvious who is the better value here.   Colavito was.   It was a dumb trade for the Indians.

 

 

Earnshaw Cookin’

 

I got this letter in "Hey, Bill"; I’ll share it with you as I received it. 

 

 

Earnshaw Cook was the one who first proposed the pinch-hit-every-time-for-the-pitcher idea.  I just looked it up in his book "Percentage Baseball".  In your 1981 Abstract you said that: "Cook knew everything about statistics and nothing at all about baseball--and for that reason, all of his answers are wrong, all of his methods useless."  Why were you so hard on his ideas?  I've read both his books and he had many good ideas.  His scoring index (DX) is almost exactly like your Runs Created formula except that the denominator is PA squared instead of just PA. (So that the number of runs is expressed as a rate rather than as a total).  And DX predates RC.  Are you telling me that you read his books yet never noticed the similarity between DX and RC? 

 

--Mark R.

 

As you will note from the previous explanation, this question has none of the characteristics of a publishable enquiry.   It’s long, argumentative, and doesn’t do anything to advance anyone’s understanding of baseball.     That’s not to say it’s a bad question or that I shouldn’t answer it; it just doesn’t fit what I’m looking for.

I discovered Earnshaw Cook in, I am guessing, early June, 1967.    I graduated from High School in 1967, and for two or three weeks I took the bus to Topeka every day, looking for work.   I was really bad at looking for work, didn’t have a car or any job skills—or, now that you mention it, any people skills—so after a few days my job pursuit consisted mostly of hanging around the Washburn University library, looking pathetically at the co-eds.   I would buy the Topeka Capitol and circle the help wanted ads that seemed promising, but the longer this went on, the less I saw that seemed promising.

After a week or so I discovered that the Washburn University library had a copy of Earnshaw Cook’s 1964 book, Percentage Baseball, which I had never heard of up to then.   At first, of course, I was tremendously excited, to discover a whole book about the exact subject that I was most interested in (other than perhaps the co-eds), and for several days after that my employment searches were conducted mostly from within the stacks of the Washburn University library.   I attempted to borrow the book from the library, but, as I was neither a student there nor a resident of the city, I was denied permission.

The book was a tremendous disappointment to me.   I couldn’t honestly say that I understood all of the math or that I read all of the book, and it’s hard to say at this distance—44 years—what it was about Cook that irritated me.   One thing I remember is that he had very strong feelings about the Colavito/Kuenn trade.    My memory is that Cook, focusing narrowly on the players’ 1959 seasons, and also ignoring defense, thought that Kuenn was a much better player than Colavito.   That may not be correct, I don’t know.

I actually don’t remember anything that Cook wrote very specifically.   I remember he made some fairly outlandish argument about gaining a lot of runs by re-ordering the lineup.    He disliked platooning, based apparently on the fact that he didn’t have any stats on the subject, and I don’t think he had any understanding at all of park effects or, now that you mention it, baseball.

No, I’m sure he made have had some good ideas, but he utterly failed to sell any of them. My impression of Cook’s book was that he would say something outrageous, follow it up with 12 pages of impenetrable math, and announce that anyone who failed to see his point was a complete moron.    I wasn’t impressed.

I pretty much forgot about Earnshaw Cook, until people started asking me about him years later.    I was drafted by the US Army in 1971.   I gather that Earnshaw wrote another book in 1972, but I was in Korea; if Earnshaw Cook had been the first man elected simultaneously as Miss America and the governor of Utah and Indiana, I would probably have missed it.    I didn’t know Earnshaw Cook had written another book until. . ..well, until Mark R. asked me about it, as much as I remember.   That can’t be right; people asked me about Earnshaw Cook once or twice a year, and somebody must have filled me in somewhere, but it didn’t stick.

Mark R. suggests that I was too hard on poor Earnie, and no doubt I was.   I was too hard on a lot of people in those days.    I still am, I suppose.

It’s difficult to explain where sabermetrics stood.  The founding of the Society for American Baseball Research was a godsend for us, but here’s a story that may paint a picture.   The entire universe of people who were seriously interested in the analysis of baseball in 1976, as far as I knew, was four people:  Pete Palmer, Dick Cramer, Dallas Adams and myself.  One year Pete Palmer wrote an article on some sabermetric topic for the SABR Research Journal.   The next year Dick Cramer submitted another article on some unrelated but also sabermetric topic.    Bob Davids explained to him very nicely that he appreciated the submission, but that SABR had published an article of this type the previous year, and he didn’t know that there was enough interest in the topic to justify an article about it every year.   The de facto quota for sabermetric publications, world wide, was one article every two years!

SABR in those days, mid 1970s, was a group of about 100, 150 people, but even within SABR we were a tiny and isolated minority.   Earnshaw Cook was nowhere around, and I didn’t know anything about him or care anything about him; as far as I knew he was just some wing nut who wrote a bad book in the early 1960s that had become an obstacle to the success of the field.

Oh yes, it was.   There was this very, very strong belief, you see, that people were not interested in this kind of material.   There was a suffocating presumption that attached to us:  that we were boring, out-of-touch geeks who never went to baseball games and didn’t know anything at all about baseball other than what the computers would tell us.  If the computer told us that Roy Oyler was a better player than Willie Mays, we were supposed to be people who would believe it.  Earnshaw Cook’s total failure as a writer and as a thinker had reinforced that presumption.  To the publishing industry, Earnshaw Cook was the proof that books like mine could never sell.

I knew that these people were wrong.   I knew it from personal experience.   I had worked it out meticulously, and I was never more certain of anything than I was that a market could be created of people who were interested in the serious, systematic analysis of baseball.   I knew 10 or 15 people in Lawrence, Kansas, who were at least a little bit interested in the subject.   Lawrence was a city of 50,000, in a country of 250,000,000.  If there was one person interested in the subject for every 50,000 that was 20 per million; that was 5,000 people, not counting Canadians.  I met people all the time who were somewhat interested in the subject, people who liked to debate who should have won the Cy Young Award in 1974, and who would listen attentively to your evidence if you explained it in terms they could understand.    Therefore, either Lawrence, Kansas was atypical by some phenomenal factor, or else there was an economic number of people in the country who were potentially interested in this subject.

That was one proof.   Another one was, how many serious, hard-core baseball fans are there?    10% of the male population, maybe?   Probably more and some women, but. . .let’s say.    How many people are there to whom mathematical analysis is natural?   5% of the population?   Let’s say it’s 5%; let’s assume that there is only a random intersection of these populations.   Still, 250 million people, 125 million males, 5% of those are hard-core baseball fans, that’s 6.25 million, 5% of those are people to whom mathematical analysis is natural, that’s. ….300,000 people.   Assume I’m wrong; assume it’s not 5%; it’s 2%.   Doesn’t matter.   The number of people who are interested in this subject has to be larger than the number of people that I need to buy my book.

What we needed to do, to go from four people to 40,000, was to demonstrate that what we had to say connected directly to the public discussion.   Yes, I was aware that people had been writing numbers-oriented baseball stuff for a hundred years, but it didn’t connect to the discussions that other baseball fans were having.   Earnshaw Cook failed because there was no way to connect what he was saying to the ongoing baseball discussion.   What did he have to say about who should win an MVP Award, that anybody could understand?   Nothing.   What did he have to say about who should be in the Hall of Fame, that anybody could understand?   Nothing.   What did he have to say about who would win the pennant this year, that anybody could understand?   Nothing.   What did he have to say about the up and coming stars, that anybody could understand?   Not a word.

There would be a Cy Young debate, and I would point out that one pitcher had been supported by 6.2 runs per start, while the other one had been supported by 4.2.   It was relevant.  It connected.   There would be a Hall of Fame debate, and Pete Palmer would produce research showing that one player had out-hit the other by 70 points in his home park, while the other had hit 50 points better on the road.   It was relevant.   It changed the debate when people knew that.

We asked small questions, and produced information that was relevant.  How many stolen bases did Johnny Bench allow?   How many runners did he throw out?   What is the oldest team in the league?   Which are the older teams?    Who are the pitchers with the best chance to win 300 games?   How do they stand, compared to the same list 10 years ago, or 20 years ago, or 30 years ago?

Look, the problem with Earnshaw Cook, and with the other pre-1970s baseball statistics guys, excepting Allan Roth, is that none of them ever did one damn thing toward the creation of a sabermetric community.   They didn’t do that because they didn’t have any concept of this being a field of knowledge that was bigger than themselves.   They weren’t interested in having a million little debates with others who were into the same topics.   They weren’t interested in creating a community.    They weren’t interested in asking a million little, tiny questions, digging out the answers to them, and slowly accruing a body of knowledge shared by a group of experts.   They weren’t interested in developing ways to measure every little part of the game that had never been measured.   They wanted to take the skeleton of readily available facts, leapfrog the discussion, answer the Great Questions, and be acclaimed as experts.   They didn’t have any concept of this being a science, a field of knowledge.   Other than Allan Roth, they didn’t have any ability to persuade the people inside the game that they had something worthwhile to contribute to the discussion.   The Elias Bureau, when they got a little, tiny foothold inside the game, did everything that they could possibly do to prevent anybody else from clambering aboard.

To me, to say that Earnshaw Cook or any of the other early Earnshaw Cooks was writing about sabermetrics is like saying that Vlad the Impaler and the Marquis de Sade were the fathers of psychology.   Theirs was a self-contained discussion.   It started with them, it ended with them.    Everybody else was a spectator.   I am being too kind; I forgot to use the word "delusional".   I didn’t actually forget; I just couldn’t figure out where to put it.

You tell me now that "His scoring index (DX) is almost exactly like your Runs Created formula except that the denominator is PA squared instead of just PA. (So that the number of runs is expressed as a rate rather than as a total)."   Well, first of all, Earnshaw’s Cook’s DX in the book that I read—I get this from Albert and Bennett—was

                        1B + BB + HBP + E                         TB + SB

DX    =         -----------------------------       X        -------------------

                                     BFP                             ​             BFP

 

A formula that bears very little resemblance to any Runs Created formula—and which, as best I understand it, Cook never actually presented;  I think Cook actually presented different pieces of this formula at different points in his book, and Albert and Bennett cleaned it up to help the reader make sense of it.

The claim that DX is almost exactly like Runs Created is based, I think, on Cook’s 1972 book—a book that had so much impact on the culture that, being one of four people in the world who was really interest in the subject, but being out of the country in the six months when the book was available, I literally never knew that the book existed.

            I said that all of Cook’s methods were useless, and you think that I am being too hard on him.   OK then, who uses any of Earnshaw Cook’s methods, today?   Can you name anyone in baseball who makes any use of anything Earnshaw Cook said or did?   I’m just asking; I don’t know.   I’m not aware of his having any influence on anyone.

Bob Davids deserves credit for his early contributions to sabermetrics; Allan Roth does, and Dallas Adams.   Earnshaw Cook deserves to be ignored.