Saturday, March 29, 2014

How to Fix the Baseball Hall: Section 1


How to Fix the Baseball Hall of Fame: Weight the Vote
Dean H. Krikorian, Ph.D.
GroupScope, Inc.
March 29, 2014


This is the first of an 8-Section series of blog posts on the National Baseball Hall of Fame. Each section is roughly 2-3 pages in length and contains graphics and links in the main body of the post.  Links open up new windows, so make sure to close them upon completion to avoid desktop clutter.  These links contain further details in the form of stories, tables, references, appendices, and relevant websites. Each section can also be accessed via the Table of Contents Page.

Section 1 – Introduction & Background

Purpose: This report is specifically directed to the The Board of Directors of the National Baseball Hall of Fame and Museum, Inc., who reserve the right to revoke, alter or amend their rules of entry at any time. I consider this an impassioned fan letter (from a statistical nerd) urging the decision-makers to update antiquated procedures in the least intrusive fashion. The system IS broke, but can be fixed easily, as you will see. Rather than sports commentate on the subject, I seek to perform a non-biased statistical analysis. I feel an adequate spokesperson - this problem was made for me because it fuses my favorite pastime with my two areas of expertise: statistics and group decision-making. My dissertation covered the gamut of group decision-making research that can be applied to cogently describe the Baseball Hall of Fame election process. Because my company specializes in visualizing data, I hope that readers can gain insight on a complex topic without an extensive background in statistics – the charts are pretty cool. But in the end, this is a comprehensive study that seeks to change policy and make the process more fair. The solution presented is to weight the vote by assigning each voter ten votes to distribute as they so choose. I think this report will aid your cause and I hope you enjoy and learn from the research presented. It was a lot of fun work for me.

Context: The BaseBall Writers Association of America (BBWAA) meets this July to discuss the “10-Vote Rule,” which specifies the procedure used to induct members into the National Baseball Hall of Fame (NBHF). This report provides an informed solution that makes the BBWAA election process more equitable – one that I have not seen considered in all I have read on the subject. This report explicitly illustrates how the current favored solution, “increasing the number of votes,” is a profoundly bad idea and how to fix the “10-Vote Rule” with a few simple instruction changes. The solution proposed is to weight the vote. This results in a fair and equitable solution with minimal changes to the current system. This report will show why and how in eight (roughly) two-page sections via posts on this blog. This first section of the report provides background information and data.

Background: What is the National Baseball Hall of Fame?

The National Baseball Hall of Fame and Museum is located in Cooperstown, a quaint village in Upstate New York. Established in 1935 and built in 1939, it is the 2nd oldest hall of fame in the US, modeled after the Hall of Fame for Great Americans at Bronx Community College in New York City (nee the campus of New York University). It is both a museum and a hall of fame, where everything baseball is celebrated. The inaugural class of Ty Cobb, Honus Wagner, Babe Ruth, Christy Mathewson, and Walter Johnson, reflect the Mount Rushmore of immortal players. Cy Young was purportedly the first snub - with his 514 wins (...and 311 losses). The Baseball Hall opened amidst great fanfare in 1939, coinciding with the reputed invention of baseball in 1839 by a young General Abner Doubleday in Cooperstown. There was a grand parade and an exhibition game between current stars of the day (two from each MLB team) - Babe Ruth pinch hit and weakly popped to the catcher.  The Bambino later led a grand tour of the Hall. Baseball had arrived. It didn't matter that Abner Doubleday didn't invent baseball and Cooperstown had, hence, nothing to do with its origins. The Hall is today a sprawling multi-wing facility with 306 inductees (67 living), located 200 miles from the nearest MLB stadium. Cooperstown is a nice place to visit, but it's in the middle of nowhere and seemingly stuck in the 1950s.

How Does One Get Inducted Into the Baseball Hall?

There are two ways to get into the Baseball Hall. The first is through the general election, held every year. This is the common popular vote that most of us are familiar with. Player Finalists (no managers) need to be on at least 75% of the submitted ballots to be inducted in any given year. Voters are asked to “vote for up to ten (10)” from a ballot of around 30 finalists narrowed down by a smaller selection committee. Eligible players must be retired for at least five (5) years and no more than twenty (20). Finalists must be on at least 5% of the ballot to be on the ensuing year's ballot.

From the beginning, the sole voters for the general election are members of the BaseBall Writers Association of America (BBWAA). Current BBWAA rules state that after 10-years, members are eligible to vote in the Baseball Hall elections (and other MLB Awards such as the Rookie of the Year, MVP, and Cy Young Award Winners). In the past this has been a very selective group.

The second way to get in is via special committee. There have been several of these committees in the past, but these are typically a group of 5-11 writers, historians, and/or living Hall-of-Famers that consider old-timers, umpires, executives, and the players that are no longer eligible for election 20 years after their retirement. In the past the Veterans Committee has the most inductions, but it has since given way in 2006 to three separate era committees that meet every third year: The Expansion Era (1973-Present); Golden Era (1946-1972); and Pre-Integration Era (1945 & Before). Thus, there are only two ways to get into the Baseball Hall of Fame: (1) BBWAA Election; or (2) Era Committee Election. Each has separate rules, here we focus for the most part on the general election process (See BBWAA Rules). For a more detailed background, I compiled a list of key dates and events (See Baseball Hall Key Events and Rule Changes Timeline).

The Baseball Hall: Past Induction Rate Charts By Source

Baseball Hall induction has changed over time. Examining the different induction sources over time provides insight into this history.

It is interesting how the Veterans Committees (Red) has actually inducted more than the BBWAA (162 vs. 114). The Negro League Committees (Yellow) were relatively marginal and heavily influenced by the 17 inductee blip in 2006. Note the linearity and relative paucity in BBWAA inductions (Blue).  Examining the first versus second half of the different Baseball Hall sources provides more insight.
The 2nd half of all Baseball Hall inductions indicates fewer inductees over time. This is mostly due to a near-40% decrease in Veterans Committee inductions since 1980. The Last 3rd of Baseball Hall inductions further illustrates a steady decline in inductions in both the BBWAA and Veterans Committees.

The final third reflected the lowest number of inductions: 1st 3rd = 106; Middle 3rd = 92; & Last 3rd = 78. This trend became more apparent over extended periods of time.
Here we see a steady decline in overall inductions: 1st Quarter (1936-58) = 80; 2nd Quarter (1959-79) = 76; 3rd Quarter (1980-97) = 68; & 4th Quarter (1998-2014) = 57.
This decline is relatively linear from 4 to 8 to 11. This trend became better understood over extended periods of time.
Here we see a more pronounced decline in recent inductions: 1st Fifth (1936-53) = 69; 2nd Fifth (1954-72) = 58; 3rd Fifth (1973-86) = 56; 4th Fifth (1987-2000) = 53; & 5th Fifth (2001-14) = 40.  The decline between fifths is from 11 to 2 to 3 to 13 (umbrella-shaped).  Note the steep drop in inductions after the first and before the last fourteen years.  Hence, the decline can be seen recently (2001-2014), but also in comparison to the original 14 election years (1936-1953).

BBWAA vs. Veterans Committees: Correlations & t-Tests

From the previous charts, it seems as if the BBWAA and Veterans Committees reflect different induction processes - one is a small committee and the other is a large group of electors.  But were their yearly inductions statistically different?  I tested this by first standardizing the number of BBWAA and Veterans inductees, converting them into z-Scores. These z-Scores were then used to determine the Pearson Product-Moment Correlations between BBWAA and Veterans.  Results indicate a significant negative correlation between BBWAA & Veterans Inductions (r = -0.274, n = 70, two-tailed).  Subsequent Student's t-test was significant: t = -2.382, n = 70, p < .02 (two-tailed).  This means that the BBWAA and Veterans Committees ran in opposite directions: When one was up, the other was down.  They are indeed different animals, so to speak.  This can be seen especially in the early years of the Baseball Hall, where, the Veterans and BBWAA complemented one another. Further tests revealed that 88.5% of the aforementioned correlation is attributed to the years prior to 1958 (more on this later). 

Summary: It is Harder to Get Into the Baseball Hall, Especially Nowadays

In summary we have seen how the number of Baseball Hall inductions has gone down over time, especially considering the last 14 years.  We also see how the BBWAA and Veterans Committees inducted complementary to one another in early years of the Baseball Hall, but not since 1958.  The next section examines how the Baseball Hall compares to other Halls of Fame.    

How to Fix the Baseball Hall - Table of Contents

How to Fix the Baseball Hall: Weight the Vote
Dean H. Krikorian, Ph.D.
 GroupScope, Inc.
March 29, 2014

TABLE OF CONTENTS

Abstract
Section 1: Introduction & Background
Purpose and Context
Background: What is the National Baseball Hall of Fame?
How Does One Get Inducted Into the NBHF?
The Baseball Hall: Past Induction Charts By Source
BBWAA vs. Veterans Committees: Correlations & t-Tests
Summary: It is Harder to Get Into the Baseball Hall, Especially Nowadays
Section 2: The Baseball Hall vs. Other Halls of Fame: History, Rules, & Statistics
Baseball Hall vs. Other Halls of Fame: Historically and By the Rules
Baseball Hall vs. Other Halls of Fame: Statistically, By Induction Year
Summary: It is Harder to Get Into the Baseball vs. Football Hall
Section 3: Potential Solutions & The Real Problem: Two Little Words...
Potential Solution1: Grey and Black Ink Tests
Potential Solution2: Hall of Fame Tiers
Potential Solution3: PinstripeAlley.com Advice
Potential Solution4: Why NOT Increase the Number of Votes per Ballot?
Past Problem Elections: Why Not Vote For As Many As You Want?
Effa Manley: Worst Person in the Hall
Protests: 2014 General Election
Fly in the Ointment: Variable Vote
Procedural Flaw: Two Little Words...The 1958 Rule Change That Changed It All
Summary:  The BBWAA Induction Process is Unfair Because of a 1958 Rule
Section 4: Baseball Hall Statistical Results
Votes per Ballot vs. Number of Ballots
Number of Ballots vs. Number of MLB Teams
Votes per Ballot and Ballots vs. Inductions
Induction Interaction Effect: Votes per Ballot Mediates Between Ballots & Induction
Summary: Stabilizing Votes per Ballot mediates effect of increasing voter pool
Section 5: The Solution: Weight the Vote (The Ken Gurnick Amendment)
How to Weight the BBWAA Vote
How to Implement the Ken Gurnick Amendment: Scarecrow Instructions
How to Score a Weighted Vote: Example 170-Ballot Sample from 2014 Ballots
Summary: How to Weight the Vote: Basic Recommendations
Section 6: Simulation: What If All Past BBWAA Elections Had 10 Votes/Ballot?
Simulation Assumptions: The Magic Weighting Equation
Preliminary Simulation Results: 2014 Sample
The First Three Tiers of the Baseball Hall (Current System)
The Next Three Tiers of the Baseball Hall (Weighted System)
Discussion on Borderline Candidates
Summary: Simulation Conclusions
Section 7: Summary 
Final Recommendations
Future Directions
Next Steps
Section 8: Appendices
List of References
List of Figures
List of Tables
Methodological Specifications
Statistical Tables

How to Fix the Baseball Hall of Fame - Abstract


How to Fix the Baseball Hall of Fame: Weight the Vote
Dean H. Krikorian, Ph.D.
GroupScope, Inc.
March 29, 2014
 
Abstract

I have discovered what is wrong with the Baseball Hall of Fame and how to easily fix it. The flaw is in a two-word rule change enacted in 1958 specifying to vote for “up to” ten candidates from “vote for ten” in the previous Baseball Hall elections. Statistical tests show that this rule has made induction significantly more difficult, especially in recent years. The fix is a weighted vote, which assigns players scores based on the number of players chosen on each elector's ballot.This will solve several problems: (1) voting for as many players as you want, realizing that more selections means less voting impact; (2) mediating the effects of an increasing voter pool; (3) reducing the recent backlog into the hall; while (4) remaining relatively difficult to gain induction.

As basic instructions, I offer one slight change to the current voting system and provide sample “scarecrow” instructions on how to easily enact this change. Then I examine a 170-Ballot sample from 2014 to demonstrate how to easily score a weighted vote using a spreadsheet. The resultant impact of this 30% sample is to induct candidates earlier that eventually got in anyway. This reduces the backlog and allows deserving hall members more enjoyable memories (and less Red Ruffing and Ron Santo moments).

The last part of the study is where it gets interesting. Here I construct a six-tier Simulated Baseball Hall using past results and Weighted Votes: (1) First Ballot; (2) BBWAA Elected; (3) Committee Elected; (4) Weighted In; (5) Weighted1 In; & (6) Weighted2 In. These results are discussed by revisiting some of the borderline candidates who unfairly lost votes in low Vote per Ballot years. In the end, Final Recommendations for the Baseball Hall Sub-Commitee are as follows: (1) revising the instructions provided for a weighted vote; (2) weighting all previous elections using actual ballot data; (3) examining the relative induction difference by weighting the actual vote; (4) presenting these results to the BBWAA in July, 2014; and (5) Adopting the Ken Gurnick Amendment or Weighting the Vote in all future BBWAA hall elections.

Citation: Krikorian, D. H. (March 29, 2014). How to fix the baseball hall: Weight the vote. Available: http://jimanddean.blogspot.com/2014/03/how-to-fix-the-baseball-hall.html.  This is part of a larger report, “The National Baseball Hall of Fame: A Statistical Analysis” or “If You Ever Wanted to Tinker, Here's Your Chance,” available from the author. 

Thursday, March 27, 2014

Effa Manley: The Worst Person in the Hall

Recently reading the the book "Veracruz Blues" by Mark Winegardner (an excellent read, btw), I came across the following quote, based upon a reputed conversation between Hall of Famer Ray Dandridge + Ex-Negro Leaguers' Burnis Wright & Theolic "Fireball" Smith:
"'I'm thinkin maybe I'll go back to Newark,' Ray said.  Burnis replied 'You won't make ten cents more playin for Effa Manley than you will down here.  Unless you Terris McDuffie.'  We had a good laugh.  Manley owned the Newark Eagles.  She was a sharp businesswoman, I'll give her that.  The crack about McDuffie, who used to be with the Eagles but who played for Torreon now--well, he used to do Effa Manley.  Don't be shocked, Frank: Owners doin it to players for years, only not so literal." (See Winegardner, 1996, p. 72).

This recalls an excerpt from a report I am preparing on the Baseball Hall (stay tuned here), subtitled "Effa Manley: The Worst Person in the Hall."  Note: Links will open up a new pop-up window, it is recommended that you close the window upon reading.

In 2006, the Special Committee on the Negro Leagues met for a one-time election.  This body of 12 eminent African-American baseball historians considered legends and pioneers of the Negro Leagues, an “Old-Timer's Committee for Negro Leaguers,” if you will.  It was expected that 4-5 members were going to be elected.  "That would be cool," I thought at the time, "I would like to know the results of such historical research when I visit the Hall."  One problem: They forgot to limit the vote. “Vote for as many as you want,” the Hall was essentially saying to the Negro League Old-Timers Committee.  The result?  17 new members.  And not to sound racist (or sexist), but several of these inductees were highly questionable.

One particular watered-down selection from 2006, Effa Manley, really bothers me.  Effa Manley was a (presumed white) team “co-owner” of the Newark Eagles.   Her husband bought the team and she served as business manager for 13 years and sole owner for less than a year after he passed (ala Georgia Frontiere).  She performed some innovative community relations in the Newark area, such as letting kids in free to games - she probably knew they wouldn't sell the seats and made money off the sale of sodas, peanuts, and Cracker Jacks. 

She was known for favoring the younger players on her team and had amorous affairs with several of them.   She would parade the players around in her ladies club.  Pitcher Terris McDuffie was a particular favorite and when her husband Abe found out he traded McDuffie to the New York Black Yankees for "two old bats and a pair of used sliding pads" (See Negro League Baseball Museum).  IMHO the bats and sliding pads deserve to be in the Baseball Hall more than Effa Manley.  The players viewed her as a distraction in the clubhouse.  She sold her team in 1949 from this sentiment – there's no whoring in baseball.

But I really have one major problem with Effa Manley.  When her players started to integrate to the Major Leagues, she opposed it.  She wanted money for her players.  And maybe she deserved that - Bill Veeck did compensate her for Larry Doby, but she got nothing for Don Newcombe (Dodgers) and Monte Irvin (Giants).  But I can't get over the fact that she obstructed in any way the integration of baseball, which set the tone for the entire country.  Effa Manley was critical of Branch Rickey (for “stealing” Newcombe) and Jackie Robinson (for forgetting his roots) – and to me that's sacrilege (btw: I loved the movie “42”).  No, I don't rightly see how Effa Manley did anything for Major League Baseball - in fact, I view her as an antagonist (and a harlot).

Jim Kaat has 283 wins and won 14 Gold Gloves [3rd all-time behind only Greg Maddux (18) and Brooks Robinson (16)].  He is not in the Hall, like Effa Manley.  Kaat has more Gold Gloves than Manley had seasons in baseball – and she only served as sole owner for less than a year.  George Steinbrenner (as much as I hate to say it as an Oriole fan), Marvin Miller, Buck O' Neil: All deserve to be in the hall more than Effa Manley – this I know for sure.  But, then again, this is what happens when you institute an unlimited vote.  Let's not learn that same lesson again, Keith Olbermann. 

This last reference is to Keith Olbermann's perceived ringleader role in issuing an unlimited vote in future BBWAA Hall elections (See Olbermann Show 1/8/14 - 6:02 Video).  Veteran BBWAA Writer Ken Rosenthal supports an increase to 15 or so votes (See Rosenthal on the Hall: It Ain't Broke, But I Can Fix It).  Current BBWAA President Le Velle E. Neal III has publicly shown support for "increasing the vote" (See Minneapolis Star-Tribune - 1/8/14 - 2:57 Video). Increasing the vote, however, is a profoundly bad idea in light of baseball history and group decision-making research.

Unlimited voting power has caused problems in the past for baseball on at least three other occasions: (1)  The 1958 All-Star Game, where Reds' fans stuffed the ballot box (and, hence, removed the fan vote for 13 years); (2) The 1944-5 Veterans Committee, who inducted 21 in two years (the so-called "dying wish" of Ex-Commissioner Kenesaw Mountain Landis); (3) The 1971-77 Veterans Committee, who "shepherded" in seven Ex-Giant/Cardinal teammates of Members Frankie Frisch & Bill Terry (as "cronyism," or group invulnerability as a symptom of Groupthink) (See Janis, 1972).  

Applying past group decision research, there are five issues associated with instituting an increased or unlimited vote: (1) watering-down the entry pool with many more finalists ( = double?); (2) rendering moot the minimum 5% carrover rule (and what do you tell previous players not qualifying in earlier years?); (3) produce a glut of mediocre finalists on every ballot (presumably not close to being elected); (4) further marginalizing selective electors (as they watch their colleagues vote for, say, 25 players); and, thus, (5) increasing Votes per Ballot variation among voters.  This last issue is the biggest problem, based upon my research.  Stay Tuned Here...I have discovered the real flaw in the voting system and how to easily fix it.

Friday, February 7, 2014

The Pitching Club


Pitchers have the 300 win club that parallels the 3,000 hit club for hitters. A sort of uncanny symmetry emerges. 24 pitchers have won 300 or more games, compared with the 28 members of the 3,000 hit club. Maybe the similarity is just a product of selecting parameters that yield a result we were already shooting for, but the 300/3,000 thing is just cool.

What's even cooler is when I add in a pitching qualifier to narrow the field even more, like when I selected members of the 3,000 hit club who had 300 home runs. Just to keep the numbers fun, how 'bout let's see how many pitchers have 300 wins and 3,000 strikeouts. See what I did there? I used a 3,000 thing so the hitters and pitchers...you get it.

Turns out, there are ten pitchers with 300 wins and 3,000 strikeouts. Do you remember how many hitters had 3,000 hits and 300 home runs? Ten. Am I blowing your mind right now? Far out.



                                                   thechive.com


Not on the list, somewhat shockingly, is Cy Young. The all-time leader in games started (815; Ryan is next with 773), and wins (511), who, even with 316 losses maintained a career winning pct. of .618, an ERA of 2.63, and a WHIP of 1.13, fell almost 200 strikeouts short of membership into this club. In almost 2,000 more innings pitched than Nolan Ryan, Cy Young struck out 2,900 fewer batters. Please re-read that last sentence. I just had to double check that, because that is bizzarre. The thing is, it's not as if Cy Young should have had, in those 2,000 more innings, 2,900 more punchouts, or 2,000, or even 1,000 more; all he needed was 200 more! One every ten innings if my math checks out. Maybe comparing Young to the all-time strikeout leader skews the picture, but Young's other percentages are so stellar that it just seems like he would have accidentally tripped over 200 more strikeouts.



Maybe Cy Young was a pitch-to-contact guy. I never saw him pitch, obviously, but I don't think "maybe" fits that statement. He had to be THE pitch-to-contact guy in baseball history. Out of the pitchers in the 300/3,000 club, I consider Greg Maddux to be the one I've seen who unabashedly pitched to contact. Maddux (.250/1.14) and Young (.252/1.13) have very similar batting avg. against and WHIP percentages. Maddux struck out 568 more batters than Young, and he did it in 2,348 fewer innings. It's just weird.


The members of this exclusive club with their career ERA and WHIP:

Walter Johnson   417/3508   2.17/1.06
Greg Maddux     355/3371   3.16/1.14
Roger Clemens   354/4672   3.12/1.17
Steve Carlton      329/4136   3.22/1.25
Nolan Ryan         324/5714   3.19/1.25                    
Don Sutton          324/3574   3.26/1.14
Phil Niekro          318/3342   3.35/1.27
Gaylord Perry      314/3534   3.11/1.18
Tom Seaver         311/3640   2.86/1.12
Randy Johnson     303/4875   3.29/1.17


Walter Johnson's resume is astonishing, and if the game was easier for pitchers back in his day, then why is he the only one from his day on the list? Seaver is the only other pitcher in the club with a sub-3 ERA, which is terrific. In fact, with all the innings these guys threw, The highest ERA and WHIP came from the knuckleballing Niekro, and those numbers are still very solid.