MLB Wins over Expected Wins pre and post adoption of Moneyball by Charles Slavik

https://public.tableau.com/views/MLB_WINS/Sheet1?:language=en-US&publish=yes&:toolbar=n&:display_count=n&:origin=viz_share_link

Who is Dick Lundy?

Dear Coaches:

Back in February 2021, our SABR Chapter tried to organize an essay contest to benefit Dick Lundy,

1. ·         a former Negro League Baseball star,
2. ·         citizen of Jacksonville, born and raised where FSCJ’s downtown location now       stands,
3. ·         an alum of Edward Waters University (back when it was functioning as a high school)
4. ·         and Bethune-Cookman when it was Cookman Institute.

Lundy will be a candidate to join Lloyd in the baseball Hall of Fame if they take another shot at adding ex-Negro League players in 2021. He is reported to have missed election the last time the Hall of Fame addressed Negro Leaguers in 2006. Lundy was not only a great ball player, but a great mentor, manager and a person who worked with Hall of Famer Effa Manley to try and salvage Negro league baseball as an entity after they were raided of talent following the Jackie Robinson signing.

We would like to elevate his profile locally and nationally through different initiatives to assist his candidacy. I have been in contact with next of kin and they feel like it is too late to affect this kind of change. Lundy is a deserving Hall of Famer on and off the field, statistically and subjectively he should be the next guy in.

I am afraid that locally, we have done too little to help, and I communicated that to a frustrated family. Please, please, please help me convince them that it is not too late. Nationally, the candidacy is on solid ground, but there are candidates whose local PR machine is greater than ours. We can change that, even now.

I wish I had essays from students to demonstrate to the family the potential of this project. They have been beaten down by the past. Now we can be involved in helping change baseball history. What could be better than that? Even a one-page essay from students would help, I dare say, especially from HBCU’s where he attended. This is a great diversity, equality and inclusion project for students. He and others were rendered invisible by Jim Crow, we cannot allow that crime to be repeated.

Please help, time is of the essence and is now more an enemy than a friend to our effort. Thank you for your consideration. I will help in any way I can.

Sincerely,

Charles Slavik

(813) 335-8678

Founder of John Henry “Pop” Lloyd SABR Chapter   

https://twitter.com/theslav1959

https://twitter.com/Floridasabr

 

Research Resources for Students:

http://sabr.org/bioproj/person/dick-lundy/

https://www.seamheads.com/NegroLgs/player.php?playerID=lundy01dic

http://www.banishedtothepen.com/who-is-dick-lundy/

http://www.cnlbr.org/Portals/0/Hero/Richard_Dick_Lundy.pdf

 

2020 MLB Mock Draft

This will be my "official" order for the 1st Annual SABR Northeast Florida Chapter Mock 2020 MLB Draft Invitational Challenge. The draft will be held June 10th, deadline for entry is midnight 6/9.

There seems to be a consensus around the top three or four picks, then after that it should get interesting. Where there appears to be duplicate entries in the rank column, signifying a virtual tie,  I broke the ties based on variance of opinions, but the person listed first on this list is the correct order for me. 

It was interesting to note that Baseball America, long regarded as one of the best resources in this type of exercise, changed their methodology somewhat. Their ranking seemed out of line with the other experts, more so than in past years. It will be interesting to see how "correct" the new methodology grades out to actual. 

The final grade of course is unknown for 5-10 years or more. 

** UPDATED 6/9/2020 10:20AM

PLAYER	SCHOOL	POSITION	RANK
Spencer Torkelson	Arizona State	1B	1
Austin Martin	Vanderbilt	OF/3B	2
Asa Lacy	Texas A&M	LHP	3
Emerson Hancock	Georgia	RHP	4
Zac Veen	Spruce Creek (FL)	OF	5
Nick Gonzales	New Mexico State	SS/2B	6
Max Meyer	Minnesota	RHP	7
Heston Kjerstad	Arkansas	OF	8
Reid Detmers	Louisville	LHP	9
Mick Abel	Jesuit (OR)	RHP	10
Garrett Mitchell	UCLA	OF	11
Austin Hendrick	West Allegheny (PA)	OF	11
Robert Hassell	Independence (TN)	OF	13
Patrick Bailey	North Carolina State	C	14
Ed Howard	Mount Carmel (IL)	SS	15
Cade Cavalli	Oklahoma	RHP	16
Garrett Crochet	Tennessee	LHP	17
Tyler Soderstrom	Turlock (CA)	C	18
Pete Crow-Armstrong	Harvard-Westlake (CA)	OF	19
Jared Kelley	Refugio (TX)	RHP	20
Cole Wilcox	Georgia	RHP	20
Nick Bitsko	Central Bucks East (PA)	RHP	22
Dillon Dingler	Ohio State	C	23
Bryce Jarvis	Duke	RHP	23
Justin Foscue	Mississippi State	2B	25
Bobby Miller	Louisville	RHP	26
Slade Cecconi	Miami	RHP	27
J.T. Ginn	Mississippi State	RHP	28
Aaron Sabato	North Carolina	1B	29
Nick Loftin	Baylor	SS	30
Carmen Mlodzinski	South Carolina	RHP	31
Jordan Westburg	Mississippi State	SS	32
Austin Wells	Arizona	C	33
Clayton Beeter	Texas Tech	RHP	34
Carson Montgomery	Windermere (FL)	RHP	34
Chris McMahon	Miami	RHP	36
Casey Martin	Arkansas	SS	37
Jordan Walker	Decatur (GA)	3B	37
Tanner Burns	Auburn	RHP	39
Daniel Cabrera	Louisiana State	OF	40
CJ Van Eyk	Florida State	RHP	41
Kevin Parada	Loyola (CA)	C	42
Carson Tucker	Mountain Pointe (AZ)	SS	43
Dax Fulton	Mustang (OK)	LHP	44
Masyn Winn	Kingwood (TX)	RHP/SS	45
Drew Romo	The Woodlands (TX)	C	46
Isaiah Greene	Corona (CA)	OF	47
Justin Lange	Llano (TX)	RHP	48
Cole Henry	Louisiana State	RHP	49
Alika Williams	Arizona State	SS	50

PLAYER	SCHOOL	POSITION	RANK
Spencer Torkelson	Arizona State	1B	1
Austin Martin	Vanderbilt	OF/3B	2
Asa Lacy	Texas A&M	LHP	3
Emerson Hancock	Georgia	RHP	4
Zac Veen	Spruce Creek (FL)	OF	5
Nick Gonzales	New Mexico State	SS/2B	6
Max Meyer	Minnesota	RHP	7
Heston Kjerstad	Arkansas	OF	8
Reid Detmers	Louisville	LHP	8
Garrett Mitchell	UCLA	OF	10
Mick Abel	Jesuit (OR)	RHP	11
Nick Bitsko	Central Bucks East (PA)	RHP	12
Patrick Bailey	North Carolina State	C	13
Austin Hendrick	West Allegheny (PA)	OF	14
Robert Hassell	Independence (TN)	OF	15
Ed Howard	Mount Carmel (IL)	SS	16
Garrett Crochet	Tennessee	LHP	17
Tyler Soderstrom	Turlock (CA)	C	18
Cade Cavalli	Oklahoma	RHP	19
Jared Kelley	Refugio (TX)	RHP	20
Pete Crow-Armstrong	Harvard-Westlake (CA)	OF	21
Cole Wilcox	Georgia	RHP	22
Dillon Dingler	Ohio State	C	23
Bryce Jarvis	Duke	RHP	24
Slade Cecconi	Miami	RHP	25
Justin Foscue	Mississippi State	2B	25
Bobby Miller	Louisville	RHP	27
Nick Loftin	Baylor	SS	28
J.T. Ginn	Mississippi State	RHP	29
Jordan Westburg	Mississippi State	SS	29
Clayton Beeter	Texas Tech	RHP	31
Aaron Sabato	North Carolina	1B	32
Carmen Mlodzinski	South Carolina	RHP	32
Austin Wells	Arizona	C	34
Carson Montgomery	Windermere (FL)	RHP	35
Chris McMahon	Miami	RHP	36
Casey Martin	Arkansas	SS	37
Jordan Walker	Decatur (GA)	3B	38
CJ Van Eyk	Florida State	RHP	39
Tanner Burns	Auburn	RHP	40
Daniel Cabrera	Louisiana State	OF	41
Drew Romo	The Woodlands (TX)	C	42
Dax Fulton	Mustang (OK)	LHP	43
Masyn Winn	Kingwood (TX)	RHP/SS	44
Justin Lange	Llano (TX)	RHP	45
Alika Williams	Arizona State	SS	45
Kevin Parada	Loyola (CA)	C	47
Cole Henry	Louisiana State	RHP	48
Burl Carraway	Dallas Baptist	LHP	49
Carson Tucker	Mountain Pointe (AZ)	SS	50

What's Cool on Campus - Data Analytics and Moneyball MLB and NCAA research

Part One: Open Letter to UNF Athletics: What's Cool on Campus - Data Analytics (UPDATE)

JACKSONVILLE, FL - After my initial research on the topic in September 2019, I continued researching data analytics adoption at the NCAA team level, adding six teams to the initial survey presented to the administration. 

The approximate 4 - 6 increase in wins over the trailing eight-year average of wins -- or wins expected -- remained in place. 

Growing number of programs finding or renewing success: (2019 record versus 2011-18 record)

Next, I looked at the final 2019 College Baseball poll from USA Today. I added a Y/N to sort for teams that had publicly announced their use of data analytics. An earlier USA Today article had mentioned Louisville and Texas Tech as College World Series participants that did not have an existing analytics program in place.  

from USA Today:
https://www.usatoday.com/sports/ncaa-baseball/polls/coaches-poll/2019/

2019 Final USA Today College baseball Poll - by adoption of data analytics

• Seventeen of the Top 25 teams had adopted data analytics (68% rate).
• Adoptees median number of Wins were 4 greater than non-adoptees, W% was 0.031 greater, which implies approximately +2 Wins. 
• Adoptees tended to play later into the post-season than non-adoptees.
• In other words, adding 2 - 4 wins while playing amongst the most competitive sub-set of teams you could select is meaningful. 

Next, I looked at adoption rate by conference. Here is where I feel the story began to expand and crystallize. 

I examined the Power-5 conferences (P-5) and selected Non P-5 conferences like the ASUN. This revealed the following:

• The Power Five conferences (P-5) adopted at a 55% rate versus 14% for Non P-5 conferences. 
• The SEC had adopted at 86% rate and the ACC at 64% rate, the highest rates observed. 
• The SEC is on track to hit 100% by 2021. 
• The Power 5 Conferences are on track to hit 100% by 2022.

     

Non P-5 schools, by contrast, are on an estimated adoption rate that mimics MLB's rate of increase (as shown below). This puts them on path to hit 100% by approximately 2032.  

That 10-year gap, if the Non P-5 status didn't do it, will relegate these schools to long-term, if not permanent, 'have-not' status in collegiate baseball. 

Year 1 = 1997 for MLB and 2017 for college baseball. 

Elitzur's MLB Timeline of Adoption (SABR teams over time)

These conclusions, if they hold true, are somewhat ironic in that one pillar of the Moneyball Theory and Dr. Elitzur's study, is that in MLB, poorer teams used data analytics to gain advantage against richer teams, in what was inherently an unfair contest between unequals. 

The analogy to college baseball flips that framework on its head in that richer teams are adopting it to further cement their advantage over less well-endowed competitors. 

 

Cum Wins v. Cum Payroll (1998-2016) - Avg. Payroll v. Diff. Wins vs. Expected Wins

Here is where things get interesting. I examined cumulative wins (from baseball-reference.com) versus cumulative payroll (from Lahman database) and applied conditional formatting to identify the top ten teams (green) and the bottom ten teams (red) by each category. The (white) cells are the middle teams per category. 

In the first category, cumulative wins and cumulative payroll, the Red Sox and Yankees (coded green-green) bludgeoned the field as top ten in payroll and top ten in wins. Not a huge surprise. 

Pittsburgh and the Rays (coded red-red) were bottom ten in both categories and that is not a huge surprise either. A 0.77 correlation between wins and payroll should come as no surprise either, it's the basis for the so-called "competitive balance tax" or de-facto salary cap. 

Cum. Wins v. Cum. Payroll (1998-2016) grouped by 10's

If you expand the conditional formatting here to show top ten-bottom ten-middle ten by category, you get some unusual pairings. 

Green-Green = BOS,LAA,LAD,NYY,SFG,STL

Green-White = ATL,CLE,TEX

Green-Red = OAK

White-Green = CHC,NYM,PHI

White-White = CHW,CIN,HOU,SEA,TOR

White-Red = ARI,MIN

Red-Green = DET

Red-White = BAL, COL, 

Red-Red = KCR,MIA,MIL,PIT,SDP,TBR 

This analysis looks at the absolute level of spending and the absolute level of wins. The top and the bottom level teams are segregated on that basis with only Oakland delivering on the upside as far as wins relative to payroll in aggregate which makes sense. They are the crown princes of Moneyball. 

Detroit finishing lower third delivering wins while spending in the upper third on the payroll side makes them the anti-Moneyball franchise so far.  

But is that the fairest way to grade franchise when the premise is to spend money efficiently rather than wastefully or recklessly?

Sum of Wins Over or Under Expected Wins and Avg Payroll Level

From the baseball-reference.com historical wins by team/year data, an "expected wins" field was created by weighting the prior three years win total, as is commonly done with projections. 

2019EW for example would be ((2018 W * 3) + (2017 W * 2) + (2016 W * 1)) / 6. 

OUW would then be the Over / Under of Actual Wins versus Expected Wins. 

I applied the same conditional formatting rules to sort in top-middle-low ten team buckets and calculated the following matrix of Wins Expected versus Average Payroll. 

Elitzur Wins -  Payroll Matrix

In my opinion, this gives a reasonable snapshot of which teams are more successfully employing the Moneyball concept of "doing more with less." 

High Wins over Expected with Low Payroll:

Oakland, Tampa Bay and Minnesota have had good success doing more with less. Arizona and Washington are also consistently on top pf their divisions. 

Low Wins over Expected with High Payroll:

Detroit and the Los Angeles Angels have had a bad run while spending at some of the highest levels. A bad combination recently.  

High Wins and High Payroll:

Boston, Chicago Cubs and Philadelphia fall into this category. High spending but high wins over expectations to match. 

High Wins with Middle range Payroll:  

Cleveland Indians with some consistently good years and the Houston Astros with some feast or famine years and a relatively new adoptee of data analytics (2012) scored well on Over/Under Wins, especially recently. 

This introduces another interesting observation which may appear to run counter to one of the earlier observations, that "if you were not in early, you were left behind."

MLB Wins over Expected Wins pre and post Adoption of Moneyball

               

click link above to see in Tableau:

Some teams who were late adopters of an analytics driven approach have had recent success in terms of wins above expectations (OUW):

Houston and Chicago Cubs (2012) and +8.63 / +3.06 OUW

Minnesota (2015) and +8.06 OUW

Philadelphia (2016) and Arizona (2017) +5.08 / +7.44 OUW

Houston and the Cubs have had good success but will regress somewhat as they add more years under adoption. Minnesota, Philadelphia and Arizona are more recent small sample size successes with only three to four years under adoption. 

Those team's recent successes, added to the previous success of the A's and the Rays, lend themselves to a belief that the teams that "need" to succeed with a Moneyball approach the task with a "have to have this work" mentality rather than a "nice to have" it work, if not throw some money at the problem. Failure is an ever-looming, existential threat to their success. This leads to a deeper commitment to the task and a greater buy in from everyone in those organizations. 

For the high payroll, high to middle success teams, data analytics is "nice to have" but not really "have to have". There is a safety net of the owners checkbook, an "in case of fire, break glass" option lesser well-endowed teams do not have, that blurs the amount of credit that should be given to data analytics for success in the W column. The low payroll teams are overcoming the 0.76 correlation between payroll and wins, the high payroll teams are surfing it to success. 

Each successive CBA defines the rules of engagement teams operate under and they naturally lend themselves to this type of stratification and perhaps always will. Organizations and staff at all levels are constrained by these rules and work with them and in some cases around them to the best of their ability in order to succeed.

On the collegiate side of the ball, college teams that lag in adoption could see some glimmer of hope from the late-adopters in MLB, however with the greater disparity and distribution of talent and the differences in the rules of engagement the NCAA and their member conferences set up between the teams, that glimmer of hope quickly morphs into a chasm of despair.  

Going forward, I would like to take more of a flow versus stock comparison, ie: change in Win Expectation versus Y-O-Y changes in payroll historically and see what that reveals.

References:

Elitzur, Ramy. “Data analytics effects in major league baseball.” (2020).