Remember me

The All or Nothing Effect

February 21, 2022
 

Strikeouts are up a record amount, and the analytics seem to indicate it doesn’t matter as long as you hit home runs. But how does this approach fare in the playoffs- when it really matters most.  Not only are the stakes higher, but the game is managed and played a bit differently.  In October, runs tend to be at a premium and as Billy Beane said, "my shit doesn’t work in the playoffs."    

To look into this further, an All or Nothing factor (AoN) was calculated by taking the team’s total number of strikeouts and home runs and dividing them by plate appearances. 

(SOs + HRs)/ PAs = AoN

From example, here’s a list of the AoN factor for the 2021 League Championship squads:

Year

Team

PAs

HRs

SOs

AoN

2021

ATL

6056

239

1453

0.2794

2021

HOU

6291

221

1222

0.2294

2021

BOS

6122

219

1386

0.2622

2021

LAD

6239

237

1408

0.2637


For comparison purposes, here are the numbers for the 1988-90 Bash Brothers A’s:

Year

Tm

PA

HR

SO

AoN

1988

OAK

6356

156

926

0.1702

1989

OAK

6110

127

855

0.1607

1990

OAK

6240

164

992

0.1853

 

The AoN factor was calculated for the forty LCS teams from the past ten years (2012-2021), and those numbers were graphed with the average number of runs scored per regular season game.

As one might expect there is a positive correlation between AoN and runs scored in the regular season.  The same exercise was done by taking that team’s regular season AoN and graphing with the average amount of runs scored per playoff game (trendline in red). 

In this graph, the specific data points were removed to make it easier to see the trend lines. The playoff data is obviously a smaller sample size but the difference in trend lines is interesting.  As expected, the average number of runs scored in the playoffs is lower.  Also compared to the regular season, as a team’s AoN increases, the bigger the divergence in the average number of runs scored in the playoffs.  The more "disciplined" teams are closer to their season average than the All or Nothing teams.  If we look at a subset of the last seven years, the difference is more drastic.

In the last seven years, almost every team has an AoN factor greater than 0.2 (the lone team is the 2015 World Champion Kansas City Royals at 0.18).  For this 7-year subset, the number of runs scored in the regular season has flattened and the runs scored are at a higher level (~5 runs per game). 


While it’s difficult to draw firm conclusions from this data given the sample size, it begs the question how did the different AoN tiers fare in the playoffs? The bar graph below shows how far teams in low-, mid-, and high-AoN tiers progressed into the postseason:

  • Low AoN- range 0.1783-0.2274 (13 teams)

  • Mid AoN- range 0.2281-0.2540 (13 teams)

  • High AoN- range 0.2543-0.3043 (14 teams) 

Over the last ten years, the High AoN group has the least number of teams that advanced past the LCS and the least number of World Champions (2).  One of the World Champion teams is this year’s Atlanta Braves with a 0.28 (third highest AoN from the 40 teams evaluated).  The other is the 2020 Los Angeles Dodgers with a 0.2543 AoN (the lowest value in the High AoN group).  


The same tally is done comparing the teams within each year.  For example, in 2012 and 2013:

Year

Tm

AoN

Group

Result

2012

SFG

0.1935

Lowest

World Champions

2012

DET

0.2069

Mid-Low

Pennant

2012

STL

0.2136

Mid-High

LCS

2012

NYY

0.2281

Highest

LCS

2013

DET

0.1955

Lowest

LCS

2013

STL

0.1991

Mid-Low

Pennant

2013

LAD

0.2090

Mid-High

LCS

2013

BOS

0.2328

Highest

World Champions


Within each year, the League Championship team with the lowest AoN was placed in the Lowest group, the second lowest AoN was placed in Mid-Low group and so on for Mid-High and Highest groups.  The bar graph (listed below) is a tally of the playoff finish of the same forty LCS teams grouped by this yearly method:


Like the previous tally, the Lowest AoN group has the greater number of teams that advanced onto the World Series.  The higher AoN groups (Mid-High and Highest) have the least number of teams that advanced past the LCS, and the lesser number of World Champions (compared to the lower AoN groups- 3 vs 7). 

 

For your reference, here’s a list of the AoN and R/G numbers for the forty LCS teams evaluated (2012-2021).  And yes, this is a shameless but certainly fruitless attempt to get more action into the game. Thanks for reading. 

Year

Tm

AoN

R/G

R/G (P)

Result

2014

KCR

0.1783

4.02

4.60

AL Pennant

2015

KCR

0.1818

4.47

5.63

World Champions

2012

SFG

0.1935

4.43

4.31

World Champions

2013

DET

0.1955

4.91

3.18

ALCS

2013

STL

0.1991

4.83

3.29

NL Pennant

2014

STL

0.2034

3.82

3.78

NLCS

2012

DET

0.2069

4.48

3.23

AL Pennant

2013

LAD

0.2090

4.01

3.90

NLCS

2017

HOU

0.2113

5.53

4.33

World Champions

2012

STL

0.2136

4.72

4.38

NLCS

2015

TOR

0.2219

5.50

4.73

ALCS

2014

SFG

0.2262

4.10

4.18

World Champions

2019

HOU

0.2274

5.68

3.94

AL Pennant

2012

NYY

0.2281

4.96

2.44

ALCS

2018

HOU

0.2281

4.92

5.25

ALCS

2020

HOU

0.2284

4.65

4.77

ALCS

2021

HOU

0.2294

5.33

5.44

AL Pennant

2018

BOS

0.2318

5.41

6.00

World Champions

2016

CLE

0.2325

4.83

3.60

AL Pennant

2013

BOS

0.2328

5.27

4.50

World Champions

2015

NYM

0.2387

4.22

4.43

NL Pennant

2016

CHC

0.2428

4.99

4.41

World Champions

2014

BAL

0.2440

4.35

4.71

ALCS

2016

LAD

0.2450

4.48

3.27

NLCS

2019

WSN

0.2456

5.39

4.59

World Champions

2016

TOR

0.2540

4.69

3.89

ALCS

2020

LAD

0.2543

5.82

5.61

World Champions

2017

NYY

0.2561

5.30

3.92

ALCS

2017

CHC

0.2585

5.07

2.50

NLCS

2017

LAD

0.2586

4.75

5.47

NL Pennant

2021

BOS

0.2622

5.12

5.45

ALCS

2018

LAD

0.2628

4.93

3.69

NL Pennant

2021

LAD

0.2637

5.12

4.00

NLCS

2019

STL

0.2643

4.72

3.56

NLCS

2018

MIL

0.2699

4.63

3.70

NLCS

2015

CHC

0.2724

4.25

3.56

NLCS

2019

NYY

0.2791

5.82

4.89

ALCS

2021

ATL

0.2794

4.91

4.06

World Champions

2020

ATL

0.2884

5.80

4.75

NLCS

2020

TBR

0.3043

4.82

3.95

AL Pennant


 
 

COMMENTS (7 Comments, most recent shown first)

hotstatrat
Yeah, it's just an immediate gut reaction, but giving HR and SO equal weight doesn't feel right. How much more of an effect on the outcome of a game does a home run have than a strike out? Whatever that is: 5 times? 8 times? Let's call it X, then it would make more sense to me to go with AoN = (X(HR) + SO) / PA. It would be interesting to see if that makes a difference in your findings.
3:49 PM Mar 4th
 
abiggoof
I had wondered about this!
2:47 PM Feb 23rd
 
pgups6
Belewfripp, thanks for your feedback.

Adding walks (and 3TO) would be another interesting analysis, but different. It would add another factor into the mix, and I was really interested in if strikeouts mattered and the way the game is played today. With all the talk of three true outcomes, the average number of walks per season hasn't really changed. In 2006, the average number of walks per team was 528. In 2010- 526, in 2021- 526.

With strikeouts, the numbers have increased, 2006- 1055, 2010- 1144, 2021- 1405. Same but not as drastic with HRs, 2006- 180, 2010- 154, 2021- 198 (2019- 226).

Taking ABs instead of PAs would take walks into account as well, but really wanted to keep it focused on the elements of today's game and if that approach may impact October.

9:33 PM Feb 22nd
 
belewfripp
Thanks for the interesting analysis. Others have already noted that a team can have a similar AoN score but have a different breakdown of offensive events (positive or negative). I would further note a couple of additional thoughts that occurred to me on reading this:

- Given you are using PA, maybe it would benefit the analysis to add walks since these are also part of the "three true outcomes" triad with SO and HR.

- Alternatively, if sticking with HR + SO, maybe using AB instead of PA to winnow the data down to only those batter appearances that were able to generate one of those two outcomes?

- Lastly, I suspect that if the effect is real, the reason it fails in the postseason is because of its all-or-nothing nature. Over the course of 162 games, you can live through the days/nights when you get a "nothing" result - it only has to work 55-60% of the time to net you a playoff spot, or be in strong contention for one. If you have a bad stretch, it's survivable.

But in a short series, it's a very different story. You have a couple of "nothing" nights instead of "all" nights and you're in a hole you might not climb back from.
5:59 PM Feb 22nd
 
pgups6
MichaelPat, thanks for your comments.

HRs and Ks are not the same, and not always directly related (you can still strike out even if you're just trying to put the ball in play) but there's no denying there is a correlation. Strikeouts and homers are way up because players are concerned with launch angle and efficiency.

And yes, AoN does not tell the full story, just like all other non-raw stats. A player with 10 seasons with 6.0 WAR has the same career WAR as a player with 20 seasons of 3.0, but they certainly didn’t have the same career. All stats need to be viewed in context.

Your proposal would be interesting in how these teams scored in the playoffs, whereas this was more about defining the team (with regular season AoN) and see how they fared overall in playoff scoring and results. Obviously, this is just one side of the ball and doesn't account for run prevention. The 2014 World Champ Giants are in the mid-high category and MadBum was no small part of that chip. With the 2021 Braves, they won it all while still scoring about a run less per game in Oct.

As stated, no firm conclusions can be drawn but I wouldn't consider it meaningless either. Thanks again!

4:41 PM Feb 22nd
 
MichaelPat

Adding home runs and strikeouts is adding unlike things. A team with 200 HR and 1600 K is not the same as a team with 250 HR and 1550 K; yet if their PAs are similar then they will have similar AoN scores.

Before I accept this as meaningful, I would like to see the same analysis with just home runs, or better yet with HR as a percentage of a team's overall run production. You might also want to consider how teams scored in the playoffs versus how they scored during the regular season.

Take the 2021 Braves, for instance. They hit 1.48 HR per game on their way to scoring 4.9 runs per game during the season, while striking out nine times per game
In the playoffs, they scored 4.1 runs per game, but hit almost as many HR (1.44/game) and saw their Ks rise to ten per game.
The Braves' AoN in the playoffs was well over .300.... (BRef summaries show 519 AB + BB, 162 K, 23 HR. Sacrifice and HBP data are not included in the team summaries, but even at two events a game they'd be over .335AoN)
12:01 PM Feb 22nd
 
djmedinah
This is seriously great ??
8:06 PM Feb 21st
 
 
©2024 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy