One interesting statistic is BABIP- Batting Average on Balls In Play.  It describes the percentage of balls hit into the field that become hits.  Through evaluation, however, BABIP appears very random from one season to the next.  This is a clear indication that too many variables, including luck, are involved when a pitcher has a stellar or a lesser than desirable one. An alternative evaluation method has become more popular and reliable due to this perceived luck… defense-independent statistics.  Walks, strikeouts, and home runs  allowed can be seen as the best predictor of success of a pitcher.

All of the maneuvering that has stemmed from a quest to replace a cornerstone since 1912.  ERA has influenced contract negotiations, CY Young awards and Hall of Fame ballots.  Since 1912, along with batting average (BA or AVG) for a hitter, ERA has been the measuring stick by which we determine who’s ace and who is not.

Now we get to our catchy title, usually reserved for a Barista to customer interaction surrounding a frappe chino or got chocolate order- WHIP or No WHIP. In my opinion, WHIP, which stands for Walks plus Hits Per Innings Pitched, is the single most important statistic to determine a pitcher’s effectiveness. This holds true in fantasy and reality.  WHIP incorporates the first of the three defense-independent pitching statistics (the walk) and it considers each pitcher-hitter outcome, even those coming after an error erased the third out  (when ERA stops tracking runs).  Pedro J. Martinez remains the standard bearer for a single season (.737 in 2000) WHIP, while the career record goes to Cleveland flamethrower Addie Joss with a WHIP (.968).  In the end, WHIP is an accurate predictor of future success, a determinant of past success, and is much easier to calculate than ERA. As a historical fanatic of baseball, it’s as close as we can get to comparing across generations.  I’ll take WHIP every time, in both fantasy baseball and reality..: but in terms of my hot beverage conundrum, it’s even easier. No whip for me- I drink coffee and I take it black.

The Hold (HLD) was invented in the mid-80’s to give middle relievers credit for their contribution to a win. The Hold is not an official stat tracked by MLB, however, and the criteria to qualify varies from one service to the next. The inclusion of Holds expands the bullpen options for a fantasy league and add value for setup guys like Dellin Betances or Adam Ottovino. Top setup guys will yield approximately 30 holds a year, compared to 40 saves. We run the risk of providing too much depth to a position that doesn’t need to be diluted.

Today’s reality offers a 2-phased approach to saves. We have the closer by committee faction vs. the set closer. However, with the fragility of millennials, the short term memory and toughness required to stick in the permanent closer role has created a revolving door at the position. So whether it’s designated as an intentional committee or not, it becomes one. If we really want to stretch, we can add the experimental opener to the mix to further water down the role in real life.

If fantasy wants to meet reality in the case of the closer, we either have to clone the Goose (Gossage) and distribute one to every team, or we have to find a value consistent with the three headed monster that is the “closer”. Can we analytically place a weighted value on saves, holds, and openers to create a real life value to a fantasy baseball league… thus far the answer is no. So if you want to win the reliever war on fantasy, you better close your eyes to reality and go get Chapman or Osuna.

Next week an examination of WHIP and BABIP.

Whether you’re a real GM or fighting for first place in your Fantasy Baseball office league, you must first prioritize the most important stat. No matter how you slice it and no matter how people want to separate the two arenas, the most coveted stat to a baseball team is the WIN. Basically the GM of the Yankees, Brian Cashman, and the GM of Chico’s Bail Bonds (My 10 yr old son’s team) are up late at sweating the same thing.

At the end of the season, it doesn’t matter how the club managed to get each win, as long as it accumulates a sufficient number of them to reach the playoffs. Every game is guaranteed to assign a win to someone, and pitchers are the ones who get the statistical credit. With 2/3 of roster spots in fantasy going to hitters and 60% of real rosters dedicated to the batsmen, you can’t miss in pitchers.

Wins are only one way to tell if a pitcher is having a productive season, but even a poor performance can earn a Win. We have John Lowe of the Philadelphia Inquirer to thank for conjuring up a derivative of the win to deepen its meaning and importance. In 1985, he defined the Quality Start (QS) as a starting pitcher completing at least six innings and allowing no more than three earned runs. There are several examples of of great games not garnering a quality start (and quality starts awarded for mediocrity), but overall this stat tends to reflect a stellar game. The average ERA for a quality start last season was 1.91, considerably lower than non-quality starts.

So, Brian Cashman and Tanner (my son) will be searching for pitchers that give you both stats. By tracking both statistics together, quality wins are worth double the cheap ones. There aren’t many Sandy Koufax type pitchers anymore (27-9 with a 1.73 ERA in his final season) but if you can monopolize the few bulldogs in the league, you’ll have a distinct advantage out of the gate.

Next article we’ll entertain closer stats that compliment the combined quality win stat- saves, holds, and the “vulture win”.

Let’s first examine Bill James and the Moneyball influence. As a special advisor to the Red Sox, James helped rewire the Breantown faithful into expecting their hitters to force opposing pitchers to throw. They traditionally took the first pitch and tried to get opposing starters to hit a certain pitch count as early as possible (ironically exploiting a Moneyball pitching staple of capping pitch counts and only seeing the lineup twice before going to the bullpen). Boston, under his influence, led the league in walks in 14 seasons since 2002. Examining that same stretch, only two teams had more plate appearances where they took the first pitch. There is a direct correlation between taking the first pitch and scoring runs (teams in the top ten are above league average in runs scored). If you follow the Moneyball pattern, taking the first pitch leads to walks, walks lead to runs, and runs lead to wins. So patience works, right?

The 80’s were a more aggressive time seeing Rickey Henderson set records for leadoff homers. Hitters saw an average of 3.5 pitches per at bat where the modern hitter sees almost 4. A half of a pitch per at bat adds up over 2 trips through the lineup… about 10 extra pitches.

Here’s some thoughts to ponder regarding the first pitch:

1. 45% are thrown for strikes.
2. BABIP .340/.350/.577 on 1st pitches
3. BABIP .319/ .330/.503 on 0-1 count
4. BABIP .338/.341/.538 on 1-0 count

Here’s a look at what happens when a batter swings at the first pitch:

1. 40% foul it off
2. 25% swing and miss
3. 26% put it in play for an out
4. 9% get on base

If you’re not good at math … you only have a 35% chance of putting the ball in play.  Yes, hitting a baseball is hard.  So, if you take the first pitch, you have a 46% chance of being down 0-1 and if you swing, you have a 65% chance of being down 0-1.  If avoiding being down 0-1 is your goal, then taking the first pitch seems to be the way to go.  BABIP for the 0-1 count is .329/.330/.503… however, the two most feared counts as a hitter are your next two options … 1-1 and 0-2.

1:1 has long sense been argued as the most pivotal count in baseball.  Let’s look at the splits when a hitter gets two strikes:

1. BABIP .149/.159/.227 on 0-2 count.
2. BABIP .161/.169/.243 on 1-2 count.
3. BABIP .181/.187/.291 on 2-2 count
4. BABIP .204/.449/.349 on 3-2 count

It’s not a promising outlook when you get two strikes on you.  I go back to the old adage my dad impressed upon me… “ If you don’t let them get two strikes on you, they can’t strike you out.”  I’d like to take that a step further as we wrap this up.  Ask yourself the question as you prepare for your at-bat: what is your approach or plan?   You have to have one.  I’ll give you one last pairing of statistics before you decide which fits you… to swing or not to swing.

1. .265/.290/.440 0-0 count if you swing
2. .242/.330/.296 0-0 count if you don’t

Here’s one last nugget to chew on before sending in your opinion.  Looking at the tipping point count of 1:1, BABIP for 2-1 count is .339 vs. .169 on a 1-2 count.  Those numbers should put fear in your heart- it’s the difference between being an all-star and being out of the league.

Let’s deduce like Columbo to solve this mystery:

1. Getting 2 strikes on results in bad outcomes.
2. Getting a 1-1 count can shift the pendulum in drastic directions… risky.
3. Being active as a hitter within the first 2 pitches appears to give the hitter an opportunity to have statistical success…

The moral of the story can be split into two genres.  The analytics guru in me tells you to avoid two strikes and the worst thing you can do as a hitter is become too predictable (Since analysts deal with using historical data to predict future performance).  The old school baseball guy in me tells you, swing at pitches in the zone and most importantly, swing at pitches in YOUR zone.  Hit it hard and keep it fair.  Cut out the swing-and-miss. Baseball is much more than homers, walks, and strikeouts.  The dichotomy of my personality is telling you to look to strike early in the count.  Both the numbers and the predictability of the rest of the league hitters are on your side.  If pitchers think you’re taking early and they want to at least be in a 1-1 count, chances are you’re going to see a strike.  Rickey, Donnie Baseball, Pete Rose… bring these types back.

Flash ahead 16 years and the balance of perspective that allows for smart people to hold each other accountable is absent. What started out as a puzzle of trying to figure out which statistics correlate to wins has turned into a mystery. Currently we are drowning in information and starving for wisdom. It was argued that walks were grossly undervalued and stolen bases, the hit and run, and the sacrifice bunt were thought too highly of. Lewis further describes Beane’s mission as one where he covets on base and slugging percentage as the holy grail that directly translate to the win column.

A lot of people bought into this and Hollywood became reality. MLB front offices started looking like Wall Street, with analysts in cubicles treating ball players like dividends. All you have to do is turn on a random game in the third inning and witness a drastic defensive shift with three guys on the same side of the infield or four guys in the outfield (see last week’s article).

Think about this. In 2002, the year before the book was released, MLB players walked 16,246 times. Since then, only two times has the walk total eclipsed that number, despite the analytics community’s enthusiasm for the stat. Weren’t they listening to Jonah Hill?!

But, there were 2 statistics not touched on by Lewis or Beane that have skyrocketed because of analytics. Home runs and strikeouts go hand in hand. In 2002 there were a shade over 5,000 homers hit and 31,000 strikeouts. Today, we’re looking at 1,200 more homers hit in a season and almost 9,000 more opportunities where the ball is not put in play.

To hit home runs, hitters have to start their swings sooner and pull the ball to the shortest part of the park- usually the corners. Starting the swing sooner means they’ll get fooled more often … thus, more swings and misses.

This is where the analyst would retort, a strikeout is an out and an out is just an out. As an analytics person, I have to question whether we should ignore the value of putting the ball in play, moving a runner over, and putting pressure on the defense. Is this now the undervalued stat? It certainly won’t get you paid.
If you drive that runner in from second you can make some good money… if you hit a homer, you can make even more.

As an industry, if a player is rewarded for home runs, yet not penalized for strikeouts, why not swing for the fences? Hit enough home runs and the strikeouts won’t matter … ask Aaron Judge who eclipsed both 50 homers and 200 strikeouts in the same campaign.

When teams apply the defensive shift, wouldn’t it make sense to lay down a bunt or hit a 27-hopper to the unprotected area on the vacant side? The single is definitely devalued. If a hitter feels like his paycheck correlates with home runs, he might let a pitch go if he can’t send it 400 feet. Taking pitches to get a favorable hitting count has resulted in pitch counts going up. But more pitches thrown and fewer balls in play means games have gotten longer… and slower.

With analytics-based offenses relying on extra base hits, why steal a base or sacrifice a runner or move a guy over? On a swing of the bat, that guy could score anyway. Ironically, standing around and waiting for the home runs has actually taken the strategy out of the game.

In making this next comment I may have to hand my SABR card and give back my analytics degree … but singles matter too.  String a couple of them together, throw in a stolen base and like magic, a run appears. Even if it doesn’t result in a run, a seemingly meaningless 2-out blooper can allow for a key hitter to get one more at bat later in the game or that starter to have to throw 5 more pitches.  Just having a runner on base who is a threat to steal changes the dynamics of the pitcher’s delivery (slide step or not), pitch selection, and his concentration is now split between batter and runner.  The defense is now altered with a first baseman holding or playing behind, a middle infield pinching for better coverage on a steal or double play, and the third baseman is creeping in to defend against the bunt.  A simple single alters the DNA of the defense adding countless variables to the decisions involved with the next batter.  All these strategies involving movement, communication, cat-and-mouse like ploys, angles, fakes … dare I say strategy, are certainly more interesting than sitting around waiting for a home run.

Analytics have provided more offense but the current methods have made the game less interesting, less intellectual…. and longer.  The all-or-nothing style of baseball has certainly resulted in some monumental power numbers, but the amount of strikeouts that come along with it has resulted in a boring product.  Bring back Billy Ball … Billy Martin that is.

I can hear Pete Rose and Ted Williams laughing all the way to the bank. Rose used bunting as an act of war on the pitcher. Of his 4256 hits, Rose had 65 bunt singles reaching base at a 69.9% clip. Do the math. Without the bunt, he would still be tied with Ty Cobb at 4,191.

In regards to the defensive shift, MLB introduced it in 1920. Ted Williams encountered it in 1940, as Lou Bourdeau and Cleveland Indians popularized it and saw the “Splendid Splinter” basically ignore it on his way to a career .344 BA and 521 home runs.

So how did we get here? Not to say correlation equals causation, but since James’ comments bunting usage has declined from 1.7% to .7%. At about the same time we saw defensive shifts increase from roughly 3,000 to 33,000 … that’s a 900% increase. The best bet in countering this defensive alignment is hitting a ground ball the opposite way, but its proven difficult to put in motion. Of the roughly 5,000 ground balls this season hit against the shift, less than 10% went to the opposite field. And of those opposite field ground balls only about 20% produced runs. If players could begin to do this on command it would prove to be ideal, because batters have produced well over a .500 average on such plays. Today’s analytic driven theories on hitting do not work that way.

Driven by launch angle, exit velocity and true outcomes, baseball analytics experts have looked past the simplest antidote: depending on the positioning of the fielders, bunt for a base hit. Depending upon the situation a soft ground ball the other way could prove to be as successful as a bunt. Since 2015 bunt attempts away from the shift have been successful to the tune of over 70% of the time, as opposed to opposite field ground balls having a success rate a shade under 50% of the time. The sample size isn’t large, but the evidence of success is at least worth a look.

To his defense, Bill James did later recognize that the “bunt could become valuable versus the shift.” Wouldn’t it be ironic if the same analytics driven approach that purest claim has ruined hitters abilities, could revitalize a new generation of hitters that handle the bat as opposed to just swing the bat. The bunt, opposite field hit, and dare I say, the hit and run could become the inefficiency that data analysts love to find. Hang on old school baseball guys, we could see the pendulum swinging back.

Base Running Run is the sum of several baserunning components.  GAR (Ground Advancment Runs, SBR (Stolen Base Runs), AAR (Air Advancement Runs), HAR (Hit Advancement Runs, and Other Advancement Runs (OAR).  BRR represents the number of runs contributed by a player’s advancement on the bases above what would be expected based on the number and quality of the base running opportunities with which the player is presented, park adjusted, and based on a multi-year run expectancy table.  When you’re looking at the average run value in terms of a base/out state you subtract average run value from actual run value on a particular play where a base runner is involved.  The result, the difference, is the run value awarded to the base runner.  Confused yet?  Here’s a basic scenario to clarify.  We have a runner on 2nd and 1 out …

A ground ball to the right side would automatically advance the runner to third.  But what about a ground ball to third?  Let’s say our same runner advances to third and our runner going to first is thrown out 25% of the time, he stays at 2nd Base 70% of the time, gets thrown out at 3rd 2% of the time and beats the throw to third 3% of the time (batter reaches on fielder’s choice).  Let’s say the average base/out run expectancy (RE) of all of those, weighted by the frequency of occurrence is .25 runs.  If the runner advances and the batter is nailed at first, the runner gets credit for .25 runs (.5-.25).  If he stays at second base, and the average RE is of a runner on second base with 2 outs is .23 runs … then he gets docked for -.02 runs (.23-.25).  A runner gets docked or credited for the resultant run value of what he does minus the average weighted resultant run value of ALL base runners in that situation.  Multiple baserunners can get positive or negative credit on a particular play.

Below are samples of situations where a runner can obtain positive of negative credit:

1. A hit, taking an extra base, not taking an extra base, getting thrown out trying to take an extra base (assuming there is no-one impeding his ability to advance).
2. A batter gets thrown out trying to take an extra base on a hit (if he is successful, we simply credit it as a ‘double’ for it.)
3. On a hit … the batter advancing, not advancing, getting thrown out while trying to advance when a runner is safe or out advancing an extra base.
4. Tagging up on a fly out.
5. Ground balls to the infield, runners stay out of a force or DP at 2nd base … whether or not the batter is out or safe on FC.
6. Runner on second advancing on GB to SS/3B or not or getting thrown out advancing.
7. Trail runner advancing, not advancing or getting thrown out when lead runner is safe or out trying to advance an extra base on a hit or an out (you can connect this scenario with #1)

Scenarios not considered …

A runner on 3B advancing, not advancing, or getting thrown out at home on a ground ball are not considered.  Runners advancing on WP or PB are also not considered.

All of the situations discussed are considered an opportunity for the base runner or batter.  The batter does enter the opportunity phase when he gets a hit unless the lead runner tries to advance an extra base and the batter has a chance to advance on the throw.

As we learned in the previous article, the value placed on base stealing players could be an inefficiency in the market, depending the success rate and frequency.  What about a great baserunner?  Could this be a piece of the pie that needs to be paid for before the rest of the baseball world catches on?  We entered the world of sabermetrics with teams paying for OBP … we correlated OBP with Runs and Runs with Wins.  Did we overlook the space between the notes.  Who are those base runners that add runs to their teams because of the ability to run the bases well?  Stay tuned to the next article- we’ll give you undervalued and overvalued players … and some prospects to keep an eye on.

Conceivably, the greatest leadoff man of all-time, Rickey Henderson’s most glaring attribute was his knack for disrupting a game with a stolen base.  Maury Wills was the first player to reach the century mark (104) in 1962 for the Dodgers.  It took another 12 years before Lou Brock, of the Cardinals, was able to eclipse the feat with 118 thefts.  The 80’s saw the likes of Rickey Henderson swipe 100 bags in a season three times, including a record 130 in 1982.  Vince Coleman ran wild from ’85-’87 totaling 326 thefts and three straight 100 steal seasons.  

As the 80’s closed out, Henderson snagged 93 bags for the Yankees and Coleman 81 for St. Louis in 1988.  Since then nobody has reached 80!  Only four men have ever reached the century mark, a total of eight times between them … and no-one has joined the club in 30 years.  Among the “sabermetricians” and “traditionalists” that make up the two political parties of baseball, the effect of the stolen base has often been debated.  The “Moneyball” approach does not implore a team to give away outs on the bases or on sacrifice situations.  It can be more station to station relying on the home run and walks.  While numerical charts in this article will support the more modern day approach, it’s hard to ignore the former World Series Champs, Kansas City Royals, and their small ball approach to the game.  The questions remains, with the tilt favoring the numbers game of going station to station, would a player like Henderson, create an inefficiency in the market place to the point where base stealers’ values would be purchased for pennies on the dollar and exploited for large value once proven statistically relevant again?

Using win and run expectancy models, the value of a person going from first to second base with no outs increases the chance of a team scoring by 19%, but going from a runner on first with no outs, to nobody on base with one out decreases the chance of a team scoring by 26%.  Another popular time to steal is with a runner on first and two outs.  The chances of a team scoring when a player goes from first to second base increases from 12.7% to 21.6% … an increase just shy of 9%.  However, remember, if the runner is thrown out the chances of scoring goes from 12.7% to 0%.  

So where are we going with this?  No statistic has a debate point of diminishing returns like the SB.  Experts have tried to place a single number on the point where a player is hurting his team if caught.  John Thorn estimates that a runner has to be successful 67% of the time to benefit his club.  I believe it varies based on the run scoring environment.  The higher the run scoring environment the less costly a “caught stealing” is.  If you’re confused, let’s look a basketball.  If two teams turn the ball over on average ten times a game, both are equally efficient, correct?  No.  It would depend on the number of possessions … or scoring environment.  A team with 100 possessions and 10 turnovers is theoretically hurt far less by the 10 turnovers than a team with 50 possessions.  Back out of the rabbit hole … Thorn was not far off.  In the next article I will explore this nuance in great detail, but for now trust me that a break even point lies somewhere between 67% and 73%.  So if Rickey Henderson is able to move from first to second at a clip greater than 73% and stealing 80 plus bases a season, his style of play would represent a huge inefficiency in the market place.

To place base stealing on an island by itself without “Baserunning Runs” metric would be irresponsible.  By themselves, they offer only a piece of the pie … but for now I am going to be irresponsible and capture the total base running picture in a following article.  To make my point, in the amazing 1982 season, where Rickey stole unfathomable 130 bases, he was caught 42 times (68%).  This season is considered less impressive in terms of base running metrics than his 1985 season in which he stole 80 bases and was caught 10 times (88,9%).  In the end stolen bases and baserunning are only a small part of the picture, but can certainly impact a win or a loss.  Stealing a base is still the most efficient way to advance, but being caught has an equally adverse effect. Guys like Beltran (86.4%), Raines (84.7), Coleman (80.9%), Lofton (79.5%), Brock (75.3%), Wills (73.8), and of course Henderson (76.2%) stole enough bases over their careers (300 or more) and had a success rate high enough to cause disruption in the physical/psychological world of baseball and in the analytic realm.  Their presence could cause a huge inefficiency in today’s market.  Today a guy with 40 SB’s will lead the league.  Who are 2019’s version of these unicorns?  As we search, don’t lose sight of the moral of the story … if you’re going to go, you better make it.

We’ve seen ERA scrapped for WHIP.  We’ve witnessed the emergence of FIP, CERA, dERA, Def Eff, DIPS, Adjusted ERA, IR, etc … as key analytic placements that decide the effectiveness of a pitcher.  It begs the question, would the Steve Carlton as we know him have existed if analytics driven decisions were as dominant in his day?

It’s amazing how we have come full circle.  In examining the modern pitcher under the premise that we are searching for inefficiencies in the market place, should we not take a closer look at something as simple as innings pitched?  Based on the numbers below, as real life or fantasy GM’s let’s take a look at placing a premium value on the pitchers that are extending their outings beyond the new norm of “two times around the order”.

Year     160.    180.    192.    200.    215

2018.    26.      12.       8.        5.         1

2017.     89.     66.      15.       15.       0

2016.     79.      56.     17.        15.       8

2015.     75.       46.    34.        28.      9

2014.    60.       35.    50.        34.      14

As a reminder, in examining the numbers above, they do not represent unique placeholders.  For example, this past season only 3 unique pitchers threw 192-199 innings.  In the upcoming week, I will balance this potential inefficiency against $ value on the market and try to ascertain the consistency of a pitcher as they move through the categories.  If a pitcher throws 160 innings how certain is it that he will be able to hit the 180 mark the following season?  Could innings be as simplistic as the BB was for Billy Beane?  What’s the most important stat and/or analytic formula for you when assessing pitching?