The general bad level of play in some amateur baseball leagues flattens out individual skills. There is a point at which the number of fielding errors, passed balls, etc., as well as gratuitous walks and all sorts of mis-judgements, become so numerous that everyone – best or worst – contributes equally to wins and losses. Systemic bad play, in other words, makes everyone equally productive and non-productive. Furthermore, individual skills and statistics seem so unreliable to predict future performance that, win or lose, there seems little use of strategy in a game where there are more errors and unworked walks than hits. How does one design a team around that?
This is the dilemma.
- Some pitchers are so easy to hit that almost everybody ends up with the same batting average by the end of the game. When they throw strikes, they get hit between .400 and .500 on average by the good hitters in the league and over .600 by the very best.
- However, the large majority of pitchers do not throw a lot of strikes. The league BA is .300 while its OBP is over .500. A .200 hitter gets on base as much as a .400 hitter, through no skill of his own. A normally .400 hitter / .500 slugger looks at so many bad pitches that he walks more than hits, thus rarely executing his fine swing or keen eye and therefore not really outshining his lesser-skilled teammates.
- Catchers and pitchers combine to do little or nothing to stop the running game. Everybody steals. A walk with 1 man on base is often followed by a double steal and a passed ball / wild pitch, thus clearing the base paths. A walk is therefore as good as a triple or a home run when adding up total bases and advanced runners.
- Fielders do not stop or handle effectively most balls put in play. There are errors and misjudgements that lead to extended innings of large amounts of unearned runs. There are few exceptional plays and hardly any double plays. And there is only a rare out made at the plate.
Is this baseball? Not sure anymore, but it’s the only game is town. That said, the point here is not to badmouth the amateur player. In fact, it’s quite the opposite: Imperfect baseball and its imperfect players need a statistical and analytical basis to their game.
What sort of statistics and player ratings can be used to choose a roster for a team who plays in such an imperfect league? How can we distinguish between good hitters and bad ones? Who walks because they work the count and who walks because the pitcher threw 4 really bad pitches? Can we really say that the best batters outperform the lesser ones when good and bad players alike are equally adept at exploiting the errors and overall weakness of the opposing team? Does the measured hitter and his careful baserunning misunderstand the chaos of unearned runs?
We need to step back from our traditional analysis of baseball, which is based on very high priced, fine-tuned professionals who practice every day and who have mastered the game to near perfection. We need instead to take a closer look at errors, passed balls, misjudgements, too many steals and few caught stealings, high batting averages, and the effects of gratuitous walks on wins and losses. We need to be able to judge imperfect players and imperfect teams on how they will perform in an imperfect league.
Take an example from an actual start of a game. Leadoff batter walks on 4 pitches. While second batter looks at his first pitch, first batter manages to reach 3rd base on a steal and a wild throw from the catcher. Second batter then hits a grounder to the 3rd baseman. Runner on 3rd doesn’t even pause, he just runs home. 3rd baseman overthrows to catcher. Runner scores and batter advances to 2nd base. Result of 6 pitches: 2 runs, 1 walk, 1 steal, 2 errors, and some bad judgements. Predictable if you’ve been playing this game long enough, but not really predictable in any measurable sense.
Now, all of this could have been avoided if some of the following had occurred:
- The first batter did not walk.
- The first batter, if walked, did not steal 2nd.
- The catcher had caught the first batter stealing.
- The catcher didn’t throw wild to 2nd base.
- The 3rd baseman had held the runner and threw 2nd batter out at first.
- The 3rd baseman threw the 2nd batter out at first and the 1st baseman threw the 1st batter out at home (if 1st batter ran to home).
- The 3rd basemen caught the 1st batter out at home.
Thing is, all of this is random. In a professional game, things are less random: once a ball is put in play, the game is fairly predictable and the execution is almost always performed without error. We can be sure that catchers will throw enough batters out to make stealing a risky, strategic device. Men on 3rd base don’t just ignore fielder excellence. Etc. Etc. If there is any room for randomness in professional baseball its found largely in what happens between the pitcher and batter – Will this be a walk or a hit or a homerun or a strikeout or an out, a grounder or a fly? True, exceptional plays and the rare error can determine a game, but the exception is just that, an exception; whereas at the amateur level, the exception is not only the rule, the number of exceptions possible is apparently endless. There is not only the batter / pitcher game and its unpredictability; at any given moment everything that follows a pitch or a swing of the bat is equally unpredictable. Baserunning mistakes and fielding errors abound; umpires invent rules; managers and players make up rules and strategies as they go. You see things in amateur baseball that you would never have imagined possible in the game of inches.
If amateur baseball is not about inches – What is its yardstick?
For example, not every leadoff batter gets on base. In fact, the level of hitting is such that any pitcher with decent control and some speed change or god forbid a curve will force many groundouts and overall poor contact. And some batters will just make lousy contact for no apparent reason. So getting on base is not a given. A leadoff hitter is therefore not always going to get on base. However, this does not stop the next set of batters to score a lot of runs. I am therefore not sure if there is any usefulness to run expectancy statistics about leadoff batters who get on base with no outs when 1 and 2 out rallies which create more than 5 runs are the rule and not the exception.
Finally, small samples of data, the constant change in players between seasons, and not to mention the terrible state of scoring a game – all this puts into question the very use of statistics at this level. There are some ways to avoid these problems:
- Use batting practice and simulation games as part of the overall statistics.
- Score your own games. Get the players involved, make it part of the game.
But the problem still remains: What to look for? What to count? And what to calculate?
Part of the answer is indeed about what to look for. Clearly, if in forming a team we want to separate the good from the bad, we will need to reconsider the use of standard baseball statistics and come up with a new set of standards. We will need to look closely at precise error statistics, and the set of skills that can best exploit the various possible error types; we will need to look at precise pitch-by-pitch observations to see who is truly working the count against those who are simply hoping and praying for a walk and are lucky enough to get it; we will need to come up with a system that classifies contact outside of the binary hit or out system (professional baseball already tracks line drives, flies and groundouts, so hitting statistics have already gone beyond the binary).
One further question to ask, which could be considered a moral question: Do we really want to measure imperfections? Or do we want to try to improve our game by playing baseball at its best, even if that means we don’t steal every time we reach base? What do we want teach a new player – that stealing is easy, do it as often as possible? or that stealing is a risk-based strategy?