This basic plot shows the fraction of game starts made by left-handed pitchers.
Before making the plot, I had expected this percentage to generally increase over time. That’s clearly not the case.
I’m particularly interested in the major dip starting in 1993. This seems to coincidence with the Steroids Era. Does this suggest that the increase in offense during this period is due, in part, to lack of availability of left-handed starters? Or were fewer left-handed starters used for some other reason?
I’m curious now about what the # (specifically FIP) for these lefty starters were during the dip period. Maybe there just wasn’t many effective lefties for the time. Altho, it does strike me really interesting too, that maybe some of the “steroid era” numbers might be a result of a lack of lefties. If so, how much of an effect did it have? Would it even be possible to calculate that?
On the twitter conversation, some have mentioned the expansion of 1993. There was also the expansion of 1998, and the labor action of 1994-95. I can remember at least one minor leaguer who quit baseball rather than become a break the strike in the majors. http://www.baseball-reference.com/minors/player.cgi?id=jordan001mic
The strike year of 1981 also produced a 12-year low that stood for another 12 years.
Who are you on Twitter?
I am known as @ghookermls on twitter
Where did you get the data? The percentages seem very high to me.
Okay, so my friend and I put together the data for the 2012 season, hopefully without too many errors…but the data in that plot seems inaccurate.
Starts by Left-Handed Pitchers By Team:
ARZ – 73, ATL – 41, BAL – 70, BOS – 71, CHC – 51, CHW – 77, CIN – 0 (yes, 0), CLE – 6, COL – 87, DET – 25, HOU – 43, KC – 74, LAA – 35, LAD = 74, MIA – 41, MIL – 26, MIN – 55, NYM – 51, NYY – 40, OAK – 52, PHI – 62, PIT – 42, SDP – 60, SEA – 33, SF – 64, STL – 20, TB – 63, TEX – 65, TOR – 63, WSN – 66
Total starts by left handers = 1530, total number of starts (2 per game): 4860 (162 * 30) – percentage of total starts = 31.48%
Remember that there are two starters per game, so I think what you must have done is divide the number of left handed pitchers that started a game by the number of games, instead of the number of starts. That would give you the 63%.
You don’t have to put together data by team. Go to the PI, select Game Finder Player Pitching, 2012, Throws R, Pitcher’s Role Starter and run the PI. The bottom of the results page shows 3305 starts. Then select Throws L and run the PI, The bottom of the results page shows 1555 games, not the 1530 games in post 4. (1555)/(1555 + 3305) yields 32.00%.
There is still something wrong with the result…still trying to figure it out.
Not all teams necssarily had exactly 162 games that counted in the statistics, although they did in 2012. In 2011, the Nationals and Dodgers only played 161. In 2009, the Cubs and Pirates played 161 while the Twins and Tigers played 163. The year before were six teams at 161 and two at 163 (Twins, ChiSox).
Brian: You missed 18 starts from the Astros and 7 from the Cubs.
Thanks, I didn’t know I could use the PI for that, I stumbled around and couldn’t figure it out. I just knew that there was something wrong with the result, and I believe it was just the factor of 2 for game starts vs. games.
In other words, each data point is off by a factor of 2, to be accurate to the actual percentage of starts made by left-handers.
You are right, of course. I got this data by using a brand new play index feature of grouping by league, which obviously is a bit buggy.