THT Mailbag: Rounding Into Shapeby Bryan Tsao
February 15, 2007
It's been kind of an exciting week here, though not for the usual reasons—we've had some continuing issues with our content management system/database, which has led to some intermittent site outages, including a major one on Wednesday that shut the site down for a couple hours. Dave Studeman, along with Bryan Donovan and the rest of our technical staff have been looking into the problem, with an assist from an old friend, and as of yesterday the issues have (hopefully) been resolved.
We apologize for any unforseen boredom this may have caused at work, and rest assured we look forward to providing our baseball content in the same, consistent manner you've come to expect in the future.
Incidentally, I particularly enjoyed the questions this week ... you guys are starting to shake off the offseason rust too.
Now, onto the questions.
Sorry, but did you just not only leave Sandy Koufax off your list... but also rate him behind Don Sutton, Dennis Eckersley, Phil Niekro (?!), Gaylord Perry (??!!), Bert Blyleven (???!!!), and Nolan (.524 WP, 292 losses) Ryan? And that's just for starters...
When are you bringing your stand-up act to Vegas?
- John D., Arcata, Calif.
Dave Studeman: Good question about Koufax, if a bit heavy on the sarcasm. As I said at the top of the article (and the previous three articles in the series), my ratings were a simple metric-based rating of pitchers by all-time WSAB. I didn't evaluate the stats any further, nor did I try to supplement the numbers with something else. I was curious about WSAB and what it said about pitchers and, based on the emails I received, so were a lot of readers.
In virtually any metric that emphasizes career contributions over a relatively low baseline, Koufax won't rank as highly as most of the other pitchers you mentioned. For a more thorough and valid comparison of pitchers, you might want to read David Gassko's article in this year's THT Annual, as well as many other lists such as Bill James' in both Historical Abstracts.
I just read through David Gassko's article on hitting pitchers, and was wondering if you folks would be willing on doing a little legwork to investigate an interesting question. When addressing pitching hitters, David says that chances are a hitting pitcher won't make up in offense what he loses in pitching, which I think would probably be the case, but when looked at from a Moneyball standpoint, could it be worth it?
To put it more simply, with the huge premium paid for pitching (even mediocre starting pitching), would a National League club come out ahead paying for recent batters due to the reduced payroll and added flexibility it provides?
David Gassko: Probably not. The fact is, a replacement-level hitter would be around 10 runs above the average pitcher with his bat. On the flip-side, that would give him about half-a-run a game of leeway on the mound, meaning that if he was only half-a-run per game worse than a replacement level pitcher, overall he would have some value. Do you think that a hitter could put up an ERA only half-a-run worse than some scrub or random Triple-A pitcher? I don't know for sure, but I tend to doubt it.
There's also the issue of there being a talent ceiling. Even Albert Pujols wouldn't have that much value with his bat if he were a pitcher, because he wouldn't even come to the plate 100 times. On the other hand, a great pitcher can have a lot of value because he can face close to 1,000 batters a year. So the potential impact is 10 or more times greater. In other words, if you want something better than replacement level or even average from your pitcher, he's not going to get there no matter how good he is with the bat; only a good arm will make him truly valuable.
It's a fun idea to think about, and it is possible that it could have some value, but my feeling is that sending a hitter out on the mound regularly would be too a extreme an experiment, with too small a potential payoff to justify it.
Bryan Tsao: For some additional insight into the value of hitting pitchers (or pitching hitters), check out this article on Brooks Kieschnick by Dave Studeman. Basically, through August of 2004, Kieschnick was able to rack up an extra win's worth of Win Shares via his hitting. Kieschnick struggled with the bat down the stretch after getting injured, which brought down his season totals, but it's not inconceivable that a better hitting pitcher could be worth an extra win or so a season, which nowadays costs in the ballpark of $4 million a year on the free agent market.
As the Ball Flies
As far as I can tell, the current ball trajectory data is fairly rudimentary, dealing with only four basic categories: fly balls, ground balls, line drives, and pop ups. It can be a judgment call classifying a batted ball as low fly ball to high line drive, short fly ball to pop up, or even hard ground ball/very low line drive. Based on that classification, we "know" the value of that batted ball.
Is anyone developing a more sophisticated system that calculates batted ball velocity/angle, and then computing approximately where (given no/little wind) that batted ball would land with a given batted ball velocity/angle?
David Gassko: Greg Rybarczyk at Hit Tracker keeps track of all that information for every home run hit in the major leagues. It's an amazing resource, and Greg is planning on tracking every batted ball for a few teams (the Red Sox, Yankees, and Cardinals) in 2007. Of course, that's a lot of work and Greg is looking for volunteers to help him out with this project. If you (or any reader) wants help, I encourage you to contact him through his website.
Also, you can look forward to an article that will analyze the Hit Tracker data in The Hardball Times 2007 Preseason Book. We'll be revealing a lot more about the book in the next couple of weeks.
John Beamer: Further to David's answer, I am in the process of interviewing Greg about exactly how Hit Tracker works and what plans he has for the site in the future. Be sure to look out for that as the new season starts.
Keep It Real
Have you done any analysis that shows which players are the most consistent throughout the season, by category? For example, a list of all hitters home run performance for each month of the season (April to September)? Or a list of pitcher's ERAs for each month?
I'm doing research for fantasy baseball and I'm working on the logical assumption that the key to a successful team isn't to stockpile it with players who have great stats by the end of September, but to stockpile it with players who have great and consistent stats every month. The problem is finding a source that has already compiled monthly stats for every player for the following categories (hitting: HR, R, RBI, SB, AVG; pitching: W, L, ERA, WHIP, K).
Can Hardball Times help point me in the right direction?
- Paul B.
David Gassko: Though to my knowledge there has never been a study specifically addressing player consistency, there have been studies of related issues, such as The Book's examination of whether hot-and-cold streaks truly exist (not really) or a study by Keith Woolner, in which Woolner examined if some players truly were warm weather hitters (they aren't).
What these studies point to is that day-to-day inconsistencies in a player's statistics are caused by random variation, and not some kind of ability to play consistently. Moreover, it's not clear that a consistent player is necessarily better for winning games (or head-to-head competitions in fantasy baseball). In other words, when you're drafting a team, don't worry about consistency; all you should care about is a player's quality.
Bryan Tsao: To answer your ultimate question in more detail, it's hard enough to predict what a player will do over the large sample of an entire season—predicting what one will do over the course of a month or a game is basically folly.
Obviously, the big question is whether you're playing rotisserie or head-to-head. Clearly, in a roto league the goal is exactly to have the best stats by the end of September, but, as David says, having the best players is pretty much a winning strategy no matter what type of league you're in. The main reason I don't think that consistency is that big a deal even in head-to-head leagues is that typically you'll have maybe nine to twelve hitters start for you.
So even if you have inconsistent hitters, the odds that they will all be slumping at once, for an entire week, is pretty slim; you'd be much better off just going for the best players, because on average, nine to twelve players who are better than their counterparts have a good chance of accumulating more fantasy points.
Bryan Tsao is the editor of The Hardball Times website. He welcomes comments, questions, and suggestions for both himself and the site via email.