THE BOOK--Playing The Percentages In Baseball

When I worked for the Cardinals, for the draft, we did metal to wood bat “translations” (I’m sure they still do).  To estimate the coefficients, we used data from college and summer league ball (in the summer leagues, they use wooden bats).  We did not adjust for weather or park effects (IOW, we assumed that the weather and parks in both leagues were the same overall).

As you would expect, the power numbers differed drastically (like multiplying the HR rate by 30% I think, and doubles and triples by 60%, or something on that order).  What this told me, was that MLB teams probably overvalued players who have lots of power in college (and high school) and undervalue players who walk a lot.

If I was MLB, I would buy wooden bats for every college in the country and insist that they are not allowed to use metal bats.

I don’t know how many bats that comes out to, say 300 colleges times 20 players times 10 bats each times $100 per bat equals 6 million$ (200,00K per team).  But, it must be worth that much to get a better prospect assessment, wouldn’t it?  Why the heck try to get a metal-to-bat conversion rate, when you can just buy them the bats.

And then do the same for high school, and spend another 20MM for them, though maybe that’s too costly.

There’s another approach other than buying wooden bats for every college. Fund research into a metal bat or a bat made of some sort of composite material that has the same characteristics as a wooden bat when it comes to hitting the ball. I find it hard to believe that it wouldn’t be possible to make such a bat at an affordable price.

The NCAA could probably be induced to adopt such a bat and, if they did, high schools and amateur leagues would probably follow.

I’ll bet you that Louisville will contribute $2MM toward NOT doing the research, if they can supply 60,000 bats every year at a discounted price.

That said, your point is well-taken.  My point is that MLB can get a better appreciation of the talent level of their prospects by spending a mere pittance.

My projection model, which has so far shown fairly accurate results when applied to college hitters who eventually became pros, assumes that each player performs in the same environment - same parks, same pitchers, same bats. Then we moves to a new enviroment, and the difference in performance is noted.

What would screw up the process is if some college hitters use wood, others use metal, and I don’t have that info. I would be measuring a factor for the body of players, not knowing that some individuals fit a different model. Even if I knew which players hit with wood, and which with metal, if they changed during the season it would still mess up the results. So far, I’m working with season level data. I guess we’d need play by play with the type of bat noted.

Many years ago I looked at the stats for one of the summer college leagues I did stats for. Metal bats started arounf 1976-77, and were near universal by the early 1980’s. “After” compared to “before” HR rates were almost doubled.

One complication though - quality of pitching (which goes directly to replacement level). 1978 and before, starting pitchers generally went 7 innings every 4th day from Memorial Day to Aug 1. The top starters went 100-110 innings, 14 to 16 starts, in about 65 days. (In 1956 the champion team only used 3 pitchers all year). There was a very visible pitching arm blowout by the best pitcher in the league in the 1978 playoffs, and the managers were immediately scared off high workloads. Aces generally were held back to one 7 inning game a week. Of course, those innings formerly pitched by the best pitchers needed to be filled by someone, and the quality of pitching just wasn’t that deep. Replacement level pitchers than got a much higher percent of total innings pitched, and batting averages skyrocketed, and this surely accounts for a good portion of the HR arte as well as metal bats.

Does anyone know if the “crack of the bat” sound is different for hickory bats, boned ash bats, ash bats that have had a lacquer sealing and maple bats?

Some comments:

1.  #3:  Such a bat exists, the so-called Baum bat.  See http://www.baumbat.com/.  It hits like wood but is much more durable.  It has never caught on as far as I can tell.

2.  Some rough comments on home run probability.  I looked into this pretty carefully a year or so ago.  Greg supplied me with his entire set of hittrackeronline data from 2007, including how far beyond the fence the ball landed.  From such data, I was able to estimate that home run probability is increased by 4% for each additional foot of distance.  Combine that with my aerodynamic rule of thumb that each mph of batted ball speed is worth about an extra 6 ft on a fly ball.  Finally, it is straightforward to estimate that nonwood bats sanctioned for use in the NCAA outperform wood bats by about 5 mph.  Putting all this together, you find about 30 extra ft on a long fly ball, and more than double the probability of hitting a home run.  The 30 ft agrees well with Greg’s measurement at the Trop.

3.  #6 For a ball-bat collision at the sweet spot, all wood bats should sound pretty much the same.  The frequency distribution of the sound field is largely determined by the collision time and not with the specific species of wood.  For balls hit off the sweet spot, the lowest frequency bending vibration will be excited, and that sound may be a bit different from one wood to another.  For aluminum, the sound field is dominated by the “ping”, which comes from the lowest hoop mode (for a baseball bat, the frequency is in the 2-3 kHz range).  That mode is responsible for the trampoline effect in aluminum bats.  See Dan Russell’s web site (http://www.kettering.edu/~drussell/bats.html) for more about the acoustics of bats.

Alan,

Your mentioning the HR probability made me think of the recent discussion I had with Eric Simon at AmazinAvenue about the Mets new ballpark.  Left field at Citi Field is deeper than Shea by an average of 9-10 feet, with a 16 foot wall vs. 8 feet at Shea; most of right field (all but very close to the line) is deeper at Citi by an average of around 18 feet.  And the Mets are suggesting it will be a bandbox because some Mets players took BP this off-season and hit some homers to left field (never mind asking yourself how deep a fence would have to be for David Wright to NOT hit a homer in BP).

Here’s a link for a comparison of CF to Shea:

http://www.hittrackeronline.com/CF_vs_Shea_comparison.jpg

One more thing I should have mentioned in my earlier posting.  Since 2000, the NCAA has enforced performance standards on non-wood bats, the so-called BESR rule.  It is this rule (along with some other limitations, such as the weight and the moment of inertia) from which I ascertained that the best performing nonwood bats outperform typical wood bats by about 5 mph (or 5%).  The NCAA has recently adopted a new standard based on the ball-bat coefficient of restitution.  When this new standard comes into effect in two years, it will essentially eliminate performance differences between wood and aluminum.  If you do a Google search on the keywords “NCAA” and “BBCOR”, you will find lots of stuff to read on the subject. It is expected that high schools will also adopt this rule.

I can’t wait to see some manager come out to argue that so and so’s bat has too high a COR - sort of like when they come out to check a hockey player’s stick curvature, only for this you’d need a bit more equipment!

Re Greg #10:  I realize you were being facetious in your comment, but let me respond anyway.  One of the big issues with slow-pitch softball bats made of composite materials (carbon fibers and such) is that the COR of the bat (i.e., the trampoline effect) can change with usage.  What happens is that small cracks develop in the material and epoxy that reduce the stiffness of the barrel with repeated use, leading to a greater trampoline effect (and therefore a greater COR).  The Amateur Softball Association has begun to implement a procedure whereby they measure the static stiffness of the barrel for every bat to be used prior to a sanctioned tournament.  If the bat fails, it cannot be used in the tournament.  The USGA has developed a test for golf drivers, the so-called “pendulum test,” whereby the collision time between the clubhead and a rigid cylinder is measured.  See http://www.usga.org/equipment/protocols/pendulum_r2.pdf.  If the clubhead face is too flexible (leading to a greater trampoline effect), the collision time increases.  By measuring the collision time, you therefore measure something they claim is highly correlated with the COR.  If the time is too long, the club cannot be used in the tournament.  The device is completely portable.  There have been efforts to develop a similar device for baseball bats.

Thus far, the NCAA has not worried about these things for baseball bats.  But I suspect they will and they will have to develop their own procedures for policing it.

Sounds like the most practical way to check a bat would be before the game (maybe not every game), have a league official inspect and somehow mark it, and then it would be up to the hp umpire to make sure that a bat being used in a game has been certified for use.