Fangraphs: Ubaldo has been worth +0.6 fWAR this year

[malkusm]

Just playing around with an idea here: For the most part we have batted ball velocity, batted ball angle, and outcome for most balls in play. Thinking about what makes a batted ball more likely to become a hit, it seems as though the following would be true:

- All things equal, higher exit velocity should lead to higher BABIP.

- All things equal, exit angles closer to zero should lead to higher BABIP.

If we accept these (maybe with some adjustment to the second statement, so that an exit angle of +10 degrees is optimal rather than 0 degrees, for example), you could create a pretty simple metric which should correlate to BABIP quite well:

Velocity - ABS(angle)

In this case, a 100 MPH screaming liner would be assigned a value of 100. A 100 MPH chopper hit off of Zach Britton (angle velocity ~-45) would be assigned a value of only 55. A 100 MPH pop-up (angle velocity ~+75) would be assigned a value of only 25. These values seem like they should correlate with how likely each is to be a hit (higher value, greater probability of a hit).

Just spitballing, and you could apply a lot more rigor to something like this, but I think any next-level pitching analysis needs to assign an expected BABIP on each ball put in play in this way, to see how much that varies from true BABIP due to luck and defense.

[Phantom]

Just spitballing, and you could apply a lot more rigor to something like this, but I think any next-level pitching analysis needs to assign an expected BABIP on each ball put in play in this way, to see how much that varies from true BABIP due to luck and defense.

Roto Fangraphs has xBABIP for batters. I think they find that it is less useful for pitchers, since they believe that pitchers have little control over their batted balls to begin with. So even if they are getting lucky on BABIP in comparison to batted balls, they would expect both the regress to the mean, instead of just regressing to the expected outcome based on batted balls.

Would be cool to look at, for sure.

http://www.fangraphs.com/fantasy/new-hitter-xbabip-based-on-bis-batted-ball-data/

[Hallas]

Roto Fangraphs has xBABIP for batters. I think they find that it is less useful for pitchers, since they believe that pitchers have little control over their batted balls to begin with. So even if they are getting lucky on BABIP in comparison to batted balls, they would expect both the regress to the mean, instead of just regressing to the expected outcome based on batted balls.

Would be cool to look at, for sure.

http://www.fangraphs.com/fantasy/new-hitter-xbabip-based-on-bis-batted-ball-data/

This is not true. Ground balls and fly balls allowed have a minimum significant sample of 70 batters faced. In fact, HR/FB, which is indirectly used by fWAR, is way, way less reliable than GB/FB rate.