Post by ericmvan on May 23, 2016 12:17:11 GMT -5
I created earlier versions of these metrics for the Sox. I used different data to re-create them.
T-Set (Table Set) measures how many RBI opportunities you provide guys hitting after you, if they were average hitters. It subtracts the opportunities you remove by knocking them in yourself. I may tweak it later to include running the bases, but right now it's just based on a standard batting line.
Kn-In (Knock-In) measures your ability to knock your teammates in. It's pretty much (H + 2B + 1.8 * (3B + HR)) / PA. (In fact, the coefficient for 3B is a little higher than that of HR, because unlike HRs, there's no bias in favor of solo 3Bs.)
Both are measured in R per 10 PA (because otherwise they all almost all start with .0). Brief derivation geekage notes later, if anyone asks.
It's important to realize that T-Set measures excess table-setting, the difference in opportunities between the start and end of your PA. If you hit enough homers versus everything else, you could be negative. That's why T-Set numbers are lower than Kn-In.
T-Set rankings 2015-2016:
But the 2016 ranking is rather different:
Here's Kn-In for 2015-16:
And for 2016:
Now, the really interesting thing you can do is take the Ratio of T-Set to Kn-In. That tells you the role the player should have. In fact, you can start your lineup construction by trying to simultaneously go from your best hitter to your weakest, while also running from the highest Ratio to lowest, with the tweak that you want a high-ratio hitter 9th (and maybe even 8th) to turn the lineup over.
2015-16 Ratio. I've included TAv, which includes AAA for Shaw and Bradley last year, as translated by Clay Davenport.
Holt is an obvious perfect 9 hitter. Ortiz and Bradley have been your two best hitters (and yes, that's a full 2015 for JBJ), so they go 2 and 4. Betts and Bogaerts have been your 3rd and 4th best hitters, and Xander's higher ratio makes him the leadoff guy and Mookie the 3 hitter. The three next best hitters have been Pedroia, Shaw, and Ramirez in that order, which is also their ratio order and a continuation of your R-L-R thing. However, Hanley is not really that hitter, and would probably not like hitting 7th. So let's look at our small 2016 sample.
Let's put Shaw 7 despite the numbers, in part because if we ran just road numbers, that's where he'd end up, and in part because he's the one guy who won't say, WTF, I'm hitting 7th? Xander is still 1, JBJ is still 2 and Ortiz is still 4, and Holt is still 9. (Actually, Hanigan's Ratio is even higher, and Holt 8 and Hanigan 9 does work.)
But with these numbers, you'd want Hanley 3rd and Mookie 5th, with Pedroia 6th.
I'm pretty convinced that Mookie should be 3 or 5 and that Pedroia should, at some point, be transitioned into 5 or 6. I do worry about him swinging at everything on the road if he's hitting 6, though, and I kind of like him hitting 5th ... after facing X, JBJ, Mookie or Hanley, and Ortiz, now you have to deal with this pest? Getting the actual four best hitters 1 through 4, when they're this good, can exhaust a pitcher, and a guy with very strong contact skills seems like a good fit for 5.
I don't think anyone can say yet where Hanley ultimately belongs.
T-Set (Table Set) measures how many RBI opportunities you provide guys hitting after you, if they were average hitters. It subtracts the opportunities you remove by knocking them in yourself. I may tweak it later to include running the bases, but right now it's just based on a standard batting line.
Kn-In (Knock-In) measures your ability to knock your teammates in. It's pretty much (H + 2B + 1.8 * (3B + HR)) / PA. (In fact, the coefficient for 3B is a little higher than that of HR, because unlike HRs, there's no bias in favor of solo 3Bs.)
Both are measured in R per 10 PA (because otherwise they all almost all start with .0). Brief derivation geekage notes later, if anyone asks.
It's important to realize that T-Set measures excess table-setting, the difference in opportunities between the start and end of your PA. If you hit enough homers versus everything else, you could be negative. That's why T-Set numbers are lower than Kn-In.
T-Set rankings 2015-2016:
Name TS
Xander Bogaerts .558
Brock Holt .554
Mookie Betts .513
Jackie Bradley .498
Dustin Pedroia .464
Travis Shaw .440
AL Average 2015 .422
David Ortiz .393
Hanley Ramirez .373
But the 2016 ranking is rather different:
Name TS
Xander Bogaerts .606
Jackie Bradley .569
Travis Shaw .551
Hanley Ramirez .540
David Ortiz .475
Mookie Betts .460
Brock Holt .439
Dustin Pedroia .430
AL Average 2015 .422
Here's Kn-In for 2015-16:
Name KI
David Ortiz .870
Jackie Bradley .860
Mookie Betts .823
Travis Shaw .820
Xander Bogaerts .795
Dustin Pedroia .744
Hanley Ramirez .724
AL Average 2015 .666
Brock Holt .650
And for 2016:
Name KI
David Ortiz 1.056
Jackie Bradley .988
Travis Shaw .885
Xander Bogaerts .868
Mookie Betts .847
Hanley Ramirez .785
Dustin Pedroia .771
AL Average 2015 .666
Brock Holt .568
Now, the really interesting thing you can do is take the Ratio of T-Set to Kn-In. That tells you the role the player should have. In fact, you can start your lineup construction by trying to simultaneously go from your best hitter to your weakest, while also running from the highest Ratio to lowest, with the tweak that you want a high-ratio hitter 9th (and maybe even 8th) to turn the lineup over.
2015-16 Ratio. I've included TAv, which includes AAA for Shaw and Bradley last year, as translated by Clay Davenport.
Name Rat TAv
Brock Holt .851 .253
Xander Bogaerts .702 .275
Dustin Pedroia .624 .269
Mookie Betts .623 .289
Jackie Bradley .579 .295
Travis Shaw .537 .263
Hanley Ramirez .515 .261
David Ortiz .451 .312
Holt is an obvious perfect 9 hitter. Ortiz and Bradley have been your two best hitters (and yes, that's a full 2015 for JBJ), so they go 2 and 4. Betts and Bogaerts have been your 3rd and 4th best hitters, and Xander's higher ratio makes him the leadoff guy and Mookie the 3 hitter. The three next best hitters have been Pedroia, Shaw, and Ramirez in that order, which is also their ratio order and a continuation of your R-L-R thing. However, Hanley is not really that hitter, and would probably not like hitting 7th. So let's look at our small 2016 sample.
Name Rat TAv
Brock Holt .774 .239
Xander Bogaerts .698 .304
Hanley Ramirez .688 .282
Travis Shaw .622 .300
Jackie Bradley .576 .333
Dustin Pedroia .558 .263
Mookie Betts .544 .283
David Ortiz .450 .341
Let's put Shaw 7 despite the numbers, in part because if we ran just road numbers, that's where he'd end up, and in part because he's the one guy who won't say, WTF, I'm hitting 7th? Xander is still 1, JBJ is still 2 and Ortiz is still 4, and Holt is still 9. (Actually, Hanigan's Ratio is even higher, and Holt 8 and Hanigan 9 does work.)
But with these numbers, you'd want Hanley 3rd and Mookie 5th, with Pedroia 6th.
I'm pretty convinced that Mookie should be 3 or 5 and that Pedroia should, at some point, be transitioned into 5 or 6. I do worry about him swinging at everything on the road if he's hitting 6, though, and I kind of like him hitting 5th ... after facing X, JBJ, Mookie or Hanley, and Ortiz, now you have to deal with this pest? Getting the actual four best hitters 1 through 4, when they're this good, can exhaust a pitcher, and a guy with very strong contact skills seems like a good fit for 5.
I don't think anyone can say yet where Hanley ultimately belongs.