This article is part of our The Z Files series.
Today we're going to take this sort of analysis to the next level and explain a little how the sausage is made, so you can get a better idea of the projected outcome of these changes. At the end, there will be a link to a downloadable Excel spreadsheet that will allow you to do exactly what is discussed. The focus will be on hitters. Next week we'll do the same for pitchers.
Everyone has their own method of doing player projections. Mine's a little different so consider it learning how to make turkey sausage. The goal is to project home runs, batting average and on base percentage. For the purpose of this discussion, sacrifice bunts and sacrifice flies will be omitted from the calculations since they add another layer that really doesn't change the end result so let's keep it as simple as possible.
Everything is based off plate appearances (PA). The walk and strikeout percentages are determined. The number of at bats (AB) is PA less walks while the number of balls in play are AB minus strikeouts.
Next up are home runs. As alluded to above, home runs are (HR/FB)
Today we're going to take this sort of analysis to the next level and explain a little how the sausage is made, so you can get a better idea of the projected outcome of these changes. At the end, there will be a link to a downloadable Excel spreadsheet that will allow you to do exactly what is discussed. The focus will be on hitters. Next week we'll do the same for pitchers.
Everyone has their own method of doing player projections. Mine's a little different so consider it learning how to make turkey sausage. The goal is to project home runs, batting average and on base percentage. For the purpose of this discussion, sacrifice bunts and sacrifice flies will be omitted from the calculations since they add another layer that really doesn't change the end result so let's keep it as simple as possible.
Everything is based off plate appearances (PA). The walk and strikeout percentages are determined. The number of at bats (AB) is PA less walks while the number of balls in play are AB minus strikeouts.
Next up are home runs. As alluded to above, home runs are (HR/FB) x (# fly balls hit). The league average for HR/FB is about 0.11.
Let's digress a bit and talk batting average on balls in play (BABIP) in general terms. The ensuing treatment is better than looking at it as a single number but falls short of breaking it into the complete array of necessary components if we were truly generating a player projection. We'll break BABIP down into four batted ball types: grounders, fly balls, line drives and bunts. Each has a league average BABIP associated with it.
Next-level analysis breaks out infield pop into a fifth subset and possibly infield grounders into a sixth. Then each of these would be broken into hard hit, medium hit and weak hit balls. At this point we're talking about 18 components to BABIP which is more than we need for the scope of this discussion.
Here's the relevant 2015 data for a league average player. Please keep in mind that this is still collected in a subjective manner so the numbers may vary from source to source. So long as you keep your source consistent, the results generated should be fine. We'll use Fangraphs.
| BATTED BALL TYPE | PERCENT | BABIP |
| Ground ball | 45 | 0.236 |
| Fly ball | 33 | 0.129 |
| Line drive | 20 | 0.678 |
| Bunt | 2 | 0.403 |
Given that a player's own component BABIP will reflect his hard, medium and weak hit percentages for each category, we'll use the league average numbers in this discussion, solely adjusting the hit distribution. The final BABIP is a weighted average of the hit distribution and the component BABIP.
Jumping back to the sausage factory, the final step is determining the number of hits. The equation for BABIP (excluding sacrifice files) is:
BABIP = (Hits – HR) / (AB – HR – strikeouts)
Solving for hits:
Hits = BABIP x (AB – HR – strikeouts) + HR
We now have everything to determine the number of projected homers, batting average and on-base percentage. As indicated, the methods are programmed into an Excel spreadsheet so you can play around with all the inputs to see what happens when you change the variables. Let's take a look at a few players with a questionable element to their projection and see what happens if it is changed.
Carlos Correa, SS, HOU
Last season, Correa hit 22 homers in 432 PA, sporting a 29 percent fly ball rate and 24.2 HR/FB. Let's give him 650 PA and keep everything else the same, resulting in a projection of 33 homers.
The party line is his HR/FB will be really hard to maintain as 24.2 was the eleventh-highest among hitters with at least 300 PA. Let's split the difference between league average and last year's mark and plug in a HR/FB of 17.6. The new projection is 24 homers – still excellent for a sophomore shortstop, though some will question whether Correa can maintain that level.
The other factor regulating homers is fly ball rate. Correa was a little below league average last year. Let's keep the 17.6 HR/FB and raise the fly ball rate to league average. Now we're talking 27 homers.
What if we keep those parameters and cut his strikeouts to 15 percent? The total is now 29 homers.
The problem with Correa is we only have two-thirds of a season of major league data. Other than the first two months of last season, he never displayed this type of pop in the minors. For all we know, his HR/FB will fall even further and he'll hit the same, or fewer fly balls while fanning more. This could easily drop his homer output into the teens.
To wit, using a 13.0 HR/FB, 20 percent strikeout rate and keeping his number of fly balls the same, Correa is now expected to smack just 17 homers. Every single one of these inputs is plausible, meaning no one should be shocked if the youngster hit anywhere between 17 and 29 homers.
Christian Yelich, OF, MIA
If only Yelich could put more balls in the air. Before we get to that, guess how many infield pop-ups he's hit in his career, spanning 1458 PA? He's hit one more than me. That's goofy. Anyway, one reason is Yelich puts 63 percent of batted balls on the ground, lofting only 15 percent fly balls. His HR/FB is a little above average at 12.5 but with the altered dimensions at Marlins Park this season, 15 is reasonable. Two years ago he had 660 PA so let's give him 650. If Yelich can nudge his fly ball rate up to 25, which is still below league average, we're looking at 18 homers. Throw in mid-twenties steals and while Yelich is still a tier below the Mookie Betts, Charlie Blackmon, Starling Marte and A.J. Pollock group in projected value, there's more profit potential relative to cost.
Billy Hamilton, OF, CIN
Coach Lou Brown would make Hamilton drop and give him 20 every time he hits the ball in the air. Last season, Hamilton's fly ball rate was 34 percent. It's not that it's excessively high, it's more that for a player of Hamilton's ilk, he wants to be stroking the ball down and hard.
In 2015, Hamilton's BABIP was .264, below his career mark. If we regress it to .295 and keep everything else the same, his average jumps to .250 with a still-poor .295 OBP. Let's take his fly ball rate to 24 percent, adding the difference to grounders. Now he's at .257 with a .302 OBP. This is still poor, but at least we've inched past the 30 percent plateau.
Hamilton's 16.5 percent strikeout rate wasn't terrible but for a slap hitter, it could get better. Dropping it to 13 percent pushes his average to .267 with a .314 OBP.
What if he became just a little more selective, nosing his walk rate from 6.2 percent to 8.0 percent? Now we're looking at a .268 average and a much more respectable .326 OBP. Is it really out of the realm of possibility for Hamilton to hit 10 percent more grounders, shave 3 percent off his strikeout rate and walk a little more?
Miguel Sano, DH, MIN
In most leagues Sano should qualify in the outfield soon enough, but he'll begin the season with DH/UT eligibility to kick things off. Those down on Sano cite his 35.5 percent strikeout rate and .396 BABIP and scream regression. His proponents point to his hard hit ball data, pointing out he should be able to support an elevated BABIP.
Let's first see what his batted ball distribution portends using league average component BABIPs. Instead of .396, it plummets to .302. But as suggested, he hits the ball hard as evidenced by this table:
| BATTED BALL TYPE | AVERAGE BABIP | SANO BABIP |
| Ground ball | 0.236 | 0.352 |
| Fly ball | 0.129 | 0.212 |
| Line drive | 0.678 | 0.711 |
Even so, there's a good chance Sano's 2015 component BABIPs will fall so let's drop them each by .020. This brings his overall BABIP to around .362 resulting in a .250 average. With his power, you'll take that. Even if each drops .030, we're still looking at a .245 hitter.
The other aspect of Sano's poor contact is that there's plenty of room for growth. Let's say he chops it down to a still-poor 30 percent. With the same .352 BABIP resulting from each component dropping .030, the added contact yields a .272 hitter.
For those wondering, if he maintains the same 25 percent HR/FB mark from last season and gets a modest 620 plate appearances, Sano projects to 37 homers.
Now you can see where the upside comes from with respect to Sano. His BABIP can drop to a still healthy .352 and he can continue to fan at a rate well above league average yet check in with a .272 average and threaten 40 homers.
Below is a link to the Excel spreadsheet where you can play around with these inputs to get a better feel for the upside and downside for players of interest. Next week we'll repeat this exercise for pitchers.
