Ready for the Big Time: Why Jack Custs are so hard to find

In 2007, the Oakland Athletics signed Mike Piazza to be their full-time DH. On May 2, a freak play sprained the AC joint in Piazza’s right shoulder, sidelining him for much of that year. The next day they traded a PTBNL for then 28-year old Jack Cust, considered a 4A journeyman at that point in his career. Originally drafted in 1997 by the Diamondbacks, Cust had been with 5 teams already and granted free agency twice. This was actually his second go-round with the A’s. Seen as a three-outcome hitter, scouts showed disdain at his slow swing and all the strikeouts. He went on to shock Major League Baseball by hitting .256/.408/.504 with 26 HR and 82 RBI, winning the DH job outright. How in the world did Billy Beane see this one coming? And would this work for any hitter in the minors with a lot of power?

Cust had a great 2006 with Triple-A Portland, hitting .293/.467/.549 with 30 HR and 77 RBI, but that didn’t earn him much playing time in the Majors. For April of 2007, his line was .300/.430/.725, but it still didn’t earn him a promotion to the punchless Padres. Given the Pads offensive woes, maybe they should have given him a chance, eh?

Teams don’t take risks on players like Cust because they have had so many experiences where they promote an older player, only to see him stumble on the big stage. The explanation is always, “he’s just not a big league player”, or “the pressure was too much for him”, etc. I posit that a good deal of the explanation lies in key rate stats, which tell us roughly what the player might do at the Major League level (this is a completely different approach than say, the Davenport Translations). In no way do I claim high accuracy, but this is a quick and dirty way to get a Major League equivalent line.

In predicting a batting line, we have to know certain key rate stats. Strikeout rate is critical to batting average, because fewer balls in play means fewer chances to get on base with a hit in the same number of at-bats. HR rate is important because that factors into average and balls in play. Along with the batting average, the walk rate is the other main determinant of OBP. And to calculate slugging percentage, you need to know a basic ISO value. And BABIP is different at the Major League level (defense is overall better), so we have to take that into account as well.

Looking at dozens of prospects over the past few years who moved up from Triple-A to the Majors in the same season (and had at least 150 PA at both levels), I derived these changes:

BB rate: down 29.3%
K rate: up 23.5%
ISO: down 12.4%

I repeated the same process for players who jump from Double-A to the bigs, and got these factors:

BB rate: down 23.5%
K rate: up 26.1%
ISO: down 13.1%

As you might expect, players walk less and strike out more against better pitching. Overall power takes a little hit, but not as much as I thought it would. This is not a perfect method, because players can overperform and underperform these percentages by quite a bit. Taking the average gives a conservative estimate, and in any case, sample size is always a problem with young players who are given limited playing time, so I’m not sure it matters too much. At the very least, this method gives you an idea of who has a shot to be good now.

In order to account for BABIP, I use GB/LD/FB data (available from the wonderful FirstInning.com). In order to account for defense, I figure the number of balls in play for a given number of PA and use the Major League averages for BABIP on each of these types of balls in play:

Ground balls: .237
Line drives: .728
Fly balls: .222

Then you must add HRs and divide by ABs. Thus Oakland’s Wes Bankston, who has roughly a 36/19/45 GB/LD/FB split, 20.8 AB/HR and a 19% K rate (increased 23.5%), is roughly a .278 hitter. His 6% BB rate (adjusted down 29.3%) puts his OBP around .303, and his .216 ISO (adjusted down 12.4%) yields a .467 SLG and .770 OPS. You can adjust PA for estimated playing time, and you get a very nice projection for the coming year.

Calculating using batted ball data helps you differentiate the real contributors from guys that just had a lot of counting stats. The 24-year old Bankston had 20 HR and 73 RBI in 405 PA for Triple-A Sacramento last year. Some people might consider Milwaukee’s Brendan Katin as a comparable. This 25-year old hit 19 HR and had 72 RBI in just 344 PA for Triple-A Nashville, but his projection stands at just .207/.242/.417 because of his extremely high strikeout rate (32%) and lower line drive percentage (16%). Bankston is clearly the better player, and will likely fare much better at the Major League level.

Incidentally, plugging in Cust’s 2006 numbers yields a prediction of .235/.327/.454 from this method. Clearly he has overperformed those numbers, but a .781 OPS does suggest that he was worth a shot for a team in need of a cheap bat.

Using this methodology, it is possible to predict breakout players who are at the cusp of the Major Leagues.