Handicapping the starting pitchers is the fundamental approach to betting on baseball. There is no player with a bigger effect on the game than the starting pitcher because of what is expected of him. Not only does the pitcher have to be effective to keep his team in the game, but he also has to pitch enough innings to spare the bullpen from a heavy workload.
Starting pitchers have always been defined by wins and losses and earned run average. ERA is a simple formula for everybody to understand and a widely-accepted way to evaluate pitchers. Take the pitcher’s number of earned runs allowed, multiply by nine, and divide by the number of innings pitched. Simple and straightforward, right?
In today’s era of advanced statistics and in-depth baseball study, we’re finding out more and more of why ERA is an overrated statistic. Pitchers are not to blame for poor defense. There are only a certain number of things that pitchers can control and everything else is just a variable. Traditionalists can make arguments like, “If the pitcher didn’t miss his spot, that wouldn’t have gone for a hit!” That’s merely conjecture. We can go back and look at video to see if a fielder should have gotten to a ball or we can use other players to determine whether or not that ball should have been fielded.
There are acceptable ranges for ERA, where average in the American League tends to be slightly above 4.00 in the current state of baseball with steroids largely out of the game. In the National League, because of the pitcher and more defense-first, bat-second types of players, the average has dipped well below 4.00 over the last few seasons.
Advanced statistics provide a wonderful set of tools for handicappers. Like ERA, most advanced statistics have a range or an average that players will eventually fall into. Any large deviations outside of the range tend to signal regression or progression. One of the most popular sabermetric stats is FIP, or fielder-independent pitching. Pitchers cannot control what happens after the ball hits the bat. It is believed, however, that they can control walks, strikeouts, home runs, and hit by pitches. A variation of FIP is xFIP, or expected FIP. This is a statistic that assumes a league average home run rate based on the number of fly balls allowed by a pitcher.
To illustrate xFIP, consider the National League HR/FB% (home run per fly ball percentage), which is 10.9 percent this season. Say a pitcher has allowed 60 fly balls. His expected number of home runs allowed is 6.54. If he has allowed 12 home runs for a 20 percent HR/FB%, the predictive value of xFIP would suggest that his home run rate will regress towards the mean.
In terms of predictive stats for pitchers, xFIP and SIERA, which stands for skill-interactive ERA, are two of the best. SIERA takes into account batted ball types (ground balls, fly balls, line drives) as well as walks and strikeouts. The difference between ERA and these advanced statistics is that the newer metrics have complicated formulas, so the common fan or handicapper doesn’t pay much attention to them because they’re difficult to understand.
With sportsbooks getting sharper and sharper, handicappers are only hurting themselves if they don’t use these statistics. They’re not infallible, but they should provide a better indicator of future performance if applied with the right context. The idea is that pitcher ERA should migrate closer to FIP, xFIP, and SIERA because they are truer evaluations of how a pitcher is pitching. That’s not always the case because pitchers may be stuck on a terrible defensive team.
Let’s apply these statistics to some pitchers to keep an eye on over the next few weeks:
Tom Koehler (MIA): Tom Koehler looks like a good bet, right? The Miami Marlins are playing better than most people expected. The Marlins are 5-4 in Koehler’s starts, so he’s on the plus side for bettors. People will look at Koehler as an underdog and see a pitcher that is 5-3 with a 2.25 ERA and think, “Wow! What a great value!”
Not so fast. Take a look at Koehler’s advanced metrics. He has a 2.25 ERA, but a 4.12 FIP, a 4.53 xFIP, and a 4.65 SIERA. What does that mean in layman’s terms?
Koehler has struck out 37 batters and has walked 24 batters in his 56 innings. That’s not a good ratio, since the National League average for starting pitchers is a ratio of 2.64. Koehler’s is 1.54.
Next, we look at Koehler’s batting average on balls in play, or BABIP. BABIP is how we evaluate luck and defensive prowess for a given pitcher. Most pitchers wind up with a BABIP between .270 and .290. The difference between BABIP and batting average is that strikeouts and home runs are not balls in play, so they are eliminated from BABIP. Obviously, batting average against is calculated with (Hits / At Bats), so BABIP is ((Hits – Home Runs) / (At Bats – Home Runs – Strikeouts)). In general, pitchers tend to fall into a personal range, which is usually within the league-wide range.
With Koehler, we have a small sample size. However, with a pitcher that strikes out only 16.2 percent of opposing hitters, which is below league average, he has a higher number of balls in play. So far this season, only 21.3 percent of balls in play are going for hits.
With the caveat that Koehler’s home run rate at home should remain lower than expected because of the park conditions, we can deduce that Koehler has gotten very lucky to this point and regression is coming. His true talent level is nowhere near a 2.25 ERA.
Mark Buehrle (TOR): Mark Buehrle has been a great story this season. At 35-years-old, the crafty lefty has put together one of his best starts to a season in a long time. He is 7-1 with a 2.11 ERA and has been the third most profitable pitcher for bettors so far this season. The Blue Jays are 8-1 in his nine starts and $100 bettors are up $735 in Buehrle’s outings.
Again, a cursory glance of Buehrle’s season will have bettors salivating. 7-1. 2.11 ERA. Up 7.35 units. What’s not to like?!
Look deeper. Buehrle has a 2.11 ERA, but a 3.12 FIP, a 4.15 xFIP, and a 4.43 SIERA. For some context, Buehrle’s career SIERA is 4.39. His career xFIP is 4.20. Remember what it says above about xFIP and SIERA being the two most accurate predictors of future pitcher results. What Buehrle’s 2014 xFIP and SIERA numbers tell us is that he is not actually pitching any better than he has over the span of his career.
His HR/FB% is anemically low at 1.8 percent. His career average is 9.8 percent. The American League average this season is 9.7 percent. Keep in mind that league average takes into account a very large sample size of pitchers from extreme ground ball pitchers to extreme fly ball pitchers.
Keep this in mind the next time you rush to back Buehrle.
Andre Rienzo (CWS): The league’s first Brazilian-born pitcher is enjoying quite a season for the White Sox. He’s 4-0 with a 4.00 ERA and has gone a perfect six-for-six for bettors this season. In those six wins for bettors, $100 players are up 8.9 units on Rienzo. A 4.00 ERA doesn’t signal the type of regression that Koehler and Buehrle are showing, however, the Rienzo money train will be coming off the tracks soon.
Rienzo has one relief appearance, which will be factored in and actually helps his cause because it was a scoreless appearance. In seven appearances and six starts, Rienzo has a 4.00 ERA, but a 5.34 FIP, a 4.82 xFIP, and a 4.75 SIERA. In 36 innings, Rienzo has recorded 25 strikeouts and 16 walks. That’s a 1.56 strikeout-to-walk ratio. The American League average is 2.35.
Rienzo, who strikes out only 16 percent of opposing batters (league average for starters is 19.3 percent), has allowed just 23 hits on 107 balls in play, for a .213 BABIP. Remember that league average is between .270-.290. High strikeout pitchers tend to have better BABIPs because they induce weak contact because of their stuff. Rienzo is a below average pitcher, so there’s no rhyme or reason for a low BABIP other than sheer dumb luck. Also, the White Sox defense is among the worst in the league in defensive runs saved.
Rienzo’s walk rate of 10.3 percent is also worse than league average. League average is 7.9 percent. In every start, Rienzo is striking out 0.825 fewer batters and walking about 0.6 more batters than average given an average of 25 batters faced per game. That may not seem significant, but over the course of the season, it is.
Rienzo has been the benefactor of run support in almost every start this season as the offense has given him more than six runs per game. That’s merely coincidental and highly unlikely to continue.
Just like there are pitchers to avoid and look for regression from, there are some pitchers that are just experiencing a lot of bad luck and should improve drastically.
David Price (TB): There aren’t a lot of chances to get value on David Price, but there could be some chances soon. Price is just 4-4 with a 4.28 ERA. The Rays are 6-4 in his 10 starts, which is on the minus side for bettors. Price’s 4.28 ERA is not at all indicative of his skill level. His FIP is 3.21, his xFIP is 2.57, and his SIERA is 2.57. Those are actually much better than Price’s career averages for xFIP and SIERA, so the argument could be made that Price is pitching better than he ever has.
The problems for Price have been home runs and balls in play. Price has allowed 11 home runs in 69 innings for a HR/FB% of 14.7 percent. His career HR/FB% is 9.5 percent, or right around league average. His BABIP is a staggering .345. His batted ball types haven’t deviated much from last season or his career numbers. His career BABIP is .284.
Expect Price to have better results for bettors very soon.
Ian Kennedy (SD): Out of 205 pitchers, Ian Kennedy ranks 198th for bettors this season. The Padres are 3-7 in his 10 starts and $100 bettors are down $568 backing Kennedy. Surprisingly, Kennedy is having one of the best seasons of his career thus far. He’s 2-6 with a 3.79 ERA, but his FIP is 2.76, his xFIP is 2.90, and his SIERA is 2.93.
The main problem for Kennedy over the course of his career has been the home run. Pitching in the National League West with a home stadium of Petco Park is going to go a long way in curbing his high home run totals. Under the tutelage of Manager Bud Black and pitching coach Darren Balsley, Kennedy has seen a rise in strikeouts, a drop in walks, and an increase in ground balls. Kennedy’s .321 BABIP is 35 points above his career average.
In 10 starts, Kennedy has only received 17 runs of support. With the way he’s pitching, his fortunes are going to turn around very soon and he’s a guy that bettors should have on their radar.
Regression and progression are not guarantees. Like all forms of analysis that handicappers use, sabermetric stats should be applied with the proper context and considered prior to making a bet. At its core, handicapping is nothing more than making predictions based on the information at a bettor’s disposal. Not using sabermetrics because they’re hard to understand or untraditional is only hurting a bettor. Oddsmakers probably don’t apply them as much as they could, if they apply them at all, and bettors should be using them to gain any edge possible. Take some time to understand the concepts and put them to use and you should see positive results.