Heath Bell and Park Effects
When Carson Cistulli asked the collective crowd of FanGraphs readers what they thought Heath Bell would receive as a free agent this winter, they picked out 2.8 years and an annual average value of $9.6 million. Rounding up, we’d get an expected contract of three years and just shy of $30 million, and yesterday, Dan Hayes reported that’s almost exactly what he’s looking for.
Bell is hoping that his resume will earn him three guaranteed seasons, and given what Jonathan Papelbon just got from Philadelphia, it seems like the market for relief pitchers is still going very strong. Given Bell’s success in the ninth inning, 3/30 might strike some as a relative bargain compared to what Papelbon signed for and Ryan Madson is reportedly seeking.
However, before any team ponies up for Bell’s services, I’d hope that they’d ask themselves just how much of Bell’s success will follow him if he leaves San Diego.
It’s no secret that Petco Park is the best place in baseball to pitch, and specifically, that it is almost impossible for left-handed batters to hit home runs there. After all, park effects aren’t a new or controversial subject, and there’s not much subjectivity in the data – pitchers who ply their trade in San Diego post numbers that they simply can’t post in other cities. Their numbers – especially in the area of home run prevention – are artificially deflated by the environment in which they pitch half of their games.
Bell is one of the signature beneficiares of how the park plays. In his career, he’s faced 791 batters in San Diego – 10 of them have managed to hit the ball over the wall, one for every 79.1 batters he faced. Away from the friendly confines, he has faced 1,182 batters and allowed 20 home runs, one for every 59.1 batters that have come up to bat against him. His home run prevention has been 34 percent better in San Diego than in all other ballparks, which of course makes perfect sense, given that he’s a right-handed pitcher and fly balls to right field in Petco have almost no chance of reaching the seats.
The park hasn’t just deflated his home run rate either – his career BABIP in San Diego is just .269, but his combined average against on balls in play in all other parks is .334. His BABIP away from San Diego is likely higher than his true talent level, and I wouldn’t suggest that teams should expect Bell to become eminently hittable upon signing with a new team, but the evidence shows that Bell has never been able to perform well on balls in play in any stadium besides Petco Park. At the minimum, that has to be concerning.
If Bell had been able to maintain his peak walk and strikeout rates in 2011, perhaps these tidbits would just be footnotes, but his K% fell from 30.0% in 2010 to 19.9% in 2011. It wasn’t just that Bell decided to throw more strikes, either – his contact rate jumped from 73.6% to 81.8%. Relievers are prone to variance in all status due to the small quantity of innings that they throw, so Bell’s regression may prove to be more outlier than sign of drastic decline, but it can’t be considered a good sign for his future that the one skill that should travel best took a big step in the wrong direction last year.
Bell’s never been a great command guy, but his combination of high strikeout rates and minuscule home run rates have more than made up for his propensity to issue a few walks. Unless a team also figures out how to sign Petco Park simultaneously, his home run prowess almost certainly won’t follow him to a new city, and so the strikeout rate would have to return to elite levels for Bell to maintain his status as a top-notch closer. It’s possible that his downturn in strikeout rate last year was a fluke, but is it really a good idea to bet a three year deal on a 34-year-old reliever hoping that his most recent K% isn’t a sign that he might be on the downside of his career?
Instead of giving Bell a multi-year deal and hoping that all the red flags turn out to not be serious problems, teams looking for a closer might be wise to check in on the budget model of this type of pitcher – Frank Francisco. Here are his numbers stacked up next to Bell’s over the last three seasons.
| Name | BB% | K% | HR/9 | BABIP | LOB% | ERA- | FIP- | xFIP- |
|---|---|---|---|---|---|---|---|---|
| Bell | 8.9% | 26.3% | 0.36 | 0.293 | 79.10% | 64 | 70 | 79 |
| Francisco | 7.9% | 26.5% | 1.06 | 0.299 | 76.00% | 85 | 79 | 79 |
Francisco’s strikeout rate is the same (just without the same dramatic decline in 2011) and his walk rate is slightly better, but his results have been a lot worse due to the dramatic differences in home run rate. Of course, Francisco has spent his career in Texas and Toronto – the anti-Petcos, if you will – and so inflated home run rates are to be somewhat expected. Francisco is also two years younger and, since he doesn’t come with the shiny proven closer label, will likely have to settle for another one year contract like he did last winter.
Even if it took a second year, the price for Francisco is going to be a fraction of what it will be for Bell, and it’s not clear that any team choosing between the two should expect to get a dramatically better performance from one than the other.
With reliever valuations continuing to reach levels that are simply hard to justify, it just doesn’t make much sense to bet on a guy whose success has been directly attributable to pitching in the best possible environment for his particular skillset. Rather than spending money for what Petco helped Bell appear to be, a smart team could likely get a similar performance from Francisco for a fraction of the cost.
Park effects aren’t anything new, but they look to be one area where teams can exploit inefficiencies in the market. And, with relievers available from both extremes of the park effects landscape, adjusting for context could save a Major League team a lot of money this winter.
Well done again, Dave. The Petco Park effect is a legitimate concern, and the numbers back it up.
But here’s a counter argument. Heath Bell is very good at keeping the ball away from hitters. Bell doesn’t throw inside much at all, and so he does a good job avoiding the middle-in area of the zone. Those are the pitches that are most likely to wind up in the bleachers. And Bell generally pitches the same way against righties and lefties, at least the location aspect of pitching – he keeps the ball away. This skill travels well. So Bell may not be as much of a product of Petco Park as the home/road splits indicate. In other words, I don’t think Petco is necessarily bailing him out. I could be wrong, because I haven’t looked at how often Bell surrenders flyball outs to the deepest parts of the ballpark – a high propensity would indicate that he is getting bailed out by the park effects. But pitchers who consistently keep the ball away from hitters will generally give up fewer home runs and other extra base hits regardless of the park, whereas the pitchers who prefer to attack the inside corner will get hurt more often when they miss and their pitches sail over the heart of the plate.
Just a thought to stir up some debate. But I agree that it is unlikely Heath Bell will have the same measure of success if he goes elsewhere. The Giants, Rays, and A’s would not be interested in what Heath Bell is looking for, so that eliminates three really good antihomer parks. Perhaps Anaheim, Seattle, or the CitiField would be good landing places?
While I agree with your main point in comparing Bell and Francisco, note that Bell has pitched 64-68 games during each of the past three seasons while Francisco has pitched only 51-56 ganes. In terms of IP, the difference is about 10-20 IP per year. Part of their likely difference in 2012 compensation may be attributed to the perception that Francisco is less durable than Bell is. Your comparison of rate stats only overlooks this difference.
With that said, for half the price, I’d take Francisco…
Don’t managers decide how long to keep in RPs?
In this case, it isn’t a matter of managerial discretion. Bell stays healthy and Francisco does not. According to the injury data base at BaseballProspectus.com, during the last three years, Bell missed one game due to injury. Meanwhile, Francisco has had 5 trips to the disabled list during the past three years plus missed another 30-game stretch during spring training 2011. There is a big difference in their past durability and it’s a factor that likely will affect their 2012 compensation.
I understand the difference in HR rates, and that Petco is unique in this respect (so only two true outcomes, really). But what about BABIP? Is this really a park effect, say, a slow infield? Might there be other parks where his BABIP would be low?
I’m not sure if Petco’s infield is slower than most, but part of the BABIP difference is likely due to the some of the same park effects that cause fewer HRs. The marine layer causes fly balls to hang up and not clear the fence. It also allows the outfielders more time to get into position and make catches on balls that might have dropped in for a hit at other parks.
I agree. Nice post Dave. I would have never guessed that Bell and Francisco had such similar profiles. The only thing I’d say about Francisco is that he was feast or famine a lot last year. He had a lot of outings where he’d be asked to get four or five outs (he was splitting saves with Rauch early on), come in and K four of the batters and look amazing. The next night he come, go 3-0 against the first batter faced, and you could just feel the five-run meltdown was brewing. I could be wrong, but I had him on my fantasy team and that’s how his year felt to me. I’m sure there are plenty of Jays fans who could confirm or deny my observations.
As a Padres fan I have been making the case for trading Heath Bell all season (with the hope that Mike Adams would take over before he got traded). Now that he is a free agent I would hat for him to accept anything with us again. Having watched Bell last season you could tell he wasn’t as good as his 2010 numbers. He very rarely seemed to have a 1, 2, 3 inning and it was always heart in mouth time when he came in with runners on base. He’s not worth $10 million a season and a three year contract is just asking for trouble given his declining numbers…
Dave, you raise a good point about his BABIP split. It will be interesting to see if that follows him should he sign with a new team. However, I feel like your portrayal of his HR split is dubious.
Since joining the Padres in 2007, Bell has allowed 10 HRs in Petco (0.43 HR/9) and only 6 HRs on the road (0.31 HR/9). There doesn’t appear to be a discrenable pattern within those five years. So, that leaves me wondering how much credit should be given to Bell working with Darren Balsley instead of Bell working at Petco.
What is your justification for including his numbers with the Mets, and how that represents what we can expect from Bell in 2012 and beyond?
this. dave’s analysis is wrong. the data he uses only tells something we already knew – that bell was bad before coming to the padres.
The Babip thing is weird. I guess it suggests……..I don’t know what it suggests. I do know this. I would not touch a PADRE reliever for 3 years 30 million. The Red Sox are gonna be sorry if they plug him in for Pap.
Bell’s splits this year we’re not so bad at all, still a 2.88 ERA on the road, what’s so bad about that?
You make the good argument that similar results can be had for less money, but when isn’t that true?
The point, which is not mentioned, is why does a payroll-restricted club, that is at least 3 years away from competing, want a 34 year old, high priced closer?
Closers are not necessary unless you think you’re team can go to the playoffs. Think about this, their value is the hardest in all of baseball to defend because it’s based on such a small sample. Not just the 50-70 innings they throw, but the less than 10 innings they MIGHT get to throw if you make the playoffs. For the regular season, it simply isn’t that important to have a star closer. But the playoffs, because of the pressure, is a different animal. This is why Mariano Rivera really is worth $15 mil a year, because of his other-wordly performance doing the most stressful job in baseball, closing payoff games for the Yankees. A player who closes games for the Padres can’t possibly be worth even half that, because he’ll never get a chance to perform in a situation where he’ll be able to earn his paycheck.
Let Bell go to the Red Sox, wake up Padres.
Pitchers also learn to pitch to their home parks. Some of the park effects, I’m sure, are real, but some of Bell’s effectiveness at home was simply his familiarity with the confines vs. lack of familiarity with road parks. When he gets a new home, he will become familiar with that park, a natural effect of playing half his games there.
Late to the party here, but I just have to point out the error in the statement that Bell’s home HR rate is 34% better than his road rate.
The difference between the two rates, which is what I think the reader (and probably the author) is after, is 14%. Still statistically significant, I’ll assume, but a MUCH smaller effect size than what the author implies.