Phranklinby Craig Burley
January 16, 2006
A number of Phillies fans clutched at their chests last week when they opened their morning papers (or their morning e-mail) and saw that the Phils had signed righthanded pitcher Ryan Franklin to a $2.6 million contract. Franklin is a longtime mediocrity, who had perhaps the lowest profile of any veteran starting pitcher in baseball until last season, when he was suspended for 10 games under the MLB drug policy for a steroid infraction. Franklin's turn-ons are long walks on the beach, syringes, and home runs allowed, which is why he managed to go 23-44 and surrender 95 dingers in the last three years in the pitcher-friendly confines of Seattle's Safeco Field. "Bad Doctor" over at Baseball Think Factory was kind enough to repeat for us some of Philly general manager Pat Gillick's defense of his signing of Franklin:
Gillick was on the radio this morning fielding questions (pretty hard-hitting questions from Cataldi, actually) about this signing. He of course cited knowing what kind of person and pitcher Franklin was, then said his numbers were misleading because he had the second worst run support in the league (methinks this was probably not adjusted for park) and that "if you remove four bad games, two against Toronto and two against Texas, you will see that his ERA was not above 5.00." Which can probably be said about any pitcher. You could probably say it about Paul Abbott in '04.
Could you say it about any pitcher? (We'll read "starter" for "pitcher" here.) If you take any starter's four worst starts out of the equation, can you make a decent pitcher of him?
We'll ignore the fact, to begin with, that a guy whose "ERA was not above 5.00" still wouldn't be much of a pitcher in Seattle. With a league-average offense and pen, a pitcher who throws 200 innings with a 4.75 ERA in Seattle will have a record of about 10-13 (Franklin actually went 8-15). If we then add in the four sure losses we took out at the beginning, we've got a 10-17 pitcher ... so Pat Gillick's defense of Franklin boils down to "we think he can go 10-17 for us." Even if it's, "We think he's a 10-13 pitcher," it isn't exactly a ringing endorsement, let alone something you want to pay money for. But anyway, Franklin only had a 5.10 ERA to begin with, so something is fishy.
The question I'm going after, though, is how much a guy's record improves if you take out his four worst performances. This isn't going to be a properly-conducted study, so if you want something rigorous, click on something on the sidebar right away and hie thee to something by Studes or Dan or Dave (or anyone else here, for that matter). I'm just doing this for some fun.
First, let's run the numbers for Franklin so we get a baseline of what Pat Gillick is talking about. Franklin's four worst starts weren't actually the four that Gillick referenced; the second Texas start wasn't as bad as an outing versus the Angels on August 13 in which Franklin gave up 12 hits and 8 runs, all earned, in 4.1 innings. Seattle lost 9-1.
In total in those four starts, Franklin gave up 44 hits in 17 innings, leading to 28 runs, 26 of them earned. He surrendered six homers but, incredibly, struck out 11 batters and only walked three. Both the strikeout and walk numbers were well better than his normal strikeout-to-walk ratio, which supports one obvious conclusion anyone can make about Franklin when they watch him pitch ... his stuff is so mediocre and so hittable that if he's grooving strikes to everyone, he's in huge trouble.
Take those four starts out and Franklin allows 168 hits in 173.2 innings, with 22 home runs allowed and 82 runs against, all earned, and strikeout-to-walk numbers of 82/59. His ERA moves from 5.10 to 4.25.
Ryan Franklin GS W-L IP H R ER HR BB SO ERA Real 30 8-15 190.2 212 110 108 28 62 93 5.10 Adjusted 26 8-11 173.2 168 82 82 22 59 82 4.25
Bad Doctor talked about Paul Abbott's reprehensible 2004 season, in which he went 3-11, with a 6.47 ERA between Tampa and Philly. Looking at Abbott's game logs from 2004, noting in passing that his ESPN.com picture makes him a dead ringer for Ron Silver ("HIS FATHER IS THE DISTRICT ATTORNEY!"), Abbott's four worst performances yielded only three losses, but he allowed eight home runs in 14 innings while doing so, with 26 hits and 30 runs allowed (29 earned), 10 walks and 6 strikeouts.
Take that padding out of his record, and Abbott looks half-decent. True, his record was just 3-8 in 15 starts, but he allowed just 80 hits in 82 innings, with 40 earned runs (46 in total) producing a trim ERA of 4.39, allowing 14 homers, although his strikeout-to-walk numbers remain ugly (40 and 48). Of course, Abbott only pitched 19 starts. Let's try it instead for some full-season guys.
In 2005, the worst pitchers among ERA qualifiers were Jose Lima (5-16, 6.99 ERA), Eric Milton (8-15, 6.47 ERA) and Mark Hendrickson (11-8, 5.90 ERA). Just to compare, Franklin was ninth from the bottom, 85th among 93 qualifiers with his 5.10 ERA. (No, these numbers are not park-adjusted).
Mark Hendrickson GS W-L IP H R ER HR BB SO ERA Real 31 11-8 178.1 227 126 117 24 49 89 5.90 Adjusted 27 11-5 167.2 195 98 91 20 46 81 4.88
Eric Milton GS W-L IP H R ER HR BB SO ERA Real 34 8-15 186.1 237 141 134 40 52 123 6.47 Adjusted 30 8-12 168.1 198 109 102 32 46 113 5.45
Jose Lima GS W-L IP H R ER HR BB SO ERA Real 32 5-16 168.2 219 140 131 31 61 80 6.99 Adjusted 28 5-14 153.2 185 110 101 21 58 74 5.92
OK, Milton and Lima still look just horrid, which is what happens when you pitch that badly. Both pitchers had way too many bad performances for them to look good by taking out just four. But Hendrickson improves by enough to make him look acceptable (and get him under a 5.00 ERA) and his record looks great. Interestingly, just like with Franklin, these guys had pretty decent strikeout-to-walk peripherals in their terrible starts. Add up the four guys whose 2005 seasons we've looked at so far, and they walked 15 batters and struck out 35 in the 16 starts at issue, in 60.2 total innings. That's a better walk rate, better strikeout-to-walk ratio and better strikeout rate than in their normal starts.
Before we move on, what's up with that? How in the heck can a group of pitchers making 16 awful starts in which their collective ERA is 16.91, post better peripheral numbers than these four guys who were bad, but not bad enough to pitch themselves out of a job?
One thing is, as I said before, none of these guys has stuff that's tough to hit. Second, remember it's not all the "peripheral" numbers that are good ... they gave up 28 homers in those 16 starts. But ultimately, most pitchers have awful outings because they give up lots of base hits tied together. It's a rare manager who will let a guy stay in long enough to give up eight or nine walks, but a guy can give up 10 or 11 hits in three-plus innings all the time.
Now from looking at the guys above, I think it's safe to conclude that a typical ERA qualifier with a bad season will improve his ERA by about a full run if you take out his four worst performances. What's interesting, of course, is that Franklin improved slightly less than the other three guys we looked at, so he may not even be a good candidate for the analysis that Gillick made on Philadelphia radio. But is Franklin improving more than some good pitchers? I picked four pitchers more or less at random, two good ones and two decent ones, to see.
A.J. Burnett GS W-L IP H R ER HR BB SO ERA Real 32 12-12 209.0 184 97 80 12 79 198 3.44 Adjusted 28 11-9 190.0 157 75 59 7 65 185 2.79
Bartolo Colon GS W-L IP H R ER HR BB SO ERA Real 33 21-8 222.2 215 93 86 26 43 157 3.48 Adjusted 29 21-4 202.0 183 62 61 19 32 143 2.72
Matt Morris GS W-L IP H R ER HR BB SO ERA Real 31 14-10 192.2 209 101 88 22 37 117 4.11 Adjusted 27 13-7 174.0 172 80 72 19 32 111 3.72
Kyle Lohse GS W-L IP H R ER HR BB SO ERA Real 30 9-13 178.2 211 85 83 22 44 86 4.18 Adjusted 26 9-10 161.1 180 63 62 14 38 76 3.46
A.J. Burnett's consistency was terrific, and it's amazing he lost as many as 12 games; he only had a Game Score of less than 30 once, and that was a 29. He had only two "Tough Losses" (as defined by Bill James; games in which he posted a Game Score of 50 or more but lost) but in most of the six games in a row he lost after the Marlins gave up on their season he pitched decently but lost anyway. Matt Morris was also very consistent, losing only 39 points of ERA. In all, these pitchers averaged 63 points of ERA better when you take away their bad starts, which is slightly less improvement than Franklin's 85. Still, it means that Franklin's 85 point improvement is in no way untypical.
What's interesting is to compare how these pitchers fared compared to the bad pitchers in their poor starts. They pitched not only much better, they pitched quite differently as well ...
GS W-L IP H R ER HR BB SO ERA 4 bad starters 16 0-12 60.2 149 118 114 28 15 35 16.91 4 good starters 16 2-13 75.2 127 96 83 23 36 43 9.87
What is immediately noticeable is that the four good pitchers had absolutely lousy strikeout-to-walk rates in their bad starts, especially compared to their usual performance. While the crappy pitchers actually had better strikeout-to-walk rates across the board in their 16 worst starts than in their regular games, the good pitchers were much worse in all three categories in their 16 worst starts. More interestingly still, the very good pitchers, Bartolo Colon and Burnett, were also tagged by home runs more often compared to their normal rates.
I am just speculating here, but I suspect that the explanation for this is a mirror image of the reasons I gave above for the bad pitchers having good strikeout-to-walk ratios in their bad games. Just as the bad pitchers' stuff is so bad that anyone can hit it if it's grooved, the good starters normally have success when they groove the ball, but when they stop hitting the strike zone and are forced to throw from behind in the count, their good stuff gets easier to hit, which it isn't normally.
I think all this just goes to show that Franklin is a bad, hittable pitcher, a lousy signing for the Phillies, and Pat Gillick's well-intentioned defense of him doesn't hold water. I think the real reason Gillick signed Franklin is that he's familiar with him from Seattle, knows his abilities and limits, and likes him as a person. He said about as much in the interview quoted above. But as nice as that is, it doesn't change the fact that Franklin has had a hard time getting major league hitters out for years now, and is one of the most homer-prone starters around, which didn't hurt him that badly in Seattle (where he only allowed 10 of his 28 home runs), but is bound to hurt badly in Philadelphia. Maybe Gillick also extracted a promise from Franklin only to use the undetectable steroids from now on. But talk about run support and taking out the bad starts doesn't really stand up to common sense.
I think a lot of the older generation of baseball men, of whom Gillick is unquestionably one of the most able, have not adjusted to the fact that baseball fans are far more sophisticated than they once were about baseball. The problem that the Philly fans have with Franklin (I haven't even asked my friend Bill Liming his opinion of Franklin yet, for fear of what he'll do) isn't that he went 8-15 ... it's that he's an 8-15 pitcher, and what's more, he's obviously unsuited for their ballpark. Now Pat Gillick may well have his reasons for wanting Ryan Franklin with the Phillies, but it isn't that he's a good pitcher marred by a few bad starts.
Craig Burley can be contacted via e-mail.