Anatomy of a Record

How do I go about finding these obscure, pointless records? Well, the brief answer is by tinkering with the Baseball-Reference Play Index (B-R PI) until it spits out the answer I seek. However, I feel like taking you on a tour of my thought process throughout.

The first step is deciding just what record to look up. There's been some agitation in Milwaukee recently about the Brewers' staff having trouble pitching six full innings in an average start, so how about we look up what pitcher holds the record for least IP/GS in various numbers of starts? Now that that's settled, we've hit our first potential roadblock: starters that don't last very long likely won't remain starters all year. The first option is to just look at guys who started every game they appeared in. That's the easiest method, though a lot of guys would likely fall through the cracks. The other option is to investigate game logs for each of the guys who may have had a worse IP/GS average, though that would limit us to players since 1957 (the extent of the database). The solution, in my mind, will be to look up first the guys who started every game they appeared in and then see if any guys who were bumped to the bullpen midseason threaten their mark.

Then there's the decision of how many starts I want to set as the minimum category. While seasons like Kevin Ritz's 1990 are fun to look/laugh at, I find it more interesting to see mediocrity spread out across most of a season. With that in mind, I think starting at 20 starts and increasing at 5 start intervals makes sense.

Firing up the Pitching Season Finder at the B-R PI and setting it to the predetermined conditions (20 starts, 100% of games pitched as starts, sorted by IP in ascending order), we get this list:

1. Kirk Rueter, 92.1 IP, 20 GS, 1994
2. Pascual Perez, 95.2 IP, 22 GS, 1985
3. Kip Wells, 98.2 IP, 20 GS, 2000
4. Oliver Perez, 103.0 IP, 20 GS, 2005
5. LaTroy Hawkins, 103.1 IP, 20 GS, 1997

It's not hard to see that Pascual Perez will hold the record for this category, as his innings pitched is second in two more starts than anyone around him. His 95.2 IP in 22 starts works out to 4.35 IP/GS, meaning he was good for 13 outs. Here's his game log for 1985. For anyone to be worse, here's the innings they'd have to throw in increasing numbers of starts:

• 20 GS: 86.2 IP
• 21 GS: 91.0 IP
• 23 GS: 100.0 IP
• 24 GS: 104.1 IP
• 25 GS: 108.2 IP

It continues, obviously, at a increase of about 4 1/3 IP for each start. Skimming down the list, the first 23 game starter (Kevin Appier, 2003) tossed 111.2 IP, so that won't cut it. No one else came close in any larger number of starts, either.

Now, like I said earlier, it's time to make things harder and attempt to find any starter/relievers that might have done worse in 20+ starts. Resetting the Pitching Season Finder to reflect the new conditions (1957-2007, 20+ starts, sorted by IP in ascending order), here's the list of guys who had at least one relief appearance mixed in their starts:

1. Steve Avery, 21 G, 20 GS, 99.0 IP, 1990
2. Mike O'Connor, 21 G, 20 GS, 105.0 IP, 2006
3. Tomokazu Ohka, 22 G, 21 GS, 107.0 IP, 2001
4. Ismael Valdez, 21 G, 20 GS, 107.0 IP, 2000
5. Josh Beckett, 23 G, 21 GS, 107.2 IP, 2002
6. Randy Jones, 28 G, 20 GS, 107.2 IP, 1982
7. Steve Arlin, 27 G, 22 GS, 107.2 IP, 1974
8. Doug Drabek, 23 G, 21 GS, 108.2 IP, 1998
9. Shawn Chacon, 26 G, 20 GS, 109.0 IP, 2006
10. Scott Scudder, 23 G, 22 GS, 109.0 IP, 1992

Referring to our chart of maximum innings for the various numbers of starts, I think it's safe to say that none of the 1-2 relief appearance guys on the list broke the record. That cuts it down to Randy Jones, Steve Arlin, and Shawn Chacon. Luckily, a quick look at their splits page (here's Jones') lets us determine whether or not they did set the record. Jones had 93.1 IP in his 20 GS, so not bad enough. Arlin's 99.2 IP in 22 GS and Chacon's 98.0 IP in 20 GS also miss the cut. Ignoring the fact there may be someone from before 1957 or a freak of nature throwing 5+ IP relief outings among his horribly short starts, it's pretty safe to say Pascual Perez's 4.35 IP/GS in 1985 is worst among starting pitchers with 20 or more starts.

Let's increase the minimum to 25 or more starts. Rather than make you read how I determined all this again, I'll just post the results. First, for guys that only started:

1. Ramon Martinez, 27 GS, 127.2 IP, 4.73 IP/GS - 2000
   
2. Sean Bergman, 28 GS, 135.1 IP, 4.83 IP/GS - 1995
3. Aaron Sele, 25 GS, 121.2 IP, 4.87 IP/GS - 2003
4. Colby Lewis, 26 GS, 127.0 IP, 4.88 IP/GS - 2003
5. Joe Niekro, 25 GS, 125.2 IP, 5.03 IP/GS - 1986

Now including the guys that started and relieved:

1. Ramon Martinez, 27 GS, 127.2 IP, 4.73 IP/GS - 2000
2. Sean Bergman, 28 GS, 135.1 IP, 4.83 IP/GS - 1995
3. Pete Redfern, 28 GS, 135.2 IP, 4.85 IP/GS - 1977
4. Ryan Vogelsong, 26 GS, 126.0 IP, 4.85 IP/GS - 2004
5. Aaron Sele, 25 GS, 121.2 IP, 4.87 IP/GS - 2003

Duly increasing the minimum now to 30 starts, here's the table of both starters and guys with at least one relief appearance:

1. Bobby Witt, 31 GS, 157.2 IP, 5.09 IP/GS - 1986
2. Jeriome Robertson, 31 GS, 159.0 IP, 5.13 IP/GS - 2003
3. Dick Drott, 31 GS, 159.0 IP, 5.13 IP/GS - 2003
4. Tony Armas, 30 GS, 154.0 IP, 5.13 IP/GS - 2006
5. Ken Cloude, 30 GS, 155.1 IP, 5.18 IP/GS - 1998

With the minimum at 35 starts, we get these guys:

1. Dennis Lamp, 37 GS, 198.2 IP, 5.37 IP/GS - 1980
2. Claude Osteen, 37 GS, 204.1 IP, 5.52 IP/GS - 1975
3. Steve Bedrosian, 37 GS, 206.2 IP, 5.59 IP/GS - 1985
4. Jim McGlothlin, 35 GS, 195.2 IP, 5.59 IP/GS - 1969
5. Sterling Hitchcock, 35 GS, 196.2 IP, 5.62 IP/GS - 1996

Finally, while I know it's not likely we'll see a 40-start pitcher anytime soon, let's look at those guys:

1. Stan Bahnsen, 41 GS, 248.2 IP, 6.07 IP/GS - 1972
2. Jim Kaat, 42 GS, 262.1 IP, 6.25 IP/GS - 1965
3. Johnny Podres, 40 GS, 255.0 IP, 6.38 IP/GS - 1962
4. Jim Bibby, 41 GS, 264.0 IP, 6.44 IP/GS - 1974
5. Tom Bradley, 40 GS, 260.0 IP, 6.50 IP/GS - 1972

Well, that was fun. To recap, here's the leaders at each minimum set:

Minimum Starts	Name	Year	IP/GS
20	Chris Knapp	1980	4.30
25	Ramon Martinez	2000	4.73
30	Bobby Witt	1986	5.09
35	Dennis Lamp	1980	5.37
40	Stan Bahnsen	1972	6.07

No one's really threatening any of these marks this year, though Matt Chico might come close to the 35 start mark if he has ten more starts. He's currently at 5.29 IP/GS through 25 starts. Edwin Jackson could threaten the 30-start mark, as he's at 5.07 IP/GS through 23 starts, though he'd have to start struggling again.

EDIT: Zeyes correctly points out that Chris Knapp in 1980 managed only 4.3 IP/GS in 20 starts, making him the all-time leader in duration futility. He also points out Ron Bryant's 1974 featured 23 starts at 4.38 IP/GS.