Saves Per Inning Pitched

In this earlier post, I stumbled across a new statistic, SIP. I have decided to write an article on this. Who knows, I may decide to submit it to Baseball Primer! Before I even consider such a thing, I would like you, my loyal readers (:wink:) to critique the article below. Rip it to shreds or modify certain paragraphs, add, subtract, do anything you think needs to be done to improve this article. Or perhaps just write off SIP as something worthless, SIP is only just an idea right now.
Saves per Inning Pitched, aka SIP – by Evan Brunell
* Please note that all SIPs are from 2003, unless otherwise noted.
I have a baseball blog called Fire Brand of the American League (firebrand.mostvaluablenetwork.com), based off of the Most Valuable Network (mostvaluablenetwork.com), owned by me, which deals with the Boston Red Sox. I was discussing the effectiveness of the Boston bullpen so far when my fingers started typing with abandon and invented a new statistic. It is not even close to a perfect statistic, but if used correctly, can show a lot about a closer.
SIP measures the true worth of a closer. Is he someone who gets cheap saves? Is he someone only used for closing situations? Or is he someone the team depends on, that is used in close games, tie games, or even situations where there is no save opportunity but there is a save-the-team opportunity. If a closer gets saves in two-thirds of his innings pitched, one can infer he is used in closing situations only – never used in tie games, to keep the game close, to get the offense going. He isn’t used in the seventh or the eighth to shut down the 3, 4, 5 hitters. He’s used when his team has a lead in the ninth inning, one or two runs up on the opponent.
SIP is far from a perfect statistic and can actually be quite misleading if you do not apply it right. For starters, it is a tool used for closers only. It cannot be used for middle relievers. How does this work, then? Who do you define as a closer? SIP only works when there is one clear-cut closer for an entire season. SIP is also team-dependent. How does the team use this closer? For example, if you took Eric Gagne, who has a SIP of 0.67 (got a save in two-thirds of his total innings pitched in 2003) and moved him to Kansas City, where the manager was more aggressive about putting him in situations to bail the team out and not just to get saves, Gagne’s SIP would decrease.
So perhaps SIP is not an individual statistic, it is a manager’s statistic. But what if this statistic shows how closers excel? Maybe some excel being a closer, not a relief ace. That would be Eric Gagne. A SIP of 0.67 (55 saves in 82.1 IP). Some excel being a relief ace, not a closer. They get the saves, but they also come into tight situations. Someone like Billy Wagner – a SIP of 0.51. Rocky Biddle has a SIP of 0.47, while Trevor Hoffman has a SIP of 0.64 (in 2002). This means that Montreal used him in tighter situations – tie games, maybe the Expos being down a run or so. Keeping the game close. Being the relief ace. Biddle may have been the team’s closer, but he was also the relief ace. He led the Expo relievers in innings. Hoffman is a closer, not a relief ace. Strictly the ninth inning, brought in for the save. Not brought in in a tie game, not brought in in the seventh to quash a rally by the Pirates. He’s brought in to ‘save’ the game.
Saves have become quite misleading now. What used to be a novel topic is now a very overrated topic. You always need someone who thrives on pressure, makes the pitches, gets the outs. But sometimes that’s not the closer. The closer was initially the relief ace, but now it is someone who comes into the ninth with a 5-3 lead and has a 1-2-3 inning. How is this a ‘save’, per se? I believe the true intention of the save was to document which pitchers had more of a direct hand in closing out games, to give relief aces their due. However, nowadays, the general notion is that a closer is to pitch the ninth and get that save – most of which are cheap. Rocky Biddle is the Expos’ relief ace. Was Gagne the Dodger’s 2003 relief ace? I say no. It was Guillermo Mota, tossing 105 innings with an ERA of 1.97. He was the go-to guy in the pen to keep the game close, to give the Dodger bats (however woeful they may have been) a chance to win the game.
SIP is a way to determine who is a closer and who is a relief ace. Which teams use their closer to get the saves, and which use them to keep the game close, to win against the opposition? Saves per Inning Pitched.
The closer is not there to get saves – he is named the closer because he is your relief ace. Any tight game or tie game in the late innings – your closer should be used. As far as I am concerned, closers should average 80 to 100 innings. When Derek Lowe was the Red Sox closer in 2000, he had 91.1 IP and 42 saves. One could argue that the Dodgers are not using Eric Gagne as much as they could. Couldn’t Gagne easily top 100 IP? I think so. Keith Foulke, current Red Sox closer, is a perfect six for six in saves. 17 innings, 6 saves. In 2000, Lowe had 42 saves in 91.1 IP. How can Gagne have 55 saves in only 82.1 IP?
Before I get into my listing of 2003 SIPs and what this tells us, how do we determine who a closer is? It was not Byung-Hyun Kim in 2003 for the Red Sox for he had 16 saves in 79.1 IP. He started five games and also lost the closing job. Isn’t Kim more of a relief ace, if one wants to say that? SIP is not used for relief aces. It is used to see if a closer IS a relief ace, or if he is just a true-blue closer. So there needs to be some limitations.
Limitation One: Said person that is being measured by SIP needs to have at least 50% of all team saves. That eliminates Kim, for the Sox had 36 saves total last year. (36!) Half of 36 is 18, so Kim narrowly misses this.
Limitation Two: The next person with the most saves on the team cannot have more than 25% of the total saves. If Kim had 18 saves in 2003, he would be ineligible for SIP if Scott Williamson had 10 saves. 25% of 36 is 9. Why is this limitation here? In this hypothetical situation, Kim was not the clear-cut closer. It was a dividing of saves between Kim and Williamson. This would reduce SIP to nothing, for it is measuring a closer. The SIP for Kim and Williamson in this hypothetical situation would be low, showing them as relief aces. While SIP can certainly be used to show that they were relief aces part of the team, then there would be an entirely new debate on how to divide up relief aces and middle relievers. For example, Chad Fox had three saves for Boston before being released. Brandon Lyon had nine saves total for Boston last year. If you wanted to use SIP for Brandon Lyon to show that he was a relief ace, how could you justify not using SIP for Chad Fox? SIP is used to measure worth of a CLOSER, a full time closer, not two people sharing closing duties. In the hypothetical situation above, Kim and Williamson would be relief aces for they have a good number of saves but also have innings in other non-save situations. To sound like a broken record, SIP is to find out if full-time closers are closers or relief aces.
The following are 2003 SIPs for some closers in the majors.

  • ANA – Troy Percival: 0.67
  • ATL – John Smoltz: 0.70
  • ARI – Matt Mantei: 0.53
  • BAL – Jorge Julio: 0.58
  • BOS – Disqualified. Kim had 16 SVs, team total was 36 SV. He falls short of the halfway mark.
  • CHN – Joe Borowski: 0.48
  • CHA – Disqualified. Tom Gordon 12 SV, Billy Koch 11 SV, Damaso Marte 11 SV.
  • CIN – Disqualified. Scott Williamson had 21 saves, Chris Reitsma had 12 saves, 30% of team total.
  • CLE – Danys Baez: 0.33
  • COL – Disqualified. Justin Speier had 26% of the saves.
  • DET – Disqualified. No one meets the 50% requirement.
  • FLA – Braden Looper: 0.35
  • HOU – Billy Wagner: 0.51
  • KCA – Mike MacDougal: 0.42
  • LAN – Eric Gagne: 0.67
  • MIL – Disqualified. Danny Kolb had 21 SV, Mike DeJean had 18.
  • MIN – Eddie Guardado: 0.63
  • MON – Rocky Biddle: 0.47
  • NYN – Armando Benitez: 0.42 (Total in Season : 0.28)
  • NYA – Mariano Rivera: 0.57
  • OAK – Keith Foulke: 0.50
  • PIT – Mike Williams: 0.67 (Take this with a grain of salt, as Julian Tavarez had 25% of the team’s saves. Limitation is OVER 25%. Total, William’s SIP was 0.44, as he was a set-up man after he was traded to PHI.)
  • PHI – Jose Mesa: 0.41
  • SDN – Trevor Hoffman (2002, inj. 2003): 0.64
  • SFN – Robb Nen (2002, inj. 2003): 0.58
  • SEA – Disqualified. No one meets the 50% standard.
  • STL – Jason Isringhausen: 0.52
  • TBD – Lance Carter: 0.32
  • TEX – Disqualified, Francisco Cordero had 35% of the saves.
  • TOR – Disqualified. Cliff Politte had 33% of the saves.
  • This needs to be taken with a grain of salt. Braden Looper lost his job, having 28 saves and 6 blown saves. Obviously, this all needs to be taken with a grain of salt, but I still believe SIP is a concept that could help identify what closers are treated as, or rather what they are capable of. This year, Keith Foulke’s SIP is 0.35. Is this an indicator of the team or the closer? I argue the closer. This is also why Troy Percival has a high SIP – he is a one inning CLOSER, not a relief ace.
    There is one glaring problem with SIP – it cannot be used on a year to year basis. While starting pitchers are always (or most always, anyways) judged on their ERA from the previous year, SIP is a continuous, career average that always needs to be watched. Eric Gagne’s career SIP (which only encompasses 2 seasons) is 0.65. Trevor Hoffman’s career SIP takes a marked downward trend to 0.56.
    That being said, the career SIP and 2004 SIP of Eric Gagne does tell a lot. His 2003 SIP was 0.67. This means that over half of the time he was in, he earned a save. Heck, it’s two-thirds! Two-thirds of the time he was brought in, he recieved a save. Therefore he rarely pitched when the game was tied. How do I know this? If you come into a game and it is tied, you cannot earn a save. You either win, lose, or have a no-decision. He also rarely came in with the team a run or two behind – other relievers were used to do that. Sure, LA had one of the best bullpens last year, but isn