SWIP: Strikeouts / Walks / Innings Pitched
Part 1: Relievers
Contributed By: Ray Flowers
Last year I introduced a stat called SWIP that I thought received little attention in the baseball world. Apparently someone at ESPN was listening as they ran a “new” stat called PFR (Player Finesse Ratio) just weeks after my first story appeared (I’ve sent correspondence to ESPN that has been unanswered regarding this development). I’m not saying ESPN “stole” my stat, perhaps we both just came up with the idea at similar times. However I will detail the fact that ESPN has nonetheless bastardized the stat, and in fact produced something that I see as fairly useless. But before I critique their work, let me detail just what all the commotion is about.
In this three part work I will begin by detailing the exploits of Relievers in PART I followed by PART II were I will move on to discuss the plight of the Starting Pitchers. PART III will focus on the shortcomings of ESPN’s version of SWIP.
But before we get to any of that, here is a primer that describes just what SWIP is and how it works.
SWIP- A Definition
WHIP , a “roto-geek” stat, has become all the rage in the past ten years. You measure WHIP by adding Walks and Hits together and then dividing by Innings Pitched.
(Walks + Hits) / Innings Pitched = WHIP
WHIP is but one of the many predictive tools that can be used to forecast the success of a pitcher. But I believe that the vagaries of hits allowed (a ball lost in the sun, a wind gust, a misstep by a fielder, the coaches positioning of players) can possible offer a “false” negative in terms or a pitchers overall effectiveness (should a pitcher be punished because any of the above events occur?). So that got me to thinking…
If you think about it the pitcher is directly in control of but a few things during the game. Two of those “events” being whether or not he strikes outs the batter or walks him (we will not worry about the vagaries of umpire’s strike zones when discussing SWIP). About the only other directly influenced event that regularly occurs is the Hit by Pitch, but since WHIP does not count it, we won’t either.
Now it stands to reason, does it not, that the fewer times a batter puts the ball in play the fewer hits he gives up? Therefore the pitcher who successfully limits the amount of balls in play stands a greater chance of given up fewer base runners, and by inference, fewer runs (this idea will be taken up at a later point when I discuss DIPS ERA which speaks directly to this phenomenon as well). So to address this issue I proposed a new stat called SWIP.
S- Strikeouts (also abbreviated as K)
W- Walks (also abbreviated as BB)
IP- Innings Pitched
Numerically speaking, the formula for SWIP works along the same lines as WHIP. Therefore SWIP is determined by the following equation:
Strikeouts minus Walks divided by Innings Pitched equals SWIP.
K – BB / IP = SWIP
Another way to look at this is to say that for each positive result, the recording of an OUT in the form of a K, the pitcher receives a (+1). For each negative encounter, in the form of a BB, he receives a (–1). Simple enough right? Here is an example so you can see what I’m talking about.
Johan Santana had 265 K and 54 BB in 228 IP in 2004
(265-54) / 228
211 / 228 = 0.93 SWIP
Though SWIP is recorded in the same manner as WHIP, the way to read the results is a bit different.
Whereas the lower the WHIP the better one has performed, SWIP works in the opposite direction…the higher the SWIP the better.
Now admittedly SWIP has its flaws, as does almost every other metric that measures anything. SWIP expresses this flaw by favoring power pitchers over soft-tossing hurlers. This can be seen as a limitation or it can be viewed as something else (as we mentioned above the fewer balls put in play, the fewer opportunities to allow base runners). Whereas SWIP might not be the greatest predictive tool if taken on its own to gauge the effectiveness of a starting pitcher, it might be an extremely useful tool when dealing with relievers.
Starting Pitchers (SP) have multiple innings to set up batters, vary pitch sequences and to work themselves out of trouble. This “freedom” allows SP to pitch with a variety of styles, all of which can be successful. On one end of the spectrum you have knuckleballers like Tim Wakfield and soft-tossers such as Jamie Moyer and Greg Maddux. Obviously these pitchers do not record strikeouts at the same rate as the power pitchers we are about to mention, but they still get OUTS and that is the job of the pitcher no matter how it happens. On the other end of the scale you have fireballers like Kerry Wood and Randy Johnson who rear back and bring the heat. All of these pitching styles can be successful whether they rely on the K or not if the pitcher has enough time to work out of “jams.”
Contrast that with Relief Pitchers (RP). RP usually don’t have multiple innings to set up batters and they often come into games when there are already runners on base. They don’t have time to find their grove and work on touch pitches like change-ups and curveballs. They need to come in and throw strikes…now. Therefore it appears that SWIP might be an even more useful tool to judge pitchers who rely mostly on “hard stuff” (fastballs, sliders and forkballs). These pitchers, as a general rule, congregate more in the bullpen, and this is where we will go next.
This first table (#1) lists all relievers who had at least 10 saves in 2004.
|M. Rivera, NYY||53||B. Lidge, Hou||29|
|F. Cordero , Tex||49||L. Hawkins, ChC||25|
|A. Benitez , Fla||47||M. Herges, SF||23|
|J. Isringhausen, StL||47||J. Julio, Bal||22|
|E. Gagne , LA||45||U. Urbina, Det||21|
|J. Smoltz, Atl||44||B. Wagner, Phi||21|
|J. Nathan, Min||44||T. Worrell, Phi||19|
|J. Mesa, Pit||43||S. Takatsu, CWS||19|
|T. Hoffman , SD||41||E. Guardado, Sea||18|
|D. Graves , Cin||41||D. Hermanson, SF||17|
|D. Kolb, Mil||39||J. Frasor, Tor||17|
|O. Dotel, Oak/Hou||36||G. Aquino, Ari||16|
|S. Chacon , Col||35||C. Cordero, Mon||14|
|T. Percival, Ana||33||B. Wickman, Cle||13|
|K. Foulke, Bos||32||J. Affeldt, KC||13|
|D. Baez, TB||30||F. Rodriguez, Ana||12|
|B. Looper, NYM||29||R. Biddle, Mon||11|
Nothing more needs to be said about Table #1, it’s self-explanatory. In Tables #2 and #3 I will list each relievers K, BB and IP totals for 2004 followed by their SWIP figures.
|M. Rivera, NYY||66||20||78.2||0.59|
|F. Cordero , Tex||79||32||71.2||0.66|
|A. Benitez , Fla||62||21||69.2||0.59|
|J. Isringhausen, StL||71||23||75.1||0.64|
|E. Gagne , LA||114||22||82.1||1.12|
|J. Smoltz, Atl||85||13||81.2||0.89|
|J. Nathan, Min||89||23||72.1||0.92|
|J. Mesa, Pit||37||20||69.1||0.25|
|T. Hoffman , SD||53||8||54.2||0.83|
|D. Graves , Cin||40||13||68.1||0.40|
|D. Kolb, Mil||21||15||57.1||0.11|
|O. Dotel, Oak/Hou||122||33||85.1||1.05|
|S. Chacon , Col||52||52||63.1||0.00|
|T. Percival, Ana||33||19||49.2||0.28|
|K. Foulke, Bos||79||15||83||0.77|
|D. Baez, TB||52||29||68||0.34|
|B. Looper, NYM||60||16||83.1||0.53|
|B. Lidge, Hou||157||30||94.2||1.35|
|L. Hawkins, ChC||69||14||82||0.67|
|M. Herges, SF||39||21||65.1||0.28|
|J. Julio, Bal||70||39||69||0.45|
|U. Urbina, Det||56||32||54||0.44|
|B. Wagner, Phi||59||6||48.1||1.10|
|T. Worrell, Phi||64||21||78.1||0.55|
|S. Takatsu, CWS||50||21||62.1||0.47|
|E. Guardado, Sea||45||14||45.1||0.69|
|D. Hermanson, SF||102||46||131||0.43|
|J. Frasor, Tor||54||36||68.1||0.26|
|G. Aquino, Ari||26||17||35.1||0.26|
|C. Cordero, Mon||83||43||82.2||0.49|
|B. Wickman, Cle||26||10||29.2||0.55|
|J. Affeldt, KC||49||32||76.1||0.22|
|F. Rodriguez, Ana||123||33||84||1.07|
|R. Biddle, Mon||51||31||78||0.26|
Something you might notice right off the bat is that the SWIP numbers appear to be in no particular order. The reason for this is that I have continued to list all the relievers in order of their Save totals from 2004. So let me digress for a moment to discuss Saves.
What are Saves? A Save is when a reliever finishes a game in which he entered with a 1, 2 or 3 run lead that he never relinquished (or if he enters the game with the tying run on-deck). So what does a save really say? I’d say that what a Save reflects is really a random sequence of events that has no empirical bearing on a pitchers year to year success. These arbitrary events include:
1- An offense that places the team ahead but not by more than 3 runs.
2- The support of a manager that he, better than any other relievers on the team, should be allowed to get the final out of a game. This is an arbitrary decision made by the manager and may not accurately reflect a pitchers contribution to the team effort (how does a guy get the final out to earn a Save if some other reliever pitched 2 1/3 without giving up a run and he gets nothing but a Hold?).
3- The Save rule is arbitrarily made up. Why 3 runs? Why the tying run on deck? Heck, aren’t games more often than not decided in the 7 th or 8 th innings when there are runners all over the bases? How many times does a “closer” come in with a 3 run lead and the bases empty in the 9 th inning? Too often to receive definitive credit in my mind.
Therefore one should not judge relievers based on their Save total (though a high premium is placed on the Save in the fantasy game…something we will discuss in a future piece). A better indication of a pitchers talent level and possible success rate in the future is to look at the rest of his stat line to determine how effective he has been (another aside: you might want to stay away from ERA when judging relievers. Since reliever’s pitch so few innings if they have one horrible outing and give up 5 runs, their season ERA could be about one full run higher than you would normally expect).
After this little diversion why don’t we list our same group of relievers not by their arbitrary save totals but by their SWIP totals (Table #4)? As I mentioned above, I think that a better predictor of a pitchers future success is to look at categories other than ERA or Saves, hence the reason for this article. Here are our pitchers listed by their 2004 SWIP totals…be prepared to see a shift.
|B. Lidge, Hou||1.35||B. Looper, NYM||0.53|
|E. Gagne , LA||1.12||C. Cordero, Mon||0.49|
|B. Wagner, Phi||1.10||S. Takatsu, CWS||0.47|
|F. Rodriguez, Ana||1.07||J. Julio, Bal||0.45|
|O. Dotel, Oak/Hou||1.05||U. Urbina, Det||0.44|
|J. Nathan, Min||0.92||D. Hermanson, SF||0.43|
|J. Smoltz, Atl||0.89||D. Graves , Cin||0.40|
|T. Hoffman , SD||0.83||D. Baez, TB||0.34|
|K. Foulke, Bos||0.77||T. Percival, Ana||0.28|
|E. Guardado, Sea||0.69||M. Herges, SF||0.28|
|L. Hawkins, ChC||0.67||J. Frasor, Tor||0.26|
|F. Cordero , Tex||0.66||G. Aquino, Ari||0.26|
|J. Isringhausen, StL||0.64||R. Biddle, Mon||0.26|
|A. Benitez , Fla||0.59||J. Mesa, Pit||0.25|
|M. Rivera, NYY||0.59||J. Affeldt, KC||0.22|
|T. Worrell, Phi||0.55||D. Kolb, Mil||0.11|
|B. Wickman, Cle||0.55||S. Chacon , Col||0.00|
LIDGE moves to the top of the list with a season of enormous magnitude according to SWIP (since 2001, the only higher SWIP season was GANGE in 2003 when he had a SWIP of 1.42). Based on past success, a pitcher whom you would expect to see at the top of any discussion of relievers, WAGNER, moves way up the list after his injury plagued 2004 season just didn’t offer him enough opportunities to accumulate a great Saves total. After mentioning those two we must also mention DOTEL who occupies the 5 th spot on our 2004 SWIP leader board. Think for a moment about how dominant the Astros bullpen could have been had it kept 3 of the top 5 SWIP pitchers in 2004 on its roster instead of trading two of them away! Another pitcher who moves way up our board is FROD. If you can somehow manage to grab this electrifying, slider throwing monster on draft day for around $20, well, you’ll be sitting’ pretty (though the hype surrounding him may preclude you from grabbing him at that level).
As for relievers who struggled according to SWIP and might have some trouble reproducing their success of last year and years past, we would list the following:
URBINA…he will be the set-up man in Detroit if he stays, a poor closer if he moves on. His SWIP has declined the last three years: 0.85 to 0.61 to 0.44.
GRAVES …amazes me ever year that anyone has faith in him. His career SWIP of 0.21 is horrible for a “closer.”
PERCIVAL…Age is catching up to this fireballer. SWIP in 02’ (0.76), 03’ (0.51) and 04’ (0.28) shows a steady decline.
MESA …Come on, you thought I would recommend this guy? Career SWIP = 0.27.
And there is one other prominent reliever who we feel we should mention….DANNY KOLB.
No one can quite figure this guy out. From a statistical point of view he really shouldn’t be as successful as he is being that he doesn’t fit the typical strikeout mold of a “closer” (think Dan Quisenberry). Here are Kolb’s stats from 2003-04.
How does he do it? Well according to SWIP he shouldn’t be able to. Apparently he is an example of pitcher who defies “normal” explanation and will be successful no matter what numbers we crunch. I know the Braves sure are hoping this is the case.
PART I – CONCLUSION
As I wrap the discussion on Relief Pitchers, I hope that you will take the time to read the next two installments of SWIP. I also hope that you see the value of this metric which measures, more directly at least, that which is directly in control of the pitcher. SWIP, when combined with other stats and metrics, can be an extremely useful tool in rating how a pitchers past performance may be indicative of his future success.
Click Here to check out the entire FantasyBaseball.com University Series!