Half of a 5x5 reliever’s value to his fantasy team comes from saves.  The rest of their value comes from superior ratios on smaller innings totals along with wins and K’s dwarfed by their rotation counterparts’ totals.  Understanding this critical stat will help you better project save totals, open your eyes to buying opportunities and improve your chances of taking home the hardware next October.

We’re going to look at the factors involved in making a save happen.  A closer’s team creates the opportunity for a save.  The closer’s manager then gives him the opportunity to earn a save.  Finally, the closer is responsible for converting his opportunity.

Before digging in here, we need some numbers.  Unfortunately the raw data provided by MLB is flawed, so we have to create our own stats.  I ask you this:  How can it be called a save opportunity when the pitcher has no realistic chance of earning a save?

Can a mid-reliever really blow a save in the sixth inning?  He can’t earn a save in the sixth inning, so how can he blow one?  Realistically, this should be called a blown hold.  If the pitcher is successful in shutting down the opponent in a mid-game situation with the lead, he earns a hold but no save opportunity is credited.  When he is unsuccessful, he is given a blown save (as well as an unsuccessful save opportunity).  Had he been successful in this sixth inning situation, what are the chances the manager leaves him in for the rest of the game to have the opportunity to earn a save?  Not since the days of the Goose have we seen such managerial tactics. 

I’m throwing out MLB’s “BS” statistic, as we know it and drilling down to the True numbers that will tell us a better story about the 2008 bullpens.  We’ll be able to better understand the relationship between team wins and closer saves for our projections in 2009. 

MLB tracks saves, blown saves and save opportunities.   I’m going to tweak each of those stats.

I’ve thrown out three inning saves in my numbers.  There were 16 three inning saves in 2008.  The only closer-type pitcher to earn one was Salomon Torres, and he earn his prior to taking over the role from Goggles in late May.  They are earned by mid-relievers in situations that don’t apply to standard save opportunities, so we don’t care about them.

I’ve thrown out all blown saves and save opportunities that were credited when the pitcher had no realistic opportunity to earn a save.  I followed one rule.  Any blown save/save opportunity credited to a mid-reliever who pitched in the 8th inning or earlier was wiped off the books.  A lot of times these mid-relievers would enter in the eighth, blow the hold and the closer would pitch the ninth inning regardless.  Either that or the reliever who blew the save would remain in the game and the closer would be saved for extra innings.  Once extra innings hit, all’s fair on the blown save front, because a successful conversion of a save opportunity would likely have resulted in a save. 

After all this is done we’ve eliminated 25% of the 2008 save opportunities (SVO) and are left with True Save Opportunities (TSVO).  Did you know Scot Shields was 4-for-4 in TSVO, not 4-for-9 in SVO with bogus 8th inning “save opportunities” with K-Rod waiting in the wings?  Let’s dig in.

Team Creates Opportunity

There are a lot of components involved here.  First, you need offense (just the right amount).  Second you need starters and mid-relievers to preserve a lead until the late innings.  Finally, it never hurts to have a little luck; if your closer’s team is going to extra innings, the only way you’ll get a save is if he’s pitching for the visiting side.  On average each team provided true save opportunity in 46 of its 162 games in 2008.

What is the relationship between team wins and true save opportunity?  While the trend does follow the intuitive pattern, there are some interesting outliers in these numbers. 

Team     W    TSVO
Los Angeles Angels 100 74
Tampa Bay Rays 97 59
Chicago Cubs  97 51
Boston Red Sox 95 52
Philadelphia Phillies 92 46
Milwaukee Brewers 90 54
New York Mets 89 54
Chicago White Sox 89 40
New York Yankees 89 44
Minnesota Twins 88 46
St. Louis Cardinals 86 56
Houston Astros 86 56
Toronto Blue Jays 86 50
Florida Marlins 84 46
Los Angeles Dodgers 84 41
Ariz Diamondbacks 82 45
Cleveland Indians 81 39
Texas Rangers  79 40
Kansas City Royals 75 49
Oakland Athletics 75 41
Detroit Tigers  74 42
Colorado Rockies 74 41
Cincinnati Reds 74 42
San Francisco Giants 72 47
Atlanta Braves  72 32
Baltimore Orioles 68 39
Pittsburgh Pirates 67 39
San Diego Padres 63 34
Seattle Mariners 61 47
Washington Nationals 59 35
Average  81 46
 

The Seattle Mariners and their 61-101 record provided their bullpen with 47 TSVO, one more than the Philadelphia Phillies and their 92 wins.  While the Philly bullpen converted all 46 of their opportunities into wins, the Mariners were reeling from the loss of JJ Putz and paraded ten different relievers to the mound, converting at a 77% rate.   Perfection is a little optimistic, even though Lidge and his pen mates were just that in ’08.  We would we have expected nine additional wins from the ’08 Mariners if JJ Putz was converting for them at his healthy 2007 40-for-42 rate. 

While there is correlation between team wins and TSVO, it would be much stronger if each team’s conversion rate were the same.  To illustrate the strength of the relationship between TSVO and team wins, let’s compare them to other standard categories:

Correlation Between Wins and x: (2008 Stats)
TSVO   0.70
R   0.59
SLG   0.56
SB   0.54
TB   0.49
HR   0.44
Run Differential 0.43
Batter BB  0.41
AVG   0.34
Batter K’s  0.05
Pitcher Ks  0.03
ERA   -0.15

A value of 1 or –1 would be perfect correlation.  Anything between 0.36 and –0.36 is regarded as not statistically significant, so team batting average, batter and pitchers strikeouts and team ERA have no statistical relationship to team wins.

With that out of the way, let’s look at the TSVO we would expect winning, .500 and losing teams to generate.
     Average TSVO
Top 10 Teams (88+ wins)  52.0
Mid 10 Teams (75-86 wins)  46.3
Bottom 10 Teams (74 wins or less) 39.8

Your average playoff team generated 52.0 true save opportunities for it’s end game bullpen, roughly 12 more opportunities than a bottom dwellar.  Of course LA’s 74 TSVOs were a record-breaking number, equal or better than ten teams’ TOTAL WINS.  For the sake of argument, we take Los Angeles out of the equation and your winning teams generate 9.8 more TSVOs than the bottom feeders. 

This is a great tool to use when evaluating the opportunity of a closer who has been traded.  Has he moved from a loser to a winner?  A lot of prep in fantasy baseball should be focused on the team, not just the player.  Despite being a highly individualistic game, six of ten standard 5x5 categories rely on team performance (R, RBI, W, ERA, WHIP, SV).  With all the free agency moves about to take place, one needs to evaluate the teams, with a specific focus on divisional opponents (pitchers for hitters and vice versa).

A word of caution before we proceed:  We’re only at 0.70 correlation, so bet on your winners all you want, but you may end up with an 89 win ChiSox team that only generates 40 TSVOs.  Bobby Jenks can do everything humanly possible, but all it takes is one lazy slider and he isn’t going to hit 40 saves.
 

Manager Provides Opportunity

Once the team has created a True Save Opportunity, it’s up to the manager to determine who he’ll put on the mound.

The average closer gets 89% of his team TSVO.  If you were to take out some of the injury riddled pens, that number would be closer to 93%.  Factors that prevent a manager from calling on his closer every time are durability, rest days, frequency of close games, schedule, match-ups and extra inning games where the closer has been used already.  There’s also the odd extra inning game where two TSVOs occur if the closer blows the save, the lead is regained and a new reliever comes in to shut the door. 

This chart shows the portion of team True Save Opportunities the closer was given:

CLOSER TEAM CLOSER 
TSVO  TSVO  TSVO% 
34  34  100% HOFFMAN
47  47  100% B WILSON
45  46  98% NATHAN
40  41  98% CORPAS/FUENTES
34  35  97% RAUCH/HANRAHAN
40  42  95% CORDERO
37  39  95% SHERRILL
47  50  94% ACCARDO/DOWNS/RYAN
69  74  93% KROD
38  41  93% SAITO/BROXTON
37  40  93% WILSON/FRANCISCO/GUARDADO
36  39  92% BOROWSKI/BETANCRT/KOBAYASHI/LEWIS
45  49  92% SORIA
51  56  91% VALVERDE
40  44  91% RIVERA
38  42  91% JONES/RODNEY
37  41  90% STREET/ZIEGLER
35  39  90% CAPPS/MARTE/GRABOW
50  56  89% IZZY/FRANKLIN/PEREZ
41  46  89% LIDGE
40  45  89% LYON/QUALLS
46  52  89% PAPELBON
47  54  87% GAGNE/TORRES
34  40  85% JENKS
45  54  83% WAGNER/AYALA
38  46  83% GREGG
47  59  80% PERCIVAL/WHEELER
40  51  78% WOOD
33  47  70% PUTZ/MORROW
20  32  63% SORIANO/GONZALEZ
1231  1381  89%

Terry Francona tends to protect his closer which led to 89% of team TSVO heading Jonathan Papelbon’s way, compared to Ron Gardenhire’s handling of durable Joe Nathan (98% of team TSVO). 

Surprisingly few teams truly went with a closer-by-committee approach.  Most pens had a timeline when the established closer went down with injury and the next guy filled in.  Even Cleveland, with four guys taking turns had a distinctive pattern.  Atlanta had a true committee for during the first half with seven different arms filling in between Soriano’s DL stints and Mike Gonzalez’s return in mid June when he assumed the role.

There were 136 complete games in 2008, many of which could have been save opportunities, but for one reason or another the skipper decided to keep his starter in the game.  These could have potentially been additional opportunities, and manager preference can add or take away a couple TSVOs that your closer gets.  With a 79.6% team conversion rate (23rd in the majors), it’s no wonder the Milwaukee bench bosses of ’08 kept their closer on the pine for the five potential TSVOs while Sheets and Sabathia twirled twelve complete games.

For fantasy players, the manager component is make-or-break for your bullpen.  Your fortunes in the save category depend heavily on what goes on inside the head of your closer’s manager.  This is the most frustrating part of investing in the lower tiers of closers, but also where the most opportunity for profit lies.  If your league allows for a bench, it’s a great place to stash Plan B relievers who enter the season as set-up men for less than reliable closers.  The best GMs key on the Torres’s, Wheeler’s, Fuentes’s and Rauch’s of the world, those guys pitching behind shaky, un-proven or injury prone closers. 

Closer Converts Opportunity

Ahh, the old Rivera-Borowski debate.. Who would you rather have on the hill in the 9th with the game on the line?  Yes it’s true the team has to create the opportunity and the manager has to provide the opportunity to your closer, but if he’s not going to convert it’s not going to do you much good (and he may not be given many more opportunities!).

This chart shows each closer(s)’s conversion rate on the True Save Opportunities they were provided by their manager:

CLOSER CLOSER CLOSER 
CONV  TSVO  CONV% 
41  41  100% LIDGE
39  40  98% RIVERA
42  45  93% SORIA
62  69  90% KROD
33  37  89% WILSON (24/28) FRANCISCO (5/5) GUARDADO (4/4)
41  46  89% PAPELBON
30  34  88% HOFFMAN
30  34  88% JENKS
35  40  88% LYON (26/31) QUALLS (9/9)
41  47  87% B WILSON
41  47  87% ACCARDO (4/5) DOWNS (5/6) RYAN (32/36)
41  47  87% PERCIVAL (28/32) WHEELER (13/15)
39  45  87% NATHAN
44  51  86% VALVERDE
30  35  86% CAPPS (21/26) MARTE (5/5) GRABOW (4/4)
34  40  85% CORPAS (4/8) FUENTES (30/32)
34  40  85% F CORDERO
34  40  85% WOOD
17  20  85% SORIANO (3/4) GONZALEZ (14/16)
32  38  84% SAITO (18/21) BROXTON (14/17)
31  37  84% SHERRILL
31  38  82% JONES (18/21) RODNEY (13/17)
29  36  81% BOROWSKI (6/10) BETCRT (4/6) KOBSHI (6/7) LEWIS (13/13)
36  45  80% WAGNER (27/34) AYALA (9/11)
37  47  79% GAGNE (10/15) TORRES (27/32)
29  37  78% STREET (18/24) ZIEGLER (11/13)
26  34  76% RAUCH (17/22) HANRAHAN (9/12)
29  38  76% GREGG
25  33  76% PUTZ (15/21) MORROW (10/12)
36  50  72% IZZY (12/17) FRANKLIN (17/23) PEREZ (7/10)
1049  1231  85% 

You can see that Tony Larussa aged a few years in 2008.  His closers converted at a league low 72% and they still won 86 games, a glaring weakness for the club.

“Closers”, in the loosest sense of the word converted 85% of the TSVOs they were given.  The 150 non-closer TSVOs were also converted at a rate of 85%.  Hence the fad in recent years for real world GMs not to overpay for closers, rather employ a closer-by-committee.  If you’ve got your ace in the pen and he can’t convert at the league average, he shouldn’t even be in the closer role.  But now that’s the manager’s call and we know what happens when we try and get inside their heads…

The key here is that good closers can create future opportunity for themselves by becoming successful at the position in the eye of their manager.  If the manager likes your closer and he doesn’t make him look like a fool too often, your chucker will continue to receive his share of the team TSVOs, converting them into championship winning saves for your squad.

Wrap Up

This chart shows the expectation of saves earned based on team wins and reliever skill.  I left out the manager component, simply giving 90% of team TSVOs to the closer.  Keep in mind the 80% closer shouldn’t keep his job very long, no matter what kind of team he plays for.

                                       TEAM
                                        TSVO's CREATED
RELIEVER SKILL       WINNER .500 LOSER
(CONVERSION RATE)        52    46      40
90%                             42   37      32
85%                             40   35      31
80%                             37   33      29

In preparing my saves projections, I won’t put as much stock in the team as some of these numbers will tell you to.  There’s too much variance in team performance to begin with, as well as the exceptions to the rule that come up every year (KC, SF and Seattle with 47 TSVOs or more).  Who knew Tampa Bay would provide the second most opportunity for their closers to earn saves in all of baseball? 

I’ll buy a reliever based on my confidence in his skill set (key ratios K/9, BB/9 etc) and roll the dice on his manager sticking with him and his team providing enough opportunity.  We’re basically talking about the bottom tier of closers, ignoring the elite stoppers.  I’ll take a skilled reliever on a bad team ahead of a mediocre reliever on a great team any day because I believe his chances of keeping the job are greater.

I won’t conclude by passing around a petition to re-write the MLB stats, but looking at these numbers should provide a clearer picture to what went wrong last year and new ways to look at the upcoming season.