wins per inning pitched: are starters necessary?

Posted: Tue Mar 21, 2006 12:34 am

i sometimes wonder about starters. you need them, they say. for wins. but do you?

in a roto league, you're working with an innings cap. if you know what you're doing, you're going to hit the cap, and you're going to hit it just about on the nose. so it doesn't exactly matter how many strikeouts your pitchers get - it matters how many strikeouts they get per inning. efficiency.

the same goes for wins, i figure. you want to cram as many wins as you can into those 1250 (or whatever) innings that make up your pitching for the year. so who gets the most wins per inning? i decided to check.

i didn't do every pitcher. i just did a sort of random sampling, of guys who seemed relevant: the SP wins leaders, a smattering of mid-range winners among the SP, the RP wins leaders, some closers, and some other random RPs who got less wins. so this is hardly scientific. but it gives you a sense of scale.

Quote:

Pitcher - Innings per Win (semi-random sample)
J Crain - 6.64
B Donnelly - 7.26
S Shields - 9.17
S Linebrink - 9.21
B Colon - 10.60
D Willis - 10.74
M Rivera - 11.19
C Lee - 11.22
M Timlin - 11.47
C Carpenter - 11.51
R Halladay - 11.80
J Beckett - 11.91
R Oswalt - 12.09
J Duchscherer - 12.24
J Rincon - 12.83
J Bonderman - 13.50
P Martinez - 14.47
J Santana - 14.48
R Madson - 14.50
M Clement - 14.69
B Backe - 14.93
D Marte - 15.10
M Prior - 15.15
D Haren - 15.50
J Peavy - 15.62
H Street - 15.66
B Sheets - 15.67
C Zambrano - 15.95
L Hernandez - 16.42
B Myers - 16.56
J Blanton - 16.78
R Betancourt - 16.93
G Maddux - 17.31
J Vazquez - 19.61
J Patterson - 22.03
D Hermanson - 28.55
A Otsuka - 31.10
D Wheeler - 36.55
B Fuentes - 37.05

as you can see, the top winning relief pitchers aren't just a "nice supplemental source of wins," they're far and away the best source of scoring in the category.

however, there is, of course, a fly in the ointment, which is that you don't know which relief pitchers are going to be the big winners in any given year. note that the bottom-feeders in this area are also a couple of perfectly serviceable relievers who just happened not to get many wins (two each, to be specific). this happens sometimes.

of course, in a similar vein, it's notoriously difficult to predict who the big winners are going to be among the starters in any given year. several SPs who had great years were terrible winners anyway. i'm looking at you, John Patterson. but based on role, you should be able to estimate which relievers are reasonably good candidates for wins. setup guys on good teams who are good pitchers, basically, are the guys you want. out of this group, even most of the ones who don't run up a fluky win total (hi crain), like duchscherer and madson, are still right there among the pedros and santanas and oswalts in terms of wins per inning.

now and then you'll crap out with an otsuka. is that enough to cancel out your ill-gotten crain/linebrink gains? probably not. it deserves a more thorough study, but i didn't see a lot of good setup guys who got fewer than about 5 wins - that's Huston Street-level on the chart, still between peavy and zambrano, who were disappointing winners based on the quality of their pitching, but it just goes to show, SPs are just as liable to crap out on you as relievers are.

of course, you need some SPs, or you're never going to hit the innings cap at all. that's what holds this strategy back from just up and taking over the fantasy world - the simple fact that you don't have 15 roster spots and 15 active P slots to run your phalanx of timlins out there every day. but it seems like you could do a lot worse than carrying a few middle relievers. if you can spare the roster room, three linebrinks look like this:

3 linebrinks: 220 IP, 24 W, 210 K, 1.83 ERA, 1.06 WHIP
Johan Santana: 230 IP, 16 W, 238 K, 2.87 ERA, 0.97 WHIP

clearly you could go into a lot more depth on this, but i just thought it'd be a good thing to chew on.

Posted: Tue Mar 21, 2006 10:36 am

some more data: the relationship between wins and holds. here are last season's holds leaders (everybody who got 20 or more), in order of total holds, with how many wins they got.

Quote:

T Gordon, NYY - 5
S Shields, LAoANA - 10
J Tavarez, STL - 2
S Eyre, SF - 2
R Madson, PHI - 6
B Howry, CLE - 7
S Linebrink, SD - 8
J Rincon, MIN - 6
M Timlin, BOS - 7
G Majewski, WAS - 4
C Politte, CWS - 7
D Marte, CWS - 3
L Ayala, WAS - 8
A Otsuka, SD - 2
C Qualls, HOU - 6
R Villone, SEA - 5
U Urbina, DET - 5
S Schoeneweis, TOR - 3
J Putz, SEA - 6
K Mercker, CIN - 3
J Borowski, CHC/TB - 1
M DeJean, NYM - 5

average # of wins: 5.1
average IP / W : 13.73

these are the guys you didn't want:

Quote:

J Tavarez, STL - 2
S Eyre, SF - 2
A Otsuka, SD - 2
S Schoeneweis, TOR - 3
K Mercker, CIN - 3
J Borowski, CHC/TB - 1

out of them, only tavarez was on what you would really call a good team.

General Manager Joined: Jan 19, 2006 Posts: 234

thats some really interesting stuff there.

I was trying to see how I could incorporate your hard researching work to benefit me in my league (10 team CBS head-2-head) but we can only start 4 SPs and 2 RPs....and saves is a category so I need to go with 2 closers so I can compete in saves each week.

I wish I could find a way to work that magic.....I just dont see how.

Posted: Tue Mar 21, 2006 11:44 am

yeah, a lot depends on your particular league parameters. it's probably most practical in a league with deep benches and something like 3 SP slots / 3 RP / and then 2 or 3 generic P slots, where you run your closers at RP and then the setup guys in the P slots, and just shuttle your SPs in and out of the SP slots as they pitch.

i never really bought into the Scot Shields thing before; i've spent a lot of time on here telling people to drop setup guys. but if you look at the numbers it works. if you can spare the roster spot. i think the ideal is to get three of them to replace one starter altogether and run a shorthanded starting staff, so that you can invest even more heavily in offense. the revelation here, to me, is that you're not sacrificing wins when you do that; you might actually be gaining wins. the starter you're replacing isn't one of your Kelvim Escobar types, you can actually build a sort of frankenstein three-headed ace who takes up three roster slots but puts up johan numbers and doesn't cost you anything in the draft.

General Manager Joined: Jan 19, 2006 Posts: 234

Youre absolutly right. I have a Yahoo league draft coming up Sunday that has 4 P spots, in addition to 2-SPs....if I get a stud closer, I may try at least a two-headed monster of Sheilds-Linebrink or something like that. It could be the small difference I need.