I've decided to post my thesis on starters vs. relievers once and for all, so I have a URL to point to when people say that relievers don't help your ERA and WHIP.
Postulate: The top closers (and middle relievers) help your ERA and WHIP approximately as much as the top starters.
To prove this, I have created stats called ERA Imp and WHIP Imp. I have not seen anyone else use these stats, and as far as I know, they are my own invention. ERA Imp measures how much a player's ERA will affect the "average fantasy team", by taking into account both ERA and IP.
For purposes of this proof, I will consider the "average fantasy pitcher" to pitch 163 and 2/3 innings, and to have an ERA of 3.355. This is based on an (admittedly) abitrary sample of the top 100 pitchers, based on last year's stats. I think these were the top 100 point-getters on Sportsline. Something like that.
For purposes of comparing any two pitchers, your staff will consist of 7 pitchers. The first 6 pitchers will be John Mediocre, a hypothetical pitcher who throws 163.67 innings and has an ERA of 3.355. The 7th pitcher is Pitcher X, that is, whoever we are trying to determine the value of.
Let's take Gagne as an example. The 6 John Mediocres will pitch a total of 982.0 innings, and will give up 366 earned runs. Gagne, last year, gave up 11 ER in 82.33 inings. Adding him to your staff will give you an ERA of 3.19, an improvement of .165.
What about Prior? Prior gave up 57 ER in 211.33 innings. Adding him to your staff of John Mediocres will also give you an ERA of 3.19, the same improvement of .165.
I computed this number for all of the top 100 pitchers, and, surprise, the top relievers are very close to the same ERA Imp and WHIP Imp as the top starters. I will post the entire list in the next post.
I believe that this proves my point.
Possible objections:
1) Most staffs have more than 7 pitchers, and you rotate in your starters to get better matchups out of them.
Answer: Yes, but your stud starters stay in for all matchups. And I am only trying to prove the relative value of stud starters versus stud closers. So this doesn't apply here.
2) My staff had a much better ERA than 3.19. OR My staff had a much worse ERA than 3.35.
Answer: The staff ERA is highly dependent on league size and pitching staff size (the bigger either one is, the higher your ERA will be, as you have to rely on worse starters). I have no data that suggests this makes a difference in relative value of closers or starters.
<pre>
ERA Imp WHIP Imp
P. Martinez -0.181 -0.025
J. Schmidt -0.177 -0.043
K. Brown -0.171 -0.010
E. Gagne -0.166 -0.039
M. Prior -0.164 -0.017
J. Smoltz -0.137 -0.020
G. Mota -0.134 -0.020
D. Marte -0.132 -0.011
R. Cormier -0.131 -0.021
S. Hasegawa -0.130 -0.007
T. Hudson -0.129 -0.025
B. Wagner -0.127 -0.026
B. Donnelly -0.124 -0.009
M. Rivera -0.113 -0.013
L. Hawkins -0.109 -0.008
K. Foulke -0.103 -0.025
R. Soriano -0.093 -0.021
E. Loaiza -0.085 -0.016
B. Webb -0.080 -0.007
O. Dotel -0.071 -0.018
S. Shields -0.066 -0.001
C. Schilling -0.059 -0.021
H. Nomo -0.048 0.010
J. Borowski -0.047 -0.010
R. Oswalt -0.044 -0.007
C. Zambrano -0.044 0.022
J.Isringhausen -0.041 -0.001
J. Beckett -0.040 0.016
U. Urbina -0.040 -0.005
J. Santana -0.039 -0.013
M. Mantei -0.039 -0.010
M. Mulder -0.036 -0.003
T. Worrell -0.036 0.008
F. Cordero -0.032 0.009
L. Hernandez -0.030 0.002
J. Nathan -0.029 -0.010
E. Guardado -0.029 -0.014
K. Wood -0.027 -0.001
A. Benitez -0.027 0.012
F. Rodriguez -0.026 -0.017
R. Halladay -0.022 -0.027
J. Vazquez -0.022 -0.017
J. Moyer -0.015 0.006
T. Gordon -0.014 0.000
K. Wells -0.012 0.009
T. Hoffman -0.012 -0.001
B. Zito -0.010 -0.003
D. Willis -0.008 0.012
B. Kim -0.005 -0.009
T. Percival 0.006 -0.003
M. Mussina 0.008 -0.021
B. Looper 0.025 0.014
M. Batista 0.030 0.022
D. Baez 0.032 -0.003
M. Redman 0.038 0.004
R. Franklin 0.038 0.006
C. Sabathia 0.041 0.017
V. Padilla 0.046 0.008
M. Morris 0.060 -0.003
B. Anderson 0.071 0.016
S. Ponson 0.071 0.011
D. May 0.073 -0.001
S. Trachsel 0.073 0.019
J. Pineiro 0.075 0.013
R. Ortiz 0.081 0.020
R. Johnson 0.094 0.014
W. Williams 0.094 0.010
R. Clemens 0.098 0.002
A. Leiter 0.099 0.045
B. Colon 0.102 0.001
G. Maddux 0.110 -0.003
A. Pettitte 0.116 0.023
K. Millwood 0.121 0.010
W. Miller 0.124 0.018
J. Peavy 0.125 0.019
T. Wakefield 0.126 0.018
M. Clement 0.128 0.006
B. Penny 0.129 0.014
T. Ohka 0.136 0.034
D. Wells 0.140 0.006
J. Suppan 0.144 0.019
K. Escobar 0.145 0.044
B. Lawrence 0.147 0.009
R. Wolf 0.148 0.012
M. Buehrle 0.149 0.029
T. Lilly 0.151 0.020
C. Pavano 0.161 0.011
B. Myers 0.177 0.043
O. Perez 0.185 0.013
J. Washburn 0.187 0.009
J. Garland 0.188 0.028
D. Lowe 0.191 0.038
F. Garcia 0.196 0.023
B. Sheets 0.201 0.010
K. Rogers 0.201 0.037
B. Radke 0.202 0.013
K. Lohse 0.213 0.012
J. Lackey 0.219 0.038
J. Thomson 0.271 0.019
R. Ortiz 0.286 0.048
</pre>
<p>
I agree with you on this one, and in roto leagues with an IP maximum, relievers are even more important, the wins and Ks categories effectively become W/9 and K/9.
A good middle reliever, who may well win about 5 games over a season in aroung 75 IP is equivalent to 12 wins over 180 innings, which is as much as some 4th or 5th starters on fantasy squads.
And by their very nature, most relievers will have a higher K/9 ratio than average starters.
In your example comparing Gagne's effect to Prior's effect, you did not point out that the reason they have the same effect is that Gagne gave up .1336 ER per inning, and Prior gave up .2697 ER per inning. The relief pitcher needs a way better ratio than a starter to make the same impact on a fantasy team's overall ERA. The closer they are together, the more the starter's impact will be.
Relievers have smaller sample sizes as well, so if the relief pitcher gives up even a few runs he didn't in the prior year, he will not provide the expected help. The starting pitcher probably has more chance of being close to his expected ERA than a reliever has.
Eric, you are absolutely correct about Gagne's effect compared to Prior's effect. And I didn't make that clear in the original post. Gagne was only as effective as Prior in lowering your team ERA because he had a much lower ERA than Prior did. Your average closer will not help much in team ERA. On the flip side, your below-average closer has a similar ERA to your below-average starter, and thus will not hurt your ERA much. A closer would have to have about a 5.5 ERA in order to hurt you as much as a 4.2 ERA starter... or something like that, I didn't really run the numbers on this.
As for the small sample size... well, you do have a point, but Gagne has been so close to perfect the last two year, that I don't expect him to screw things up any more than I expect Prior to screw things up. But again, you do have a point on this.
Thanks, Eugene, I wanted to know if my logic was off somewhere. And by the way, I tend to agree with you on Gagne. There are just so few relievers that can have that kind of impact, and the history to back up a continuation of that prowess. Remember, Gagne's only done it for two years, but what a two years!
I think you posted this info in another thread where a guy was basically arguing that Gagne was a one category player and wouldn't accept that relievers help a ton in ERA and WHIP.
Anyway, thanks for the stats. Hopefully this will stop people from talking about relievers not being as useful as starters.
EugeneStyles wrote:Any more? I will post results in the next post.
first-awesome post. what a pleasure to read!
a question-I assume that if your league pitches less pitchers, then you would argue a reliever has even more of an impact? My league has 5 Pitchers, 4 starters, 1 reliever.
To me the real question here is does a Middle Reliever positively impact a Fantasy team, in that while a Dotel or Donnelly can positively impact Whip and Era, is it worth the tradeoff in Saves?
larry
EugeneStyles wrote:Postulate: The top closers (and middle relievers) help your ERA and WHIP approximately as much as the top starters.
To prove this, I have created stats called ERA Imp and WHIP Imp. I have not seen anyone else use these stats, and as far as I know, they are my own invention. ERA Imp measures how much a player's ERA will affect the "average fantasy team", by taking into account both ERA and IP.
oh oh The Cat's out of the bag now. I've been using your postulate and the same type calculations for years (it's ERA"F"(factor) and WHIPF on my spreadsheet) For the most part, my leaguemates never believed me when I'd try and tell them that elite closers do indeed help the ratios THAT MUCH.
Great post EStyles, but it just got a little harder to beat those guys