Top Starter or Top Closer?

In response to my post last week, a commenter at The Hardball Times mentioned that he doesn’t need the help in ERA and WHIP benefits of a top closer because he gets that from a top starter. I’m going to explore that approach in this post. I determined my top starters by taking the same league and using the first ten starters by draft position, which is the same way I determined the top closers last week. Those starters were Johan Santana, Jake Peavy, Josh Beckett, CC Sabathia, Erik Bedard, Brandon Webb, Justin Verlander, Cole Hamels, John Smoltz, and Felix Hernandez.

One thing I noticed immediately is that the risk among SPs is close to the risk among RPs, which flies against part of the argument for drafting a top starter and not drafting a top closer (i.e. the idea that top closers are riskier than top starters). Bedard and Smoltz weren’t able to stay healthy like J.J. Putz, Billy Wagner, and Takashi Saito. Beckett, Verlander, and King Felix didn’t live up to performance expectations like Bobby Jenks. Now, this is only one year’s worth of data, but it seems to me that drafting top starters is just as risky as drafting top closers. When you consider that the top starters averaged a third-round pick versus the top closers’ average sixth-round pick, you could argue that the top starters are riskier than the top closers, but remember that it’s only one year’s worth of data.

To make the statistical comparison, I want to directly compare the top closers to the top starters, but first, I need to even out the innings pitched column. I’m going to do so by adding a replacement-level starter’s stats to the top closer. (The IP didn’t completely add up, so I normalized the replacement-level starter’s stats from 114.4 IP to 118.0 IP.) Here are the stat lines for the three player types and the sum of the top closer and replacement-level starter.

Top Starter: 180.0 IP, 12.0 W, 0.0 SV, 162.0 K, 3.31 ERA, 1.21 WHIP

Top Closer: 62.0 IP, 3.8 W, 34.9 SV, 67.2 K, 2.50 ERA, 1.11 WHIP

Replacement-Level Starter: 118.0 IP, 6.9 W, 0.0 SV, 88.9 K, 4.40 ERA, 1.38 WHIP

Closer + Replacement Starter: 180.0 IP, 10.7 W, 34.9 SV, 156.1 K, 3.74 ERA, 1.29 WHIP

So, in the same 180 IP, the top starter gets 1.3 extra wins, 5.9 extra strikeouts, a much better ERA, and a much better WHIP while giving up 34.9 saves. You can determine the value there for yourself, but I’ll continue on to show what my draft system suggests their values are.

I’m going to repeat my strategy from last week of replacing an average player on an average team with each option. Using a maximum inning limit of 1400, here are the results.

Average team with top starting pitcher: 83.5 W, 77.6 SV, 1086.6 K, 3.83 ERA, 1.30 WHIP

Average team with top CL and replacement SP: 82.2 W, 112.5 SV, 1080.7, 3.89 ERA, 1.31 WHIP

The differences in wins, strikeouts, and saves stay the same as you’d expect. The ERA and WHIP differences come closer together as you’d expect as well. (Note: If your IP limit is smaller, these differences would expand, increasing the value of the SP in these categories.) Now for the point values compared to the top closer, the top SP is worth 0.42 more points in wins, -3.36 in saves, 0.19 in strikeouts, 0.79 in ERA, and 0.77 in WHIP. As you can see, the difference in saves far outweighs the other differences, a margin of 1.19 in favor of the top closer and replacement-level starter approach. When you add in the fact that the top closers went 3 rounds later on average, it’s obvious that this is the better approach. Now, if you are extremely confident in your ability to tell apart the good, cheap closers from the poor, cheap closers, then I’d say you should continue to ignore the expensive closers. On the other hand, if you end up selecting guys like C.J. Wilson last year, you might want to pay for saves.

 ----------------------------------------------------------------------

For full disclosure, I usually concentrate on hitters for the first 6 rounds before taking a pitcher at all, but I feel like I can pick the good, cheap closers and the good, cheap starters out of free agency throughout the season. Obviously, this approach varies based on the league settings. For example, I just joined a head-to-head league that uses holds instead of WHIP, so I plan on ignoring pitchers altogether until my lineup’s full. Consider that you can win the W, SV, K, and HLD categories with a large quantity of pitchers, so by sacrificing ERA, I can win without quality pitchers. If I had to sacrifice ERA and WHIP, I couldn’t follow this strategy. I also plan on following the same strategy on offense by sacrificing steals because steals either have a negative or a poor positive correlation with the other categories. By ignoring steals and ERA, I can build up a strong team in the other 8 categories and shoot for an 8-2 victory every week.