I'm a friend posting under JFerg's account.
First, we need to make sure we understand how to evaluate baseball players. Wins are the primary unit of measurement, with runs being the secondary unit of measurement. Basically, in your typical baseball season, every 10 runs or so will create one win for your team. This depends on the run scoring environment though. The run scoring environment for this league seems to be a little higher than normal, but this doesn't matter when comparing two players within one season, but instead when comparing two players in two different seasons.
So now we're looking for how many runs each pitcher saved. The most basic stat to look at is ERA. If you want to attribute a pitcher's ERA fully to himself, you have every right to do so. But let's break down what ERA is comprised of. Basically, it involves three things. They are:
1. Balls not in play (K, BB, HBP, HR) These are almost 100% under the pitcher's control.
2. Balls in play (All non-HR hits and outs recorded by fieders) I don't think anyone would argue that the pitcher entirely controls these plays
3. Sequencing (Out, out, 2B, BB, BB, out vs. BB, BB, 2B, out, out, out)
It's up to everyone individually to determine how much of these three to use in evaluating pitchers. If you're just using FIP, you're ignoring #2 and #3. If you're using ERA, you're using all three. If you use ERA but try to factor out the differences of defense behind each pitcher, your using #1, #3, and some of #2.
For #1, Ducey did outpitch Sheldon. You can just use FIP and see that this is true. My guess is that Sheldon's ERA advantage comes from #2, and maybe #3. I don't have the BABIP numbers, but my guess is Sheldon did better on balls in play. Here's the problem with that in real life. At the seasonal level, most of the difference between BABIP is due to luck. Or if the luck doesn't sit well with you, call it random variation. Here is a link to see what I mean: http://www.insidethebook.com/ee/index.php/site/article/do_you_want_to_bet_rob_neyer/
As you can see, there is a tiny amount of skill in terms of balls in play, but it takes a large amount of data to accurately see that skill. Now, I've been told that BABIP differences are larger than real life in this simulation. I'm in a good mood though, so I'm going to give 100% of that difference to the pitchers. Which means I'm not even going to consider the differences in defense for both teams. Same thing goes for sequencing. Whatever difference there was, I'm going to fully attribute it to each pitcher. Thus, I'm solely going to use ERA in comparing both pitchers.
Another poster pointed how many runs each saved compared to average. And they were identical. What this means is, when using an average baseline, both pitchers saved the same amount of runs for their team. However, in real life, baseball talent is NOT evenly distributed. There is not a bell curve when in comes to MLB talent. What this means, is that there are far, far more below average players than above average players. There are far, far more below average players than exactly average players. What this means is that being below average still has value. I'm going to assume that this non-even distribution of talent also exists in this league.
What this means, is that the average baseline is too high. You're missing out on some of their value. The replacement level for a starting pitchers is about 70% below average. Using this baseline, Ducey actually has saved about 10 more runs than Sheldon through out the season. For anyone trying to remove Ducey's IP advantage, you are wrong, at least in real life. Basically, Sheldon plus 81 average innings = Ducey. But, if this league is similar to real MLB, average is hard to come by. Meaning there is value in those 81 extra innings.
Last thing we need to look at is home parks. From what I'm told Shledon pitches in a pitcher's park and Ducey pitches in a hitter's park. I don't have the exact increases in scoring for each park, so I'm just going to assume that it's 5% for both parks. That is, one park increases scoring by 5% and the other decreases scoring by 5%. This may not seem huge, but it actually creates another 10-difference between the two players. Thus, Ducey is now 20 runs better than Sheldon. Or approximately 2 wins above replacement better.
I used only four things in making this evaluation by the way. They were ERA, IP, replacement level, and differences in their home parks. Bottom line, Sheldon saved runs at a better rate, but pitched less innings, and did so in a pitcher's park. The math says Ducey had a better season. You can't argue with the math, only with the assumptions. If you knock out using replacement level, then Ducey still wins based on being equal compared to average but doing so in a tougher park. If you knock out the park advantage, they're equal. So you have to knock out the innings advantage, and look only at ERA to get Sheldon being more valuable than Ducey.
I'll be more than happy to answer any questions.