How I Stream Pitchers

In the past month I've covered what teams to stream pitchers against and how to stream hitters in your fantasy playoffs/roto final weeks. Today, I turn the focus on my personal method of selecting streaming pitchers. In one of my leagues, I have been streaming a starter every day this week, with mixed results. Needless to say, I have been looking at a lot of potential streaming starters, so I thought I would take you step by step through my process.

I should say that there are tons of ways to skin a streamer (or cat) and my method is by no means the "right" method. Our weekend streaming expert Phil and our daily fantasy writers can certainly chime in with their own methods/thoughts. Feel free to let everyone know your technique in the comments, too, unless you are worried your leaguemates will read it.

Ok, on to my method of streaming.

I'm going to use Wednesday's decision in my league as an example for this method.

Step 1: Make a list

This is pretty simple: get a list of all the available pitchers starting today. I just went to my league player list and filtered by "probable pitchers."

Here is a short version of the list that came up:

Step 2: Cross off guys with really bad matchups and guys that are just terrible pitchers

The title of this step says it all. This is still the easy part. I tend to include guys with injury concerns or who are just coming back from injuries on this list. Also, I stay away from rookies making their debuts.

In our example, this step eliminates Morton (matchup), Davies (bad pitcher), Perez (bad pitcher), Tillman (bad pitcher), Finnegan (debut), and Montas (debut). I don't have the guts to start guys with ERAs, FIPs, and xFIPs in the mid-4s or higher. That's who I label "bad" pitchers.

Step 3: Look up the stats

Now, I look at the key stats for each of the remaining starters. I like to use ERA, FIP, xFIP, SIERA, K%, BB%, K%-BB%, SwSTR%, and GB%. A K% under 18% or so, a BB% over 10%, a K%-BB% under 12% (league average for starters is 12.3), a SwStr% under 7%, and a groundball % under 45% (league average) are all grounds for removal, especially if a pitchers hits more than one of those categories.

If it is just one of them and they are good in the other stats, I won't yet eliminate them. An xFIP or SIERA (or both) over 4.5 will also result in removal, as that is too much risk. ERA isn't as important since we are trying to look forward to today's start, not at their results so far. That's why the other stats are all skills-based, not luck-based. Groundballs are in there because they don't have the homerun risk of flyballs. Heavy flyball pitchers can be successful, but only in spacious ballparks and often with a little luck.

Here's our example list after Step 2 with the relevant season-long stats. Note that these stats (from Fangraphs) are all rates and don't favor guys who have thrown lots of innings over guys with just a few.

Applying this step to this new list eliminates: Jake Peavy (K%, GB%), Colby Lewis (K%, GB%, K%-BB%, xFIP), Chase Anderson (K%, GB%, K%-BB%), Phil Hughes (K%, K%-BB%, SwStr%, GB%), and Ivan Nova (K%, K%-BB%, xFIP, SIERA). Now, were getting somewhere.

Step 4: Look at xFIP and SIERA along with Park factors and the opposing lineup

Now that the list is getting short, we can dig deeper. In Step 3, we only used xFIP and SIERA (good ERA predictors) to eliminate the very worst pitchers with a value over 4.5. In this step, we will sort them by one of these two values and start to use those two as determining factors. Let's combine that with looking at the ballpark they are playing in (how pitcher friendly is it?) and the strength of the opposing lineup. Is that lineup good against lefties but bad against righties? Are they just a great offense overall? These things matter.

Here's are new short list with the relevant stats again.

Sorted by xFIP, you can see that Hendricks is clearly the best of this group skills-wise. Fiers is a little further back, but still above average. After that, the xFIPs start to get ugly. Park factors aren't a big factor here since all of these guys are in pitcher-friendly parks, with Miami having a slight edge over the others. Again, the lineups strengths against the handedness of the pitcher don't show much differentiation. Tropeano has the toughest matchup for this category and Fiers the easiest. There aren't any top-10 or bottom-10 matchups here.

Step 5: Make your pick(s)

At this point, there may not always be a clear difference between the remaining options. In our example, Hendricks is clearly the best and is, in fact, the one I started in my league. Often, you will be choosing from Elias, Conley, and Tropeano-types. In that case, I would look hard at the park and lineup factors and select Conley because he is playing in the most friendly park against the worst lineup. If it came down to Elias and Tropeano, I would go back to the stats in step 3 and see that Tropeano is heavy on fly balls and Houston has a short left field porch, so I would choose Elias. It can come down to a coin flip in those cases, sometimes.

Well, that's my method. It works well sometimes and it fails miserably sometimes, but I can sleep at night knowing I had a plan and based my decision on the numbers and just got unlucky. Whatever method you use, stick with it and make sure you can make peace with every outcome knowing you made the best decision you could at the time. Sometimes, you just can't predict ball. Tschus!