The Fantasy Baseball Sherpa

In leagues that allow weekly lineup changes many fantasy baseball team owners choose to rotate different starting pitchers each week through the same roster spot. Commonly known as streaming, this practice usually involves looking for pitchers with either two starts or one very favorable match-up during the upcoming week.

How do you assess which pitchers have the most favorable match-ups? There are three main criteria to consider: (1) the quality of the pitcher you’re considering, (2) the quality of the pitcher’s probable pitching opponent(s), and (3) the opposing team’s strength in the areas relevant to your league’s pitching categories.

Most fantasy team owners are fairly confident in their ability to assess the first two factors. However, most owners don’t give sufficient consideration to the opposing team’s capabilities. Some owners make a quick, qualitative judgment (e.g. – the Royals’ offense stinks) and move on. Others focus on a statistics that either don’t tell the whole story by themselves (e.g. – batting average, number of runs scored), or worse yet, may have limited relevance (e.g. – number of home runs hit, unless HR Allowed is one of your pitching categories).

Fortunately, teams’ strength in the areas relevant to your league’s pitching categories can be quantified and used to your advantage. Let’s assume that you’re playing in a league that uses the standard five pitching categories (Wins, Saves, ERA, WHIP, and Strikeouts). Obviously, Saves aren’t a consideration in selecting your starting pitchers. We want to find teams that combine a low number of wins, runs scored, and hits + walks with a high number of strikeouts.

The Sherpa Points system involves assigning each team a score between 0 and 1.00 in each category, then summing across categories to get an overall score. The calculations are straightforward for counting categories in which we want the opposing team to have a low score (e.g. – Wins, Runs Scored, Walks+Hits). The team with the highest total in each category receives a score of 1.00; a team with 75% of the leading team’s total would receive a score of .75, a team with 50% of the leading team’s total would receive a score of .50, etc. For counting categories such as Ks (where lower values are better for hitting teams) or average-based categories such as K/BB or K/9, you can choose a proxy statistic that can be minimized. Again, we want to be able to minimize all the opponents’ statistics that we’re considering.

I performed season-to-date rankings based on this system for the 30 teams through the games of Thursday, 6/5/08. If you’re trying to find the best teams to stream a starting pitcher against, you want the teams with the lowest total scores. Here are the results of my study (ranked from most favorable team to pitch against to least favorable):

1. Kansas City Royals 2.96

2. Seattle Mariners 2.98

3. Washington Nationals 3.00

4. Colorado Rockies 3.07

5. San Francisco Giants 3.08

6. San Diego Padres 3.10

7. Cleveland Indians 3.15

8. Baltimore Orioles 3.18

9. Detroit Tigers 3.19

10. Los Angeles Dodgers 3.26

11. Florida Marlins 3.33

12. Milwaukee Brewers 3.33

13. Cincinnati Reds 3.34

14. Houston Astros 3.37

15. Arizona Diamondbacks 3.37

16. New York Mets 3.37

17. Chicago White Sox 3.39

18. New York Yankees 3.39

19. Tampa Bay Rays 3.40

20. Minnesota Twins 3.41

21. Pittsburgh Pirates 3.42

22. Oakland A’s 3.43

23. Toronto Blue Jays 3.45

24. Los Angeles Angels 3.46

25. Atlanta Braves 3.59

26. St. Louis Cardinals 3.74

27. Philadelphia Phillies 3.75

28. Texas Rangers 3.78

29. Boston Red Sox 3.87

30. Chicago Cubs 3.97

There are a number of surprises on this list (e.g. – Twins, Pirates, A’s, and Angels), but it can serve as a fairly informative guide if you’re trying to decide among several starting pitchers that you have rated fairly evenly for the upcoming week.

This approach can be used for either points-based leagues or roto-style leagues; just be sure that the categories you choose are relevant to your league. For example, if your league also uses Home Runs Allowed as a pitching category, Baltimore, Florida, and Milwaukee would become less desirable opponents.

You can update this analysis as often as you see fit. I don’t expect that the results would change significantly from week to week, but I would expect meaningful changes over time periods of a month or longer.

And oh yes, the Royals’ offense does indeed stink.

 

Until next time,

The Sherpa

Many fantasy baseball team owners engage in the practice of streaming - i.e. - identifying and placing players in their lineups who have favorable match-ups for the next scoring period rather than going with the same lineup each week. While streaming can be used for hitters, it is used most often with starting pitchers, especially with those who have two starts scheduled during the upcoming week.

Some owners are content to go enter the season with a staff of mediocre starting pitchers; these owners make a decision to spend most of their auction dollars or early-round draft picks on hitters. For these owners streaming their starting pitchers is a deliberate strategy. However, many owners are forced to stream, either due to injuries or the trade of a pitcher out of the AL or NL in an “only” league.

Either way, owners who stream at least one of their starting pitching spots are seeking to optimize the production from that spot. How should you go about doing this? Obviously, the underlying ability of the available pitchers is one key criterion, but depending on how deep your league is, there often isn’t much to choose from on the free agent list/waiver wire.

The second criterion, which doesn’t get as much attention as it should, and typically varies much more than the ability of the available starting pitchers, is the underlying ability of a pitcher’s upcoming opponents. Owners often make quick, qualitative stabs in this direction (e.g. - “Great! This pitcher has two starts this week, and they’re against Kansas City and Seattle!”), but rarely do owners actually seek to quantify their intuition.

One method to rank teams’ offenses involves identifying key offensive categories, such as Runs Scored (R), Home Runs (HR), Stolen Bases (SB), On Base Percentage (OBP), and Strikeouts (K), then seeing how the 30 teams compare in each of these categories and in aggregate. The Sherpa Points system involves assigning each team a score between 0 and 1.00 in each category, then summing across categories to get an overall score. I prefer to weight the categories evenly, but you could easily modify this method to give the categories varying weights.

The calculations are straightforward for the “counting categories” such as HR, R, and SB: the team with the highest total in each category receives a score of 1.00; a team with 75% of the leading team’s total would receive a score of .75, a team with 50% of the leading team’s total would receive a score of .50, etc. For counting categories such as Ks (where lower scores are better) or average-based categories such as OBP, a proxy statistic must be used. I realize that there is a fair amount of overlap among the 5 categories I’ve chosen, but I have not made any adjustments to correct for these correlations.

I performed the year-to-date rankings based on this system for the 30 teams through the games of Saturday, 5/24/08. If you’re trying to find the best teams to stream a starting pitcher against, you want the teams with the lowest total scores. Here are the results of my study (ranked from most favorable team to pitch against to least favorable):

1. Washington Nationals 3.10
2. San Diego Padres 3.18
3. Kansas City Royals 3.19
4. Cleveland Indians 3.35
5. Oakland A’s 3.49
6. New York Yankees 3.53
7. Baltimore Orioles 3.58
8. Pittsburgh Pirates 3.59
9. Chicago White Sox 3.60
10. Toronto Blue Jays 3.61
11. San Francisco Giants 3.63
12. Minnesota Twins 3.66
13. Atlanta Braves 3.68
14. Arizona Diamondbacks 3.69
15. Detroit Tigers 3.74
16. Seattle Mariners 3.76
17. St. Louis Cardinals 3.77
18. Florida Marlins 3.82
19. Colorado Rockies 3.83
20. Milwaukee Brewers 3.85
21. Los Angeles Dodgers 3.86
22. Cincinnati Reds 3.92
23. New York Mets 3.92
24. Los Angeles Angels 3.96
25. Tampa Bay Rays 4.05
26. Texas Rangers 4.23
27. Chicago Cubs 4.24
28. Philadelphia Phillies 4.32
29. Boston Red Sox 4.43
30. Houston Astros 4.46

There are a number of surprises on this list, but generally, it can serve as a fairly informative guide if you’re looking to select a starting pitcher to stream for the coming week or two. The results can be used to guide your decisions in points-based leagues as well as roto-style leagues.

I’ll update this analysis several times during the course of the season, and as promised in my previous post, during the coming week I will begin forecasting remainder of season stats for a number of players off to surprisingly good/bad starts.

Enjoy your holiday weekend!

The Sherpa