The Z Files: Weekly Rankings Primer

Once again this season, I'll be generating the site's Weekly Pitcher Rankings and Weekly Hitter Rankings. They'll be posted Saturday with a refresh based on pitching changes late Sunday night. Here is a primer, explaining the process for each so you can maximize utility.

Weekly Pitcher Rankings

The rankings are designed for standard 5x5 rotisserie scoring. They reflect how each pitcher helps a team that week relative to everyone else scheduled. New this season is the inclusion of an average reliever, so you can gauge when it's best to insert a reliever into your lineup in lieu of a starter with a potentially damaging matchup.

From a strategic perspective, it makes sense to be conservative early, building up a foundation then streaming more liberally down the stretch if you need to chase points in wins and strikeouts. However, that is opposite of the approach dictated by history. Here is a chart showing starter's ERA by month from 2017-2019:

Clearly, being aggressive early is the most efficient plan of attack. It's not a coincidence the ERA trend dovetails with average temperature by month. The caveat is the lack of trustworthy player and team data by which to identify the lower-level arms worth acquiring.

Another consideration is the number of two-start options will likely be lower than usual, shifting the focus to single-start hurlers with superior matchups. Since 2018 (obviously excluding last year's special circumstances), the season has started on a Thursday, adding several more off days to the schedule. This decreases the number of seven-game weeks, reducing the two-start candidates. In addition, several clubs are deploying six-man rotations and/or bullpen games, further depleting the two-start inventory.

The rankings begin with my personal expectation for each pitcher. This is continually updated throughout the season. This baseline then goes through a series of adjustments based on matchup.

Venue

Projections assume half home and half away starts. The first step is neutralizing the projection, then adjusting to reflect the site of the outing(s).

Home/Away

This effect has been dwindling as umpire scrutiny increases, but there is still a small home field advantage.

The baseline projection is altered accordingly.

Opponent

The final step with respect to adjusting the projection is accounting for the opposing lineup. Each stat is adjusted based on the quality of opponent, including hits, homers, walks, strikeouts and runs. This is done by handedness, so the opponent's numbers versus left-handers and right-handers are incorporated. Early in the season, this is more projection than based on actual performance, but as the season progresses, the in-season data takes over.

Length of Start

This is the only subjective element to the process, as I project the number of innings for each start. It's based on pitcher history and quality of opponent.

Wins

Once the innings are projected, the balance is assigned to the team's bullpen and the total runs allowed are calculated. Runs scored by the starting pitcher's team is just the runs allowed by the opposing pitcher. A modified Bill James Pythagorean calculation predicts the team's chances for a win. The deeper the pitcher works into the game, the better his chances of garnering a win. If his chance for a win is 54%, he is assigned .54 wins.

Once I have a projection for each scheduled start, the individual numbers are summed to handle those with two starts. Let's say there are 150 pitchers scheduled to work in a given week. I have 150 projections for innings, wins, strikeouts, hits allowed, walks and runs allowed – everything involved in 5x5 scoring. While it is a little more complicated than this, in essence, the five categories are ranked roto-style with the pitcher totaling the most points being the best option for the week.

As mentioned earlier, an average reliever will be included in the rankings. The idea is that rostering a non-closing reliever could be more valuable than a risky starter. Mixed leagues will include a 12-team and 15-team proxy, while the American and National League rankings will contain an average reliever inherent to those formats.

Finally, providing rankings for points leagues is impractical for couple of reasons, the most important being the sheer number of different scoring systems. Plus, all you need to do is estimate how each pitcher will do, then calculate the representative number of points and compare.

Weekly Hitter Rankings

Teams are ranked in several categories, detailed below. The system assigns a value of 100 (neutral) to the average team relative to the others. Anything above 100 favors batters, while scores below are unfavorable.

The final three columns are averages: left-handed batters, right-handed batters and overall. It's a weighted average of all the components, designed to incorporate the relevance of each in 5x5 scoring.

Full disclosure, I never use these final rankings. They were added by popular demand, because

1. The customer is always right

2. When the customer is wrong, see No. 1

Admittedly, "wrong" is a bit harsh. Personally, I prefer to use the ranking specific to the type of players I'm considering. For example, if debating between Myles Straw and Bobby Dalbec, I don't care about the Astros' home run park factors and the quality of catchers the Red Sox will face. I want to know how easy it will be for Straw to swipe some bags and Dalbec to leave the yard. To facilitate this analysis, the hitting tables are sortable by column heading.

Games

Very straightforward, the total number of scheduled games is presented, also broken down by home/away and facing LHP/RHP. As simple as it is, this could be the most useful information on the chart.

HR Index

The weighted average for the week's venues is presented for left-handed and right-handed batters. For example, if a team plays four games in a 110 park and three in a 95 venue, the index is (4 x 110 + 3 x 95) divided by 7, or 104.

SB Index

The historical and in-season data for the opposing team's starting pitchers and catchers is crunched into an index. Each season, there are a handful of batteries to target when looking to supplement the steals category.

Quality of Opponent

The opposing teams weighted on base average (wOBA) facing righties and lefties is employed as a measure of quality of opponent. Initially, team projections are used but as the season wears on, in-season performance takes over. Bullpens are also factored into this index.

A grid showing opposing pitchers is also provided. Please note the site's Projected Starters grid also has this information, and it is updated throughout the week.

Well, there you have it. Hopefully, gaining an idea of how the rankings and indices are generated will facilitate lineup decisions. As always, I'm happy to embellish anything here in the comments section as well as answer specific lineup question in the appropriate rankings column. However, I do have one favor. I pull the schedule from the Projected Starters grid, so please direct specific rotation questions to that page.