The Z Files: Using Composite Player Rankings

One of the flaws with conventional valuation is the assumption each player occupies an active roster spot all season. Clearly, this is not the case for a significant portion of the pool. Injury-prone players are an obvious example, as are players beginning the season on suspension such as Fernando Tatis. Latent examples are pitchers drafted to be deployed in a streaming capacity.

When rostering a player, you're not just drafting the player, but rather the lineup spot. The expectation is for the player to occupy the spot the majority of the season, but as mentioned, in many instances it will be shared.

This is not a new concept, as the common advice is to move such players up your rankings or to pay a little more than their projected earnings in a salary-cap draft. Those with their own valuation program can adjust the projections and generate the expected earnings from the roster spot, which can drive when to draft a player or how much to pay for them. However, not everyone has taken the time to code their own valuation program.

Here is where the NFBC Best Ball scoring system enters the chat. By way of history, several years ago, the NFBC wanted to introduce a best ball competition called The Cutline that involved a regular season, then a series of playoff rounds where only the highest-scoring teams made the cut (hence the name.) However, they feared their clientele would shy away from a contest involving rankings that greatly differed from their standard roto 5x5 scoring. The solution was devising a scoring system that generated rankings matching their 5x5 ADP. I didn't do the heavy lifting, but I was a consultant and helped tweak the system to maximize the correlation between the roto and points rankings. After combining hitters and pitchers into the same pool, and adjusting raw points for positional replacement levels, the correlation coefficient is between .92 and .95, depending on the valuation system.

It's not perfect, but a workaround for revaluing players is converting the projections to points and adding the expected points supplied by those sharing the spot. This new total is then compared to the rest of the player pool to gauge where this composite player should be ranked. Some Excel acumen is needed, but not nearly the level required to code a valuation spreadsheet.

Here is the Cutline scoring system:

HITTERS

• At Bat: -1 point
• Hits: 4 points
• Runs: 2 points
• RBI: 2 points
• Stolen Bases: 5 points

PITCHERS

• Innings Pitched: 3 points
• Hits Allowed: -1 point
• Earned Runs Allowed: -2 points
• Walks Issued: -1 point
• Strikeouts: 1 point
• Wins: 6 points
• Saves: 8 points

The first step is downloading the projections to an Excel spreadsheet. Let's start with hitting. On the menu bar, go to Projections --> Full Season. The batters should be displayed. At the bottom, you'll see Export Table Data, click Excel.

Please pardon me if you're familiar with Excel and can do this on your own.

In Cell X3, type =2*G3+4*H3+4*I3+4*J3+10*K3+2*L3+5*O3-F3

This converts the projection to points. Next, fill down by left clicking and holding your mouse in Cell X3, then dragging down through the rest of the hitters. When you reach the bottom, release and hit CTRL D. This fills down the formula.

The final step is clicking on Cell X3, then sorting by using the descending icon. This can be located in the Data tab in the top menu bar. The icon is near the middle, just to the left of SORT. Use the bottom one, with Z over A, next to a down arrow.

You now have the hitters sorted from best to worst. Let's use Tatis to demonstrate the process. He's not eligible to return until Thursday, April 20. Your ability to replace him depends on the league format. In weekly leagues, you can activate him Monday, April 17 and take a few zeroes, or you can wait until the following Monday and use another batter all week. In the NFBC, with midweek moves for hitters, you're probably going to activate Tatis on April 21. Decide when you will likely activate him and determine how many days/weeks you'll be able to use a replacement player.

The NFBC is easy. You get the short week to begin the season, two full weeks and then the Monday-Thursday period with a replacement player. This is essentially three full weeks.

If you're in a weekly league, you get either two and a half, or three and a half weeks with another player in your lineup. If you choose to use a substitute for the entire week, you'll need to subtract the points Tatis likely would have scored in the four games he was active, but on your bench. Don't sweat this, ballparking will do the job.

Let's use the NFBC for our example. An advantage of replacing a player to open the season is injuries have not thinned the pool. If you're able to have Tatis occupy the Utility spot, you can replace him with any position, perhaps slightly increasing the added stats. Let's assume his replacement is drafted somewhere in the 200 to 220 range on the hitter's rankings. There are 210 active hitters, so you're using your last pick, or your first reserve for the player. Using this range accounts for some pitching drafted in these rounds. Catchers will be excluded from the replacement pool.

The average number of points scored for the replacement player is 292. Accounting for the All-Star break, the season is 26 weeks long. Prorating the 292 points to three weeks of action yields 34 points from the replacement player.

Tatis is slated to generate 881 points, 17th most among hitters. Adding 34 means the roster spot is estimated to garner 915 points. This composite Tatis-plus player now ranks 11th among batters. if you really want to get into the weeds with this, include the pitching points and determine how much Tatis moves in the overall rankings. That said, it really isn't necessary. Seeing where Tatis moves in the hitting pool should give you an idea where you can draft him and collect a profit on the draft spot, assuming everyone else isn't accounting for the replacement stats. For example, Tatis is currently the 16th hitter off the board, with a 19.62 ADP. Since the composite player is the 11th ranked batter, making Tatis the 12th-15th batter taken bakes in some profit potential. It may not be much, but every little bit counts, especially early in a draft where profit isn't the primary objective.

The algebra changes a bit for NL-only. Downloading the NL-only pool and calculating projected points renders Tatis as the eighth-best hitter. Most NL-only leagues are weekly, so it's necessary to activate him for the entire week or use a replacement player for the week Tatis is first eligible. Let's calculate the projected points for each scenario.

Using the same process as above with hitters ranked 135-154 gives a replacement player expected to score 169 points, or 6.5 per week.

Tatis is projected for 681 points over 23 weeks, for an average of about 30 per week. Even if he plays just half the week, it's better to activate Tatis on Monday, April 17 and take a few goose eggs. This means the replacement player is active for two and a half weeks, adding 16 points to the composite player for a total of 697. This jumps Tatis from eighth to sixth among hitters. Again, not much, but if he can be drafted as the seventh batter, a small profit is expected.

This same process can be applied to a player expected to miss time due to injury, although of course you may not be able to pinpoint an exact return date that way you can with a suspension. The big difference is, when the injury occurs will affect how many days he misses and the quality of the replacement player. Again, don't get too caught up in the exact number of games. It's an estimate, used to help re-rank the composite player.

The same procedure can be used for pitching. As you can stream hurlers with more favorable matchups, it isn't necessarily the stats of a replacement-level player you're adding, but rather those you honestly feel can be collected over the time the pitcher is absent. This is different per format and is more subjective than the objective manner used for hitters. The other main difference is there are maybe 50 starting pitchers who will never be healthy but inactive. Some may only be reserved for a start or two, but the point is, more finesse is necessary when determining composite pitching points, especially for starters drafted as streamers. That said, using the described method on a spot-starter type may provide a sweet spot to draft them ahead of their ADP, but still at a point the composite pitcher can produce profit relative to the draft spot.

For those needing a little Excel assistance, on the projections page, choose PITCHING and download to Excel. In Cell Z3 type =3*I3-L3-2*N3-O3+P3+6*J3+8*R3. The computes the projected points.

I'll do my best to answer questions pertaining to this method, or the Excel skills needed to employ it.