Points Leagues: The Doubt About Trout

What is there not to love about Mike Trout? I know. The fact that I don’t own him in any of my keeper leagues and the chances of me owning him in any of my other leagues is slim. In order to do so I’d have to have the first pick of the first round. Or would the second pick be good enough to land me Trout? Dare I suggest the idea of drafting Clayton Kershaw ahead of Trout in head-to-head points leagues? The thought of not taking Trout with the first pick is one that only Dom Cobb could plant in your subconscious. But would it be such a terrible decision? Let’s look at the numbers.

The only number that matters in points leagues is points. I’ve said it a million times (ok maybe only 15), but you get the point. Pun intended! Mike Trout is projected to score 509 points this season. That’s more than any other hitter. But not more than every pitcher. As a matter of fact, there are 17 pitchers that are projected to out-point Trout. I think I just coined a new term with my use of “out-point”. I wonder if this post counts as an official trademarking agreement. Grey, will this hold up in court?

Clayton Kershaw is projected to score 728 points. That’s 219 more points than everyone’s favorite fish named outfielder whose first name is Mike. If you thought I was referring to Mike Carp, guess again. I know nothing about fishing but based on my knowledge of baseball and love for analogies I’d say that a trout is awesome and a carp is not. And that is me using my SAT prep skills 20 years later. Ok, enough about that, let’s get to it.

Well just because Kershaw will score more points than Trout, does that automatically mean he’s the clear cut first pick? How does position diversity and scarcity factor into the equation? A useful mathematical method for ranking players in roto leagues is to calculate each players’ z-score. What’s a z-score? It sounds like something a German dude would say. “Hallo. Does anyone know zee score of zis game?” Sorry to any German people that might have offended. A z-score is the measurement of how many standard deviations away from average a data point is. I bet you wish you had paid attention in calculus and statistics class now. Can’t I just use the Pythagorean theorem? Not today junior. In roto leagues you calculate the z-score for each player’s stats in each of the 5×5 categories and then add those scores up to determine the player’s total score.

In head-to-head points leagues we have only one category. Points. That makes things a lot simpler. The first step was to split up players by position, but in order to do so I needed to determine what the player pool should be. How many players at each position should be included in the analysis? Well after many late nights of research I have come up with the following. The goal is to determine how many players at each position that will be drafted will be more valuable than the average player at the same position. This “average player” is a likely candidate to be found on the waiver wire.

This post assumes the following lineup for 12 teams:
C, 1B, 2B, 3B, SS, Util IF, OF (4), SP (4), RP (2), Bench (7) = 23 players per team
12 teams times 23 players per team = 276 total players

My research dictates the following. In a 12-team league we only care about the top 12 catchers. That part was simple. The rest took a little bit of playing around with and relying upon the work of several others that have embarked on a similar endeavor. Here is the breakdown of players per position that I have decided upon.

C (12), 1B (20), 2B (12), 3B (16), SS (14), OF (60), SP (62), RP (36) = 232 players

The next step is to fill these positions with the players that are projected to score the most points. For each position I included one extra player. This player represents the replacement player at that position. A replacement player is defined as the average player at a position whose shoes (cleats) can generally be filled with another player from the waiver wire. This year’s replacement players are:

C: Wilson Ramos
1B: Matt Adams
2B: Dee Gordon
3B: Trevor Plouffe
SS: Jhonny Peralta
OF: Dominic Brown
SP: Brandon McCarthy
RP: Brett Cecil

Now that we have our player pool broken down by positions we can calculate z-scores.

Z-SCORE = (Player’s Projected Points – Average Projected Points For All Players At That Position In Our Pool) / Standard Deviation For That Data Set

Each position is calculated separately. We now have z-scores for each player at each position, but this alone doesn’t help us rank players any more than projected points does. That’s because we only have one category. Points. Had there been more than one category, as is the case in roto leagues, we would now be able to rank players within their position by adding up the z-scores for each stat category.

Is it time to dance yet? Not quite. We now need to take the last player in each positional list (the average player), let’s call him Joe Bloggs, and zero him out. Anyone remember Joe Bloggs? Actually we can call him one of the replacement players listed above if we want to make this official. What we want to do is bring each of their z-scores to zero. To do this, for each position take the average replacement player and subtract his z-score value from every player at that position, including the replacement player. I most cases (all of ours) you will be subtracting a negative number which will raise everyone’s z-score. Once you have done this you now have each players’ fantasy value above replacement at said position. Now do this for all positions and we can start comparing players across positions. Go ahead, moonwalk across the room like it’s 1984 and you think you’re really good at it and nobody’s watching.

Now let’s look at the numbers. I should mention that when splitting the players up by position, I included a multi-position player at the position he ranked this highest. Here are the top 100 players according to FVARz. The complete list is included in the Excel spreadsheet attached at the bottom on this post.

Well look who we have at the top of this list. Clayton Kershaw! Were you expecting Yu Darvish? Ok, so let me start with a little disclaimer. This entire exercise has been somewhat of a experiment. While grounded in a lot of research and math, I do not proclaim these results to be the end-all be-all. I will say this, however, I have a reasonable amount of confidence in the results. While controversial, I do believe that Kershaw is the appropriate first pick. The amount of points he will give you each week is just too many to pass up. But… Kershaw’s FVARz is barely greater than Trout’s. And given a different points system or lineup configuration, Trout could easily have come out on top. It’s practically a photo finish, so you really can’t go wrong either way.

A few other things to point out. While many suggest punting on relievers, this list shows no such bias. I’m not sure how I feel about that. After Craig Kimbrel, Aroldis Chapman and Greg Holland, you don’t know what you are going to get. But does that make them worth an early pick? Probably not as early as the list above dictates, but I wouldn’t hesitate taking them sooner than later to lock up a stud. Remember, the goal is to fill up each position in your starting lineup with the available player that will score the most points at that position while keeping track of the players still available at your position(s) of need.

So who are the top dogs at each position? Let’s take a quick look at the 2015 Preseason All-malamoney Team:

Some of you might be thinking “how the fudge do I actually interpret all this potential nonsense?” I will try and clear up as much as I can. While FVARz allows you to compare players across all positions, a players’ FVARz defines how valuable that player is in relation to the rest of the players at their primary position. Let’s look at an example. Who would you draft between Jose Reyes (2.602 FVARz, 379 points) and Anthony Rizzo (2.571 FVARz, 442 points)? Rizzo is projected to out-point™ Reyes by 53 points. It’s a points league, no? We want the points, no? Assuming I don’t already have a shortstop I’d take Reyes with the higher FVARz  because of the greater impact he will make at his position relative to my other options at shortstop. BUT… if I already have both a 1B and SS and I am drafting for Infield Utility, I would likely take Rizzo because in this situation I want the points. Make sense? One more example. Buster Posey (3.372 FVARz, 396 points) or Cole Hamels (1.892 FVARz, 541 points)? What do you call a catcher that raps? Buster Rhymes!

A note about the attached spreadsheet. You can update the “Points System” worksheet to reflect your league’s scoring system and it will automatically update the “Hitters” and “Pitchers” projections worksheets. It will not, however, update the “Positional Rankings” and “Overall Rankings” sheets. Unfortunately generating those sheets is a manual process. If your scoring system is drastically different than the one this post is based upon, hit me up in the comments and I will try and generate a version for your league. Eventually I will have a web app that makes this much easier. No promises, but I’ll see what I can do. As long as you are not in any of my leagues!

The scoring system I am using for the basis of this post is as follows:
RUN (+1), RBI (+1), 1B (+1), 2B (+2), 3B (+3), HR (+4), BB (+1), KO (-1), HBP (+1), SB (+1), CS (-1), SF (+1)
WIN (+7), LOSS (-5), IP (+3), K (+1), BB (-1), SAVE (+7), BLOWN SAVE (-1), ER (-1), HIT (-1), HBP (-1)

Excel Spreadsheet

Find me on twitter at @malamoney