Fantasy Rankings Prep (Part 1 of 3)

About four times a season, Eno unleashes the shocky monkeys and a few of us slow-footed writers are forced to enjoy ranking all the players. For the next few days, I am going to go over how I prepared my rankings.

Note: I am trying to keep the amount of math to a minimum. If somewhere you get lost in the procedure let me know and I can explain the procedure in more detail.

The first item to remember is all leagues are not even close to being the same. In my three keeper leagues, two are points based and the other is an AL only league with one pitcher category being Wins+Saves+Holds. Additionally, some leagues have keepers. How the keeper’s “salary” is set determines a their value. Other league options have innings pitched limits (good rates stats needed) or as in the case of my league with W+S+H, an IP minimum is set to keep owners from only using relief pitchers. Catcher rankings can vary quite a bit from a one catcher to  two catcher leagues or even two catcher slots with a 162 game limit as in Ottoneu. For my rankings, I did them off a basic 5×5 12-team league with 23 positions (14 position players, 9 pitchers).

If possible, I like to use the standing points gained method (a couple of explanations on the methodology). Some other methods may or may not be better, but I like it because it shows the league’s biases and gives me good baselines for final league positioning. I will not go into every gory math detail of the ranking method, so you may want to go back read the two articles previously reference or read Art McGee’s book, How to Value Players for Rotisserie Baseball or Larry Schechter’s new book, Winning Fantasy Baseball.

To get my values, I calculated the average final 2013 values from the last 20 teams to draft at NFBC last year. Here are the final totals and averages for each place in the standings along with the value (slope method) it takes to jump up one position in the standings.


1st 0.280 1109 293 1075 191
2nd 0.276 1078 272 1039 176
3rd 0.273 1048 266 1016 167
4th 0.271 1035 260 996 159
5th 0.269 1020 252 977 153
6th 0.268 1007 247 965 145
7th 0.267 987 241 949 139
8th 0.265 975 237 931 130
9th 0.263 960 229 917 120
10th 0.261 926 220 897 113
11th 0.259 899 208 854 107
12th 0.256 840 191 796 93
Average 0.267 990 243 951 141
Change to move up 0.00195 20.8 7.8 21.4 8.2


Rank ERA Wins WHIP Strikeouts Saves
1st 3.11 109 1.13 1496 122
2nd 3.27 102 1.16 1446 110
3rd 3.34 100 1.18 1404 102
4th 3.40 97 1.19 1370 95
5th 3.48 94 1.21 1347 90
6th 3.55 92 1.22 1323 84
7th 3.62 90 1.23 1303 80
8th 3.68 88 1.25 1281 75
9th 3.76 85 1.26 1256 69
10th 3.85 82 1.27 1221 61
11th 3.95 76 1.29 1149 52
12th 4.10 70 1.32 1063 35
Average 3.59 90 1.23 1305 81
Change to move up 0.081 3.0 0.015 33.3 6.8

These tables can be used for two purposes.

Number 1:  Knowing where your teams stands during drafts and auctions.

I keep track of a couple of items (in the points leagues, just one, total points) during the draft to make sure I am not getting too unbalanced, SB and HR (possibly AVG). Once a player is picked, just total their projected stats. You can then know how close you are at getting to a value which will put you too high in the rankings and over killing a stat.

One item to notice is how the top and bottom one or two teams are further away from the pack. These teams won by too much and wasted the stats they accumulated. When you are building a team, don’t aim for the previous top value, aim for just more than second place.

Number 2:  Value Players.

Well, I am going to go stop here and go over this step in detail in my article tomorrow. Please let me know if you have any questions so far.

Print This Post

Jeff writes for FanGraphs, The Hardball Times and Royals Review, as well as his own website, Baseball Heat Maps with his brother Darrell. In tandem with Bill Petti, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Twitter @jeffwzimmerman.

18 Responses to “Fantasy Rankings Prep (Part 1 of 3)”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. srpst23 says:

    I am in the process of ranking my pool of fantasy players, using the standings gain points (SGP) method. My problem is that my league went from 11 to 12 teams this year. We have been at 11 teams for the last 7 years so I have a ton of SGP data for our 11 team league. Is there a way to correlate that data to a 12 person league, as our league only has 13 hitters and 8 pitchers per team (we only use 1 catcher, and have 5 util slots instead of CI/MI/Util) and all of the generic SGP data I can find is based on 14 hitters / 9 pitcher leagues (generally with 2 catchers). This matters because I really don’t have to adjust for positional scarcity with only 12 catchers being drafted, and don’t necessaryily have to draft an additional MI. If anyone can help, it would be much appreciated.

    Vote -1 Vote +1

    • Jeff Zimmerman says:

      I think your formula may be pretty close. You will basically be adding in a set of replacement level players to the all the teams.

      Here are some changes I could see.
      Instead of your #6 values being around your average, it will be the values half way between #6 and #7. For example, if your average ERA was 3.50 (close to #6) and the 7th value was 3.58, I would move your ERA average from your previous equation to 3.54.

      In a shallow-ish league like yours, the number of HR will be down, but the slope(difference) will be the same. Free agent players will be available to keep the numbers up which are unavailable in deeper leagues.

      Vote -1 Vote +1

      • srpst23 says:

        That makes sense, as I never thought of it as adding 13 replacement hitters and 8 replacement level pitchers (relatively speaking) from the 11 team league. Thanks

        Vote -1 Vote +1

  2. baruchk says:

    I have a question to everyone: Are random public 50 dollar leagues on here generally more active than random yahoo public money leagues? I feel like they’d have to be since its all year around and everything, but im just making sure

    Vote -1 Vote +1

  3. JNB says:

    When totaling up projected stats during a draft, I’m often worried about comparing those totals to previous year team stat totals. The worry I have is that if the projections I use are systematically biased up or down in any category then I will not accurately weigh my pursuit of various categories. I’m not sure what the alternative is though. I’ve tried to simulate player distributions to teams and then performance variation from the player projections but that seems potentially flawed as well, perhaps more so.

    The other problem I’ve run into is trying to properly weight bench picks and projected playing time when deciding which player stats to total when comparing to previous year totals/goals.

    Vote -1 Vote +1

  4. ML says:

    Good post. These are always helpful refreshers. Didn’t like schechters chapter on this…left much to be desired.

    Vote -1 Vote +1

  5. Blue says:

    Don’t like SGP approach because I am wary of the inherent assumptions about the distribution of the statistic in question across the league. What I much prefer is a two-stage process where I determine the total aggregate amount of a given statistic that will be generated over the course of a season (given league settings) and then determine what proportion of that production will be contributed by a given player (batting average in this approach is converted to number of hits above/below league average). The value of a player is determined by the summed share of production.

    Vote -1 Vote +1

    • ML says:

      Don’t you run the risk of ranking a player too high based on disproportionate production. Guy like billy Hamilton comes to mind

      Vote -1 Vote +1

      • Blue says:

        Yeah, you have to be careful with steals–I have a mod to the system that takes that into account. I mainly use it for the four big hitting categories.

        Vote -1 Vote +1

  6. Cybo says:

    I’m way too lazy and don’t have near the time for all that. I just look at as many different rankings I can find, weed out the ones I don’t like and average them out. Like an ADP but with all the experts rankings.

    Vote -1 Vote +1

  7. onSALE17 says:

    my strategy is to aim just above one standard deviation above the mean. If you do this, you are basically aiming for 8.5 in all categories. Let’s say you do a 5×5 old school, you basically draft a 85 total point team. then scourer the waiver wire.

    Vote -1 Vote +1

  8. dscottncc says:

    I have a calculated system that probably isn’t the most efficient, but it is fun and passes the time.

    First I get the Bill James book with projections.
    Second I get a magazine with their projections.
    Third I make my own projections on players I like.

    I average those three projections and base my rankings off of the average of the three, taking out any bias I might have.

    Vote -1 Vote +1

  9. Dave Shovein says:

    Jeff, to calculate your SGP values, you say that you used the last 20 teams to draft at the NFBC last year. Why 20 teams and which 20 did you select?

    When calculating mine or this year, I used all 435 teams that drafted in the Main Event last year.

    Vote -1 Vote +1

  10. Sean says:

    How do you calculate the “Change to move up” values? Thanks!

    Vote -1 Vote +1

    • Sean says:

      nvm, found my answer…

      “The first place finisher in Home Runs accumulated 291 HR for the year. The last place finisher accumulated 191. That leaves a spread of 100 HR from the first place team to the last place team (291 – 191 = 100).

      In this 12-team league, there are 11 standings points that can be gained (even the last place team gets 1 point, the first place team 12, so 11 points can be gained).

      This means that it takes roughly 9.1 home runs to move up one position in the home run standings (100 HR gap / 11 positions to move). Pick just about any position in the home run list, add 9 HR, and you can see that you do move up about 1 spot.”

      Vote -1 Vote +1

  11. Sean says:

    Any suggestions on where to find past data to make my tables? My yahoo league is 12 teams, 9 hitters (no CI or MI).

    Vote -1 Vote +1

  12. Mark says:

    Are the “average final 2013 values” from over an entire 162 MLB season?

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *