ROTISSERIE: 2017 SGP Denominators

For those new to the concept, Standings Gain Points (SGPs) estimate the amount of a given statistic you would need to move up one point in the standings, based on the distribution of each category. For example, depending on the league and the season, 18 runs, on average, would move you up one place in the final standings. A player who scores 54 RBI above replacement would then be worth a three SGPs in runs. They have the advantage of measuring every category on the same scale, so you can simply add up a player's SGPs to gauge his relative value.

To determine SGPs you need to know the distribution (or "lumpiness") of the standings in order to calculate the correct denominators. The best source of SGP data is from your league's history, especially if you've had the same scoring system and number of teams for a while. You can take the standings for each category and use the SLOPE function in Excel (or LINEST function in Excel or OpenOffice) to calculate the best fit between the individual stats totals and the points for each rank (use points as the X value and the stat totals as the Y value; this may be counter-intuitive to those who speak math, but it works). Check your numbers against the tables below for reasonability.

If you don't have good data for your league, are in a new league, or just don't trust your math skills, you can use these denominators here. Just be aware that there's good and bad in using "generic" numbers like these. The bad part is that valuations can be very sensitive to the denominators (i.e., small changes in a denominator, especially in the "average" categories can cause large changes in valuations). However, that may dispel the illusion of precision that dollar values sometimes give us. You're much better off thinking of a player in terms of dollar ranges ($25-$30) rather than a specific number ($27). Better yet, think in terms of value ("second tier") rather than as a priceable quantity.

The Denominators

SGP denominators fluctuate from year to year and league to league. In order to mitigate the outliers, we take the average SGPs across many leagues and use a weighted average of the past three seasons. By averaging across a large sample, we hope to produce the most likely (note: not the "correct") distribution. Note that the dollar values you generate from SGP are not likely to match the projections on BaseballHQ.com or in the Baseball ForecasterSince standings gain points are specific to a particular format, they require historical data to estimate accurately. BaseballHQ.com, therefore, uses PVM (percentage value method) for its player values.

Below are the values for the three most common league types: 12-team AL- and NL-only, and 15-team mixed. For 2017, we've also added values for 10-team leagues. The "formula" shows you how to calculate the SGP for each player. The ideal method would be to calculate performance above replacement level. This is usually found by running a preliminary valuation and taking the average of the first 5 or 10 players who fall below the draftable line.

In 2016, we saw significant jumps in the SGPs for HR, RBI, and SB. This is not necessarily due to an overall increase in offense (though there was that, too). What it most likely means is that there was more of a gap between the top players and the bottom, though this is a generalization. Note, too, that ERA and WHIP increased in all formats, which magnifies the value of players who are well below the average in those categories. Of course, things could regress to the mean all around, which is why we use smoothing to generate the values for 2017.

Things you should know

These numbers are based on leagues with 14 hitters and 9 pitchers. While we allow for minor variations when calculating the denominators, the AB and IP totals are strictly based on leagues with a 14/9 configuration.

The AB and IP totals have been updated from 2016, based on recent research. In previous years, we used the same totals for both types of unmixed leagues, but this year, AL-only and NL-only AB and IP totals may differ. This makes sense intuitively, as the AL has 15 more every day hitters, and they tend to do less pinch-hitting than the NL.

Note also that we only have two years of data on 10-team leagues. So don't take these numbers as gospel (for goodness' sake, don't ever take ANY numbers as gospel!).

12-team NL-only

Category    SGP     Formula
========  =======  ==========
Runs        27.4    Runs/27.4
HR           8.8    HR/8.8
RBI         27.3    RBI/27.3
SB           8.8    SB/8.8
AVG       0.0022    (((1494 + Hits) / (5700 + AB)) - 0.262) / 0.0022

W            3.5    W/3.5
Sv           6.8    Sv/6.8
K           40.9    K/40.9
ERA       0.0845    (4.03 - ((537 + ER) / ((1200 + IP) / 9))) / 0.0845
WHIP      0.0159    (1.292 - ((1550 + H + BB) / (1200 + IP))) / 0.0159

12-team AL-only

Category    SGP     Formula
========  =======  ==========
Runs        26.2    Runs/26.2
HR           9.4    HR/9.4
RBI         27.0    RBI/27.0
SB           6.5    SB/6.5
AVG       0.0022    (((1609 + Hits) / (6200 + AB)) - 0.259) / 0.0022

W            3.1    W/3.1
Sv           6.2    Sv/6.2
K           36.5    K/36.5
ERA       0.0815    (4.13 - ((551 + ER) / ((1200 + IP) / 9))) / 0.0815
WHIP      0.0143    (1.290 - ((1548 + H + BB) / (1200 + IP))) / 0.0143

15-team Mixed

Category    SGP     Formula
========  =======  ==========
Runs        17.8    Runs/17.8
HR           7.5    HR/7.5
RBI         19.7    RBI/19.7
SB           7.0    SB/7.0
AVG       0.0016    (((1868 + Hits) / (7000 + AB)) - 0.267) / 0.0016

W            2.7    W/2.7
Sv           6.0    Sv/6.0
K           27.9    K/27.9
ERA       0.0637    (3.81 - ((550 + ER) / ((1300 + IP) / 9))) / 0.0637
WHIP      0.0120    (1.247 - ((1621 + H + BB) / (1300 + IP))) / 0.0120

10-team NL-Only

Category    SGP     Formula
========  =======  ==========
Runs        28.8    Runs/28.8
HR           9.7    HR/9.7
RBI         27.7    RBI/27.7
SB          10.4    SB/10.4
AVG       0.0024    (((1638 + Hits) / (6200 + AB)) - 0.264) / 0.0024

W            4.2    W/4.2
Sv           9.5    Sv/9.5
K           48.2    K/48.2
ERA       0.1087    (3.98 - ((530 + ER) / ((1200 + IP) / 9))) / 0.1087
WHIP      0.0204    (1.284 - ((1541 + H + BB) / (1200 + IP))) / 0.0204

10-team AL-Only

Category    SGP     Formula
========  =======  ==========
Runs        26.6    Runs/26.6
HR          10.2    HR/10.2
RBI         26.6    RBI/26.6
SB           7.4    SB/7.4
AVG       0.0022    (((1724 + Hits) / (6600 + AB)) - 0.261) / 0.0022

W            3.4    W/3.4
Sv           8.1    Sv/8.1
K           41.8    K/41.8
ERA       0.0855    (4.10 - ((593 + ER) / ((1300 + IP) / 9))) / 0.0855
WHIP      0.0140    (1.285 - ((1671 + H + BB) / (1300 + IP))) / 0.0140

10-team Mixed

Category    SGP     Formula
========  =======  ==========
Runs        21.0    Runs/21.0
HR           8.7    HR/8.7
RBI         20.1    RBI/20.1
SB           8.6    SB/8.6
AVG       0.0020    (((2021 + Hits) / (7500 + AB)) - 0.269) / 0.0020

W            3.4    W/3.4
Sv           8.1    Sv/8.1
K           41.4    K/41.4
ERA       0.0830    (3.80 - ((633 + ER) / ((1500 + IP) / 9))) / 0.0830
WHIP      0.0146    (1.242 - ((1863 + H + BB) / (1500 + IP))) / 0.0146

4x4 Leagues and other mythical creatures

The denominators shouldn't change in standard 4x4 leagues; they would be the same within each category. But if your league doesn't exactly match the parameters of one of the above choices, you can make some adjustments. In general, you should adjust upward for each statistic in smaller leagues. The smaller the league, the greater the adjustment. And the adjustments should be bigger in the lumpier categories (SB, W, Sv). If you're unsure, you can experiment with different settings until a handful of test players look "right" to you.

If you use categories other than the canonical 5x5, Ed DeCaria has some suggestions:

For those that use categories beyond the typical 5x5, the best solution is to use your own league's historical data (to the degree possible), or match up your non-standard category to the 5x5 category that it most closely resembles in terms of statistical distribution (e.g., OBP should behave similarly to batting average, triples should behave similarly to stolen bases, holds might fall somewhere between wins and saves).

Here's one last tip: don't be bound by the numbers. Make adjustments until they look reasonable to you overall (don't worry if a few players seem out of line). And remember that the dollar values are just a guideline. Other factors, including potential for changes in playing time, injury risk, and regression should play large roles in your buying decisions. And most of all, have fun.

A thousand thanks to the folks at OnRoto for providing the league data that made this analysis possible.


Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.