Don't be shellfish...Share on FacebookTweet about this on TwitterShare on Google+

For the ‘season to date’ Fantasy Baseball Player Rater (everything except the ‘Last 7 day’ and ‘Last 20 day’), I’ve made a methodology change that improves the Point Share accuracy of AVG, ERA, and WHIP.

As some of you may know, creating Point Shares requires that I create a ‘rostered universe’ for each league format – e.g., I think these 18 2nd basemen would be rostered in a 12-team mixed league.  This is all done with formulas that I’ve tweaked over the years to reflect how fantasy baseballers tend to select players.  The Razzball Commenter League provides a great reference check to see how well my formulas work for a 12-team MLB league.

It was clear even on Player Rater launch that my hypothetical rosters have better AVG, ERA, and WHIP than the average league.  The reason is pretty clear – my rostering is based on hindsight.  Most rostering decisions are driven by counting stats vs ratios – particularly HRs, SBs, SVs, and Ks.  So a pitcher like R.A. Dickey – who doesn’t wow with counting stats (at least historically) – might not make a roster until late in the season whereas my formulas would count his full-season stats.

I’ve corrected this by comparing my modeled AVG/ERA/WHIP vs. the RCL averages at the 1/3 mark of the year and then adjusting the averages accordingly.  It ends up being about 10% less hits for hitters, 9% more ER for pitchers, and 5% more H+BB for pitchers.

The impact on the rankings and $ values is relatively minor.  It does correct what was a large difference between ERA and WHIP Point Shares though – e.g., Justin Verlander went from 1.0 / 2.8 for ERA/WHIP to 1.8 / 2.4, Brandon Beachy went from 1.8 / 2.5 to 2.6 / 2.2, etc.

For those of you who use the Other Hitting Stats Player Rater for OBP/OPS/SLG, I can update these ratios with a little help.  If you could provide the team ratios for everyone in your league (in the comments) for one or more of these stats, I’ll be able to tweak the ratios accordingly.  For s’s and g’s, here are my current estimates for best/worst in a 12 team MLB league:  OBP (.363/.320), OPS (.862/.732), SLG (.505/.406).

4 Responses

  1. NS says:
    (link)

    I don’t know if it helps, but I’m in a 15 team 6×6 that uses OPS (same active offensive lineup as the RCL). We are looking at .8198 and .7072 for the best/worst ratios at the moment.

  2. Wgolds says:
    (link)

    We have 12 team 5×5
    Current Hi/Lo
    OBP .360 / .324
    SLG .406 / .478

    • @Wgolds, Thanks. This looks pretty close to my #s. I think for now I’ll keep it these %s as is (even if my modeled slugging % is a bit high).

Comments are closed.