Player valuation and SGP
Hey folks, sorry it’s been a while since my last post but I
was pretty jammed up with the Winter Meetings. Now that that’s out of the way
I’m catching up on things, so let’s pick up where we left off last time, which
was illustrating the value of stats in relative context.
To sum it up, Andre Ethier (below) didn’t hit 31 homers in a bubble last year, he hit them
in the context of an overall offensive environment, and therefore, the value of
those homers is dependant upon the overall environment.
However, I didn’t really do much for your fantasy planning
by putting it in the context of recent MLB league stats… knowing that there 651
fewer homers hit in 2009 than in 2000 is helpful, but how much does that effect
the value of any player? And besides, we don’t compete with the Dodgers or
Yankees or Tigers, we compete in 12- and 15-team fantasy leagues. So let’s put
things into a fantasy context and see what that means.
My main league, the one in which I invest the most of my
time and effort, is my 15-team mixed NFBC league. I know more people do 12-team
leagues than 15, but hey, this is the data I have available and it’s still
useful for illustrative purposes. There were 26 NFBC leagues last year, a total
of 390 teams. I dropped the top and bottom five teams in each category, since outliers
don’t help much, then took the top and bottom 10 teams in each category and averaged them
out. From that, here’s what we get for average stats of the first and last place teams in
each of the 26 NFBC leagues last year:
AVG AB R H HR RBI SB W SV ERA WHIP IP
(ERA) IP (WHIP) SO
.2850 7524 1142 2144 302 1127 197 106 99 3.617 1.248 1520.9 1433.2 1367
.2611 6936 912 1811 205 876 96 68 21 4.696 1.433 1408.3 1407.6 959
(The first thing you should notice right away is the
importance of playing time… the average first place team had nearly 600 more
at-bats than the average last place time, which is why we say, never take a day
off when it comes to managing your roster and maximizing playing time! Note
though that there are two IP values for ERA and WHIP, since the top and bottom
teams in each category were not necessarily the same.)
Let’s stick with homers for our example on how to use these stats
to help figure out player values. The average top team hit 97 more homers than
the average last place team, or an average of about 6.9 homers for each of the
14 roster spots. This means that on average, for each 6.9 homers a particular
player hit, he moved his team up one point in the standings in that category.
Well, not so fast… because the last place team had 205
homers, you had to do better than that to get your first point and move up from
there. This is where the notion of replacement value comes in. The average
player (roster spot) on the last-place team hit 14.6 homers, so in fact, the
“replacement value” – the average player on the worst place team – was actually
just under 14.6 homers, and for each 6.9 homers above THAT, another player earned his team
one place in the standings.
This is the beginning of developing what is known as
standings gain points, or SGP. The concept is that you figure out what each player
was worth in each category over the replacement level – the worst player on a
roster, or slightly below – and then add it up in each of the five (batting or
pitching) categories to get his overall value.
So our friend Andre, who hit 31 homers, was worth
about 2.4 SGP’s in homers, the number of points in that category he was worth
over a replacement-level player. Do this exercise for each player in each
category, then add ‘em up for their overall value.
But wait… Joe Mauer (right) hit 28 homers and surely those were more
valuable than the 31 Ethier hit as an outfielder, right? After all, Mauer is a
catcher and we know catchers stink, so plus production from one of them is
worth more than a good outfield season, right? Exactly… and that’s why
replacement value is different at each position. But more on that next time.
Looking forward to your feedback!