So, Division Dog off a win vs. sub .500 team.
I like that. “Division” makes things closer to a coin toss.
Other team being below .500 makes things closer to a coin toss.
Coin toss parameters are what I always want to play with on dog systems.
Here’s what I’d do with this one:
#1 – Eliminate April – You have win percent as a parameter. Most stats mean nothing in the first month. April profits, but the roi is extremely low. Filtering out the first month makes this:
697-723 +135.64 units, +9.6% roi.
#2 – No odd totals – Bring it in to 10.5 or less
675-702 +10.0% roi
#3 – Avoid blowouts – Nothing makes a team cocky and in for a let down than a huge win they just had. Previous margin is under 8 runs.
625-624 +154.64 units, +12.4% roi SU
#4 – Our dog is never more than 10 games behind opponent – Keep things competitive. It doesn’t matter if our team is behind or ahead of the other divisional opponent, but avoiding situations where our dog is way way far behind improves this to:
598-588 +153.06 +12.9% roi SU
p:W and division=o:division and D and o:WP<50 and month>4 and total<11 and p:margin<8 and wins-o:wins>=-10
This profited 11 out of 12 seasons. Not bad at all.
Might seem like a lot of extra parameters, but I only felt comfortable doing that because you presented a good thing in its simpler form.