hi jeff acorn here. could you elaborate on shelf life of udms. does it make sense to make inactive, udms that no longer show a profit. what seems to happen is they work fine for me then the roi eventually dips, even though the win rate stays around the same. i read a book by a mr folgen i believe that was statistically based. upon forward testing, he had a few catagories of races that were sub par. his statistical tests validated that he just happened to run into a bad "patch" of low payoffs. in most catagories he maintained at least an 8 percent profit or so. and it just happened in one or two classifications of races that there was sub par performance. do you account for this. or do you throw out the " baby with the bathwater" and just put the udm on the inactive list or dump it altogether?
After I create them I need to see them 'validate' (by that I mean see them perform reasonably well on fresh races after the tag date or date I created them.)
Once they 'validate' (and only about half of what I create does) I make an effort to rotate them into live play - and monitor performance going forward from there.
If I notice the avg odds start to drop - I don't hesitate to deactivate them.
A select few (based on concepts very few players ever think about) end up having long shelf lives (even though they can cycle back and forth between high and low roi.)
But most - especially the ones based on concepts generally followed by a lot of players that only showed up on my radar after noticing a handful of unusually big payoffs that ended soon after my validation period?
I deactivate those without hesitation once I see the avg odds drop.