Even if Rufus were using 0.2 Kelly I think his $500 would be reasonable considering we know his bankroll to be in the millions. I think he has enough data at this point to fairly accurately estimate his edge, so I don’t think the difference in model vs real prob applies as much versus someone who is just randomly grabbing outrights based on DG’s model.
Well if I had a $10M bankroll I’d just let the wager ride and not think twice
Thats totally reasonable to say the longshots in the model don’t hit as often as they should (although there’s probably some sort of bias because a 1% prediction will lose 99% of the time). Where I respectfully disagree is to arbitrarily assign a $50 max bet to any outright with a 1mil bankroll. That completely disregards the 60 year proven track record of kelly criterion. If you said I use 1/10th kelly and I take the implied probability and multiply by 0.8, that at least has some sort of rational behind it IMO
I didn’t assign a $50 max bet, I said I probably would have bet $50 at 150-1 if I had a bankroll of $1M
If I was reasonably sure that I had a big edge and I got 300-1 odds I might have bet $100-200, hard to say since I’m not in that situation right now
As for me I’m just trying to grow my bankroll as quickly as I can and trying to refine my methodology given actual results. I’m fine tuning the bet sizes in real time
In general I’ve found that you want to cut your bet sizes down to the point where 20% downswings are rare
My only point is there is a literal proven solution for this and its betting using kelly criterion. cheers
2 Likes
I use it myself so we’re on the same side of the argument
The thing is that Kelly advantage is only known when you are 100% certain of the actual probability
If your model says something has a 1.20 ROI but in reality you’re returning 0.90 then you’re going to have problems
I see this a lot with models especially when you bet longshots
I had Mito at 300-1 too. It was on Circa Sports like a week and half before the event started. You can always get very good numbers on there in the futures market for the majors. For example I have Wise at 200 -1 this week and had found Lowry at 125-1 three or four weeks before the Masters.
2 Likes
Do you know if Circa LV in person odds are same as Circa Colorado online odds?