Red Flags Part 2
Social situations are not always my bag, never have been, particularly when the group gets larger than 6-10 or so and I’m forced to engage in the whole mingling thing. Given time, I can usually find enough commonality to have a substantive conversation with just about anybody, but the small talk phase is of particular discomfort to me, so my party trick is generally to go find obvious sports guys and talk about stuff like this. I’m actually a pretty good fella to meet at one of these parties too, because if you’re at a setting with local sports interests, but you happen to be from some other city or went to some other school, no problem, I’ll give you a break from Mahomes worship and we can just talk about your Brewers or your Florida Atlantic football team. Often times I even find that with kids and work and whatever else the distraction may be, people haven’t been keeping up with their own team as much as I have, but at least I tried.
I also find that people seem to have interest in this system as well, and that is it doesn’t always matter whether or not the person is the gambling type, which is why I’m writing this, but inevitably my system tends to betray me when its application is not as inclusive to every game as I am otherwise amenable to. Quite often, the system will shut down teams for entire seasons, so I find myself time and time again explaining to people why I have interest in a game no one at all cares about, but the big conference game that they’re excited about, I just don’t have an offering for that. It can be a little dissatisfying, but the reasons, of course, are these red flags I’ve been discussing. I already got into the process a bit and while I’m certain to return to this as this blog rolls on, today I’d like to get into the specifics, the primary reasons my system rejects a game that it very much has an opinion on, but has some classification that has statistically sent my winning percentage in the wrong direction over the course of several years.
The first is tricky to explain, but also the most significantly damaging if not adhered to. My system is a value play, and while the scale I discussed is not at all dependent on the spreads set by the oddsmakers, it does look to take advantage of the variance. In other words, I’m not ranking the teams as a definitive this team is better than the next equation, but rather, an algorithm of determining that I should take whatever number of points I’m offered for one team against another, and in the event that I have to give some up, how many I should be willing to go. But what I’ve found is that it doesn’t work in the opposite direction, by determining some number of points that I need to be given to take one team over another. So when someone suggests to me that a spread is too high on the Kansas game, so take Kansas, what my system usually thinks is that Kansas can lose by as much as the Vegas people they think they will, and the question I should be asking is how many points I’m willing to give up to take Oklahoma over Kansas, and if the number exceeds that, just don’t bet the game. I’m constantly testing concepts that might enable better plays on more underdogs, but for now, I haven’t found reason to take points unless the system thinks the team is the better play in the first place. This can be eye popping at times, and the 40% of the time it loses can look embarrassing, but just as often my underdog actually wins the game, not to mention that I was getting points in the first place, which gives me more wins than only the outright variety.
The second red flag involves home field advantage. These games are always close calls in the first place, but you might be surprised how much they affect my winning percentage. Essentially, what I found is that while I should certainly use home field advantage as a reason to not bet a close decision, I should never use it as a reason to make the bet. Remember that I use the 3.8 average in my calculations, so if that difference is enough to change a neutral field play into another at a home site, it is the only scenario that this red flag is applicable. Let’s say I am willing to give up 6 points on Ohio State on a neutral field game against Georgia. OK that’s cool for a postseason scenario, but what if it’s a gem we’re given in the early part of the schedule and the game is a home game for Ohio State? Should I use my 3.8 and then be willing to go up to 9.5 to take Ohio State? Nope. Still 6. But if the game is at Georgia I will use the 3.8 to determine that I probably shouldn’t give up more than 2. This works out more than you’d think and the why probably has something to do with better teams winning most of the games in the first place, but who knows, I’m not as interested in obsessing over the why, only the data and it helps.
Speaking of “the why,” you can use up a lot of time thinking about red flag number 3, and whatever you come up with would be a good enough explanation to me. But for whatever the reason, don’t bet on first year coaches, for or against them. I should preface this by clarifying that I mean first year in a program, not just young guys getting their first head coaching gig. Take Florida State this year. Beat North Carolina, lost to just about everyone else. Weird, right? Well, it’s Covid year, so everything is weird, but what’s notable in the data I’ve tracked is that a new coach with a new system at a new school makes for just about the least predictable outcomes imaginable. Remember that upward or downward mobility are most prominent during these times, so let the team find themselves once a coaching change has been made, for better or for worse, and don’t try to use data to predict what they’re going to do.
The 4th is related to the 3rd, but the data here still gives me room for pause and it involves 2nd year coaches, the North Carolina part of the Florida State vs North Carolina equation. Basically, give it 2 years as opposed to just one. The problem here is that while the red flag has also helped my winning percentage, it’s not as significant as the others and recently (not Covid year) I’d be leaving money on the table, even though my winning percentage demonstrates a better play. It’s simple math, 54% brings down a 60% average, but the subset would still be a money maker. I still use it, primarily because the numbers have been less consistent and consistency is what I’m looking for, but it remains the one I still offer plenty of scrutiny to and am open to changing in the future.
Red flag number 5: a worldwide pandemic. Very painful. Anyway, I will have a number of other considerations in part 3, but it’s time to go ahead with this week’s games. The best news is that the system is showing signs of recovery, having made 2 weeks of .500 picks and then a nice 5-2 week last week. The early games this week that I posted on the Facebook page went 1-1 with one push, so we’ll see what happens today before we get too excited. Here are the games.
Wednesday
Eastern Michigan +7 at home vs Toledo
Thursday
Tulane +6 on the road vs Tulsa
Friday
Minnesota ++3 at home vs Purdue
Saturday
Army -4 at home vs Georgia Southern
North Texas +1 at home vs Rice
Wisconsin -7 on the road vs Northwestern
Oregon State 3.5 at home vs California
Cincinnati -6 on the road vs Central Florida
Iowa -2.5 on the road vs Penn State
Alabama -30 at home vs Kentucky
Georgia State -3.5 on the road vs South Alabama
Auburn -11 at home vs Tennessee
Oklahoma -7 at home vs Oklahoma State
USC -3 on the road vs Utah
Oregon -13.5 at home vs UCLA
Remember that I freeze the lines at some point midweek, just as you would if you made bets at some given time. Some of the lines have definitely moved, but this is where I had them and I have no interest in trying to use the arbitrary time to my advantage, so I’ll take the negative with the positive where they happen as well.
