Bowl Mania Football!!!

Song of the Day

First, today’s song is called Summer On You, from Sam Feldt, a European (Dutch) deep house DJ. It’s a nice mellow and fun tune. Reminds me of summer, which seems like a million miles away these days. Enjoy.

Capital One-ESPN Bowl Mania (NCAA College Football)!

Well, thank God for Bowl season to give me something to do! I registered for the Capital One Bowl Mania on ESPN.COM for fun. If you pick all 42 bowl games correct, you win $1 million. Of course, it’s impossible to correctly pick all 42 games (the odds are horrendous). There are no perfect entries left, not surprisingly. But, the top prize remaining is tickets to next year’s NCAA football playoffs package and $5,000 Amazon gift card. More importantly, bragging rights are worth a whole lot more! Haha. I think if I win the top prize, I’ll give it away, maybe to a charity like Make-A-Wish foundation.

As of last night, I was back in the 99% bracket. Following this morning’s game of South Florida vs South Carolina, I’m in the 99% bracket – top 10,000 nationally – but far outside of first place, unfortunately. Not bad for a contest with at least a million entries, perhaps millions.

I was comfortably in the 99% bracket for days, then the day before, I lost all 4 of my games which was shocking and devastating, and pretty much ruined my chances of winning, most likely. It got me down to 93% – holy crap! But, yesterday I got all my picks right and now am back to the top 99%. And after winning today’s first game (South Florida over South Carolina), I’m now in the top 10,000 bracket. My remaining picks for today are Virginia Tech over Arkansas, and Colorado over Oklahoma State.

If I don’t at least finish the contest at 99% or higher, I will consider myself an abject failure in life. I will not be mediocre at anything ;-)

This bowl season has been quite surprising. A lot of upsets; more than normal. I model about a 30-40% upset ratio, where the underdog beats the favorite team. Of course, trying to predict which underdogs will win isn’t easy. As of the end of yesterday, 11 out of a total of 25 games were upsets (44%), above my most optimistic case of 40%. And some were huge underdogs (13.5 point spread).

So far, on my underdog picks, I’m 6-2. I got 6 out of 8 underdog picks correct (based on popularity picks on ESPN). Not bad.

Betting On Football

I’m just doing this for fun. I don’t bet any money on these games. I’ve never spent so much time watching and analyzing football games as I have these past few months – for obvious reason. LOL.

I think if I could bet again next year and make my picks, my chances and performance would be significantly improved based on what I’ve learned. Of course, there is always about a 10% element of chance and luck in any sporting event. Not even God can overcome that uncertainty factor, because a lot of it is based on intangibles like will and determination. Plus, sometimes the ball just bounces the right way for certain teams on a given day, especially an oblong ball like a leather football.

So, the absolute best anyone could hope for on a long term sustainable basis is 90% accuracy in picking winners. Given nobody is perfect, realistically, adding another 10% error factor, if we can achieve 80% winners correctly, you are a god. It’s virtually impossible to do consistently. If you simply picked all favorites, you would likely end up with about 55% to 65% success rate but would never go higher over the course of a season (or corresponding 35-45% success rate picking all underdogs). And you would never make money.

But if you actually want to bet on football games – and make money, since making money is always preferable to losing it – then I highly suggest the following strategy:

  1. Only bet money line bets, i.e. pick winners instead of against the spread. There is far less manipulation and point shaving.
  2. Pick underdogs and stay away from favorites.
  3. Bet unevenly across all your underdog wagers, with a linear coefficient based on only the spread. Don’t go high on one bet that you have a higher confidence level unless it is spread based. Only pick underdogs that you think has a reasonable chance to upset in the first place. You can slowly increase the amount per wager as your account grows, but week to week, the bets should all be equal for equal spreads to mitigate risk.
  4. Stay away from sentimental picks, like your favorite team.

The reason is simple math. The risk/reward is far better betting the money line bet (MLB) of an underdog winning straight out, versus against the spread.

If you get an underdog correct, you 1X bet can yield somewhere between 1.1X to 20X (or more) return on a single game bet, depending on the spread. However, if you pick a favorite to win, the best you can hope for is 1X (100%) return, or even barely a 10% return. The risk/reward is clearly figuring out which underdogs will win. So, if you can get 45-50% of your underdog picks correct, you will always make money. If you get 60% or more underdog picks correct, you’re a damn rock-star and you will do very well – like you would eventually own the casino!

Here is my basic and simple strategy of picking teams:

  1. Pick the teams with significantly better records. The gap should be +2 wins or more. For match-ups pitting Power Conference teams (SEC, BIG-10, Big 12, ACC, PAC-12) verses non-Power Conference teams, the 10+ win non-power team will most likely beat a mediocre power conference team. However, when the spread is too large, say over 10 points in favor of the non-power team, there is a good chance of an upset. So be wary of large spreads in these situations. (Feelings of disrespect can fire up an opponent.)
  2. Weigh the final 6 games of the season far more heavily than the first 6 games. Teams with end of season winning streaks are likely to win. A massive inflection point of winning usually indicates something fundamental has changed – personnel changes, such as at quarterback, impressive learning curve, etc.
  3. Don’t pay any attention to national poll rankings in making your decision. An un-ranked team can easily upset a ranked top 25 team.
  4. Strength of conference is often a distraction that leads to the wrong picks. For two Power Conference teams, when unsure, pick the team with the better record. For instance, the last night’s game pitting Texas A&M (SEC) vs Kansas State (Big-12), 79% of the people picked Texas A&M, based on the SEC being a superior conference vs the Big-12. Better conferences simply means the top teams are superior, while mediocre teams are nothing special. A&M had 8 total wins and went 4-4 in the conference, with a 6 game losing streak to end the season, largely due to injuries (the month long time-off can sometimes heal these injuries however). Kansas St had 9 wins and went 6-3 within conference. Given the solid and consistent run game and solid defense, coupled with the slightly better record, I picked Kansas over Texas A&M. Kansas won, despite A&M being a 3.5 point favorite and a landslide popular favorite.
  5. Reversion to the mean. If the upset ratio of bowl games gets over 40%, I believe the pending games will likely increase the probability of the favorites to win, so I may factor that into choosing the projected winner.
  6. Any Vegas odds over 13.5 is a significant barrier in college. Go with the favorite. Assuming a 35% upset ratio, the probability is proportional to the spread gap. Over 13.5 points is too unlikely, stay away from these and simply pick the favorites. Generally speaking, I broadly categorize into buckets: 0-3 points 40% probability of upset, 3-6.5 points <30% chance, 7-9.5 points is <20% probability, 10-13.5 points <10% probability. If you have a third of your 10-13.5 spread match-ups as being upsets, you’re going to be disappointed.
  7. Analyzing statistics is mostly useless. But there can be some nuggets of useful information. I like the TD/Int ratio for QBs, and the average Yards Per Completion. I like teams with balanced run/pass offenses. Defensively, I like 3rd down percentages and TFL (tackles for loss) as well as QB hurries and hits (sacks are less meaningful). And net takeaways are critical. But statistics don’t often translate well between conferences, so you can’t rely on these too much. Average points allowed and margins of victory are significant, normalized for strength of opponent.
  8. I like to look at the schedules of each team and look at the losses, much more so than the wins. I try to understand why they lost, and what weakness the opponent exploited. Then I ask myself, how likely is the upcoming opponent likely to be able to exploit the same weakness. Sometimes, each team gets a mulligan and we just have a bad day, so we need to understand that. For instance if a team committed 7 unusual turnovers in a single game and lost, it’s unlikely to happen again. Call it a bad day and throw out that data point. If it happens more than once (high turnovers in a single game), then investigate further, because there is more to it. This was the case with Army vs North Texas. NT had beaten Army earlier in the season when Army committed 7 turnovers. The bowl rematch, I picked Army to win. They won in a thrilling overtime game.
  9. Strength and maturity of the QB and the caliber of the coach are important criteria as well.
  10. I also look at previous bowl history and recent match-ups.
  11. I probably forgot a few things, but who cares, I’m out of time!

In the NFL, pro-football, 7 points is the equivalent of 14 points in college. Anything over a 7 point spread is highly unlikely to be an upset.

Anyway, I love football! And if you do this for fun, and don’t bet hundreds or thousands of dollars, it adds to the enjoyment of watching the game. You can use this strategy to maybe win next year’s $1 million Bowl Mania grand prize! Good luck!