College Football Week 10 Pick’Em Strategy: Does The Public Put Too Much Emphasis On Wins And Losses?

November 2, 2011 - by David Hess

This post covers picking strategies for college football pick’em contests for Week 10 of the 2011-12 season, based on college football predictions from our algorithmic models and other relevant data such as public picking trends. It also complements the analysis and projections presented in our College Football Office Pool Analysis pages.

This commentary reflects predictions and public picking trends as of Wednesday afternoon. The pick recommendations on our Office Pool Picks pages will continue to update based on the latest available data until shortly before the first AP Top 25 game or official Yahoo! Pick’em game of the week. Make sure to check our Game Winner Pick’em and Spread Based Pick’Em pages for the latest.

Week 9 Highlighted Picks Recap

Our Odds-On Value and Odds-On Contrarian picks had a fantastic week, going a perfect 5-0 combined. Given the public pick percentages, that should have gained you about 2 games over the average contestant.

For those of you who needed to take big risks, the week went even better. That’s because 2 of our 3 Long Shot Upsets pulled off wins, as Georgia Tech beat Clemson, while Ohio State upset Wisconsin. It was a good week to take some big risks.

Unfortunately, our High Risk Upsets didn’t fare so well, as all three took losses. Still, even if you picked ALL of our highlighted games — which would have been more upsets than we recommended — you should have gained about a half game over your competition, on average.

Overall, our method of mostly staying conservative — but throwing in a couple carefully chosen upsets for those in larger pools — has worked out well, as our Conservative game winner picks are now beating 95.8% of all ESPN college football pick’em contestants. Our Aggressive picks are slightly behind, at the 85.1% level. Our Very Aggressive picks are further back, topping only 61.3% of competitors.

For spread-based pools, our advice last week was simple. For Conservative picks, use our projected cover odds to rank your teams. For Very Aggressive picks, go against the public in every game. And for a middle ground, those strategies are blended in our Aggressive picks. It was a mediocre week all of our picks, as the Conservative and Very Aggressive ATS picks both went 13-13, with the Aggressive picks a bit behind at 11-15.

Those records, keep in mind, were against actual Vegas spreads. Most likely, your lines were a bit softer than those we compare against. Hopefully you picked up an extra game or two by using late line movement to your advantage, as detailed in our preseason pick’em strategy post.

College Football Week 10 Pick’em Strategy: Game Winners

It’s time for the Possible Game Winner Picks table. This valuable tool summarizes the info we use to make our recommendations. It’s based on our detailed pick’em analysis page, with only a couple tweaks.

First, we only show you the games where the public is undervaluing a team. Second, we add a “Value” column that shows the difference between projected win odds and public pick percentage — this is a key data point that tells us which teams give us the most bang for our buck. Third, we label the picks with a type that corresponds to the summary lists from our analysis page.

Team	Opponent	Win Odds	Public %	Value	Spread	Pick Type
Boise State	at UNLV	99.0%	~95%	4.0%	-41	Odds-On Value
Georgia	vs. N Mex State	98.6%	~95%	3.6%	-34	Odds-On Value
Houston	at UAB	95.2%	~90%	5.2%	-27.5	Odds-On Value
Wisconsin	vs. Purdue	94.2%	~90%	4.2%	-25	Odds-On Value
USC	at Colorado	93.8%	~90%	3.8%	-21	Odds-On Value
Alabama	vs. LSU	64.5%	~60%	4.5%	-5	Odds-On Value

NC State	vs. N Carolina	45.3%	~25%	20.3%	+3.5	Low Risk Upset
Connecticut	vs. Syracuse	47.2%	~35%	12.2%	-1.5	Low Risk Upset

Iowa	vs. Michigan	38.5%	~5%	33.5%	+4	High Risk Upset
Pittsburgh	vs. Cincinnati	43.5%	~20%	23.5%	+2.5	High Risk Upset
Missouri	at Baylor	42.5%	~25%	17.5%	+2.5	High Risk Upset

UCLA	vs. Arizona St	26.9%	~5%	21.9%	+8.5	Long Shot Upset
E Carolina	vs. S Mississippi	29.5%	~15%	14.5%	+8.5	Long Shot Upset
S Carolina	at Arkansas	29.3%	~15%	14.3%	+5.5	Long Shot Upset
Wash State	at California	24.0%	~5%	19.0%	+9.5	Long Shot Upset

We have six Odds-On Value picks this week, but none of them have a ton of value. All five games feature favorites where the public pick % is roughly 5% less than the projected win odds. That means these are solid picks, but nothing that is going to really move the needle in terms of shooting you up the standings. Still, we do suggest you choose the favorite in all five highlighted games, and slightly boost their confidence points.

Our upset picks are more interesting. For the past couple weeks, we’ve had no Low Risk Upset picks, because the public was doing a good job of evaluating the close matchups. This week, we finally have a couple. For those of you who are more than about 4 points behind, we’d suggest picking one or both of these:

North Carolina State vs North Carolina. While Vegas favors the Tar Heels, our models see this as very close to a toss up. That means the smart move according to game theory is to pick the less popular Wolfpack.
Connecticut vs Syracuse. Connecticut is actually a 1.5 point Vegas favorite, though our models give a slight edge to Syracuse. The public, however, gives an even bigger edge to ‘Cuse, meaning that your best bet to pass a bunch of opponents is to take the Vegas favorite here. That’s not a very scary play at all.

We also have several High Risk and Long Shot Upset picks. These are designed for those of you who are either in large pools (hundreds to thousands) or are far behind the leaders — basically, people that need to take some gambles. If we simulate the week 1,000 times, these picks will have a lower projected AVERAGE score, but they have higher highs and lower lows, and those highs have the best chance of giving you a significant boost over the competition. It’s a risky boom-or-bust strategy, but when you’re behind at this stage, you need to take some risks.

If you’re one of those that needs to gamble (behind by 5+), add one or two of those more risky upset picks, whichever strike your fancy. If you’re down by 8 or more points, you should probably choose several, and hope that you get extremely lucky. Here are the choices that look most appealing to us:

Iowa vs Michigan. The public is giving Iowa essentially no chance to win this game, with 95% of Yahoo contestants choosing Michigan. The Wolverines are definitely the better team, but they are only 4 point favorites. We project Iowa with nearly a 40% chance to win this game. Not great, but not a bad risk when the reward is so great.
Pittsburgh vs Cincinnati. Cincinnati has won 5 straight, which probably contributes to them being slightly overvalued here.
Missouri at Baylor. We got a support email this week that does a good job of illustrating why Missouri is underrated. The basic question was “How can Missouri be a top 15 team when they have 4 losses?” There was no mention of who they lost to, or where the games were played. If you check the Missouri football schedule, you can see that at home they’ve lost to our #1 team, and beaten #32 and #91. About all that tells us is that they’re decent. But their road performance is more revealing. They beat #13, and lost to #7, #14, and #15 — but the key point here is that their net scoring margin was only -4.25 points per game. That’s roughly on par with home field advantage (remember, these were all road games), meaning they seem to belong in the #7 to #15 range. The public simply sees a 4 in the loss clumn, however, and thinks they’re mediocre.

College Football Week 10 Pick’em Strategy: Against the Spread

As with the last two weeks, our Conservative College Football Spread-Based Pick’em Picks are simply chosen based on our predicted cover odds. These should be fine for small or medium pools. Our highest confidence picks this week are:

Remember, though, what we mentioned in our last preseason football Pick’em strategy post — if your pool has a different spread than what we list, be sure to use that to your advantage, by picking the team that the line has moved towards. As an example, if the line in your pool is Oklahoma State -12, but the current line in Vegas is -14, the smart pick for you would be Oklahoma State. You’d be getting two “free points” compared to true market value.

In huge pools (thousands of people), the small difference in our projected cover odds (generally ranging only from 45% to 55%) should probably be outweighed by the need to pick against your opponents, especially when some teams are being backed by only ~5% of the public.

As a result, our Very Aggressive picks this week are simply going to assume that the Vegas lines are efficient, and then pick whichever team has less public support. That gives you the maximum differentiation from the crowd, while costing you relatively little in win odds.

If you can see the number of people who have chosen each team in your actual pool, pick the less popular squad. Otherwise, you can use our picks for guidance, since the same teams should be popular in most pools.

Our Aggressive picks are a midway point between these two strategies. This is a good pick set if you’re in a pool with several hundred to a thousand people, or if you want to place a little bit less emphasis on our projected cover odds.

We hope this has been helpful. Feel free to ask for specific advice in the comments!