I do an annual college pick'em with some friends. $20 buy in for the year (so nothing crazy), no rake so it should be on the up & up. $10 weekly winner prize, at the end of the year 1-2-3 pays out 50%-30%-20%. Usually 25-35 people play. Did it for 5 years on yahoo but there were a few things I did not like about yahoo:

* As admin, you could not pick the games you wanted the group to choose from. They had top 25 games only but most o the guys in my league are Big 12 grads so I liked having Big 12 games in there especially if they were interesting. I don't think that Tech/A&M last year was in there and I would have wanted.
* As a participant, you had to choose every game. I do not like that as if we go to Vegas, you don't need to bet on every game; only the ones you are interested in. But obviously you need to have a minimum # of games that everyone needs to pick on over the year. This also helps solve the problem if people forget to get picks in for a week.
* You can never choose over/unders just against the spread with yahoo. I normally don't think O/U but would like that option.
* If the spread changes with yahoo you are locked with the last spread even if you made you picks earlier in the week with another spread. That's weak.
* Many games yahoo would have a spread as OFF and you basically just need to take the winner (say Florida v Citadel). This is weak b/c 1) I wouldn't want that in my list to pick from and 2) it skews how well you did.
* Finally the scoring. Yahoo based it on # of correct picks over the year. This hurts you if you don't get a pick in time,... but it also never shows your record. And the OFF thing skews it.

Ok, so a friend of mine created the website www.FadeThePublic.com. I asked him if he could add the functionality to host my pick'em group and he agreed. He put it in there and let me do all of the above. Which is cool. And now we can track how people are doing W-L-Push as well as # of units up or down and winning %. Very cool. People will now have more transparency in how they are doing. And this is only games they actually picked too. I would not be surprised if the median is less than 50% correct due to bias. But we'll see. Excited to see how the year plays out.

Here is the problem and the point of this thread. There is some resistance with some preferring how yahoo used to do it for the people that play. Been doing it for over 5 years. Many like the added functionality and the ability to pick & choose your games (I made it min. 140 games over the course of the year which is basically 10 picks * 14 regular season weeks).

But the way I have decided to rank everyone is # of units up or down. That is how it would be in Vegas except you can only bet 1 unit per pick. But if you want to take Texas against the spread and the over/under in the same game you can do so. Some have asked "how do you keep people from getting an early lead and quitting". That is why I have said you have to do a minimum of 140 picks on the year.

In the case of tiebreak at the end of the year for those that are up the same # of units, I have said the person who made the most # of picks. Obviously they will have a worse % than the person who made less picks but they also tried more and the more datapoints, the more things regress to the mean.

I have a weekly winner as well to keep people interested in case they are not doing well. You can win 1/2 your buy-in back with a hot week. I have also decided to this based on # of units up or down with the tiebreak going to the person who took the most picks (of those that tie). I might need to put in the caveat that you need to make 10 picks to be considered for the weekly winner prize.

So a few questions -> thoughts on how I am scoring this? Am I putting in enough incentive to make sure people don't play just a couple of games and actually make picks? Thoughts on the weekly winner prize also?

At this point, I can make a change to the scoring but I better do it sooner than later (to let everyone know). You guys are degenerate gamblers (or follow it closely) so I think you should have some good feedback.