Posted:
#1
While I have frequented the covers site for quite a while, today, I decided to take a deeper look at the consensus picks and the varioius contests. I am always wary to make a bunch of comments and suggestions, as soon as I join, but then I go ahead and do it anyway. Here goes...
If we are to determine who the "experts" are, I think a couple of changes in
- The choice of metrics displayed in the columns of the consensus picks
- The choice of tie breakers in the contests
- The methodology of the user interface when making picks.
My focus here is specifically on money line contests. Presumably, the main goal of the site is to reward expertise. There are reasons to do things the way they are done, for a disingenuous purpose. It is my hope that this isn't the case, as there is a thriving community of sports betting enthusiasts here whose primary goal is to make money. I bring this up in conjunction with my first suggestion:
1. Get Rid of the Won-loss Column
I am reluctant to even bring this up, as I realize that the site's bread and butter comes from the people who sell their picks. But winning at sports betting and a person's won-loss record are two separate things. It is a far too common and IMO rather shady practice, to boast about one's impressive winning percentage, when your "tips" involve choosing the biggest favorite on the slate. So, while touting the -240 teams each day will yield a nice won-loss record, it has little to do with the maximization of profits. Contest participants, not affiliated with the site staff may also like this arrangement - they can rack up a nice won loss record and then advertise their own "service", bragging about their amazing winning percentage.
"Looky here! I have a 35-20 won loss record! It's documented right here on Covers.com! " That's all well and good, if you enjoy ripping people off, but if you actually want to demonstrate prowess, to someone in the know, you had better show a nice profit. If you amassed that won-loss record by touting 55 -240 bets, your marks er customers are going to wind up being down overall. A sharp sports bettor who picked the dogs with +200 on those same hypothetical 35 games will wind up ahead, in spite of her 20-35 record. Winning percentage is a very nice metric, when dealing with sports, in which there is a spread, as opposed to a money line, but in baseball and hockey it should be eliminated, unless your motives are less than genuine.
If the goal is to cultivate true winners, recognizing the sheer uselessness of winning percentages in money line contests is an important step in the process.
this observation leads to two suggestions:
A much more telling and useful (in the sense of determining how successful a bettor is, as opposed to conning or confusing people) is to replace the winning percentage with the average win per wager.
In conjunction with the above mentioned of a won-loss-tie record show the total number of wagers made. Call it expectation or return on investment, but the bottom line is what percent profit are we making per bet.
2. Once a significant sample size is achieved, your return on investment is much more telling than than your total winnings.
It's true one should be more impressed with a player who has made 1000 bucks, over a long season betting one dollar at a time, as compared to someone who won one bet for 1000. But, if one guy made that 1000 after 3000 wagers and another guy made 1000 after 1500, I know which guy I want to pay more attention to - the guy that makes me rich twice as fast!
This observation leads to two suggestions for metrics, which are much more telling and useful (in the sense of determining how successful a bettor is, as opposed to conning or confusing people) than the winning percentage:
- the average win per wager. (ROI or EV)
- The total number of wagers
If we want the contest(s) to mirror real life sports betting we should also take a look at the tie-breakers.
The first tie-breaker is wins. If I haven't driven my point across yet about why this isn't such a great choice of tie-breaker I never will.
The second tie-breaker is number of picks. See the argument for the "average win per bet" column to understand why I believe that that should also be the first tie-breaker. If you choose to break the tie by giving the person, who makes less money per bet, as in the example above, you are exactly rewarding the wrong person.
You do need to institute a minimum number of wagers, so as to make it statistically significant, but since we are all in agreement that the total units won is the best metric to determine the winner, it is unlikely (if not impossible) that someone would tie in this department without a statistically significant amount of wagers.
3. There is a strong and useless bias towards favorites.
In addition to the already discussed emphasis on winning percentage causing this bias, this happens because of the choice of user interface. Currently, when one selects to submit a betting action, the amount in the list box reflects the amount to be won instead of the amount to be wagered. This has the net effect of making a person's average bet size larger for favorites than for underdogs.
At the very least, Covers.com should change from amount to be won to amount wagered. I would suggest doing like the sports betting sites do and just have a text box for each, or at least offer more values in the listbox.
Making these changes would, in my opinion, more accurately reflect real life, where there are just as many opportunities to make a favorable bet on an underdog as on a favorite, and it also would make the "consensus picks" more accurate.
That's it! Other than that the site is perfect! 
Hot Forum Topics
